Citation

## Material Information

Title:
Observation of pulsars and other sources at low radio frequencies
Creator:
Reyes, Francisco, 1941-
Publication Date:
Language:
English
Physical Description:
viii, 136 leaves : ill. ; 28 cm.

## Subjects

Subjects / Keywords:
Antennas ( jstor )
Bandwidth ( jstor )
Calibration ( jstor )
Flux density ( jstor )
Linear polarization ( jstor )
Pulsars ( jstor )
Pulse duration ( jstor )
Signals ( jstor )
Astronomy thesis M.S ( lcsh )
Dissertations, Academic -- Astronomy -- UF ( lcsh )
Emission spectroscopy ( fast )
Pulsars ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

## Notes

Thesis:
Thesis (M.S.)--University of Florida, 1980.
Bibliography:
Includes bibliographical references (leaves 131-134).
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Francisco Reyes.

## Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Francisco Reyes. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
023445811 ( ALEPH )
07367606 ( OCLC )

Full Text

OBSERVATION OF PULSARS AND OTHER SOURCES AT LOW RADIO FREQUENCIES

BY

FRANCISCO REYES

A THESIS PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF MASTER OF SCIENCE

UNIVERSITY OF FLORIDA 1980

to Maria Angelica
Carolina
and Daniela

ACKNOWLEDGMENTS

I wish to express my deepest gratitude to my research

advisor, Dr. Thomas D. Carr, for his guidance and encouragement, and his patience in reading and correcting the manuscript. It has been a privilege to work under him in this project, and I have benefited in many ways as his student.

I am also grateful to Dr. Alex G. Smith for providing me with financial support and for serving as a committee member. I give my appreciation to Dr. John P. Oliver for the development of the necessary hardware and software, for his advice and for serving as a committee member.

My deep appreciation goes to Dean Claudio Anguita and Professor Hugo Moreno of Facultad de Ciencias Fisicas y Matematicas of University of Chile for their encouragement and interest in my career and for providing me with financial support. I am also grateful to Jorge May and Juan Aparici for their encouragement and optimism and for providing me with the data at 45 MHz.

I wish to thank Jorge Levy for his help in the operation and calibration of the 26.3 MHz array, Chris St. Cyr for his encouragement and help in the computational part and Wesley Greenman and Richard Flagg for their help with the electronics.

iii

My thanks goes to Hector Alvarez, Mauricio Bitrin and Fernando Olmos of Maipd Radio Observatory for their help with the 45 MHz data.

I am grateful to the Astronomy Department and the Physics Department and in particular to Drs. Heinrich Eichhorn, Robert J. Leacock, F. Eugene Dunnam, Charles F. Hooper and Richard E. Garrett for their financial support in the form of teaching assistantships and fee waivers.

My gratitude goes to Mrs. Brenda Dobson for her

patience in correcting my English and for her devotion in typing and editing the manuscript.

The realization of this project has been possible thanks to the continuous support provided to the radio astronomy program of the Department of Astronomy of the University of Florida and to the Maipd Radio Observatory through a grant from National Science Foundation (principal investigator Dr. Alex G. Smith).

iv

PAGE

ACKNOWLEDGMENTS..... ...................... ............iii

ABSTRACT .............. .............................. vii

CHAPTER 1 INTRODUCTION.............................. 1

CHAPTER 2 DATA ACQUISITION AND REDUCTION............. 5

2.1 Data Acquisition of Pulsar PSR1133+16... 6 2.2 Data Acquisition of Pulsar PSR2045-16... 10 2.3 Data Reduction........................... 12
2.4 Block Designation for Each Pulsar....... 16 2.5 Computer Programs for' Data Reduction.... 19

CHAPTER 3 DATA ANALYSIS .... ........................ 22

3.1 PSR1133+16 Data Analysis.............. 22
3.1.a Peak Flux Density, Mean Flux Density
and Energy for Pulsar PSR1133+16...... 35
3.1.b PSR1133+16 Pulse Width................. 44
3.1.c Tests Performed with the Pulse
Obtained on May 25, 1979............... 48
3.2 PSR2045-16 Data Analysis............. 63
3.2.a Peak Flux Density, Mean Flux Density
and Energy for Pulsar PSR2045-16...... 77
3.2.b Pulse Width for Pulsar PSR2045-16..... 83
3.2.c Period of Pulsar PSR2045-16 from
Observations on Two Consecutive
Days................................... 87
3.2.d PSR2045-16 Pulse Intensity as
Function of Time ...................... 91

CHAPTER 4 CONCLUSIONS ............................... 93

APPENDIX A LIST OF PULSARS............................... 103

APPENDIX B SAMPLING RATE, TIMING PULSE FREQUENCY
AND DATA POINTS PER SECOND.................. 107

APPENDIX C SCHEMATICS..................... ............ 110

v

APPENDIX D COMPUTER PROGRAMS............ .............113

APPENDIX E 26.3 MHz ARRAY CALIBRATIONS .......... 123

LIST OF REFERENCES ................................... 131

BIOGRAPHICAL SKETCH................................... 135

vi

Abstract of Thesis Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

OBSERVATION OF PULSARS AND OTHER SOURCES

By

Francisco Reyes

December 1980

Chairman: Thomas D. Carr
Major Department: Astronomy

The investigation of the radio emission spectra of

pulsars at low frequencies is important since it helps to clarify the mechanism responsible for this emission. At the present, the low frequency end of the spectra is poorly determined compared with more precise measurements available at higher frequencies. Also, measurements of pulsar intensity variations at low frequencies due to scintillation are of value in the study of the interstellar medium.

Detection of pulsars at decametric wavelengths is

difficult because of the deterioration of the signal-tonoise ratio caused by the high galactic background temperatures and the turn-over of the spectra of most pulsars. Also the dispersion of the pulses imposes the use of narrow bandwidth, reducing the sensitivity of the system.

vii

Two pulsars have been observed at low radio frequencies using narrow bandwidth to avoid pulse dispersion problems. Pulsar PSR1133+16 was observed at 26.3 MHz using the University of Florida large array; pulsar PSR2045-16 was observed at 45 MHz using the Maipu Radiobservatory large array (Chile). Long term integration techniques were used to reduce the data.

Both pulsars were successfully detected and the average pulse intensity and integrated pulse profile were obtained. Time resolutions of P/38 were obtained for PSR1133+16 and P/62 for PSR2045-16, where P is the pulse repetition period. The two observed pulsars were selected because of their low dispersion measure, long period (>1 sec), high pulse intensity and their location with respect to the galactic background.

Several other radio sources were also observed and were used to calibrate the effective areaof the 26.3 MHz antenna.

Chairman

vii

CHAPTER 1
INTRODUCTION

Pulsars were discovered in 1967 by S. Jocelyn Bell and Anthony Hewish. The first pulsar paper was published in Nature in February 1968 by Hewish et al. (1) telling about the discovery of pulsar PSRI919+21 and giving some of its parameters. Since then, about 321 pulsars have been detected and cataloged.

According to currently accepted models, pulsars are rapidly spinning neutron stars originating from supernova explosions. They are believed to have strong magnetic fields on the order of 1012 gauss, probably the strongest known magnetic fields in the Universe. The neutron star is believed to emit continuously a narrow beam of radio waves which corotates with the star, producing the observed pulse each time the beam sweeps past the earth. The periods of the pulses which have been observed range from about 0.03 to 4.3 seconds. So far, for only 2 pulsars has it been possible to find the optical counterpart of a pulsar.

Several mechanisms have been proposed to explain the origin of the radio emission and how the rotational energy of the neutron star is converted into electromagnetic wave pulses; this subject still remains one of the least understood aspects of pulsars. Some of the important parameters

1

of a pulsar that need to be measured before realistic models of the radio emission mechanism can be developed are the spectrum of the radiation, the pulse shape, peak intensity, and the time variations of these parameters. A large number of measurements have been made in the range of frequencies between 100 and 2,000 MHz, but only a few below 100 MHz (2). The low frequency part of the spectrum is particularly interesting because most pulsars present a turn-over of the spectrum at metric and decametric wavelengths, and also because pulse shape and widths show some peculiarities in this region. The spectral maximum for most pulsars is in the range 60 to 250 MHz; this peak is nearly symmetrical with respect to the frequency at which the maximum occurs (2). Measurements of the low frequency part of the pulsar spectra have been made by Bash et al. at 38 MHz (3), Bruck and Ustimenko at 10, 16.7, 20 and 25 MHz (4, 5), Craft and Comella at 40 MHz (6) and more recently by Izvekova et al. at 61 and 102.5 MHz (7, 8).

Most pulsars measured at low frequencies are in that part of the sky which is observable from northern hemisphere observatories. Observations at the lowest frequencies have been carried out in the southern hemisphere only by Australian observatories, at 80 and 150 MHz (9, 10). Detection of pulsars at low radio frequencies is difficult because of the deterioration of the signal-to-noise ratio caused by the high galactic background temperatures, and the low frequency spectral turn-over exhibited by most

1 3

pulsars. Also pulse dispersion in the interstellar medium becomes more severe the lower the frequency. At higher frequencies, compensation can be made for dispersion, permitting the use of relatively high bandwidths. However, at the lower frequencies, it is generally necessary to restrict the bandwidth because of dispersion, further reducing the sensitivity of the detection system. Due to the intrinsic short pulse durations of most pulsars (5 150 msec), short time constants must be used, imposing still another limitation on the sensitivity of the system.

Nevertheless, the sensitivity of the system can be

improved by averaging many independent records of the same source. If N records are averaged, the signal-to-noise ratio (S/N) is improved by a factor of /N For pulsars, this can be accomplished by averaging groups of consecutive periods. Digital sampling is generally used in such averaging. Due to the relatively short pulsar periods, the signal needs to be sampled at a fast rate in order to have enough time resolution to meet the requirements of the sampling theorem. Since it is not feasible to manipulate the large amounts of data generated in the digitizing process by hand, digital computing techniques must be used, providing an efficient way to average, process, store and plot that data.

In the past few years, the development of small and inexpensive microprocessors has made possible improved methods of pulsar data handling and processing. If the

4

period of a pulsar is known to be a good accuracy, it is possible to average many periods (N typically in the range of a thousand periods), greatly improving the S/N ratio.

The University of Florida Radio Observatory, located

in Dixie County about 50 miles west of Gainesville, and the Maip6 Radioastronomical Observatory of the University of Chile, located 20 miles southwest of Santiago, have long been engaged in a cooperative program of low frequency radio astronomy. Although most of the effort over the years has gone into the study of Jupiter's radio emission, other decameter-wavelength sources are also investigated. Each of the two observatories has a very large antenna array, in addition to many smaller radio telescope antennas. The Florida array consists of 640 dipoles operating at a center frequency of 26.3 MHz, while the Chile array has 528 dipoles and operates at a center frequency of 45 MHz. Both arrays were believed to be suitable for observing pulsars, although previous attempts to use them for this purpose had not been successful.

The goal of the research on which this thesis is based was to develop a low frequency pulsar data acquisition and reduction system, and to use it to detect and measure the important parameters of at least one pulsar with each of the two large arrays (i.e. the one in Florida and the one in Chile). This goal was achieved.

CHAPTER 2
DATA ACQUISITION AND REDUCTION

Two pulsars, PSR1133+16 and PSR2045-16, were selected (from the lists in Appendix A) for observation because of their low dispersion measure, relative long period (>1 second), high pulse intensity and favorable location with respect to the galactic background. Pulsar PSR1133+16 was observed at 26.3 MHz using a bandwidth of 8 kHz, and pulsar PSR2045-16 was observed at 45 MHz with a bandwidth of 10 kHz.

The detected IF signals of the pulsar were recorded on one channel of a magnetic tape, and a time code generator signal on the other channel. The tapes were played back (making a 4:1 slow down), the pulsar signal sampled, and the resulting values stored in the NERDC Amdahl 470 computer. The time code generator signal provides the timing pulses for the sampling and also indicates Universal Time (U.T.). The sampling of the signal was accomplished by a 12 bit analog/digital converter (A/D) and a KIM I microprocessor. Pulsar pulse averaging is carried out by a computer program which averages consecutive periods of the sampled pulsar signal, previously stored in the Amdahl computer.

5

6

2.1 Data Acquisition of Pulsar PSR1133+16

The pulsar PSR1133+16 was observed for 20 nights

between April 25, 1979, and June 8, 1979, using the University of Florida 26.3 MHz large array, located at the University of Florida Radio Observatory in Dixie County. The 26.3 MHz array is a nearly filled rectangular array of 640 half-wave dipoles that covers about 30,000 m2. The half power beamwidth (HPBW) is 2.50 north-to-south by 6.00 east-to-west. It is a phase steered array with full N-S directional control of the beam in quarter-beamwidth increments, and a selection of 8 E-W beam directions spaced one beamwidth apart. Phasing is accomplished by means of a network of Butler matrices, hybrid rings and plug-in phasing cables.

North-south phasing of the array for a source having a given declination is accomplished in two ways, a coarse adjustment in large increments of declination and a fine adjustment providing small steps. A change in the coarse phasing requires the replacement of many plug-in cables located throughout the array, and requires about 2 or 3 manhours of labor. Fine phasing for the desired declination is quickly accomplished by means of switches at a single location. The beam pointing can be altered in small increments to any declination within a strip roughly 100 wide by means of the fine phasing controls. Positioning the beam to a new declination outside this strip requires a coarse phasing readjustment. The right ascension of the beam is of course

7

controlled by the rotation of the earth. Scans by any or all of the 8 selectable east-west beams (designated 4E, 3E, 2E, 1E, 1W, 2W, 3W, 4W) are made by starting each scan at the appropriate sideral time. A computer program called ARRAY developed by Dr. M. D. Desch, is used to obtain information on cable lengths required for coarse phasing, switch positions for fine phasing, and beam transit time across the desired source (11). A more detailed description of the antenna and its adjustment has been given by Desch, (12) and Desch et al. (13).

In order to use the maximum gain of the array, only the central E-W beam positions 2E, lE, 1W and 2W were used, and within each beam only the central portion of it was recorded (%8 minutes each beam). A maximum of 32 minutes was recorded each night. Pulsar PSR1133+16 was selected because of its relatively high pulse energy (%0.75x10-26 Wm-2Hz1 at 61 MHz), long period (1.188 s), and low dispersion (DM = 5
-3
parsec cm-3). It was also easy to reach this pulsar using the phasing the antenna had at that time (phased to make Jupiter observations), without having to completely rephase the array. Only the fine phasing needed readjustment. During the period in which the pulsar was observed, it was transiting at about midnight.

Only 8 nights (out of the 20 observed) gave reliable

data, free from static and stations. The end of the observation period coincided with the beginning of the thunderstorm

season, at which time it was necessary to shut the array down for the summer.

A block diagram for the interconnections of the equipment used to observe the pulsar is shown in Figure 2-1. The equipment basically included a Collins R 390 receiver, a Magnecord 1022 magnetic tape recorder, a Tracor 5D frequency standard, and a Systron Donner 8154 time code generator/reader.

The time code generator (TCG) was synchronized against

WWV; this was accomplished by triggering an oscilloscope with the 1 pulse per second (1 PPS) output of the TCG and measuring the delay of WWV 1 PPS. The 1 PPS of the TCG was brought as close as possible to the 1 PPS of WWV (usually between

5 to 20 ms) by making short interruptions of the 1 MHz signal from the Tracor frequency standard. The delay between the two 1-PPS pulse trains was measured to an accuracy of about 0.5 ms, and from that information the U.T. was obtained, except for the propagation delay between Boulder, Colorado, and Dixie County Radio Observatory. No corrections have been made for this delay because no attempt was made to relate the epoch of this observation with any other observations.

The measured drift of the Tracor time standard is about 80 Ps/h; this causes a drift in the epoch of about 100 ps between pulsar transits by beams 2E and 2W (lhl16m); this drift is small enough to be noticed, considering the time

26.3 MHz ARRAY

MAGNECORD
COLLINS 1022
R 390 N CHI RECEIVER CH2

SYSTRON
TRACOR 5 D SYSTRON
FREQUENCY DONNER 8154
STANDARD TCG TCG

TRIGGER
WWV INPUT

Figure 2-1. Block diagram of the equipment
interconnections used to record
pulsar PSR1133+16.

10

resolution of %31.5 ms (P/38; P = period) obtained for this pulsar.

The detected IF output of the receiver was recorded on channel 1 of the Magnecord at 15 IPS and the recording level for the galactic background was set to about 7 VU on the recording level monitor, to prevent any saturation during signal peaks.

The time code generator was recorded on channel 2 of the Magnecord. This signal serves a double purpose: it provides a time reference for the pulsar signal, and later on, in the data reduction process, provides the timing signal for the KIM I microprocessor.

As a back-up to obtain the correct time, at the beginning of each observing session the WWV signals were recorded on channel 1 (the TCG also on channel 2) for about 2.5 minutes. The correct time can be obtained by measuring the delay between the 1 PPS of WWV and the 1 PPS of the TCG.

The receiver was tuned to a center frequency of 26.3 MHz and the bandwidth set to 8 kHz. Using data for higher frequencies an extrapolated pulse width (at the base) of about 80 ms was obtained. Pulse broadening due to dispersion
-3
(DM = 5 parsec cm-3) is about 18 ms (for a bandwidth Av = 8 kHz), which was considered acceptable.

2.2 Data Acquisition of Pulsar PSR2045-16

Pulsar PSR2045-16 was observed for 11 nights between

May 15 and May 30, 1979, using the Maipd Radio Astronomical

11

Observatory 45 MHz large array, located about 20 miles south-west of Santiago, Chile. The 45 MHz array is a filled rectangular array of 528 full-wave dipoles with an effective area of 12,000 m It is a transit instrument with 6 independently phased subsections in the N-S direction and can be oriented to a zenith angle of about 450 in the meridian plane. The HPBW is 2.10 north-to-south by 4.60 east-to-west. A multiple beam system is generated by a Butler matrix. A more detailed description of the array has been given by Reyes (14) and May et al. (15).

Pulsar PSR2045-16 was recorded for 20 minutes each night, which is the time it takes to drift between the E-W 3 dB points of the beam. This pulsar was selected because of its relatively high pulse intensity (al.0x10-26 Wm-2Hz at 34 MHz), long period (1.96 ms) and its low dispersion measure (DM = 12 parsec cm-3). A zenith angle of only 180 at transit was also another favorable factor considered.

The equipment used for the data acquisition of this pulsar consisted of a modified General Dynamics telemetry receiver, a Magnecord 1022 magnetic tape recorder, a Systron Donner 8220 TCG and a Rhode and Schwartz quartz time standard. The interconnections are similar to those shown in Figure 2-1. During the observing sessions the pulsar was transiting at dawn, and no interference was noticed. Nine of the eleven nights gave reliable data; the other two nights were discarded because of equipment malfunctions.

12

A bandwidth of 10 kHz was used; this is the minimum bandwidth of the receiver. Using data obtained at higher frequencies, an extrapolated pulsar pulse width at the base of about 130 ms was obtained at 45 MHz. The pulse broadening due to the 10 kHz bandwidth is 11 ms. This was considered acceptable.

The quartz time standard is checked daily against the

Chilean National Observatory cesium time standard; the epoch is believed to be known with an accuracy of about +10 ps. The drift in time is about 3.6 ps/h (or about 0.00190 ps/P, where P = 1.96 s, the pulsar period). The drift is 1.2 ps in the 20 minutes between the east-to-west HPBW. For all practical purposes considered in this research, this is negligible.

2.3 Data Reduction

Data reduction for both pulsars was performed using the

facilities of the Department of Astronomy of the University of Florida in Gainesville. The basic configuration and interconnections of the equipment for data reduction are shown in Figure 2-2. The schematics for the most relevant blocks are presented in Appendix C.

As stated earlier, the IF signal in the receiver is

band limited to 8 or 10 kHz and appropriately smoothed. The resulting audio frequency signal is recorded in analog form on channel 1 of the Magnecord tape recorder at a tape speed of 15 IPS. The time code generator is recorded on channel 2.

13

MAGNECORD 1022

CH1l CH2

DETECTOR SYSTRONS PHASE LOCKED
TIME CONSTANT DONNER LOOP
DC AMPLIFIER 8154 DC
AND OFFSET TCG OUTPUT

1 PPS '250 Hz

A/D SYNCHRONIZER
CONVERTER AND
(12 BITS) DIVIDER : 2 125 Hz

TIMING
PULSES BRUSH
KIM I CHART
RECORDER

SYNCHRONIZING
SWITCH

TO
HAZELTINE AMDAHL 370 TERMINAL COMPUTER

Figure 2-2. Block diagram of the data reduction equipment.

14

In order to achieve the desired time resolution and to be able to sample the data at an appropriate rate, the magnetic tapes were played back at 3 3/4 IPS, providing a slow down of 4:1.

Pulsar signals from channel 1 of the Magnecord are sent through a diode detector and smoothing filter to the 12 bits A/D converter. The filter time constant was adjusted to 45 ms for both the rising and falling edges of the noise pulses.

The TCG signal from channel 2 of the Magnecord is connected to the input of the TCG (working now as a reader) to provide both a visual display of the time and a 1 PPS pulse (taken from a modified output). The latter gives the synchronizing signal to start the sampling of the data at a precisely known time. It is particularly important to know the time at which the sampling is started in order to be able to combine in phase 2 or more blocks of data.

In order to obtain the timing signal for the KIM I, the TCG signal from channel 2 of the Magnecord is connected also to a phase locked loop (PLL) circuit. The PLL circuit locks at a particular frequency of the signal, filtering out the rest of the frequencies present in the complex wave form of the TCG signal. The 1000 Hz signal of the TCG is transformed to 250 Hz due to the 4:1 slow down. The PLL was carefully tuned to 250 Hz, and the output then divided by 2. The resultant 125 Hz constitutes the timing signal for the KIM I.

A gated DC output from the PLL circuit was monitored on a Brush Mark II chart recorder. This provides a good means

15

to detect any loss of synchronism or dropout of the TCG playback signal. If there is any loss of synchronism between the input and PLL signals, then the gated DC output falls to zero, giving a visual indication to stop the sampling process. If synchronization were lost, the pulsar signals would be averaged out of phase, wiping out any pulse at the end of the integration process. A means was provided for starting the sampling at the next 1 PPS pulse from the TCG occurring after a switch was thrown. This made it possible to know the exact time at which sampling started. This was accomplished by connecting the start switch to the D input and the 1-PPS signal to the clock (c) input of a D-type flip-flop. The output of the flip flop gates the 125 Hz signal in a NAND gate (see schematic in Appendix C). When the start switch is turned on, the next 1 PPS of the TCG enables the 125 Hz signal to go into the KIM I and the sampling is started.

The sampled values (coded in hexadecimal) are sent

through the Hazeltine terminal to the Amdahl 470 computer and stored there for further processing and analysis.

Since the maximum number of cards that can be stored under a new data file name is 493, it was necessary to break up the data into several "blocks" to meet this requirement. The 8 minutes of data obtained in one beam for pulsar PSR1133+16 were broken up in 2 blocks of 4 minutes each, generating.about 312 cards for each block (for a TSAMP= 02 and timing signal frequency = 125 Hz; see Appendix B). The

16

20 minutes of data obtained each night for pulsar PSR2045-16 was broken up into 4 blocks of 5 minutes each, generating about 390 cards for each block.

The method of breaking up the data into blocks proved to be useful also because one can inspect the results of the average of each block separately, making it possible to search by eye for a possible pulse that might appear at the same location in the different blocks. It is highly unlikely that interference transients would appear in the same location in each of several blocks.

Monitoring of the DC gated output of the PLL circuit is very useful since synchronism is sometimes lost when a tape is played back. If that happens, then that portion of the tape is skipped and a new block started.

The KIM I program which is necessary to digitize the

data is loaded into the microprocessor from a cassette tape recorder.

2.4 Block Designation for Each Pulsar

Many blocks of data were generated in the data reduction process. In order to avoid any confusion with the data and to be able to distinguish quickly between blocks, a designation scheme was developed for referring to a specific data block. The designation format contains in abbreviated form the pertinent information for that block. For pulsar PSR1133+16, the following block designation was adopted:

17

P 1 1 XX Y

Number that designates the night when the pulsar was recorded Letter that designates the block within the beam

Beam in which the pulsar was recorded Designates pulsar PSR1133+16 Antenna beam designations are as follows:

2E 2nd beam east of transit lE ist beam east of transit 1W ist beam west of transit 2W 2nd beam west of transit

In order to distinguish between the different blocks

within a given beam, a capital letter was used, usually A and B. In case it became necessary to break up the data into more than two blocks (i.e. due to loss of synchronism of the TCG signal) C was also used.

The indicator for the date of the night on which the pulsar was observed was an integer, as follows:

Night Date

1 April 25, 1979 2 April 27, 1979 3 April 28, 1979 4 April 30, 1979

5 May 20, 1979 6 May 25, 1979 7 May 26, 1979 8 May 30, 1979

18

The designation scheme for the blocks of data from pulsar PSR2045-16 is slightly different, as follows:

P 2 04 5 X Y

Number that designates the night when the pulsar was recorded Letter that designates the block within that night

Designates pulsar PSR2045-16 The block designation is usually one of the letters A, B, C, D, E or F.

The nights were numbered as follows:

Night Date

May 15, 1979

1 May 16, 1979 2 May 17, 1979 3 May 18, 1979 4 May 19, 1979 5 May 22, 1979 6 May 23, 1979 7 May 24, 1979 8 May 29, 1979 9 May 30, 1979

The reduced data for each block was stored on a disk

called UF.B0030601.S5.XXXXX, where XXXXX is the name of the block. The data have been recorded also on a digital magnetic tape called PULSAR.

19

2.5 Computer Programs for Data Reduction

In order to analyze the data and to plot the values, a set of two computer programs called PULSAR and MCAPGM are used. The basic parameters necessary to analyze the data must be given to the PULSAR program; they are:

N = Number of bits digitized (=12)

TPP = Timing pulse period (ms)

NPA = Number of pulse intervals averaged

P = Pulsar period (s)

PLOFF = Plot offset

MAG = Plot magnification factor

The MCAPGM program (using the PULSAR program parameters) takes the first group of pulse intervals to be averaged (NPA) from the data of a particular block and calculates the overall average and the standard deviation. These two quantities, and the average for each bin, are plotted. The program next takes the second set of intervals from the data and does the same as before. Also the first set of intervals is averaged with the second set, computing a new average (cumulative) for each bin, an overall average, and a new standard deviation and plots the values for the bins. This process continues including each time the next group of intervals (NPA).

The first set of programs is run once only to check the quality of the data,to detect any problem and to obtain the total overall average for a block which will be used in a second run.

The second set of programs called PULSITO and MCAPGM1 is similar to the already described PULSAR and MCAPGM, but

20

this time, the data is taken from disk (where it has been previously stored), produces a magnified plot and punches cards with the final values for each bin, the total block average and the final standard deviation. This set of cards is used later on, to stack 2 or more blocks.

Usually after the process already described, the digitized data is recorded on the PULSAR magnetic tape and released from the data files.

In order to stack 2 or more blocks, the sets of cards of each block (except the first, which is used as reference) is shifted the proper number of bins so that the bins can be averaged in phase. The number of bins to be shifted in each block is given by the following relationship:

Nbi {- [Frac( TP -T) +1 (2-1)
bins

where Nbins = Number of bins the block i needs to be shiftedwith

respect to block 1

T1 = Time start sampling block 1 T = Time start sampling block i

P = Pulsar period
P
R = Time resolution (= ber bins) Number bins
Frac = Fractional part
Ti-T
The fractional part of 2 1 corresponds to the fraction of the period between the time of beginning of block i and the time of beginning of the first block. The relationship given by 2-1 was evaluated using an HP-29C calculator.

21

A last program called P2045 (or P1133) reads the cards

of the blocks to be averaged (previously shifted), calculates the average for each bin (weighting by the number of periods in each bin), computes the final average of the combined blocks, the standard deviation a, and plots the data. Also plotted is the 3 a reference line, which is used as a reference to aid in deciding whether or not a pulse was present at the end of the averaging.

The KIM I program to digitize the data and the MCAPGM program were developed by Dr. John P. Oliver; MCAPGM1 is a modification, made to adapt it to the use previously explained.

A program called DOPVEL was used to obtain the apparent pulsar period for the night and time when the pulsar was observed. This program gives the pulsar period at a particular site and at a given time, correcting for Doppler shift due to the earth's rotational velocity and its motion around the sun. This program was developed at the NRAO by Drs. R. N. Manchester and R. A. Gordon and made available by Dr. Steve T. Gottesman.

Finally a summary of the times spent observing and reducing pulsar data and the total time of terminal use is presented in Table 2-1.

Table 2-1. Summary of time spent with pulsar data.

Total Time Total Time Total Time Observing of Digitizing
Pulsar Pulsar Data Reduced Data

PSR1133+16 10h 40m 4h 10m 16h 04m PSR2045-16 3h 40m 3h 12h

CHAPTER 3
DATA ANALYSIS

3.1 PSR1133+16 Data Analysis

From the 8 nights of data reduced for pulsar

PSR1133+16 (26.3 MHz, Florida), pulses were found with no ambiguity on 2 nights. Pulses were found on April 28 and May 25, 1979, over averages of 340 and 390 periods respectively; the results are plotted in figures 3-1 through 3-7.

On April 28, after averaging over 340 periods (%6m44s a pulse appears in bin 33 with an amplitude of %3 o above the mean value (Figure 3-1). The 2 data blocks P112EA3 and P112EB3 used to obtain this final block have been plotted separately in figures 3-2 and 3-3. It can be seen that a pulse appears in both blocks at the same location (bin 33). Even though the pulse appearing in each block would not alone be considered to be statistically significant, the fact that they appear in the same location for two separated sets of data provides an excellent test of the validity of the pulse.

On May 25, after averaging over 390 periods, a prominent pulse was obtained in bin 25 with an amplitude of

3.4 a; this pulse is shown in figure 3-4. The same test for validity as before was made. Blocks P112WA6 and P112WB6 are 22

DULSA4 PS1133+16 DEIOD=1.1874I9 SEC FWEQUENCY=26.3 MHi

P112EA3 APRIL 28 7') TSTART 02H 03M 0)5 STOP 02H 05M 50S TSAMP 02 Pil2t83 APRIL 28 79 TSTAQT 02H 05M 55S STOP 02H 10 005 TSAMP 02

BLOCK AVERAGE= 120.lbi SIGMA= 0.089 3 SIG4A= 0.?68 MAGNIFICATION= t00

NUMILUA IF PLOCKS= 2 AVIEZAGING TIME= '?q4.6 SFC

NT. AVEAGF EIODO9 TNTNSITY ', HE; v ~AGE)
I 1C.0D24 343 I
2 123.156 340 +
3 120.123 340 I
4 123.197 140
5 120.197 340
6 I). 20 340 7 123. '9 34 I + 9 123.174 340 I + o 120.71 340 1 + ID 123.120 140 I II 120.12 340
12 120 .332 I 5
13 120.1J i 340 I 14 120.1)7 340 I
15 120 ,0 343 1 + 16 12:.0?o 340 17 120.23e 340 TIME + IS 12 174 343 + 1 1 2,.?00 340 + 2) 120.138 340 1
1l.)97 340 I + 22 120.13 340 1
23 120 .291 340 + 24 123.259 340 25 120.353 340 1 26 120.121 340 I + 27 123.0r4 240 ( 2b 123.224 343
29 123.JE 340 ( + 3 12C.141 340 1 31 123.103 340 1
2? 120.091 340
12.7 30 9 1----0 33 3.0 SIGMA 33 123.4P7 340 +
34 120.273 340 I +
35 10.1e 340 I + 36 123.173 340 I
37 123.097 340 I+ 32 120.350 219 +

Figure 3-1. Pulsar PSR1133+16 (Blocks Pll2EA3 and Pll2EB3) recorded at 26.3 MHz

on April 28, 1979. Average over 340 periods.

,trULSAC 5 IPt133I, D-C R =lt.I ?7' SIC fRFUVUE NCY=26.3 MHZ

0DIC2EA APE1L 2M To TSTAR 021 O31 00OS STCP 02H Ot m 50S TSAMP 02

IPLCK AVEIAGr= 12C0.160 SIGMA= 0. 137 3 SIGMA- 0.4tt MAGNIFICATION= 100

NU*4 SF POF 1ILCCKS= I AV1rAGI(NG T I MlE= 36?.5 SIC

sIN AVE AGF rF ICDS INTENSITY
I E k A I 1A ,1, 0

2 I. 1)7 lg140 3 123. 193 (40
4 1 3 .32 1 ) 10
5 12C.150 140
6 t 2 ).2 9 140 I .
7 1L3 .314 143 1 .
1 120.150 140 I S 120 J.337 140

12 120.1 0 140 I 13 120.193 140 14 120 3 33 140.* 15 120. 32 140 I* If 120.143 140 17 123.?57 140 I 18 120 .243 140 1 + 19 123.'07 Il 1TINE .) 219 123.237 140 20 ?0.3I 140 21 120.)?9 140 ( 22 123.357 140 23 120.33 140 + 24 1 34 1 140* 25 11.97 9 1 40 I 26 120.27q 140 27 123.'?9 140 1 28 120.329 140 I
2 1230 7 ; 140 33 I 1 .971 14J 31 120.314 140 22 12214 140 *33 33 123.479 140 I 34 120.314 140 + 35 12) .136 140
36' 120.037 140 I 37 120.107 140 1 39 12j.356 19 I I

Figure 3-2. Pulsar PSR1133+16 (Block P112EA3) recorded at 26.3 MHz on April 28,

1979. Average over 140 periods.

PUL SAW P 1 1i33+Ib PRIf)OD=1.IA7999 FEC FrEOUENCY=(6.3 MHZ

P112ED3 APRIL 28 79 START 02H 05M 55S STCP 02H 10M 005 TSAMP 02
1LCCK AVE;AGEF= 119 620 SIGMA= 0.114 3 SIGAA= 0.342 MAGNIFICATION= 100

Nh4RER OF BLOCKS= 1 AVERAGING TIME= 232.1 SEC

FIN AVEWAGF DOE~ LurS NTENSTTY '^oER AVfEAGED
I 119.I15 2'00
2 li .N50 200 I
3 119*.545 200 I
4 119.7PC 200 I 11 .f90 ?00 + 6 119.*5) 200 I
7 ,11 .?7C 2CC I t I I.- 230
l 119.2C 200
13 119.615 203
i 1I'19 .30 200 + 12 119.120 25 1' 114~P65 200 ) 14 1IJ19 77 200 .
1 11 4.545 2CO
It 11i;.515 7100
17 119. q5 200
18 119.58 2C0 ( 19 19. 655 200 IE 20 ',119.'70 200 21 119.435 203 22 11,.44.5 203 + 23 1I1.720 2C0 I 24 119.730 20 25 119.5t0 200 I 2(- ll .47C 200 I + 27 114. "QI 200 I 28 llv 513 2003 2,i 1 C'..'C 2C3 I

31 115.670 200 I 0' 11**465 200 I ,3 33 11'. 50 200 I 33 '4 I11.705 2CC +

36 1L.75C0 00 I + 27 I 0 200 I I1 19.I10 200 I

Figure 3-3. Pulsar PSR1133+16 (Block P112EB3) recorded at 26.3MHz on April 28,

1979. Average over 200 periods.

DULSAR P5~1133+16 PRS I3 I .I3IO SEC FPEQUENCY=26.3 MHZ

PII?%A6 MAY 25 79 TSTAFT 01t 11OM 10S STOP 01H 34M 105 TSAMP 02 0112wF6b MAY 25 79 TSTART D01N 34M 10S STOP 01H 38M 04S ISAMP 02

LOCK AVTDAGE= 122.609 SIGMAI 0.081 3 SIGMA= 0.243 MAGNIFICATION= 100

NUMBER OF BLOCKS= 2 AVERAGING TIME= 452.7 S"C

sIN AVEJ AGE ~E, lOO INTENSITY RJMBIER AVE AGED
I 123.081 224 I + 1?2.b'2 35o0 I S 122.722 390 I. +
4 122.(10 30 I*
5 127.6 02 3'0 1*
6 122.RSQ 390 1
7 122.,13 3 0 I + 12.7 92 30 *
S 122.7be 390 I
1 122.t4e 350 I 11 1?2..3C 350 I
12 122.5320 3-10 + 13 122.543 390 1 14 122.510 9 I *02 15 12'.5 3 0 + 16 172.20 3.0 I. 17 122.505 350
18 122.607 390 .
1 122.546 390 (TIME 25 122.4L4 350 I 21 122.0i 350
22 122.582 350 I
23 122.497 3;0 I 24 122.700 350 I 25 3.4 SIGMA 25 122.894 390 I 25 122.728 390 I 27 122.t400 30
4 122.52C 3,0 I .
P3 122.515 390 I. 14 122.597 390 .
21 122. (3M 3 0 1 +* 32 122.572 31,0 33 122.f43 3S0 1 *
4 122.533 350
35 ?2.;5 1 3,0 *
36 122.55J 30 I '7 122.595 30 39 12?.555 216 ( *

Figure 3-4. Pulsar PSR1133+16 (Blocks P112WA6 and P112WB6) recorded at 26.3 MHz

on May 25, 1979. Average over 390 periods.

uQ7 11 7"
Mm
(D
CD --- (To A

m Ln

( ))0 ,0000 0 t 0 0U 70400 o 07 00 )) 0 U a, 0 0 3, C2 r A
A ) n.

D U) ~
a x

<11 Cn J > (D r
J 0 .

A A

OD to

oo >

*0
+l 1 0

(D J.
~JH '7- on0

Zt E I

N N

t m
0"1~

N
tQ OX I

OH

L

LZ
-i w
t'Q N 74 37+ I NIS C 00n
GO: O 0
N

t-S ~ ~ ~ ~ ~ ~ ~ I 0 I IfI \I Or.. 00\ I I. ,c
(D A La (A4

to h

C)PO

Z L
NM ON

VXX

ii 7

PULSAP PSkII33+16 OFR1IO=II 10?3 SEC FEQUENCY=26.3 MHZ

P112W86 MAY ?S 79 TSTART 01H 34M 10S STOP 0IN 38M 045 TSAMP 02

ELOCK AVERAGE= 121.800 SIGMA= 0. 105 3 SIG4A= 0.315 MAGNIFICATION= 100

NUMBlEF CF BLOCKS= 1 AVERAGING TIME= 220.5 SEC

RIN AVErAGE PFA1OS INTENSITY NJ W E R AVEWACD
S 121.7'2 24 I*

3 121.P95 190 I
4 121.753 193 1

1 121.761 190 I
7 121.300 1 003 1 9
1 121 3005 I. II 121.92 I. 0 12 121.742 10 .
13 121.768 0 + 1I 121.705 190 I 15 121.975 191 r 16 121. 42 I O I. 17 121.821 I 0 I 19 121.942 190 1 TIME 19 121.762 19 I 2C 121 .637 19 3 + 21 121.737 IS3 I 00 29 121 .589 190 I + 23 121.595 190 1 + 24 121.926 190 I 25 25 122.042 190 I + 26 121.874 1sO I + 27 1-l. 63 193 I + 25 121.5 5 190 1
2 121.747 19 I + 33 121.795 193 I + 31 i12170 1SC0 + 3' 121.70' 1 0 I + 33 121.755 190 I
24 121.637 190 I +* 35 121.21 I 10 I
30 121 .16 1 10+
3 121.12 10 1 31 121.726 190

Figure 3-6. Pulsar PSR1133+16 (Block P112WB6) recorded at 26.3 MHz on May 25,

1979. Average over 190 periods.

29

t(

r-4

0

N

0 e4

-, NN I i IlNA 3

I 1I >01
e. 0000000

of .. ........... ....... .... **

C----- ---------.----F---- ---------- ....a
4
- 00- -0

- - - - - - -OI lr cca

2O
1. L 2 >C r -7 "1
L n CL .a t < L F C O N

1---- A 1r .
H

C--

4hr-~h~.1-I

-.+u i S -'t if UN C C-C.I 7U t 31 l LUi~ih i U*Ci i0ilit~ O T

30

plotted separately in figure 3-5 and 3-6. A pulse appears in bin 25 in block P112WA6, with an intensity 2.6 o above the mean value; also in block P112WB6, the pulse is weaker, being only 2.3 o, but is at the same location (bin 25). Once again we have the fact that 2 pulses appear in the same location in two independent sets of data.

From a precise knowledge of the pulsar period and the elapsed time between the April 28 and May 25 appearances of the pulse, it should be possible to preserve its phase so that it would reappear in the same bin on the second date as it occupied on the first. However, this could not be done because of the period uncertainty. This problem will be addressed later when analyzing PSR2045-16 data.

The pulsar was detected in only one beam on each of the two nights, disappearing when the beam was switched to the next position. This phenomenon will be discussed later in the conclusions, when polarization effects are considered.

Before other tests made on this pulse are discussed, it is interesting to note that in bins 9 and 21 of block P112WA6 (figure 3-5) two more small pulses can be distinguished. Although neither of them alone would be considered statistically significant, they both seem to appear again in block P112WB6 (figure 3-6) in the same position. The pulse in bin 21 leads the main pulse in bin 25 by 125 msec (resolution %31 msec/bin) or approximately 380 of phase.

The pulse in bin 9 is separated by 496 msec from the

main pulse, which is about 1500 (or 0.42 P, P = 1.88023 sec).

31

The pulse in bin 21 suggests the existence of a

subpulse leading the main pulse. Several authors have indeed reported for this pulsar a complex main pulse composed of 2 subpulses, the leading subpulse decreasing in intensity with respect to the trailing subpulse as the frequency decreases (6, 7, 8). The separation between the two subpulses has been measured at higher frequencies and extrapolation from that data suggests a much closer separation for low frequencies than the separation obtained at 26.3 MHz between the pulses in bins 21 and 25. In 1972 Bruck and Ustimenko (4) reported some observations made at 25 MHz with a linearly polarized antenna; they found that this pulsar shows a variety of pulse shapes with rapid transitions from one to another. In one plot, they show a pulse composed of 2 subpulses and separated by 1148 msec. Despite the poor resolution used (P/16 or %74 msec), the 2 subpulses are clearly defined.

The separation of the subpulses as a function of frequency has been plotted in figure 3-8, including the result by Bruck and Ustimenko and our result (i.e. for the bin 21 and 25 subpulses). The plot suggests that either the pulse separation increases very rapidly toward lower frequencies, or the subpulse detected at 26.3 and 25 MHz is a new one which appears only at low frequencies. In figure 3-9 the average pulse profile obtained in this thesis is compared with that obtained by Bruck and Ustimenko. The phase of the

200

ISO I THIS THESIS

BRUCK AND USTIMENKO (4) 100
A CRAFT AND COMELLA (6)

SEPARATION A IZVEKOVA et al. (7)
(ms)
o IZVEKOVA et al. (8) 50 AND LYNE et al. (22)

30

20
20 0

I I I I I I I t I I I I I I I 10 20 30 40 50 100 500 1000 FREQUENCY
(MHz)
Figure 3-8. Subpulse separation as a function of frequency for the pulsar PSR1133+16.

20 BRUCK AND USTIMENKO (4), 25 MHz OCTOBER 11, 1972 15
INTENSITY (ARBITRARY SCALE)

5

P I

21 25 9

90- I THIS THESIS, 26.3 MHz MAY 25, 1979
80
INTENSITY (ARBITRARY SCALE)

60

50

Figure 3-9. Comparison of the pulse profile for pulsar PSR1133+16 obtained in this thesis (average of 340 periods) with that obtained by Bruck and Ustimenko (4) (average of ,212 periods). The bins of the 26.3 MHz pulse profile have been shifted in order to align the pulses.

PUI. SA PS1 I J It, "r))= I=.18J?3 SEC FrFOUENCY=26.3 MIZ

PII2~A6 MAY 25 7 TSTAPT 01H 30' 10S STCO OIH 34'4 )5 TSAMP 02

BLOCK AVERAGE= 122.690 SIGMA= 0.113 3 SIGMA= 0.339 MAGNIFICATION= 100

NUMLER OF BLACKS= I AVEPAGINL TIME= 162.5 SEC

etN AVEr4GE PrProS INTENSITY NUM-ER AdFRAGLD
1 122.714 14) 1
? 122.629 140 I
3 122.657 140 +*
4 122.700 140 I
5 ??22.571 140 I + 6 122.593 140 I +
7 122.t36 140 .
9 12'.65 140
Q 122.-21 140 .
10 122.721 140
II 122. 2' 140 12 122.571 1.0 I I1 122.664 140 I 14 122.614 140 I 15 122.14 140 16 122.779 140 I + 17 122.o93 140 (TIME
1- 122.u5C 140 * I1 122.636 140 + 2) 122.743 140 21 62 SIA 21 122.9U6 140 I 21 62 SIO 22 122. 14 140 I. + 23 122.664 140 I 24 12 2, 07 140 , ?5 122.993 140 1 ---- 25 2.68 SIG' 26 122 .76 140 I Z7 127.571 140 28 I?2.5T 14' I 2; 122.4 6 140 3? 122.693 140 31 122.796 O 4 I 32 122. 736 140 I + '3 122.664 140 1 +* 34 122.654 140 I 25 122.779 140 I ;6 122.721 140 ( I 37 122.957 140 . 38 123.000 18 + Figure 3-10. Pulsar PSR1133+16 (Block P112WA6) recorded at 26.3 MHz on May 25, 1979. Average over 140 periods. 35 pulse profile at 26.3 MHz has been shifted, to align the pulses in bin 21 and 25 with the pulses at 25 MHz. Although the subpulse in bin 21 is much weaker than that in bin 25 for the 390-period average, the two are almost the same strength in the 140-period average shown in figure 3-10 (block P112EA6). In the latter case, the heights are 2.68 a and 2.62 a for the bin 25 and 21 pulses, respectively. A possible implication is that the bin 21 pulse varies in intensity much more than does that of bin 25. A small pulse appears in bin 9, when blocks P112WA6 and P112WB6 are averaged together. This pulse is also present in each block average separately (figure 3-5 and 3-6), suggesting the possible existence of an interpulse leading the main pulse by 0.42 P (or 1500). In Bruck and Ustimenko (4) a broad and poorly defined pulse seems to appear at 0.5 P from the main pulse, in the same plot where the double main pulse appears (figure 3-9). So far, no interpulse has been reported for this pulsar. 3.1.a Peak Flux Density, Mean Flux Density and Energy for Pulsar PSR1133+16 Several continuous radio sources were observed in order to calibrate the array, to obtain the minimum detectable flux density, and then to obtain an estimate of the peak flux density for pulsar PSR1133+16. The radio sources observed and their main parameters are presented in Table 3-1. 36 Table 3-1. Radio sources used for calibration purposes. Flux Antenna Density Beam Radio Source R.A. Declination (Jy) Used 3C310 15h02.9m 26015'6" 382 1W, 2W 3C315 15h11.7m 26016'8" 133 1W, 2W 3C409 20h12.2m 23025'15" 381 1E 3C433 21h21.4m 24054'6" 267 1W Positions and flux densities were taken from the Clark Lake Survey at 26.3 MHz (16). Of these four sources, only 3C409 and 3C433 were used in the calibrations; the 3C310 and 3C315 measurements were discarded because their difference in m R.A. is only 8.8 (corresponding to 2.20). Since the HPBW of the radiotelescope is 6.00 east-to-west, these two sources cannot be resolved. The parameters of the radiotelescope for observing the calibration sources were: Bandwidth (Av) = 250 kHz Time constant (T) = 1.5 s Frequency (v) = 26.3 MHz The minimum detectable flux density Smin' deduced from calibrations using these sources, is presented in Table 3-2. 37 Table 3-2. Sm. and Galactic background temperatures deduced from calibrations. S min Galactic background Radio Source (Jy) temp. oK 3C409 101 38,975 3C433 97 27,840 The galactic background temperature for pulsar PSR1133+16 is: Tbackgr. = 18,5140K All the galactic background temperatures given are actual brightness temperatures, corrected for losses in the antenna. A Hewlett Packard noise generator was used for the calibrations, and a pair of type 5722 noise diodes served as the standard of comparison for the noise generator. Details of the calibrations are given in Appendix E. The minimum detectable flux density has the following expression (17): 1 AS. m C xT (3-1) m nAvn syst where n = number of averaged records Av = pre-detection bandwidth of the receiver T = post-detection time constant Tsyst = system temperature In this case, the temperature of the system is completely dominated by the galactic background temperature, and in 38 equation (3-1), it is possible to substitute Tsyst by: syst Tgalactic background T Assigning subscripts p to the pulsar and r to a continuous radio source, it is possible by using equation (3-1), to write two similar expressions, one for the pulsar and the other for the radio source. By dividing one by the other, the following expression can be obtained: n Av T T S = r r p S K (3-2) ASmin p n Av T T min r p pp r The constant K takes into account that the pulsar signal, instead of being continuous, is being sampled. Using the values obtained for the pulsar on May 25, 1979: n = 390 periods (453 s) Av = 8 kHz p T = 12.5 ms p T = 18,5140K and the values for the radio source of: n =1 r Av = 250 kHz r Tr = 1.5 s ASr = 101 Jy T = 38,9750K the following value is obtained: Smin p = 149 Jy (considering K = 1). To obtain the peak flux density of the pulsar, it is necessary to make several assumptions: 39 A. In order to be consistent it has been considered that the value S for both the continuous source and the Mil pulsar is that which would cause a deflection of 3 a (i.e., 3 standard deviations) from the mean value of the galactic noise background. The 3 a criterion was used to determine whether or not a pulsar pulse was present at the end of an integration run. B. It is assumed that the sample rate is fast enough compared with the post-detection time constant that very little information is lost in this process; then K can be considered %1. (Details about sampling rates can be found in Appendix B.) C. It is assumed that the pulsar signal is unpolarized, as in the calibration source. This is probably not correct; a strong linearly polarized component has been reported (18). The effect of variable Faraday rotation in the terrestrial ionosphere would then be to cause the plane of the linearly polarized component to be oriented (at different times) at any angles from parallel to perpendicular with respect to the antenna dipoles. As a result, if the polarization is actually 100% linear, the value of Smin as given by (3-2) may range anywhere from too low by a factor of two to too high by a factor of infinity. Thus, no signal at all would be detected if the linear polarization is perpendicular to the dipoles, while according to (3-2) S min would not be zero. 40 By making these assumptions, once the pulsar reaches a value slightly higher than 3 a at the end of an averaging process, the peak flux density can then be considered equal to Smin (calculated for the corresponding number of periods averaged). Therefore, Speak p = Smin p = 150 Jy Pulse energy is usually defined as: E = S(t) dt (3-3) where P = pulsar period S(t) = flux density as a function of time For pulsars with no interpulse, the limits of the integral are restricted to the duration of the pulse (base time duration). In order to make the computations simpler, it is usually assumed that the pulse has a relatively simple geometrical shape. Izvekova et al. (7) assumed a triangular shape for most pulsars (including PSR1133+16). By looking at the pulse shape obtained on May 25, 1979, one can see that this is a reasonably good approximation, considering only the pulse centered at bin 25. Using this approximation, the pulse energy in equation (3-3) can be written as: E =- 1S W (3-4) p 2 peak p base where Wbase = pulse width at base, corrected for dispersion and receiver time constant. 41 Substituting the values of: Speak p = 150 Jy Wbase = 71 ms the following value for E is obtained: E = 5.3x10-26 Jm-2Hz-1 p The mean flux density of the pulse is defined as: S = E /P (3-5) mean p Using the value obtained for E and the value for the p period P = 1.188 sec, the following value for the mean flux density is obtained: S = 4.45 Jy mean It is interesting,now, to compare these values with values obtained by other authors. In 1973, Bruck and Ustimenko (4) observed the same pulsar at frequencies of 16.7, 20 and 25 MHz using a linearly polarized broadband array and reported peak flux densities in the range of 10 to 40 Jy. In 1968, Drake and Craft (19) observed the same pulsar, using the Arecibo 1000-foot radio telescope at frequencies of 111.5 and 195 Mhz and reported peak flux densities of 150 Jy at 111.5 MHz. Since these observations were made at a much higher frequency, they cannot be directly compared with the one at 26.3 MHz, nevertheless, the value obtained at the two frequencies gives an idea of the order of magnitude for the peak flux densities. A value of 38 Jy can be deduced from the data at 61 MHz given by Izvekova et al. (7). 42 Regarding the pulse energy and mean flux density, more recently, in 1979, Izvekova et al. (7), have given some results at 61 MHz. They obtained a value for the pulse energy of 0.89x0-26 Jm-2 Hz and a mean flux density of -26 0.75x10 Jy, using a linearly polarized antenna. In 1971, Slee and Hill (9), using a circularly polarized antenna, reported at 80 MHz a value for the pulse energy of 2.2x10-26 Jm-2 Hz- for one night (with a median over 6 nights of 1.0x10-26 Jm-2Hz ). It is not easy to compare the values obtained in this thesis with values obtained by other authors, because of the polarization used and the different averaging times (not always given). Nevertheless, the value of pulse energy -26 -2 -1 E = 5.3x10 Jm Hz obtained at 26.3 MHz, looks at least p -26 -2 -1 comparable with the value of E = 2.2x10 m26 Hz obtained p at 80 MHz (which also corresponds to a peak flux density of %110 Jy). According to Izvekova et al. (7), PSR1133+16 has a maximum in its average spectrum at vmax = 100+40 MHz. Kuz'minetal. (20), making simultaneous observations at several frequencies, found that the instantaneous spectrum changes from day to day. For several days it doesn't show a maximum in its spectrum but rather the intensity seems to increase towards lower frequencies. This can be considered as part of the explanation of why the peak flux density, the mean flux density and the pulse energy obtained at 2i.3 MHz are higher than the values 43 10 THIS THESIS SIEBER (29) & SLEE AND HILL (9) 5 o BRUCK AND USTIMENKO (4) 4 o IZVEKOVA et al (8) 3 2 ENERGY (Jm-2Hz-1 x10-26 .5 .4 .3 .2 I I 1 1 f I I I 10 20 30 40 50 100 200 FREQUENCY (MH z) Figure 3-11. Low frequency energy spectrum for pulsar P'S'R1133+16. 44 obtained at higher frequencies: the pulsar seems to show abnormally strong pulses at low frequencies during some particular nights, as it may have done in our case. Another interesting aspect of this has been pointed out by Ekers and Moffet (21). They observed PSR1133+16 at a wave length of 23 cm (1300 Mhz). From a histogram of energy and number of pulses with a given energy, they noticed a long tail in the distribution, towards the high energy end. The ratio of the mean value (0.020x10-26 Jm-2 Hz- ) to the highest energy pulse (%0.090x0-26 Jm-2 Hz- ) is about 4.5. If one takes the value for the mean energy of 5.3x10-26 Hz-1 at 26.3 MHz as an average over a relative short period (%390 P), and compares it with the calculated mean energy of 1.6x10-26 Jm-2Hz obtained from the observations of Bruck and Ustimenko at 25 MHz (4) a value of 3.3 is obtained for the ratio. Such a ratio is not improbable according to the distribution function of Ekers and Moffet. Pulse energies obtained reported in this thesis and by several other authors are plotted as a function of frequency in Figure 3-11. 3.1.b PSR1133+16 Pulse Width Several factors can distort the actual pulse shape as observed; the factors that need to be considered are: A. Effect of the dispersion over the finite bandwidth of the receiver. B. Effect of the post detection time constant of the receiver. 45 C. Polarization of the pulses and antenna. The first two factors will make the pulse wider than the actual pulse, but their effects can be corrected for if the pre-detection bandwidth, post-detection time constant of the telescope and the dispersion measure for the pulsar are known; a simple form needs also to be assumed for the pulse. In order to see the effect of dispersion let us consider the basic formula for the time of propagation through a plasma with a dispersion measure DM at a frequency v. The time of propagation is given by: t = kDM 12 (3-6) where k is a constant depending on the plasma. Differentiating this equation with respect to v, taking finite At and Av (instead of dt and dv), and expressing the parameters in consistent units, the following expression can be obtained: At = 8.3xl0-3 DMAv(vo) ms (3-7) where DM = dispersion measure (parsec cm-3) Av = bandwidth (kHz) VO = center frequency (hundreds of MHz) This relationship gives the time it takes for a dispersed pulse to sweep through a bandwidth Av centered at frequency v.o If a triangular shape for the pulse is assumed, (3-7) is transformed into (3-8) for the half intensity level: At0.5 = 4.15x10 -DMAv(vo)-3 ms (3-8) 46 The correction due to the time constant of the receiver (for the half intensity level and assuming a triangular pulse shape) is given by (8): T = 0.693 (3-9) c where T is the post detection time constant of the receiver. No correction can be made due to the effect of polarization but some discussion about its effect will be given in the conclusions. Consequently, the corrected half intensity width of the pulse is: W0.5 0.5 ob At0.5 Tc (3-10) where W0.5 ob is the observed half intensity width. Substituting in equations (3-8) and (3-9) the values -3 DM = 4.8 pc cm Av = 8 kHz vo = 26.3 MHz and T = 11.25 ms, the following values are obtained for the corrections: At0.5 = 8.8 ms T = 7.8 ms c Table 3-3 gives the values for the observed half intensity width and the half intensity corrected width for the pulses obtained on April 28 and May 25, 1979. The pulse width at the base is the width assuming a triangular shape, and is twice the corrected half intensity width. I00 THIS THESIS PULSE WIDTH O CRAFT AND (ms) COMELLA (6) 50 A IZVEKOVA et al 50 O (18) 40 O 30 20 I I II I I i i I I I II 20 30 40 50 100 200 500 FREQUENCY (MHz) Figure 3-12. Pulsar PSR1133+16 half intensity pulse width as a function of frequency. 48 Table 3-3. Observed half intensity width, corrected half intensity width, and pulse width at base, for pulsar PSR1133+16 Date W0.5 observed W0.5 corrected Wbase (ms) (ms) (ms) April 28, 1979 38 21 42 May 25, 1979 53 36 72 The graph of Figure 3-12 is a plot of the values of the corrected pulse half-intensity width at several frequencies (6, 8), including the values of this thesis obtained at 26.3 MHz on May 25, 1979; this is the more reliable of the two values presented in Table 3-3. 3.1.c Tests Performed with the Pulse Obtained on May 25, 1979 An independent and simple method for testing the pulses to see if they are indeed from a pulsar having the assumed period is to repeat the averaging process using both shorter and longer periods than that given by the DOPVEL program. If the observed pulses were spurious ones which happened to have approximately (but not exactly) the same period as that of the assumed pulsar, then the averaged intensity should either increase or decrease for the shorter period, and do the opposite for the longer period. If the averaged intensity is a maximum at the assumed pulsar period, then 49 it is almost a certainty that the pulses are indeed due to the pulsar. May 25, 1979, the date on which the best pulse was obtained, was chosen to make the test. Data were analyzed using periods longer and shorter than the assumed period by 40, 80, 160,320, and 640 ps. Periods shorter or longer than the assumed period P by 80 ps should cause a shift of 1 bin at the end of the 390 period average. The result of this test is shown in Figure 3-13, where the relative averaged pulse intensities have been plotted as a function of the assumed period length, taking as the base the periodP = 1.188023 s given by the DOPVEL program. It can be seen that the pulse nicely peaks with the assumed pulsar period, and decreases sharply for shorter and longer periods. The lack of symmetry of the plot might be due to the unsymetrical pulse shape. The computer plots of the results for the different periods used in this test are presented in Figures 3-14 to 3-24. The pulse intensity as a function of time has been plotted in figure 3-25 for the pulse obtained on May 25, 1979. This pulse was chosen because it was the better of the two and is well defined. This plot is useful not only to visualize the temporal behavior of the intensity of the pulse but also to check that no interference or isolated transient pulse is mistakenly taken as a pulsar pulse. 300 200 RELATIVE INTENSITY OF I.ZRAGED PULSE (ARBITRARY SCALE) 0 100 -800 -600 -400 -200 P +200 +400 +600 +800 (Us) AVERAGING PERIOD (us) Figure 3-13 Relative pulse intensity as a function of the averaging period for pulsar PSR1133+16 on May 25, 1979. P = 1.188023 s. J. A 113 1 F)= 1. 11 17 3 I 1i3 7E r U 'Je Y7, 3 *47 %ls?wA6 -AY 2, 79 TSTT'T 011 OM LOS105 TOP 01H 34M 1)S T0AM@ 02 PiI?wof v 5 7 '4-A IOlI 34" 10' OTO Ol-1 311 045 TSIAM4 02 A.3C( AVI?%AGI' 132.609 S' IA = 3'5 1 )'l -MA= 1.>DS AGIFICATION= 103 NjMirm OF rLOCKST AVCE AG IFG I ME- 4 2.; AVCary5 mO0anSr INTENSITY I I> .407 103 I S 1.?2 443 10) 1 5 122 ,F4 ' 6 12'.579 '03 I 7 12'.- 86 Cq I 3 12' -' 30 I 9 122.6k? 190 I 1) 122.'"' 315 11 1 2? A54 OO I 122.7?3 193 I t 12.3 '3 4I 15 122.564 -SO I I' I ''.~l 1~) 1 12>,517 0 3 F 1Q 122.58? 50O ITIME 2 1 2? 10 a 3 I 1 122587 i3) I 2 122.541 303 24 122 603 100 ?7 1 27 ma2 191 27 123 5 7 33 I9 122170 303 I 33 122.4 5 50 I 'I 12?.03 37'1 I tP 773 Too 1 32 122 nge 31)) 36 122 6,16 35 10 3 122 .33 212 1 + Figure 3-14. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a period P = 1.187383 s (P-640 us). 7? 246 vAY 2r~ TT. T Ol 33" 0 10S STOP 31H 34AM IS TS4mP '2 I7W'If AY 2%, 70 T4 '- C1, 34' 10% STP 0I- 3R4 3 S 'TAMP 37 IL 3OCK AVERAGE= 122.6 9 5! M2= O. Ct 5 3 ST (.A= 0 173 MAGNIFICATION 100 NutM'rre fF tOrKS- A VFr4AGING T I F= A ?.7 5FC "' AVn-.F P r IOj TNTENSITY t2 I?? ''t3 I 3 12 ?.*I ''2 I I?'.' 73 113 5 122 '-3 TC3 I 122.5378 TSO 12- 3., 303 12 122.f77 0 SO i 12? Ass' my3 15 t?;.An me 1 "4 17 12I 548 153 S 122., TIME 2' 12? IA 39 + ?1 ?12?.7 3 I o 2' I 2~? 7 ,Aq 3Q ) 23 122.F05 1 SO ? 127. 7 3 2" 27, 54 30 I 8 12'7 f, 393 + 27 20 122.658 'O + 'I t?,? s i 100 I '' 122t 303 2' 12'.'?,t 2 4 * '9 12 '.1 1W) ? 35 122.685 350 1I 17 122 0P2 13 I ?" 122.f54. 14 I Figure 3-15. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a period P = 1.187702 s (P-320 us). UL*AC 'iPll+Ill '9711 .197863 TrC r 0UENCY=26,3 m<4Z I12.A6 'tY 25 79 TST4IT 311H 11M 101 STOP 01H 3AM 105 TSAP ? IIl2w3 ',Y 79 "TJL0A 01+ IA 10' 'TnD 01 3 3 45 TSqAP 32 BLO-t A/FPAGE= 2?.'.609 ST "MA= .07 3 1 .MA= 0.209 MAGNIFICAT I1=. 130 NU* 'rl nF PLO0KS AVEPAGING 'IME= 11's,6 SIC A'' V AV PC rI cy 11 INTENSITY I. 5Tr; *VAIS; 3 172?E59 1t 3 I 2 12' 72 j9 S 122 A T 1O) I S 2? 1 .5 1 10 I S 122.649 3lO I tl 127 054 1 1 14 122 5,54 1) I S I ?;," 7, -j' )I 12 .1 O 13 5 122 470 IC8 IT 17 12? 3 T1 q I I' 1 T' I 1 22 ? 1 I 2p 122.818 1 I 12' .3) '9; I P1 1 2' j4 12 I 32 122.750 110 I 34 121 ;1 10 I 8 12'**564 393 I 3? 122.660 2315 I Figure 3-16. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a period P = 1.187863 s (P-160 is). PULS4' PoI1lIT' 1FFI 'O=I.lI rm' 1At5C rRUJFNCYac,3 AHZ it AA "AY 7 T- A4 71 1 ) '.' 't ) 3 -l -1AM 1 30- TSA. 92 12?W'6 MkY >5 7 T',TAPT I1Hi 1Am IOS STOP )IH 3J1 OS45 TS4"' 12 FILOCK AVrPA4= 12.9 IG Ar 3. 7 r'i = .?17 uAqGNIFICAT1?N= 100 NJMRCP CF BLOC-S ? AV'nt59! NG '"- 45', 6 5SC I' AV A94GE I^1- ) INTENSITY N Il~rR AVE AGED I 122 654 ?4 I + 1 ?r4 1 39 00 4 1P2 650 190 122.574 3) 7 122 62 390 I 122.#12 3 101 + 1 ? 0 0 lwO I' 172. 4 1 1 0 * I' 12' '13 ) T 13 122.564 390 I l??.92" 3M' 1- 122 .12 3q0 1S 122.561) 1~9 2'.' 30 IME 19 122 7 3) TIME 19 122.51? 1A 0 + n 1 I2?.556 Tp + 12?.f17 190 I 122.554 93 ? I''.5I" 191 24 122C92 TO I + 1 7?2.t 1 sq 27 1 22 ,?A 90 I 1 2.12 3,) 12, s won I 3) 122.549 3qO I33 t?*6A9 I1 I 127.925 390 3, 122 587 3S2 I+ 1 2254 3G0 9 122 "'3 215 I Figure 3-17. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a period P = 1.187943 s (P-80 uis). r 11H44A 44y ', 70 r TA; OlH '' I9 S1 1'' 1-4 34A 135 TSAMP 0' 12n6 MAY 75 7) T'-TAQRT 3H 'O4 105 STOO O1H 3AM 045 7TSWu ^2 (-OCK AVr*VC A 12. 00 'G1(-MA- 1 98 I ".4 1N 2,234 MA N I FIC ATIIN 10 3 N )MqfC ('F 5.OCK%- 2 AVT AGING I 49 7 S:C Il Av"4GF ora I INTENSITYr NI 'FP AVERAGED I 122 733 224 I -'4 I' '.' '4 To Ip 4 122 071? o) I 12 ?2. 'o 3 I 1 7 122 618 "90 I I' 122.4',? 9U0 I' 122 579 13' I 1' 1?.2 91 390 1* 1 12 12 lqo IS 12? Ii o3 I 122.579 19 17 122.f,20 TO ( I I'??eC' 3co 19 127,'l1 c3 I TIE I?2.523 390 I 01 'I 122 619 T3S I 22 122'3 C 15 I1
122.470 305
2 12? 710 JC3
122 4o 3 I
25 132? 1u 3 d I I

127.477 3 i s
'3 I>> 397 I')I
31 122 638 353 I
lT 1 4T 49) I
'' 122.531 '93F

'6 12?>.53? 193 I
37 122 579 50 I *
122.530 2152

Figure 3-18. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187983 s (P-40 us).

nULSAP PSrll33*'16 PFIOP=l.IR0?;'1 SEC FREQUENCY=26.3 YMZ

PI2*A f MAY 25 7(o TSTACT 0lH 1)" IOS STOP OlH 34m 10S TSAMP 02 o112186 AY 25 7S TSTAnT 01it 34M1 lOS SFCP 0OH 38f 04S ISAMP 02

PLOCK AVEAGEF 122.609 SIIcMA= 0.3RI 3 SIGMA= 0.243 IAGNIFICATION= 100

NUMIAE fF BLOCKS= 2 AVERAGING TIME= 45?.7 S7C

L' IN AVE8AGL 1F0, 1UOE INTENSITY
NJM E p AVfAGED
I 122. 81 224 I
? 1'.t 2 3r0 I 122.723 ItO I
4 l22.tO1 390 5 172.60? 3'0 I 5 12?,ws 390 7 122.13 30

9 129.76 T) 0 (
13 12d.t4 3SO I
It 12 .2f30 350
17 122.30 J0 IS20J
13 122.943 393 I
14 1'22.51 34 0 I .
15 122.700 390 I
1 6 12. 20 3 03
7 122. 595 30
18 172.037 33 I
I 122.546 3g0 I TIME
20 122.484 3 _0 21 I22.(tl 340 ? c
22 172?.482 340
73 12?.490 390 I
24 122.700 390 25 3.

27 122.0c 350

?) 122.5297 30 (

2t 122.t3e 390 I 32 I12.?5 3.0 I 33 122*43 340 I '14 122.451' 390 I
35 l?2'.tt 390 I
3 1 22.'53 340 + '7 122.45 390 1 3 12?'.S55 216 I+

Figure 3-19. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a

period P = 1.188023 s (DOPVEL period).

ULSAP PS l 1 31+! PFRIIOD= I. 03 SCC F E3U'I4Y=Z?6. I -t

Ptt2wA6 4*Y '5 79 TST4cT 31H nlm 10' STD' 0OIH 344 10OS TA"P 02 rlI2Wl6 4AY 25 7? TST ART 01H 144 10S 37' 01H 3A14 04S TS49' 02

BL0rK AVEPAGE= 122.619 F IG' A 0070 3 StI M- 0, 36 MANIFICATI-= 133

Ntl4IE f F L C rKS= 2 Vf AGINr TIMF= 4,".?7 SrC

'In VuA Ge "e = I INTENSITY
0 RE: VF A 3;
12? 1Q) I
2 122 677 SO0 I +
12.133 39' I
122 36, 190
5 122 536 SO
,12 0' 39D
122? I5 '9) I
9 122 700 390 I+
122.'4 3'00 I
1 122 729 393
1 12? 549 I9
12?.R2 300 1
I I 42 k% 3 0 I
'1 122 ,0 3 '93 TIME

i 122.623 390
I 7 7 3Q:)
1 122 556 103 I
122.298 390 +
"1 12;' 951 30) I
22 122 99 7'9
122. I4 393 S 12 755 19) I
25 122 782 190 1 + 25
12.71' 301
2' 122 51 C)

O 1?'. 151 1) 1
122 633 390 +
12?.~A1 390 I
31 12? 92n %QO I
'4 12'9.56C 190
1 2 ,.5 390
65 12? '54 393 37 122. 497 .30
1 2.R4A 51

Figure 3-20. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a

period P = 1.188063 s (P+40 us).

JL"A -,LI3T415 'I'D=l. 181L33 E "FNOU CY=75.3 4Z

III2,6 ','Y 25 79 TSTAPT O)H 33m IOS STOP OIH 344 IDS TSAMO 32 lIPOP', 'A Y !5 70 -a AI 7I1 14'4 tC' Tnr 01- 3AM4 045 Tr A-o 0PLOC< AvrPAG: 1'2.609 IGmA %79 3 SI 74= 0'1b 4MAGNIFICATION= tOO

NUMA4r nF 'LO'KS= P AVErAGING 'IME= 4'** SFEC

nyF,:rr 'F~IO n INTENSITY
I E ?: V-4so I
12' 7 C) I.
3 122.589 350

12? 559 '59 I S
S 1 27.5 30 I
1?? 63 T 1
A 122 73 T 0 1I 5 122.674 3SO I I) 12 577 I9) I
I1 12?2 10 1V?
12 1 2?. 1 I )
I1 12?.P0 P 91
14 122.6Ms 390
P 2.cTT 10) I
17 122,' T0
1 I. 11 330 1

2 122 554 350 TIME
I ? ?-5?! 3Qq O

'3 10?.718 393 I
It I??.4 1 10 I
?5 12? 7AO 3c3 I 25
25 127.676 310 I
77 1277*6 3on I

12 505, 19)

32 12?.t33 310
1 .7 ,4 0 j5 3

5 12.44 190 I
17 3f'. 30)
7 I??. 56 i
33 123.044 51 + *

Figure 3-21. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a

period P = 1.188103 s (P+80 us)

Ptl A. A I -AAY p. e ***4=' '1 14 v "5 T111 I1 1 34m 10'; TS1,D 02
"11i 2w16 *Y ?5 79 TSTArT OlH 14 10S S1OP 01H 3'8M )~S TSA P 12

9rLO K A IFr- A" I 509 "I(.-A= (I : ta=- P. U MAGNIIFICATICN= 100

N I'JAF I-F LO AVr 5 -C, G TI F- 4i? l ST 5

PIN AVERAGE 'E IrJl INTENSTT
NJ -IF NV FA ON Gli)
S1 1T2POOql

4 122 5 66 5T II
l?2 .' 3 9 *

122 700 390 1I
127 2 49 390 2.'4' q i
I 122 33 350 I
It 127.133
1 12? 46 193 13 122,5O 390 I I 12?, 70 303 I IR IP? 4Q97 3 ) I

1" 122 525 190 I
1, 12?. 77 390 I
) law,5mm yO) tTIME
2 122.643 190 I
12?'5 30
T3 I'?7? 7> To I 23
24 12 D 710 '9) I
12 '43 393 I

27 122.532 6a50
12o2. '93 3 16

3 1?2. 7 1900 1

122 6'' '93 I
33 122.53R 50 I3
1 2?I,"51 3 0 TB 12> 97 ses
5 122.'l I I
2.3 393 I3 3-7
38 122 909 > I

Figure 3-22. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a

period P = 1.188183 s (P+160 us).

nULRAP P CII33#I6 ;rF l flr)=1:=I lsA' 42 cr rR O0IFNCY 76.3 HLZ

PIt?Af. -AY -, 79 1crc '1H 7' 10 FTO 01H 34- 105 TS4"fa 02 'I 2w4f M'Y ?5 72 TSTART D 0 14M 13S STOP 01H 74 14AS TSA4 22

r1j9K AVFPAGF= 1??. 13 -' 149 4 A 0.IQ rIA 9 M'AGN IFIC(AT1IN= 100

NIJMPE F n rC'S: 7 AVF AGING T IF 4 7. 9 S'C

"IN rAFR4G r trIinD INTENSITY
N M R AV fr AGED
S 122 951" 10

7 122 635 1q0 I
12P..I10 1o1 1
1 I 7 rT 13 n I
122 0t 190 1

t? 12' '-1 1 I

I'; 12, 22 7 I'9 0
13 122. 4 89 90
14 1 22 .'1 ) 390
1 I" 12 W) 1 I

I 12,R 44 19 I+
7 I I 75 I I l +
19 122 .5 40 TIME

21 122 774 193 I 21 0
122.' 7 390 1+
1 2' 1,1 3J +
24 1 22.64 19 0 3
122.717 33 I
? 122 -11 1'O

122.3 '4 TO I

2 122 574 33 1+I t??.3' 33)

33 ?- 19
1?2 56 3 43 I

36 122 633 390
S 122."25 22 I *
34 L?2 A6: 1 9 I I + "

Figure 3-23. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a

period P = 1.188342 s (P+320 us).

D112WA MY7-v ?7 TT AR OlIt 1I 105 0 T OlH 34 I )S T7S14 02 l12 46b uAY 25 70 T.STAT 01', 314" 1S S TnP Olt 1"M 34; TSAMP 0 ?
RLSC' ~VFrA(;F= 122 613 FIM= 0-7A64 1 SIG4A= ,77 4AGNIFICATI314= 103

PNJuBER or scqq s a 4GING T ?= 453.2 e'C

HI'4 yCr'V 7 .r iD INTENSITY

1 ? 7 8 TIM

22.'" 3 ?

12? 151 IQ3

12?"45o 3" I

127. 41 '

7 12'.907 a I 3' 122.459 3O I

15 122. 1 0 O
I 12. '7 31)
17 122 107 1 Qoo to 122 f$391+ 21 1 22 .71 8 () I2+ 1 22.502 3o 1 ? 1 2' 7 19) ( 24 122.65' '50 I S 122.'"' 3;- I 25 121 t 111 ,7 1' I2'54 Two I 3) 122.620 TSO 12'.'") 35c 3' IP2.5 A -Nco I '4 12'1 22A I 35 12'.41 109 36 122.t2I 3,0 I 37 t??.'43 'o3 * 34 12' 635 ? 3 Figure 3-24. Pulsar PSR1133+16 data recorded onMay 25, 1979, analyzed with a period P = 1.188553 s (P+640 is). +1.5 +1.0 +.5 RELATIVE INTENSITY (ARBITRARY SCALE) 0 -------- - -- -- -- --------.5 100 200 300 400 TIME (PERIODS) Figure 3-25. Relative pulse intensity as a function of time (periods) for pulsar PSR1133+16 on May 15, 1979. Each point represents an average over 10 periods. P = 1.188023 s. 63 It can be regarded as a double check of the test previously described. Each point on the graph represents the intensity with respect to its corresponding average value of 10 consecutive periods, for bin 25 (peak of the pulse at the end of the averaging process). The intensity scale is in arbitrary units and the time scale is in periods P given by the DOPVEL program (P = 1.188023 s). The plot shows that the intensity in bin 25 is almost always positive with respect to its corresponding 10-periods average value, and no clear periodic variations appear over the 390 periods plotted. 3.2 PSR2045-16 Data Analysis Of the 10 nights of data reduced for pulsar PSR2045-16, pulses were found on 4 nights: May 15, 16, 18 and 19, 1979. A summary of the dates, averaged pulse deflection in terms of standard deviations 0, and number of periods averaged is presented in Table 3-4. Table 3-4. Summary of the date, pulse deflection and number of periods averaged each night for pulsar PSR2045-16 Pulse deflection Date in terms of o Periods averaged May 15, 1979 3.16 540 May 16, 1979 3.48 440 May 18, 1979 3.28 610 May 19, 1979 3.30 415 64 Plots of the data for these dates are presented in Figures 3-26 to 3-30. Figure 3-31 corresponds to a plot of a typical day when only noise was present and no pulse appeared. In general the pulses were of different shapes, except for the last two days, for which the shapes were similar. Some have steep leading and trailing edges (as on May 16, Figure 3-28); others have only a steep trailing edge and a not so well defined leading edge (as on May 15, Figure 3-26). On May 18 and 19 (Figures 3-29 and 3-30) both pulses show a steep leading edge and much less steep trailing edge. On May 16 the pulse was clearly present with an intensity of 3.48 a above the mean value over an average of only 180 periods (Figure 3-27). On the other 3 days, pulses were not seen clearly until the end of the averaging process for that night. For partial averages in these cases, some indication of a pulse can be seen at the location where the pulse appears after completing the average, but its intensity never rises significantly above the noise level. No indication of the existence of a complex pulse or interpulse can be deduced from the plotted data. Pulsar PSR2045-16 presents at higher frequencies a complex pulse, composed of 3 subpulses, with the trailing subpulse decreasing dramatically in intensity with decreasing frequency (22); at 610 MHz the trailing pulse Figure 3-26. Pulsar PSR2045-16 recorded on May 15, 1979, at 45 MHz. Average over 540 periods. 66 N *I* ** ***. ** * * O alh~Lt 000000 laaOaaa z N 777217 C =<4 -4 4< 4000000 14dC4 If Inc Vo~l IL 0007 oooo-V n u 4 TIT277) V, 0 Lo L h -iIII a' Z In q~Aoq o 0 noocoo 4 0000-O --- -> a u NCMNM 41 0 0 0 0 . 0 t 0oaEo0 0 h~- ~ II - 4z~r U OraL UC~I 0 i * 47j430 A Ir 1111 1~ 0 0000.J a40Ulr1 QCZC -rtr 0' II z L 4 --A0- U 4 44n4<4 z, In I1 I r oul ,Ii ,L) 7D~7 a=~~, .000000 I aaCICD 0 9 a U, -----_-_-__,_____,_4 0 1000 n10000CCQo000n000ooon0ooo0nooonooo..o0o~oO. U 1U4 44 .--LC0U43 '4 40444.1404.0Q 0*0004444440fCOfllflfllQ444 > C 4 u 'a~0 OrCJPCC,~O~O~~h14Q*~~Q-SC~~Cra 0 01I ; Lt -h~ ~ rov ?o;c ,Neh dc EI0~ -~uiro eon .4 *^-- ^--~- _,-,I-,,, I 1 a; -CL~'-" ~ C~-E~COnNJ~ ~ ~hlO 0 ----1V(rrrY FIV i~IrrrCr9JJO~Qn ui~~u ~n.ni0 Figure 3-27. Pulsar PSR2045-16 recorded on May 16, 1979, at 45 MHz. Average over 180 periods. mm C 7 b ,Ci ri rWii ~ ~~u-P~ffC O *P~~ttP ttP 1. j Pt W t e pPni P;;iutt<..t -r C rL .. .. . .. .. .. .. ... .. . ....... 0 o a,0. ~ O c.C C 99 9 A C.2 -I o I -- - - ---frl-actO~CatrLD,bC.. Xr,V~9y.,,O :orVncr r,:WCrnjtnra rot o Jarn;,rr r frctLavu c rnr r ---- ---- - -- -------------- -- ------L-------- --- --N000N00YI JtJL~lU t 00 OOC C (iCiQ~k0OU 0 0.)00CJoC 00Q0L,0oU000u OULIc,0OoO cso). 00i- C 9C r 9 C 7C 0 Z It #C, 0a, (11 ..... ...... .............. ......... .. ............ ...2 '4 B~ C 00 Z 99a io 0 us K t -I 9 9 > OO O NW Ln w M OO 89, Imp I CC 0 0 II Li vit z 0 (p (0 P it 9 S. *.r (4 if V 899 LU ff0) SC P Vi9 (( Tf 09 I Ii ViV 9 r ~ H -4 44 K vu, O i -- -----'-----------5---'-'-----7 89 Figure 3-28. Pulsar PSR2045-16 recorded on May 16, 1979, at 45 MHz. Average over 440 periods. >t f 0090O000000000000000 0000in000000NN 000N00 000 000000 C1. >X 00 On o O C 00000 r Z - r O 4 -j 3.> m L 0 NN- 94 0 C 3 .3 . *00000 T -c-OL Figure 3-29. Pulsar PSR2045-16 recorded on May 18, 1979, at 45 MHz. Average over 610 periods. 72 2:222 8 ...... ........................................*..********** N(VN14N 0 S0000 O 444<4 Z N U 0 Ln 4(I1IV( 4 w 4000 U 4 4u ) 4) In Ln o ilti> z O OI~~~l4 + **,~++. +. +++ *J+*++++ + W**~ +++ t*+++*4*1* .**~ ** ++* W 00000 U 0009 -N oouo 0 0 0 0 mi l 2,~(2 V, 0000O O It 00000 a: Iaoto I i LI ) it 0T U * IIIII n 00000 0 0 * - 122'n Z 44<44 Y O Ii 0 NNr-NN *cl0 4 -) In > Oc 0 4 Z o >>>>> < N 44444 I 0 .00000 0 N(V1NNNN 'I a S------- --------.4q( : U U -.- -. - -- . . . 4 0 < eyeC r-.flC.r'00100 e00C'Of00000C :0r1(,0041J00 00000000000 v000x-4.0 me0oa ne '4 >24-CCi.C ."1' 4 ,-r-tC.r-C 1-rrl,.C. O-NMWp :oN~o r e-. Nr dPONOr-rr o-NO4 tt Cve70-eameroN 0 Zr rrd o >IUII -o l.) -~' -rr-t..-ftf~),~~ f-r.r-'1l,-( -f.v~-r-.p.rr b. ~LICO r 4 1VN -' '0 I N3 44 N' 0 n0 'N0 V0 n 0 4I NO~0~ lIZ 3C( NVV I0l,.(0 14 4 4 4 4 VIr. C~,.. Figure 3-30. Pulsar PSR2045-16 recorded on May 19, 1979, at 45 MHz Average over 415 periods. 74 ----------------------------..... .............................**...*.**..**************** 0 nNNNn 0 0000 It pflwnAM I III Z *400 44 o2 0 0000 6 Lm, C a ooSo U ~I~ Z 2 4"O00 0 n 4r) 4 z 1 0rrr+~rt~Ir ,I 1r i i\ XII *i1 4*4'. V. ** 4 **tl 4 *, ** +*.r *. *4 *t4* *.4* 44 'C z 0 00 0300 0 * 2292 2................................ .......*.....*..**.*********** "0100 U C@4 L 4 0LOEdGp C4 .c O'n n to a OONLaO 0 U I u 444 49 44 ct 44 44 45444 441tW4944M t4 44 44 44e444.4 4 4444l64 4 ..4 - no i Cu 0I o~ 00 0 000 0 0 noonno 0oCoono noon onco one oooor o-co eo no3 N NN N 0 4 -4 i )-r)-) 1 0I rI h '4 4 O4 4 42 4444 ~ 0. Q O > 099 0 U .4 04 4 0 4 IL -cOrjLp0CzOI)4.~~,nlnniVN0CiU tcS 4tv'c1V)N 42LX-4 orr0Jrj40aol 2 t OON7jq J~~O QJj~ .2?Q 1 02 ------------ tO fl12 tO~0-04 -t 1 CtCCCIr0X4V-tttf0 tI4 44444fUil~~lil0 Figure 3-31. Pulsar PSR2045-16 recorded on May 17, 1979, at 45 MHz. Average over 390 periods. No pulse present. 76 ----------------'----------------------------------.. . .. .. . CCOO O <4C0 7 i lr (I N a0V0 C 7 F ~Lu~rVnu a: 0c~o u 0000 4 > 7727 U f' C I 2 Z inM 0 C 0 a :pc o oo ... ........ ...... . *i 4.t H 0000 F. LaL cCC 'd r(-0 ,S ( -a. 0 o D>O L << 4 U) tr, In el C OC v n O >>> * :7T2 . 2~ ccacc 'Y 40444 O; 3aC U 4C o )->" em M: 7 T C.0 QMNN MC. DCq <.Q~44 a.d t>&O >NNN C L C. ICu 7 0T a ., (l O C C a. N a: I. a r a a a Co :: a: e : k a a a: a: a a a a: ~ a : v a trC r (r c nz( > NN LCJOOC.CC' IcCO ,N O N NNmI m I .n(4 a It 4 a v CIn 4 ---.---------_--- -~I 0 0(.ti 77 is the most intense, but the intensity decreases rapidly, and at 151 MHz is about 2/3 of the central subpulse. The intensity of the central and leading subpulses seems to remain almost constant with frequency. No indication of resolved subpulses can be obtained from the data at 45 MHz. Nevertheless, the difference in steepness of the leading and trailing edges might suggest the existence of an unresolved much less intense subpulse accompanying the main subpulse. This aspect will be discussed later when discussing the pulse width and the effect of polarization. 3.2.a Peak Flux Density, Mean Flux Density and Energy for Pulsar PSR2045-16 In order to estimate the peak flux density of this pulsar, several radio sources were observed for calibration purposes. Radio source PKS2034-17, one of the closest to the pulsar, was clearly recorded despite the steepness of the galactic background (23). This radio source has apparently not been observed before at any frequency lower than 80 MHz. However, an extrapolated flux density at 45 MHz was obtained from the spectral index derived from flux density data at higher frequencies, as reported in the literature. Spectral index can be defined (17) by the relation: S ) (3-11) 78 where S = flux density v = frequency n = spectral index For any two frequencies on the linear part of the relationship between S and , SuV1 2 (3-12) v2 The best available values of the flux density for this source (24, 25) are at frequencies of 160 and 80 MHz: S160 = 8.8 Jy S80 = 15 Jy Substituting these values into equation (3-12), a value of n = 0.76 is obtained. Assuming a linear relationship for S as a function of 1 - down to 45 MHz, the flux density at this frequency is found to be: S45 = 23 Jy Applying the previously used criteria, the following value for the minimum detectable flux density at 45 MHz was obtained for the region near source PKS2045-17: Smin = 6 Jy The radiotelescope parameters for observing the radio source PKS2045-17 were: Av = 3 MHz T=3S n =1 r 79 where Av = bandwidth, t = post detection time constant and nr = number of records averaged. It can be shown that the brightness temperature T and the frequency v are related in a similar way as S and v are related in equation (3-12): T (v2 (3-13) Using the galactic background temperature at 38 MHz given by Blythe (26): T = 12,0000K T = 7,0000K r where T = galactic background temperature near the pulsar Tr = galactic background temperature near the radio source PKS2034-17 and the value for the galactic background spectral index given by Findlay (27), valid for the range 20 to 400 MHz: n = 2.65 the galactic background temperatures T and T at p r 45 MHz were found to be: T = 7,6300K Tr = 4,4520K Using equation (3-2) and the appropriate parameters of the radio telescope, the minimum detectable flux density can be obtained for the pulsar observations. 80 The radio telescope parameters when observing PKS2034-17 are: Av = 3 MHz r T =3s r n =1 r T = 4,4520K Smin = 8 Jy The parameters when observing the pulsar are: Av = 10 kHz p T = 11.25 ms p n = variable, depends on the day T = 7,6300K The peak flux density (which is also the minimum detectable flux density) for the different days is shown in Table 3-5; this table also contains the values of the mean pulse energy and mean flux density computed using equations (3-4), (3-5) and the values for the pulse width given in Table 3-6. No known measurements of flux density and energy at low radiofrequencies exist for pulsar PSR2045-16. The lowest frequency data on this pulsar are the 80 MHz observations made by Slee and Hill (9). They attempted to observe the pulsar on 3 nights, failing to detect it and giving an upper limit for the pulse energy of -26 jm-2 -1 <0.17x1026 JmHz. The failure to detect this pulsar is not surprising since it has been reported to have large Table 3-5. Summary of Peak flux densities, Mean Pulse energies and Mean flux densities for Pulsar PSR2045-16 at 45 MHz. Peak Flux Density Mean Pulse Mean Flux Date Periods Averaged Jy Energy Jm-2Hz-1 Density Jy May 15, 1979 540 125 3.8 1.95 May 16, 1979 440 139 3.4 1.73 May 18, 1979 610 118 3.5 1.8 May 19, 1979 415 143 4.4 2.25 Unweighted Mean 3.775 82 4 W THIS THESIS 3 0 BACKER (30) A SIEBER (29) 2 O SLEE AND HILL (9) AVERAGE PULSE ENERGY (Jm- 2Hz-I1 x0-26 .5 .4 .3 .2 20 30 40 50 100 200 300 400 FREQUENCY (MHz) Figure 3-32. Pulsar PSR2045-16 low frequency energy spectrum. 83 intensity variations, being often undetectable (28, 29). Our experiences at 45 MHz confirm this behavior also since it was possible to detect this pulsar only in the first nights of the observing period; it faded out for the rest of the nights. An unconfirmed pulse energy of %1.9 Jm -2Hz-1 has been given by Backer at 34 MHz (30). The lowest frequency at which published data on pulse energy is available is 410 MHz; the energy is 0.120x10-26 Jm -2Hz (29, 31). Our measurement at 45 MHz seems rather high (perhaps suggesting an undetected systematic error), although no direct comparison is possible due to the lack of published information at low frequencies on this pulsar. Pulse energy as a function of the frequency has been plotted in Figure 3-32, including the data mentioned in the last paragraphs. 3.2.b Pulse Width for Pulsar PSR2045-16 In order to obtain the true half intensity pulse width of PSR2045-16 at 45 MHz, the measured half intensity pulse width has been corrected for pulse dispersion and post detection time constant using equations (3-8) and (3-9) respectively. In order to correct for pulse dispersion, the following values for the parameters were used: Av = 10 kHz vo = 45 MHz -3 DM = 11.5 pc cm 84 The correction for pulse dispersion At0.5, using the above values, is: At0.5 = 10.5 ms The correction due to the post detection time constant of = 11.25 ms is: c T = 7.8 ms c The true (or corrected) half intensity pulse width W0.5 is as before (equation 3-10): 0.5 0.5 obs. At0.5 c The result for the 4 days is shown in Table 3-6. Table 3-6. Observed and corrected half intensity pulse width and pulse width at base (assuming triangular shape). W.5 obs W0.5 Width at base Date (ms) (ms) (ms) May 15, 1979 44 31 62 May 16, 1979 38 25 50 May 18, 1979 44 31 62 May 19, 1979 44 31 62 The lowest frequencies at which pulse shape and width measurements have been given for this pulsar are 150 and 151 MHz (10, 22). Lyne et al. (22) have given the separation between subpulses at a number of frequencies; the values are plotted in Figure 3-33. The separation between 200 N THIS THESIS 0 SEPARATION BETWEEN LEADING AND CENTRAL SUBPULSES. LYNE et al 100 O SEPARATION BETWEEN (22) TRAILING AND CENTRAL SUBPULSES. LYNE et al (22) PULSE SEPARATION (ms) 50 40 40 0 30 00 00 20 10 20 30 40 50 100 200 300 400 500 1000 FREQUENCY (MHz) Figure 3-33. Pulsar PSR2045-16 subpulse separation as a function of frequency, including the pulse width at the base obtained at 45 MHz in this thesis. 86 the leading and central subpulses and the trailing and central subpulses is plotted separately. As in a previously mentioned case, the trailing pulse seems to decrease very rapidly in intensity toward lower frequencies (22), and it is likely that at 45 MHz it no longer exists or at least it can be very weak. The mean value at 45 MHz for the pulse width at its base (obtained by measuring the width at half intensity and assuming a triangular shape) of 61 ms has also been plotted in Figure 3-33. The plotted value for the pulse width at its base, lies close to the extrapolation of a straight line passing through the points representing the separation between the leading and central subpulses. A possible conclusion of this is that the pulse observed at 45 MHz is composed of two unresolved pulses which might correspond to the leading and central subpulses at higher frequencies; the pulse width at its base would correspond to the separation between the two subpulses. No polarization measurements have been made for this pulsar at low frequency. The lowest frequency at which polarization measurements are available is 409 MHz (32, 33). The mean percentage of linear polarization at 409 MHz is 40%, and it changes during the course of the pulse, reaching about 70% for the first subpulse. The stability of the linear polarization position angle is low for the wings of the integrated pulse profile, due to the occasional 87 occurrence of pulses with orthogonal polarization with respect to the mean, but is high at the center of the pulse (32). If the degree of linear polarization of the pulse is higher at lower frequencies, there will probably be a distortion in the pulse shape when it is received with a linearly polarized antenna. This distortion will depend on the variation of the polarization angle within the pulse. Assuming that the linearly polarized component in the central part of the pulse is parallel (or nearly parallel) to the antenna, then the integrated pulse profile will appear strong in the center; however the slopes of the leading and trailing edges will be weaker and the averaged pulse profile will be narrow. More details about the effect of the degree of linearly polarized component on the pulse profile and apparent intensity variations are given in the conclusions. 3.2.c Period of Pulsar PSR2045-16 from Observations on Two Consecutive Days The apparent mean pulse period of a pulsar can be calculated from the elapsed time between two pulses; in order to increase the accuracy it is desirable that the elapsed time between the two pulses be much longer than the pulsar period. An attempt was made to obtain the apparent mean pulse period from observations made 24 apart (elapsed time 124h) for pulsar PSR2045-16. On two occasions the pulsar was detected on two consecutive days; 88 the first occasion was May 15 and 16, 1979, and the second May 18 and 19, 1979. The pulse period can be obtained by subtracting the epochof apparition of the pulse E on two consecutive days and dividing by an integer number N. The epoch of the pulse can be calculated from the relationship: E = T + nxR (3-14) start where: Tstart = time start digitizing n = bin number where the pulse occurs R = plotting time resolution The information related to the two pairs of days has been summarized in Table 3-7. The mean apparent period can be calculated from the relationship: P AE (3-15) N where AE: difference in epoch of pulse apparition on two consecutive days (elapsed time between pulse apparitions) The mean apparent periods obtained with this method are compared with the DOPVEL periods in Table 3-8. Table 3-7. Epoch and difference in epoch of pulse apparition for Pulsar PSR2045-16. R AE Date n (s) E (s) N May 15, 1979 27 0.0316 09h 49m 20s66434 May 16, 1979 8 0.0316 09h 44m 34s85415 May 18, 1979 8 0.0316 09h 38m 00 25308 86189s9367 43944 May 19, 1979 6 0.0316 09h 34m 30s18981 or 43943 90 Table 3-8. Comparison of the mean apparent pulsar periods for PSR2045-16 obtained from observations with the DOPVEL periods. Mean apparent period DOPVEL period AP Date (s) (s) (Ps) May 15, 1979 1.96137547 1.961383 -8 May 16, 1979 May 18, 1979 1.96135847 1.961386 -28 May 19, 1979 or 1.96140311 +14 It can be seen that the periods do not agree. Unfortunately, a jump in the epoch was noticed in the quartz time standard of Maipu, after the first pair of days; that makes the figures for the last pair of days (May 18 and May 19, 1979) very unreliable. In any case, this discrepancy, at least for May 15 and May 16, demonstrates that the DOPVEL period cannot be used to bring in phase two consecutive days of observations. In order to bring in phase pulses detected on two consecutive days (within +0.5 bin), the apparent mean period for each day needs to be known with an accuracy of at least A = /N (3-16) where N = Integer number of periods between two apparitions. Expressed in terms of the elapsed time T and the period P, (3-16) becomes: RxP Ac = (3-17) c 2 T 91 where N = T P Assuming an elapsed time of 24h (86400 s), a resolution of 0.0315 s and a period of 1.961386 s, the following accuracy for the period is obtained: Ac = 0.36 is This accuracy should bring the two pulses in phase within 0.5 bin. It is not clear what caused the discrepancies between the DOPVEL period and the period deduced from the observations; it could be an instrumental problem or a precision problem in the calculations in the DOPVEL program. Both possibilities were checked but the result was inconclusive; further investigations need to be made. 3.2.d Pulsar PSR2045-16 Pulse Intensity as Function of Time Figure 3-34 is a plot of the relative pulse intensity as a function of time for the pulse obtained on May 16, 1979; this pulse is the strongest and best defined. Each point plotted represents the relative intensity with respect to the noise average, averaged over 10 periods; the time axis scale is in pulsar periods. It can be seen that the pulse has large intensity variations. Most of the time it has a positive value with respect to the noise average, and it seems to show an almost periodic intensity modulation of %140 periods (peaks at periods 20, 160 and 320). RELATIVE INTENSITY (ARBITRARY SCALE) +1.0 4- 0.6 -F 0.4 4-0.2 -c - 02 - 0.4 100 200 300 40O TIME (PERIODS) Figure 3-34. Relative pulse intensity as a function of time (periods) for pulsar PSR2045-16 on May 16, 1979. Each point represents an average over 10 periods. P = 1.961383 s. Full Text Table 3-7. Epoch and difference in epoch of pulse apparition for Pulsar PSR20 45-16. R Date n (s) E May 15, 1979 27 0.0316 0 9h 4 9m 20?66434 May 16, 1979 8 0.0316 0 9h 44m 34 f8 5415 May 18, 1979 8 0.0316 0 9h 3 8m OOf25308 May 19, 1979 6 0.0316 09h 34m 30 fl 8 981 AE (s) 86114?190 N 43905 86189?9367 43944 or 43943 00 VO 79 where Av = bandwidth, r = post detection time constant and n = number of records averaged. It can be shown that the brightness temperature T and the frequency v are related in a similar way as S and v are related in equation (3-12): Using the galactic background temperature at 38 MHz given by Blythe (26): T = 12,000K P T = 7,000K r where T = galactic background temperature near the pulsar IT = galactic background temperature near the radio source PKS2034-17 and the value for the galactic background spectral index given by Findlay (27), valid for the range 20 to 400 MHz: n = 2.65 the galactic background temperatures and at 45 MHz were found to be: T = 7,630 K P T = 4 ,452K r Using equation (3-2) and the appropriate parameters of the radio telescope, the minimum detectable flux density can be obtained for the pulsar observations. 80 The radio telescope parameters when observing PKS2034-17 are: Av =3 MHz r T = 3 s r n =1 r T = 4,452K r S = 8 Jy min J The parameters when observing the pulsar are: Av = 10 kHz P n P T P = 11.25 ms = variable, depends on the day = 7,630 K The peak flux density (which is also the minimum detectable flux density) for the different days is shown in Table 3-5; this table also contains the values of the mean pulse energy and mean flux density computed using equations (3-4), (3-5) and the values for the pulse width given in Table 3-6. No known measurements of flux density and energy at low radiofrequencies exist for pulsar PSR2045-16. The lowest frequency data on this pulsar are the 80 MHz obser vations made by Slee and Hill (9). They attempted to observe the pulsar on 3 nights, failing to detect it and giving an upper limit for the pulse energy of <0.17x10 Jm Hz The failure to detect this pulsar is not surprising since it has been reported to have large 45 C. Polarization of the pulses and antenna. The first two factors will make the pulse wider than the actual pulse, but their effects can be corrected for if the pre-detection bandwidth, post-detection time constant of the telescope and the dispersion measure for the pulsar are known; a simple form needs also to be assumed for the pulse. In order to see the effect of dispersion let us con sider the basic formula for the time of propagation through a plasma with a dispersion measure DM at a frequency v. The time of propagation is given by: t = kDM ~2 (3-6) v where k is a constant depending on the plasma. Differentiating this equation with respect to v, taking finite At and Av (instead of dt and dv), and expressing the parameters in consistent units, the following expression can be obtained: At = 8.3x10_3DMAv(v0)~3 ms (3-7) _ 3 where DM = dispersion measure (parsec cm ) Av = bandwidth (kHz) vQ = center frequency (hundreds of MHz) This relationship gives the time it takes for a dispersed pulse to sweep through a bandwidth Av centered at frequency vQ. If a triangular shape for the pulse is assumed, (3-7) is transformed into (3-8) for the half intensity level: AtQ = 4.15x10-3DMAv(v0) ms (3-8) 2 of a pulsar that need to be measured before realistic models of the radio emission mechanism can be developed are the spectrum of the radiation, the pulse shape, peak intensity, and the time variations of these parameters. A large number of measurements have been made in the range of frequencies between 100 and 2,000 MHz, but only a few below 100 MHz (2). The low frequency part of the spectrum is particularly interesting because most pulsars present a turn-over of the spectrum at metric and decametric wave lengths, and also because pulse shape and widths show some peculiarities in this region. The spectral maximum for most pulsars is in the range 60 to 250 MHz; this peak is nearly symmetrical with respect to the frequency at which the maximum occurs (2). Measurements of the low frequency part of the pulsar spectra have been made by Bash et al. at 38 MHz (3), Bruck and Ustimenko at 10, 16.7, 20 and 25 MHz (4, 5), Craft and Cornelia at 40 MHz (6) and more recently by Izvekova et al. at 61 and 102.5 MHz (7, 8). Most pulsars measured at low frequencies are in that part of the sky which is observable from northern hemis phere observatories. Observations at the lowest frequen cies have been carried out in the southern hemisphere only by Australian observatories, at 80 and 150 MHz (9, 10). Detection of pulsars at low radio frequencies is difficult because of the deterioration of the signal-to-noise ratio caused by the high galactic background temperatures, and the low frequency spectral turn-over exhibited by most PUL SAP PSR I 1 33* 16 PERIf30=l 107999 SEC F RE OUENCV=26. J MHZ p 1 1 2ED3 APRIL 28 79 T ST ART 02 H 05M 55S STOP 02H | OM OOS TSAMP 02 HLCCK AvEkAGE= l 1 9.ft?0 S1GMA = 0. 1 1 A 3 SIGMA= 0.342 MAGNIFICATIONS 100 NUMBER OF BLOCKS= l AVERAGING TI 232. 1 SEC PIN NUMBER REfIUOS AVERAGED 1 1 1 9 1 6 2 ro 2 1 IR.50 200 3 1 19 545 2 00 A 1 19. 7P C 200 5 l 19 69 0 200 6 1 19.65 0 200 7 1 l9 .6 7C 2 CC 6 1 19.650 2 CO 9 119.620 2 00 1 0 1 19.6 15 200 1 1 1 19.630 2 GO 12 1 19. 120 25 1 3 1l9.665 2 CO 1 A 119.770 2 00 15 119.5AS 2 CO 1 t 1 15.5 15 2 00 1 7 119.635 2 00 13 119.586 2 CO 19 119.655 200 ?0 119.670 200 2 l 119.435 200 22 119.445 200 23 1 19. 720 2 CC 24 1 19.73 0 2 CO 25 l 19.560 200 26 1 1 9.4 7 C 2 00 27 119.390 2 00 28 119.610 200 2V 1l9.490 2 CO 33 1 l9.7? 0 2 00 31 1 l 9.6 ? 0 2 00 22 1 lv.465 200 33 l 19,650 200 24 115.705 2 CC 35 1 19.7 05 2 CO 36 119.7 6 C 2 00 37 119.550 2 00 23 1 19.5 10 200 TIME INTENSITY NJ OI Figure 3-3. Pulsar PSR1133+16 (Block P112EB3) recorded at 26.3 MHz on April 28, 1979. Average over 200 periods. FORTRAN IV G LEVEL 21 MAIN DATE 79206 18/19/42 PAGE 007 6 0077 0078 0079 0080 008 1 0082 008 3 0084 0085 230 0086 ooe 7 0088 0089 0090 009 l 2 35 0092 009 3 5 30 0094 0095 0096 0097 0098 0099 0100 0 10 1 0102 0103 0 104 01 05 0106 0107 0 10 8 511 0 109 0 110 0111 0 112 512 0 113 0 114 0 115 0116 0 117 0 118 0119 0 120 0 12 1 0 122 0 12 3 0 124 0 125 0126 520 012 7 0 128 240 0129 400 0 130 4 06 0 13 1 408 0132 4 10 0 133 4 1 1 0 13 4 4 15 0135 420 0 136 4 25 C C COL S C c c c c c DO 230 1 = 1 NO MX SINK I ) = 6 IN I ( 1) +U IN ( 1 ) NBINI ( 1 ) SNUINI ( H NO IN( I ) SUM= SUM-f B 1 N(I) SUM I =SUM I 4-0 IN I ( I) SUMNsSUMN*KBlN( I) SUMN 1 = SUMNINO I Ni C I ) CONT INUE A VE = SUM/SUMN A VE I S UM1/SUMNI DO 235 1*1# NO MX SUMSQ = SUMSOH01N< I ) A VE NBIN( I ) )*? SUMSOI* SUMSQIC BIN I( IJ-AVE INB IN I ( I ) ) **2 CONTINUF RMS= S0RT( SUMSQ/SUMN)/SORT(SUMN/FLOAT (NBMX) I RMS I SORT (SUMSQ I/SUMNI ) /SORT( SUMNI/FLGAT(NBMX)) WRI Tf(6.530 >AVE .RMS .AVEl .RMS I FORMAT(30X. 'BLOCK A VERAGE Z ,F8. 2. SIGMA:' ,F8.3. 12 X INTEGRAL AVERAGE: '.FB.2. SIGMA: .FQ.3I DO 2 40 I = 1 .NBMX I P= 1 BINA =0.0 IF(NOIN B IN* =B I N( I ) /F L O AT ( N B I N ( I ) ) IFUBIT.F0.I2) I Ps ( B INC I ) / FLO AT ( NB IN ( I ) )-PZ ERO )/ 2040. *25. FLOATt IPSCAL) t26 IF( IOIT.FQ.8) IP = (BIN{ I )/FLOAT(NBIN( I ) )-PZERO)/I28.*25. FLOAT( I PSC AL) 2 6 IFC IP.LT.l ) IP*l IF( IP.GT.51 ) 1P*51 P LOT C IP)=ALFX I PI* 1 BINI A = 0 0 IF BINIA=BINI(I)/FL3AT(NBINI( I)) IF(IB1T.EQ.12)IPI = * FLOAT ( I P SC A L ) 2 6 IF(IBIT.E0.8MPI=(BINI( I )/FLOAT ( NB I N I ( I ) ) -PZ ERO ) / 128. *25. F LOAT( IPSC AL)26 IFC IP I .LT.1 ) IPI*1 IF( IPI .GT.51 ) IP I = 5l PLOT I( I PI ) ALF X BKAVEC I ) O INlA NUBIN(I)=NBINI(I) R I TE( 6.52 0 )B l NA, NB I N( I ) OI NI A NB I Ni ( I ) PL CJT PL OT I FORMAT (1X.F9.3. I4.F10.3. 15. IX. 102A1 ) PLOT( IP)=ALF B PLOT IC IP I ) *ALFB PLOT CI ) = ALFC PLOT(26)=ALFP PLOT (51 )= ALFC PLOT I (1 ) = ALFC PLOT 1C 26)=ALFP PLOT I (5 1 ) = ALFC B INC I )=0.0 NOINC I ) = 0 CONT I NUE GO TO 200 VRITE7. 406 ) T I TLE FORM AT( 9X.20A4 ) DO 410 1=I.NBMX * RITE(7,411 )DK AVE( I ).NUBIN( I ) I FORMAT (F10.3.I 4.T76 13) *PITE(7.420)AVEI.RMSI FORMAT CF8.2.F8.3) STOP THE FOLLOWING ARE SAMPLES OF THE CONTROL CARDS 0 I 5 1 2 I 6.0000 1 0 I .000000 0 1 (COMMENTS DO NOT NEED TO OE ON CAROS) NUMBER OF BITS OIGITIZFD (8 OR 12) TIMING PULSE PERIOD (MSEC) .2MSEC4TS AMP 4N A VEC *8) NUMBER OF PULSE PCRIODS AVERAGED PULSAR PERIOD (SECONDS) PLOT OF F SC T PLOT MAGNIFICATION FACTOR 0002 0 137 END 119 *jlcat 3'ri t uih **3r-u i7702 see fptour.NCV?26 3 HZ M 1 ? *46 v A Y 26 73 T5T Act 0 1 H 3 3 m 1 OS STOP 01 H TAM 13S TSAMp 02 * 11 y 25 70 7 r-T AO * C IH 34 t OS STOP 01 - 3M ORS T?AVD 0? OL 0C< AVFFAGF= 122.609 SIGMA o. C53 3 s T G M A s 0 1 7? MAGNIFICAT ION* 1 00 m 4rc r>* nt odes- VF- AGING T 1 **F = 46?.6 SFC * f N A Vr C A F P CC1OOS N *ME3 A V F r A G r 1 | ?? ' 3 9 1 90 2 12? S *3 3C? 1 1 2? f>6* 7?? A l 29 ? 71 3 93 S 12? 1C3 6 1 22.538 3 SO 7 l??.661 VO 9 12? *74 393 l?7,5in ico l 0 1 ??.' 3 109 11 12? 555 303 1 ? 122.677 3 CO 1 3 1 ? ^ r. 3 0 1 A 12? 554 303 1 S l ?? 54 8 -C3 1 6 P?."31 3 03 1 7 12? 548 3 CO 1 5 l 2? *5.3 a 3 CO !Q l. 5 n 2 ico 2 ' 12? 61 3 CO "* 1 12?.5 1 7 700 ?*> 122 559 300 23 12? .605 1 CO ?4 1 2? .5 7 A 3C3 T*c 12? 5 A 3 303 26 1 2?.565 3 CO *7 1 2*.7?e 303 8 I? 5SO 303 ?C 1 22.650 3 CO ?0 I * .7 0 3 93 1 12? 641 3C3 T * 1 22.6 87 303 3 1 1 2 ? 6 l c 3 00 A 12? *20 3 CO 35 1 22.685 3 CO 3* 127.606 3-0 17 122 S?5 ico ?** 12?.546 ? 1 A TIME INTENSITY. (J1 to Figure 3-15. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a period P = 1.187702 s (P-320 ys) . 46 T = O.6 93t c The correction due to the time constant of the receiver (for the half intensity level and assuming a triangular pulse shape) is given by (8): (3-9) where x is the post detection time constant of the receiver. No correction can be made due to the effect of polariza tion but some discussion about its effect will be given in the conclusions. Consequently, the corrected half intensity width of the pulse is: W. c = At r T (3-10) 0.5 0.5 ob 0.5 c where WQ ob is the observed half intensity width. Substituting in equations (3-8) and (3-9) the values DM = 4.8 pc cm ^ Av = 8 kHz vQ = 26.3 MHz and T = 11.25 ms, the following values are obtained for the corrections: Atg = 8.8 ms T =7.8 ms c Table 3-3 gives the values for the observed half intensity width and the half intensity corrected width for the pulses obtained on April 28 and May 25, 1979. The pulse width at the base is the width assuming a triangular shape, and is twice the corrected half intensity width. 3 pulsars. Also pulse dispersion in the interstellar medium becomes more severe the lower the frequency. At higher frequencies, compensation can be made for dispersion, per mitting the use of relatively high bandwidths. However, at the lower frequencies, it is generally necessary to restrict the bandwidth because of dispersion, further reducing the sensitivity of the detection system. Due to the intrinsic short pulse durations of most pulsars (< 150 msec), short time constants must be used, imposing still another limitation on the sensitivity of the system. Nevertheless, the sensitivity of the system can be improved by averaging many independent records of the same source. If N records are averaged, the signal-to-noise ratio (S/N) is improved by a factor of / For pulsars, this can be accomplished by averaging groups of consecutive periods. Digital sampling is generally used in such aver aging. Due to the relatively short pulsar periods, the signal needs to be sampled at a fast rate in order to have enough time resolution to meet the requirements of the sampling theorem. Since it is not feasible to manipulate the large amounts of data generated in the digitizing proc ess by hand, digital computing techniques must be used, providing an efficient way to average, process, store and plot that data. In the past few years, the development of small and inexpensive microprocessors has made possible improved methods of pulsar data handling and processing. If the I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree of Master of Science. Thomas D. Carr, Chairman Professor of Astronomy and Physics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree of Master of Science. Alex G. Smith Professor of Astronomy and Physics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree/of ifast^ of fkpience. Jfyh P. Oliver A'sociate Professor of Astronomy This thesis was submitted to the Graduate Faculty of the Department of Astronomy in the College of Liberal Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Master of Science. December, 1980 Dean, Graduate School PULSAR PSRI 1 33 16 PERIOP-1 1 8 M 0 2 3 SEC F RE QUENCY* 26. 3 MHZ P 1 I 2* A h M AY 25 7 9 T ST AC 7 0 1 H 10M I OS STOP 0 1 H 34M lOS T S AMP 02 o112*BO MAY 25 75 TST APT 0 IN 34 M 1 OS STOP 0 1 H 38M 0 AS T SAMP 02 PL OCK AVERAGE* 122 60S SI GMA* 0. OR 1 3 SIGMAs 0. 243 MAGNIF1C ATlONs NUMBER OF BLOC K S = 2 AVERAGING IIMEs 452.7 SEC IN NJMRFR AVERAGE NENIOOS AV FRAGED 1 122.tBi 2 24 2 122.632 300 3 122.723 3 90 4 122.t 1 0 3 50 5 I 2?.6 02 390 6 122.RR5 3 90 7 12?A 13 390 n 1 ?*>. 592 3 90 9 122.766 .190 13 1 2 2. t 4 e 390 1 I I 22.63C 3 90 1 2 122.630 J 90 1 3 122.543 390 1 4 122.5 | 0 39C IS 122.560 3 90 16 1 22. (.2 0 390 1 7 l 22.595 190 1 8 1 22.60 7 390 1= 122.546 3 90 2D 122.484 390 2 I I 22.5 1 390 22 122 .se? 390 23 122.490 390 24 122.700 3 90 25 1 22 R54 390 26 l22.728 350 27 12 2. 00 390 2 6 1 2252C 3 SC 20 122.515 3 50 30 1 22 567 390 2 1 122.636 390 32 122.672 350 33 122. 43 390 74 1 22 .533 360 35 122.651 3 50 2t l22.653 350 7 122.595 350 33 l22.555 2 16 TIME INTENSITY OI Figure 3-19. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a period P = 1.188023 s (DOPVEL period). 11 Observatory 45 MHz large array, located about 20 miles south-west of Santiago, Chile. The 45 MHz array is a filled rectangular array of 528 full-wave dipoles with an effective 2 ... area of 12,000 m It is a transit instrument with 6 inde pendently phased subsections in the N-S direction and can be oriented to a zenith angle of about 45 in the meridian plane. The HPBW is 2.1 north-to-south by 4.6 east-to-west. A multiple beam system is generated by a Butler matrix. A more detailed description of the array has been given by Reyes (14) and May et al. (15). Pulsar PSR2045-16 was recorded for 20 minutes each night, which is the time it takes to drift between the E-W 3 dB points of the beam. This pulsar was selected because of its relatively high pulse intensity (^1.0*10 2^Wm 2Hz 1 at 34 MHz), long period (1.96 ms) and its low dispersion measure (DM = 12 -3 parsec cm ). A zenith angle of only 18 at transit was also another favorable factor considered. The equipment used for the data acquisition of this pulsar consisted of a modified General Dynamics telemetry receiver, a Magnecord 1022 magnetic tape recorder, a Systron Donner 8220 TCG and a Rhode and Schwartz quartz time standard. The interconnections are similar to those shown in Figure 2-1. During the observing sessions the pulsar was transiting at dawn, and no interference was noticed. Nine of the eleven nights gave reliable data; the other two nights were discarded because of equipment malfunctions. r>UL ^ap pjici | 31 1 6 FDinosi i 8* a ? *; rC F Rc OJF NC YI ?6. 3 'AHZ pi i r* a+ may ?o ?Q T CAC 0 1 H 105 STDO 01H 3AM 105 TSA mo 0? 'll 2Rf. M Y ? 5 7? T5T A9 T 0| M "A A M IDS STOP 0 1 H 3 MM 94S TS*3 02 AvrCAGFs 1 ??.i 13 I r'* A? 0 0*.Q 3 ; 1 OMA r CV 208 MAGMIF|CAT IDN = o o N ) MHEP of ni ncK3 ? AVF* AGING TJHF= 45? 9 S^C n i s Ai/FPAGP I'fr 100 8 n;m*gr av re AGEO 1 122 ftl ? 3 90 1 ?? .ft,? ' 300 "A 1 ?? ft?0 3 90 4 12? 661 3 90 > l 2?*44f) 390 ft l ? ?l 7 3 QC 7 122 635 3 90 1 2? ft 1 IQft 9 12? ft A 1 Q0 1 3 122 5ft! 3 90 1 1 1 ??.8?5 3r0 1 2 1 2? '*1 3 3 90 I 3 122.589 3 90 1 4 1 22 .ft 3 > 3 90 1 5 l ?? ftO-* 3 *0 1 ft !2?,5HT 3 90 * * 1 ?>.Sft4 \ Oft 1 8 1 2? -505 3 90 1 9 I 22.5?6 90 > 1 1 7. ? 8 s i 1 >0 2 1 122 774 3 90 5 p 122.^07 390 > 4 1 ?"*, fto> A 90 24 122.684 3 90 ?. 1 22 71 7 3 90 122 -l3 390 27 1 ??,ft 7 6 A 90 * 122.4 i4 3 90 pn l ?* 57ft 0 o? 33 122 574 3 90 * ! 1 22 .ft 3 3 390 1 > 1 ?? ft 1 T 3 40 33 1 2? -607 390 1 1 22 ft56 390 71; I ?? 56Q 30Q 3ft 1 22 6 3 3 3 90 p, 122.525 226 38 122 86? 2 18 TIME INTENSITY 03 O Figure 3-23. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a period P = 1.188342 s (P+320 ys). 2 00 - 150 THIS THESIS BRUCK AND USTIMENKO (4) too CRAFT AND C0MELLA (6) SEPARATION A IZVEKOVA et al. (7) (ms) - o IZVEKOVA et al. (8) 50 AND LYNE et al. (22) 40 A 0 30 A o 0 20 o 1 1 .... i 1 1 1 lJ L_ i i 1 1 1 i i 10 20 30 40 50 00 500 1000 u> to FREQUENCY (MHz) Subpulse separation as a function of frequency for the pulsar PSR1133+16. Figure 3-8. 13 HAZELT1NE AMDAHL 370 TERMINAL COMPUTER Figure 2-2 Block diagram of the data reduction equipment 42 Regarding the pulse energy and mean flux density, more recently, in 1979, Izvekova et al. (7), have given some results at 61 MHz. They obtained a value for the pulse 26 2 1 energy of 0.89x10 Jm Hz and a mean flux density of ~26 0.75x10 Jy, using a linearly polarized antenna. In 1971, Slee and Hill (9), using a circularly polarized antenna, reported at 80 MHz a value for the pulse energy of 26 2 1 2.2x10 Jm Hz for one night (with a median over 6 nights of l.OxlO-26 Jm_2Hz-1). It is not easy to compare the values obtained in this thesis with values obtained by other authors, because of the polarization used and the different averaging times (not always given). Nevertheless, the value of pulse energy o r __ o __ i E =5.3x10 Jm Hz obtained at 26.3 MHz, looks at least P 26 2 1 comparable with the value of E = 2.2x10 Jm ^Hz obtained P at 80 MHz (which also corresponds to a peak flux density of ^110 Jy). According to Izvekova et al. (7), PSR1133+16 has a maximum in its average spectrum at v = 10040 MHz. Kuz'min etal. (20), making simultaneous observations at several frequencies, found that the instantaneous spectrum changes from day to day. For several days it doesn't show a maximum in its spectrum but rather the intensity seems to increase towards lower frequencies. This can be considered as part of the explanation of why the peak flux density, the mean flux density and the pulse energy obtained at 26.3 MHz are higher than the values 100 200 300 400 TIME (PERIODS) Figure 3-34. Relative pulse intensity as a function of time (periods) for pulsar PSR2045-16 on May 16, 1979. Each point represents an average over 10 periods. P = 1.961383 s. 98 radiation is relatively highly linearly polarized, this should produce a significant drop in the intensity of the received signal when using a linearly polarized antenna. This last effect can provide a possible explanation of the disappearance of the pulsar signal when switching to a different beam of the 26.3 MHz array. Regarding the effect of the polarization throughout the pulse, the pulse shape can be modified when using a linearly polarized antenna; the signal will appear more intense for the part of the pulse with polarization that matches the polarization of the antenna, and abnormally weak for the part of the pulse with position angle unfavor able for the antenna. Also if a given part of the pulse presents random changes in the position angle from normal to orthogonal (which led to a low integrated polarization for that part of the pulse), the intensity of the averaged pulse will be abnormally weak in that part. This seems to be the case with PSR1133+16, in which the observed intensity at 26.3 MHz of the second subpulse is usually stronger than the intensity of the first subpulse; the latter is often undetectable or appears only occasionally. The separation of these two subpulses at 26.3 MHz is much larger (^125 ms) than the predicted separation for the two subpulses (which compose the main pulse) obtained from data at higher fre quencies; however the separation at 26.3 MHz does agree with the results obtained at 25 MHz (4). 19 2.5 Computer Programs for Data Reduction In order to analyze the data and to plot the values, a set of two computer programs called PULSAR and MCAPGM are used. The basic parameters necessary to analyze the data must be given to the PULSAR program; they are: N = Number of bits digitized (=12) TPP = Timing pulse period (ms) NPA = Number of pulse intervals averaged P = Pulsar period (s) PLOFF = Plot offset MAG = Plot magnification factor The MCAPGM program (using the PULSAR program parameters) takes the first group of pulse intervals to be averaged (NPA) from the data of a particular block and calculates the over all average and the standard deviation. These two quantities, and the average for each bin, are plotted. The program next takes the second set of intervals from the data and does the same as before. Also the first set of intervals is averaged with the second set, computing a new average (cumulative) for each bin, an overall average, and a new standard deviation and plots the values for the bins. This process continues including each time the next group of intervals (NPA). The first set of programs is run once only to check the quality of the data,to detect any problem and to obtain the total overall average for a block which will be used in a second run. The second set of programs called PULSITO and MCAPGM1 is similar to the already described PULSAR and MCAPGM, but 39 A. In order to be consistent it has been considered that the value S for both the continuous source and the mm pulsar is that which would cause a deflection of 3 a (i.e., 3 standard deviations) from the mean value of the galactic noise background. The 3 a criterion was used to determine whether or not a pulsar pulse was present at the end of an integration run. B. It is assumed that the sample rate is fast enough compared with the post-detection time constant that very little information is lost in this process; then K can be considered 'll. (Details about sampling rates can be found in Appendix B.) C. It is assumed that the pulsar signal is unpolarized as in the calibration source. This is probably not correct; a strong linearly polarized component has been reported (18) The effect of variable Faraday rotation in the terrestrial ionosphere would then be to cause the plane of the linearly polarized component to be oriented (at different times) at any angles from parallel to perpendicular with respect to the antenna dipoles. As a result, if the polarization is actually 100% linear, the value of as given by (3-2) may range anywhere from too low by a factor of two to too high by a factor of infinity. Thus, no signal at all would be detected if the linear polarization is perpendicular to the dipoles, while according to (3-2) S would not be min p zero. 37 Table 3-2. S and Galactic background temperatures min ^ deduced from calibrations. S . min (Jy) Galactic background Radio Source temp. K 3C409 101 38,975 3C433 97 27,840 The galactic background temperature for pulsar PSR1133+16 is: T, = 18,514K backgr. All the galactic background temperatures given are actual brightness temperatures, corrected for losses in the antenna. A Hewlett Packard noise generator was used for the cali brations, and a pair of type 5722 noise diodes served as the standard of comparison for the noise generator. Details of the calibrations are given in Appendix E. The minimum detectable flux density has the following expression (17) : AS oc min x t syst (3-1) nAvi where n = number of averaged records Av = pre-detection bandwidth of the receiver T = post-detection time constant Tsyst = sYstem temperature In this case, the temperature of the system is completely dominated by the galactic background temperature, and in 94 nearby frequencies. They are also larger than the extrapolated flux densities from published data at higher frequencies. Also, the pulse shape and half intensity width seem to differ from extrapolated values based on data at higher frequencies. In order to find a satisfactory explanation of the intensity variations of the two pulsars, it is necessary to consider both previously reported intensity variations and the data available on polarization at higher frequencies. Pulsar PSR1133+16 and several other pulsars have been reported to show extremely variable flux densities both from pulse to pulse and on a longer time scale, when observed at 81.5 MHz (34). These intensity variations are not believed to be caused by interplanetary scintillation, and therefore must be attributed to the source. Pulsar PSR2045-16 has been reported as often being undetectable for several days at 408 MHz, but when it is active it can produce pulses stronger than any other of the first pulsars discovered (28) . Besides these intensity variations, the intensity of the received signal will show larger and more frequent variations when received by a linearly polarized antenna, if the pulses are strongly linearly polarized. Manchester et al. have found that most pulsars present a relatively high degree of linear polarization (32). PSR1133+16 was found to have a degree of linear polarization of 33% at "NIL 6 4 n PSW 1 1 33! 6 vnirnai.nmi >rr r u- ojescy-?6* 3 mhz r> 11 ?M 46 MAY *r, 70 r TA9- 01M 70M 106 STH9 01H 34M IOS T5AMD 0? nl 1 2WH6 MAY ?5 70 T5 T A 9 7 0 1 H 7 AM 10S 6 T O7 01 H 3M 04 S T 54 w 9 A2 Ow oc* 4 vrc A r,** 12? 60 9 f i gma* o. ? 7 M 3 6|5MA* 0.214 MAGNIFICATinM- 1 0 0 N)MHCP OF BLOCK 6- 2 Avr 'AGING 7 I Mr 46? 7 SrC N M-*FP AV cc AGED 1 12? 733 2?A >* 1 1 >*.' 7 4 3 00 1 T l 703 ~ 1 4 1 2? 50? 3 09 . 1 1 ? ? 6 9 390 . * 1 6 12? *0* *90 1 7 1 22 618 * 90 y 1 12? .* 1 5 3 90 * * 1 o 12? 7 7 5 790 ~~ 1 l ' 1 22.6?n 390 * 1 1 1 122.623 7 90 1 1 *> 122.579 3 90 * 1 1 7 l ?? .561 3 90 v 1 1 4 12?.^?3 390 * *-"tr~ ! 1 5 122 6 | 7 10 1 1 1 22.5 79 790 , <^T 1 1 7 122.620 390 1 1 8 1 2? .*= >? jco , 1 1 9 1 2? MH 3 r0 1 TIME V- * I "' 1?? .523 3 90 , * 1 1 12? 578 390 ~ * 1 2? 122.620 3 90 1 " t 122.4 70 390 , 1 ?4 122 71 J 9 0 -- 1 25 1 22 0 31 3 90 . 25 i 2- 12% 7 33 3 c0 , ~ i 27 122 *73 390 1 12? 4 9 3 90 ^ 1 > , 122.497 3 90 1 7 0 t ? 7 7 3 7 190 ~~ <_ 1 3 l 122 636 3 90 * 1 V 1 ?* ' 1 2 7o7 . 1 3 3 12? 643 390 , \ 3 a 122.53? 7 90 . . 1 75 1 2? .*69 390 . 1 36 1 2? ,64 3 390 , 1 37 122 579 7 90 i 3&! l22.539 2 15 1 ai Ln Figure 3-18. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a period P = 1.187983 s (P-40 ys). 17 P 1 X X Y Z Number that designates the night when the pulsar was recorded Letter that designates the block within the beam Beam in which the pulsar was recorded Designates pulsar PSR1133+16 Antenna beam designations are as follows: 2E 2nd beam east of transit IE 1st beam east of transit 1W 1st beam west of transit 2W 2nd beam west of transit In order to distinguish between the different blocks within a given beam, a capital letter was used, usually A and B. In case it became necessary to break up the data into more than two blocks (i.e. due to loss of synchronism of the TCG signal) C was also used. The indicator for the date of the night on which the pulsar was observed was an integer, as follows: Night Date 1 2 3 4 5 6 7 April 25, 1979 April 27, 1979 April 28, 1979 April 30, 1979 May 20, 1979 May 25, 1979 May 26, 1979 May 30, 1979 8 121 < .D3S I XI T *2.3 4 = 3 WI i 9NI9VajA9 4X 9 49 I * SXDOU JO a3HwON. C 1 i 4 / ) i 9tNM3 01 oz 3N | 1* XieNt O I 94 >>311 6* < 9I .a NOIj. X 3I3 I NO V * X 5eLJ1 9H9IS Gi*XSC*J* = VW01S.*XC t'BJ* .sdOVdilAv XDO 1U C I 1 /) i V *M I 00 fe* 9 V* 4 9 r SMi 101*3 IS 4 A vi DI < 00 9* ) 3 II ' Z* (f02M2mVh*JOJ OVC <* C 02* 1* I I ( 1 I 1 )1N3 alus *t = i ose oo >* < / ZH* 9:A3Nl anoaad Das 49#34 =ooia3d osi4 .vi-svozasd vs too *c ei )i 9*30.3 cce c* oji die ( ce e *> an a .* 2* < coft ) Ji i n oze i N < 1* I)AVX-U = < T)3AV OIC C9 ( T l )N!r.N = ( !> a3dN 6-.. NIbN M al O 1 C 00 65 ( t n*i )NiunN4^3dns = a3cjns boc ss N ItN* 1= IN 90C 30 SS o#o=i3cjn ts ioi9is*soi-i eb < l >9IS =1019IS ZS m avMi = Avioi soe is 02c ni no es UiJlSt.sOlSHi 6 9 c ((I-NI9N ) iNiflN/Hins > /os3arb ) iaos*inioi s 99 Oi i]o(NiLN/Hi*nq )s j/,ii z 9 (xM3dN4nmn=Hi*ns 00 e 99 < x> 3d N < *1033 >1* es Jtir.S = 0S3fS s 9 2*$ O ) S Jd = ( > )0S 33 9*
N ) 3 AV-AV 101= I N) 53*3 f
ni kjN i = x oce 00 cv
0*0-OS 3JOS 19
o *o =hi*ps c 9
IN30/1WON=AV101 6e
( I ) ladN+ IN 30 = IN3Q 052 bC
( I )U3dN9( I )3AV4 IhONsIrtPH Â£
N I GN *1 I O S 2 U O *C
o*c=in 30 se
o *c= Ikon 9e
iDa3dN/wnN = ir 3av ore v.e
N3G-(r ) jddN 2C
< r4i )ni triM n3o=n33 osi le
ir*1>niqon*iri)ava i4nN:hnN ce
mQNMsl OSI OC 62
O O =N 3CJ 92
o*c=wns lz
NI BN Iaf 002 OU 52
( !M I3JDQ* CN* IN) AVN1H* XlfeN *2=IN II 00 92
N!bN 41 =N 2 1 I OO e
19 ) A9X*l-< I ) AV X Jl= ) J3j 30 0 11 22
XlN C-ti C I I OC 12
50C OI 09 < l 3"1* >3QN> 3 I C 2
SON I1N0 3 02I 61
(G*3 Z^J) iVhdUJ 51 fe I
( ( I IAVXbl)St9 =( I )AVXi Zl
< I >9! S4 ( I ) AV>91< 5 l *5)0933 *t
< 9| Â£ C 13) 19*303 CI SI
< (r i ) av a i ) suv = c r i ) ax> ih coi 91
c r 1) n mnN ir 1) Avxiui o* 5 lux j 9 ei
MOKl=r COI QU 21
(9902 )19*933 6 l l
<02 4 l = 1 I 4< I 1 4 I > 1N3C! ) < 6 4 5)09 3 3 OI
X WN 1 =1 C2 I OO A
< *6 3 4 9 I C ) i vwnld 9 9
O01.33d 49 9* *NI UN 4X1UN< fe4 5>C9 33 Z
/ Hl/)(Nin*/ |M I/HI1'/W1 /IVA'/'Hl/'U H/*H l/XVAV 9190 *
99 4 I W9 I S4ld9 I S4 d.l AV 4 JO 19A d3931N| S
< oo i ) a3dN < oo i 4o i )MnrN mD93ini 9
WII4 1 330 4 <0 I >3 JU30* 13S J 3 4 I kON'kON TV3d L
<0 01 )f*S 3a <00 1 )S 3d4 < 00 l )3 A9 < C I >9 1 4 < O 1 ) AV x* X 4 < GO l 4 t l ) AVX 1u IV 3*J c
ucr *

95
403 MHz, 50% at 86 MHz and 60% at 63 MHz, which is the
lowest frequency at which polarization has been measured
for this pulsar (35, 36). There is a clear dependence of
percentage of linear polarization upon wavelength, and the
tendency is for the percentage of linear polarization to
increase with increasing wavelength. Pulsar PSR1133+16
has a complex pulse, composed of two subpulses; the per
centage of linear polarization is variable throughout the
pulse and reaches about 70% in the middle of it at 403 MHz.
For individual pulses, the position angle of the plane of
linear polarization is highly stable for all but the first
component, which usually shows two preferred positions
perpendicular to each other. The result of this is that
the degree of linear polarization of the integrated pulse is
low for the first subpulse and high for the rest (32).
Pulsar PSR2045-16 has a complex pulse composed of 3
subpulses. Manchester et al. (32) measured the degree of
linear polarization to be 40% at 403 MHz, with the polariza
tion being even larger for the first subpulse (^70%). The
position angle of the plane of linear polarization varies
smoothly throughout the pulse; a change of about 30 occurs
between the first and third subpulses. This pattern of
position angle variation within a pulse is stable for many
consecutive pulses. No known polarization measurements are
available at lower frequencies for this pulsar.
Even if the linear polarization position angle were
stable when measured at a point above the terrestrial

PAGE 1

OBSERVATION OF PULSARS AND OTHER SOURCES AT LOW RADIO FREQUENCIES BY FRANCISCO REYES A THESIS PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 1980

PAGE 2

to Maria Angelica Carolina and Daniela

PAGE 3

ACKNOWLEDGMENTS I wish to express my deepest gratitude to my research advisor. Dr. Thomas D. Carr for his guidance and encouragement, and his patience in reading and correcting the manuscript. It has been a privilege to work under him in this project, and I have benefited in many ways as his student. I am also grateful to Dr. Alex G. Smith for providing me with financial support and for serving as a committee member. I give my appreciation to Dr. John P. Oliver for the development of the necessary hardware and software, for his advice and for serving as a committee member. My deep appreciation goes to Dean Claudio Anguita and Professor Hugo Moreno of Facultad de Ciencias Fisicas y Matematicas of University of Chile for their encouragement and interest in my career and for providing me with financial support. I am also grateful to Jorge May and Juan Aparici for their encouragement and optimism and for providing me with the data at 45 MHz. I wish to thank Jorge Levy for his help in the operation and calibration of the 26.3 MHz array, Chris St. Cyr for his encouragement and help in the computational part and Wesley Greenman and Richard Flagg for their help with the electronics.

PAGE 4

My thanks goes to Hector Alvarez, Mauricio Bitran and Fernando Olmos of Maipu Radio Observatory for their help with the 45 MHz data. I am grateful to the Astronomy Department and the Physics Department and in particular to Drs. Heinrich Eichhorn, Robert J. Leacock, F. Eugene Dunnam, Charles F. Hooper and Richard E. Garrett for their financial support in the form of teaching assi stant ships and fee waivers. My gratitude goes to Mrs. Brenda Dobson for her patience in correcting my English and for her devotion in typing and editing the manuscript. The realization of this project has been possible thanks to the continuous support provided to the radio astronomy program of the Department of Astronomy of the University of Florida and to the Maipu Radio Observatory through a grant from National Science Foundation (principal investigator Dr. Alex G. Smith). iv

PAGE 5

TABLE OF CONTENTS PAGE ACKNOWLEDGMENTS iii ABSTRACT vii CHAPTER 1 INTRODUCTION 1 CHAPTER 2 DATA ACQUISITION AND REDUCTION 5 2.1 Data Acquisition of Pulsar PSR1133+16 6 2.2 Data Acquisition of Pulsar PSR2045-16 10 2.3 Data Reduction 12 2.4 Block Designation for Each Pulsar 16 2.5 Computer Programs for' Data Reduction.... 19 CHAPTER 3 DATA ANALYSIS 22 3.1 PSR1133+16 Data Analysis 22 3.1. a Peak Flux Density, Mean Flux Density and Energy for Pulsar PSR1133+16 35 3.1.b PSR1133+16 Pulse Width 44 3.1. C Tests Performed with the Pulse Obtained on May 25, 1979 48 3.2 PSR2045-16 Data Analysis 63 3. 2. a Peak Flux Density, Mean Flux Density and Energy for Pulsar PSR2045-16 77 3.2.b Pulse Width for Pulsar PSR2045-16 83 3.2.C Period of Pulsar PSR2045-16 from Observations on Two Consecutive Days 87 3.2.d PSR2045-16 Pulse Intensity as Function of Time 91 CHAPTER 4 CONCLUSIONS 93 APPENDIX A LIST OF PULSARS 103 APPENDIX B SAMPLING RATE, TIMING PULSE FREQUENCY AND DATA POINTS PER SECOND 10 7 APPENDIX C SCHEMATICS HO V

PAGE 6

APPENDIX D COMPUTER PROGRAMS 113 APPENDIX E 26.3 MHz ARRAY CALIBRATIONS 123 LIST OF REFERENCES 131 BIOGRAPHICAL SKETCH 135 vi

PAGE 7

Abstract of Thesis Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science OBSERVATION OF PULSARS AND OTHER SOURCES AT LOW RADIO FREQUENCIES By Francisco Reyes December 1980 Chairman: Thomas D. Carr Major Department: Astronomy The investigation of the radio emission spectra of pulsars at low frequencies is important since it helps to clarify the mechanism responsible for this emission. At the present, the low frequency end of the spectra is poorly determined compared with more precise measurements available at higher frequencies. Also, measurements of pulsar intensity variations at low frequencies due to scintillation are of value in the study of the interstellar medium. Detection of pulsars at decametric wavelengths is difficult because of the deterioration of the signal-tonoise ratio caused by the high galactic background temperatures and the turn-over of the spectra of most pulsars. Also the dispersion of the pulses imposes the use of narrow bandwidth, reducing the sensitivity of the system. vii

PAGE 8

Two pulsars have been observed at low radio frequencies using narrow bandwidth to avoid pulse dispersion problems. Pulsar PSR1133+16 was observed at 26.3 MHz using the University of Florida large array; pulsar PSR2045-16 was observed at 45 MHz using the Maipu Radiobservatory large array (Chile) Long term integration techniques were used to reduce the data. Both pulsars were successfully detected and the average pulse intensity and integrated pulse profile were obtained. Time resolutions of P/38 were obtained for PSR1133+16 and P/62 for PSR2045-16, where P is the pulse repetition period. The two observed pulsars were selected because of their low dispersion measure, long period (>1 sec), high pulse intensity and their location with respect to the galactic background. Several other radio sources were also observed and were used to calibrate the effective area of the 26.3 MHz antenna. Chairman viii

PAGE 9

CHAPTER 1 INTRODUCTION Pulsars were discovered in 1967 by S. Jocelyn Bell and Anthony Hewish. The first pulsar paper was published in Nature in February 1968 by Hewish et al. (1) telling about the discovery of pulsar PSR1919+21 and giving some of its parameters. Since then, about 321 pulsars have been detected and cataloged. According to currently accepted models, pulsars are rapidly spinning neutron stars originating from supernova explosions. They are believed to have strong magnetic 1 O fields on the order of 10 gauss, probably the strongest known magnetic fields in the Universe. The neutron star is believed to emit continuously a narrow beam of radio waves which corotates with the star, producing the observed pulse each time the beam sweeps past the earth. The periods of the pulses which have been observed range from about 0.0 3 to 4.3 seconds. So far, for only 2 pulsars has it been possible to find the optical counterpart of a pulsar. Several mechanisms have been proposed to explain the origin of the radio emission and how the rotational energy of the neutron star is converted into electromagnetic wave pulses; this subject still remains one of the least understood aspects of pulsars. Some of the important parameters

PAGE 10

2 of a pulsar that need to be measured before realistic models of the radio emission mechanism can be developed are the spectrum of the radiation, the pulse shape, peak intensity, and the time variations of these parameters. A large number of measurements have been made in the range of frequencies between 100 and 2,000 MHz, but only a few below 100 MHz (2) The low frequency part of the spectrum is particularly interesting because most pulsars present a turn-over of the spectrum at metric and decametric wavelengths, and also because pulse shape and widths show some peculiarities in this region. The spectral maximum for most pulsars is in the range 60 to 250 MHz; this peak is nearly symmetrical with respect to the frequency at which the maximum occurs (2) Measurements of the low frequency part of the pulsar spectra have been made by Bash et al. at 38 MHz (3), Bruck and Ustimenko at 10, 16.7, 20 and 25 MHz (4, 5), Craft and Comella at 40 MHz (6) and more recently by Izvekova et al. at 61 and 102.5 MHz (7, 8). Most pulsars measured at low frequencies are in that part of the sky which is observable from northern hemisphere observatories. Observations at the lowest frequencies have been carried out in the southern hemisphere only by Australian observatories, at 80 and 150 MHz (9, 10) Detection of pulsars at low radio frequencies is difficult because of the deterioration of the signal-to-noise ratio caused by the high galactic background temperatures, and the low frequency spectral turn-over exhibited by most

PAGE 11

3 pulsars. Also pulse dispersion in the interstellar medium becomes more severe the lower the frequency. At higher frequencies, compensation can be made for dispersion, permitting the use of relatively high bandwidths. However, at the lower frequencies, it is generally necessary to restrict the bandwidth because of dispersion, further reducing the sensitivity of the detection system. Due to the intrinsic short pulse durations of most pulsars (< 150 msec), short time constants must be used, imposing still another limitation on the sensitivity of the system. Nevertheless, the sensitivity of the system can be improved by averaging many independent records of the same sovirce. If N records are averaged, the signal-to-noise ratio (S/N) is improved by a factor of /n For pulsars, this can be accomplished by averaging groups of consecutive periods. Digital sampling is generally used in such averaging. Due to the relatively short pulsar periods, the signal needs to be sampled at a fast rate in order to have enough time resolution to meet the requirements of the sampling theorem. Since it is not feasible to manipulate the large amounts of data generated in the digitizing process by hand, digital computing techniques must be used, providing an efficient way to average, process, store and plot that data. In the past few years, the development of small and inexpensive microprocessors has made possible improved methods of pulsar data handling and processing. If the

PAGE 12

4 period of a pulsar is known to be a good accuracy, it is possible to average many periods (N typically in the range of a thousand periods), greatly improving the S/N ratio. The University of Florida Radio Observatory, located in Dixie County about 50 miles west of Gainesville, and the Maipu Radioastronomical Observatory of the University of Chile, located 20 miles southwest of Santiago, have long been engaged in a cooperative program of low frequency radio astronomy. Although most of the effort over the years has gone into the study of Jupiter's radio emission, other decameter-wavelength sources are also investigated. Each of the two observatories has a very large antenna array, in addition to many smaller radio telescope antennas. The Florida array consists of 640 dipoles operating at a center frequency of 26.3 MHz, while the Chile array has 528 dipoles and operates at a center frequency of 45 MHz. Both arrays were believed to be suitable for observing pulsars, although previous attempts to use them for this purpose had not been successful. The goal of the research on which this thesis is based was to develop a low frequency pulsar data acquisition and reduction system, and to use it to detect and measure the important parameters of at least one pulsar with each of the two large arrays (i.e. the one in Florida and the one in Chile). This goal was achieved. <

PAGE 13

CHAPTER 2 DATA ACQUISITION AND REDUCTION Two pulsars, PSR1133+16 and PSR2045-16, were selected (from the lists in Appendix A) for observation because of their low dispersion measure, relative long period (>1 second) high pulse intensity and favorable location with respect to the galactic background. Pulsar PSR1133+16 was observed at 26.3 MHz using a bandwidth of 8 kHz, and pulsar PSR2045-16 was observed at 45 MHz with a bandwidth of 10 kHz. The detected IF signals of the pulsar were recorded on one channel of a magnetic tape, and a time code generator signal on the other channel. The tapes were played back (making a 4:1 slow down), the pulsar signal sampled, and the resulting values stored in the NERDC Amdahl 470 computer, The time code generator signal provides the timing pulses for the sampling and also indicates Universal Time (U.T.). The sampling of the signal was accomplished by a 12 bit analog/digital converter (A/D) and a KIM I microprocessor. Pulsar pulse averaging is carried out by a computer program which averages consecutive periods of the sampled pulsar signal, previously stored in the Amdahl computer. 5

PAGE 14

6 2.1 Data Acquisition of Pulsar PSR1133+16 The pulsar PSR1133+16 was observed for 20 nights between April 25, 1979, and June 8, 1979, using the University of Florida 26.3 MHz large array, located at the University of Florida Radio Observatory in Dixie County. The 26.3 MHz array is a nearly filled rectangular array of 2 640 half-wave dipoles that covers about 30,000 m The half power beamwidth (HPBW) is 2.5 north-to-south by 6.0 east-to-west. It is a phase steered array with full N-S directional control of the beam in quarter-beamwidth increments, and a selection of 8 E-W beam directions spaced one beamwidth apart. Phasing is accomplished by means of a network of Butler matrices, hybrid rings and plug-in phasing cables North-south phasing of the array for a source having a given declination is accomplished in two ways, a coarse adjustment in large increments of declination and a fine adjustment providing small steps. A change in the coarse phasing requires the replacement of many plug-in cables located throughout the array, and requires about 2 or 3 manhours of labor. Fine phasing for the desired declination is quickly accomplished by means of switches at a single location. The beam pointing can be altered in small increments to any declination within a strip roughly 10 wide by means of the fine phasing controls. Positioning the beam to a new declination outside this strip requires a coarse phasing readjustment. The right ascension of the beam is of course

PAGE 15

7 controlled by the rotation of the earth. Scans by any or all of the 8 selectable east-west beams (designated 4E 3E, 2E, IE, IW, 2W, 3W, are made by starting each scan at the appropriate sideral time. A computer program called ARRAY developed by Dr. M. D, Desch, is used to obtain information on cable lengths required for coarse phasing, switch positions for fine phasing, and beam transit time across the desired source (11) A more detailed description of the antenna and its adjustment has been given by Desch, (12) and Desch et al. (13) In order to use the maximum gain of the array, only the central E-W beam positions 2E IE, IW and 2W were used, and within each beam only the central portion of it was recorded ('V'8 minutes each beam) A maximum of 32 minutes was recorded each night. Pulsar PSR1133+16 was selected because of its relatively high pulse energy (^^^0.75x10"^^ Wm'^Hz""*" at 61 MHz), long period (1.188 s) and low dispersion (DM s 5 -3, parsec cm ) it was also easy to reach this pulsar using the phasing the antenna had at that time (phased to make Jupiter observations), without having to completely rephase the array. Only the fine phasing needed readjustment. During the period in which the pulsar was observed, it was transiting at about midnight. Only 8 nights (out of the 20 observed) gave reliable data, free from static and stations. The end of the observation period coincided with the beginning of the thunderstorm

PAGE 16

8 season, at which time it was necessary to shut the array down for the summer. A block diagram for the interconnections of the equipment used to observe the pulsar is shown in Figure 2-1. The equipment basically included a Collins R 390 receiver, a Magnecord 1022 magnetic tape recorder, a Tracer 5D frequency standard, and a Systron Donner 8154 time code generator/reader The time code generator (TCG) was synchronized against WWV; this was accomplished by triggering an oscilloscope with the 1 pulse per second (1 PPS) output of the TCG and measuring the delay of WWV 1 PPS. The 1 PPS of the TCG was brought as close as possible to the 1 PPS of WWV (usually between 5 to 20 ms) by making short interruptions of the 1 MHz signal from the Tracer frequency standard. The delay between the two 1-PPS pulse trains was measured to an accuracy of about 0.5 ms, and from that information the U.T. was obtained, except for the propagation delay between Boulder, Colorado, and Dixie County Radio Observatory. No corrections have been made for this delay because no attempt was made to relate the epoch of this observation with any other observations The measured drift of the Tracer time standard is about 80 ys/h; this causes a drift in the epoch of about 100 ys between pulsar transits by beams 2E and 2W Cv^l^lG^) ; this drift is small enough to be noticed, considering the time

PAGE 17

9 V 26.3 MHz ARRAY COLLINS R 390 RECEIVER MAGNECORD 1022 CHI CH2 TRACOR 5 D FREQUENCY STANDARD SYSTRON DONNER 8154 TCG TRIGGER wwv INPUT RECEIVER OSCILLOSCOPE Figure 2-1. Block diagram of the equipment interconnections used to record pulsar PSR1133+16.

PAGE 18

10 resolution of '^^31.5 ms (P/38; P = period) obtained for this pulsar. The detected IF output of the receiver was recorded on channel 1 of the Magnecord at 15 IPS and the recording level for the galactic background was set to about 7 VU on the recording level monitor, to prevent any saturation during signal peaks. The time code generator was recorded on channel 2 of the Magnecord. This signal serves a double purpose: it provides a time reference for the pulsar signal, and later on, in the data reduction process, provides the timing signal for the KIM I microprocessor. As a back-up to obtain the correct time, at the beginning of each observing session the WWV signals were recorded on channel 1 (the TCG also on channel 2) for about 2.5 minutes. The correct time can be obtained by measuring the delay between the 1 PPS of WWV and the 1 PPS of the TCG. The receiver was tuned to a center frequency of 26.3 MHz and the bandwidth set to 8 kHz. Using data for higher frequencies an extrapolated pulse width (at the base) of about 80 ms was obtained. Pulse broadening due to dispersion (DM = 5 parsec cm ) is about 18 ms (for a bandwidth Av = 8 kHz) which was considered acceptable. 2.2 Data Acquisition of Pulsar PSR2045-16 Pulsar PSR2045-16 was observed for 11 nights between May 15 and May 30, 1979, using the Maipu Radio Astronomical

PAGE 19

11 Observatory 45 MHz large array, located about 20 miles south-west of Santiago, Chile. The 45 MHz array is a filled rectangular array of 528 full-wave dipoles with an effective 2 area of 12,000 m It is a transit instrument with 6 independently phased subsections in the N-S direction and can be oriented to a zenith angle of about 45 in the meridian plane. The HPBW is 2.1 north-to-south by 4.6 east-to-west. A multiple beam system is generated by a Butler matrix. A more detailed description of the array has been given by Reyes (14) and May et al. (15) Pulsar PSR2045-16 was recorded for 20 minutes each night, which is the time it takes to drift between the E-W 3 dB points of the beam. This pulsar was selected because of its relatively high pulse intensity (-^^1. OxlO~^^Wm~^Hz~''' at 34 MHz), long period (1.96 ms) and its low dispersion measure (DM = 12 -3 parsec cm ) A zenith angle of only 18 at transit was also another favorable factor considered. The equipment used for the data acquisition of this pulsar consisted of a modified General Dynamics telemetry receiver, a Magnecord 1022 magnetic tape recorder, a Systron Donner 8220 TCG and a Rhode and Schwartz quartz time standard. The interconnections are similar to those shown in Figure 2-1. During the observing sessions the pulsar was transiting at dawn, and no interference was noticed. Nine of the eleven nights gave reliable data; the other two nights were discarded because of equipment malfunctions.

PAGE 20

12 A bandwidth of 10 kHz was used; this is the minimum bandwidth of the receiver. Using data obtained at higher frequencies, an extrapolated pulsar pulse width at the base of about 130 ms was obtained at 45 MHz. The pulse broadening due to the 10 kHz bandwidth is 11 ms. This was considered acceptable. The quartz time standard is checked daily against the Chilean National Observatory cesium time standard; the epoch is believed to be known with an accuracy of about +10 ys. The drift in time is about 3.6 ys/h (or about 0.00190 ys/P, where P = 1.96 s, the pulsar period). The drift is 1.2 ps in the 20 minutes between the east-to-west HPBW. For all practical purposes considered in this research, this is negligible. 2.3 Data Reduction Data reduction for both pulsars was performed using the facilities of the Department of Astronomy of the University of Florida in Gainesville. The basic configuration and interconnections of the equipment for data reduction are shown in Figure 2-2. The schematics for the most relevant blocks are presented in Appendix C. As stated earlier, the IF signal in the receiver is band limited to 8 or 10 kHz and appropriately smoothed. The resulting audio frequency signal is recorded in analog form on channel 1 of the Magnecord tape recorder at a tape spead of 15 IPS. The time code generator is recorded on channel 2.

PAGE 21

13 MAGNECORD 1022 CHI CH2 DETECTOR TIME CONSTANT DC AMPLIFIER AND OFFSET SYSTRONS DONNER 8154 TCG A/D CONVERTER (12 BITS) PILASE LOCKED LOOP DC OUTPUT 1 PPS KIM I 125 Hz 250 Hz SYNCHRONIZER AND DIVIDER : 2 TIMING PULSES SYNCHRONIZING SWITCH BRUSH CHART RECORDER HAZELTINE TERMINAL TO AMDAHL 370 COMPUTER Figure 2-2. Block diagram of the data reduction equipment,

PAGE 22

14 In order to achieve the desired time resolution and to be able to sample the data at an appropriate rate, the magnetic tapes were played back at 3 3/4 IPS, providing a slow down of 4:1. Pulsar signals from channel 1 of the Magnecord are sent through a diode detector and smoothing filter to the 12 bits A/D converter. The filter time constant was adjusted to 4 5 ms for both the rising and falling edges of the noise pulses. The TCG signal from channel 2 of the Magnecord is connected to the input of the TCG (working now as a reader) to provide both a visual display of the time and a 1 PPS pulse (taken from a modified output) The latter gives the synchronizing signal to start the sampling of the data at a precisely known time. It is particularly important to know the time at which the sampling is started in order to be able to combine in phase 2 or more blocks of data. In order to obtain the timing signal for the KIM I, the TCG signal from channel 2 of the Magnecord is connected also to a phase locked loop (PLL) circuit. The PLL circuit locks at a particular frequency of the signal, filtering out the rest of the frequencies present in the complex wave form of the TCG signal. The 1000 Hz signal of the TCG is transformed to 250 Hz due to the 4:1 slow down. The PLL was carefully tuned to 250 Hz, and the output then divided by 2. The resultant 125 Hz constitutes the timing signal for the KIM I. A gated DC output from the PLL circuit was monitored on a Brush Mark II chart recorder. This provides a good means

PAGE 23

15 to detect any loss of synchronism or dropout of the TCG playback signal. If there is any loss of synchronism between the input and PLL signals, then the gated DC output falls to zero, giving a visual indication to stop the sampling process. If synchronization were lost, the pulsar signals would be averaged out of phase, wiping out any pulse at the end of the integration process. A means was provided for starting the sampling at the next 1 PPS pulse from the TCG occurring after a switch was thrown. This made it possible to know the exact time at which sampling started. This was accomplished by connecting the start switch to the D input and the 1-PPS signal to the clock (c) input of a D-type flip-flop. The output of the flip flop gates the 125 Hz signal in a NAND gate (see schematic in Appendix C) When the start switch is turned on, the next 1 PPS of the TCG enables the 125 Hz signal to go into the KIM I and the sampling is started. The sampled values (coded in hexadecimal) are sent through the Hazeltine terminal to the Amdahl 4 70 computer and stored there for further processing and analysis. Since the maximum number of cards that can be stored under a new data file name is 493, it was necessary to break up the data into several "blocks" to meet this requirement. The 8 minutes of data obtained in one beam for pulsar PSR1133+16 were broken up in 2 blocks of 4 minutes each, generating .about 312 cards for each block (for a TSAMP= 02 and timing signal frequency 125 Hz; see Appendix B) The

PAGE 24

16 20 minutes of data obtained each night for pulsar PSR2045-16 was broken up into 4 blocks of 5 minutes each, generating about 390 cards for each block. The method of breaking up the data into blocks proved to be useful also because one can inspect the results of the average of each block separately, making it possible to search by eye for a possible pulse that might appear at the same location in the different blocks. It is highly unlikely that interference transients would appear in the same location in each of several blocks. Monitoring of the DC gated output of the PLL circuit is very useful since synchronism is sometimes lost when a tape is played back. If that happens, then that portion of the tape is skipped and a new block started. The KIM I program which is necessary to digitize the data is loaded into the microprocessor from a cassette tape recorder. 2.4 Block Designation for Each Pulsar Many blocks of data were generated in the data reduction process. In order to avoid any confusion with the data and to be able to distinguish quickly between blocks, a designation scheme was developed for referring to a specific data block. The designation format contains in abbreviated form the pertinent information for that block. For pulsar PSR1133+16, the following block designation was adopted:

PAGE 25

17 P 1 1 XX Y Z Number that designates the night when the pulsar was recorded Letter that designates the block within the beam Beam in which the pulsar was recorded Designates pulsar PSR1133+16 Antenna beam designations are as follows: 2E 2nd beam east of transit IE 1st beam east of transit IW 1st beam west of transit 2W 2nd beam west of transit In order to distinguish between the different blocks within a given beam, a capital letter was used, usually A and B. In case it became necessary to break up the data into more than two blocks (i.e. due to loss of synchronism of the TCG signal) C was also used. The indicator for the date of the night on which the pulsar was observed was an integer, as follows: Night Date 1 April 25, 1979 2 April 27, 1979 3 April 28, 1979 4 April 30, 1979 5 May 20, 1979 6 May 25, 1979 7 May 26, 1979 8 May 30, 1979

PAGE 26

18 I The designation scheme for the blocks of data from pulsar PSR2045-16 is slightly different, as follows: P 2 0 4 5 X Y Number that designates the night when the pulsar was recorded -Letter that designates the block within that night Designates pulsar PSR2045-16 The block designation is usually one of the letters A, B, C, D, E or F. The nights were numbered as follows: Night Date May 15, 1979 1 May 16, 1979 2 May 17, 1979 3 May 18, 1979 4 May 19, 1979 5 May 22, 1979 6 May 23, 1979 7 May 24, 1979 8 May 29, 1979 9 May 30 1979 The reduced data for each block was stored on a disk called UF.B0030601.S5.XXXXX, where XXXXX is the name of the block. The data have been recorded also on a digital magnetic tape called PULSAR.

PAGE 27

19 2.5 Computer Programs for Data Reduction In order to analyze the data and to plot the values, a set of two computer programs called PULSAR and MCAPGM are used. The basic parameters necessary to analyze the data must be given to the PULSAR program; they are: N = Number of bits digitized (=12) TPP = Timing pulse period (ms) NPA = Number of pulse intervals averaged P = Pulsar period (s) PLOFF = Plot offset MAG = Plot magnification factor The MCAPGM program (using the PULSAR program parameters) takes the first group of pulse intervals to be averaged (NPA) from the data of a particular block and calculates the overall average and the standard deviation. These two quantities, and the average for each bin, are plotted. The program next takes the second set of intervals from the data and does the same as before. Also the first set of intervals is averaged with the second set, computing a new average (cumulative) for each bin, an overall average, and a new standard deviation and plots the values for the bins. This process continues including each time the next group of intervals (NPA) The first set of programs is run once only to check the quality of the data, to detect any problem and to obtain the total overall average for a block which will be used in a second run. The second set of programs called PULSITO and MCAPGMl is similar to the already described PULSAR and MCAPGM, but

PAGE 28

20 this time, the data is taken from disk (where it has been previously stored) produces a magnified plot and punches cards with the final values for each bin, the total block average and the final standard deviation. This set of cards is used later on, to stack 2 or more blocks. Usually after the process already described, the digitized data is recorded on the PULSAR magnetic tape and released from the data files. In order to stack 2 or more blocks, the sets of cards of each block (except the first, which is used as reference) is shifted the proper number of bins so that the bins can be averaged in phase. The number of bins to be shifted in each block is given by the following relationship: T_.-T \-] 1 -M J (2-1) bins Frac where N bins T i P R Frac R = Number of bins the block i needs to be shifted with respect to block 1 = Time start sampling block 1 = Time start sampling block i = Pulsar period = Time resolution (= Number bins Fractional part T -T^ The fractional part of corresponds to the fraction of the period between the time of beginning of block i and the time of beginning of the first block. The relationship given by 2-1 was evaluated using an HP-29C calculator.

PAGE 29

21 A last program called P2045 (or P1133) reads the cards of the blocks to be averaged (previously shifted) calculates the average for each bin (weighting by the number of periods in each bin) computes the final average of the combined blocks, the standard deviation a, and plots the data. Also plotted is the 3 a reference line, which is used as a reference to aid in deciding whether or not a pulse was present at the end of the averaging. The KIM I program to digitize the data and the MCAPGM program were developed by Dr. John P. Oliver; MCAPGMl is a modification, made to adapt it to the use previously explained. A program called DOPVEL was used to obtain the apparent pulsar period for the night and time when the pulsar was observed. This program gives the pulsar period at a particular site and at a given time, correcting for Doppler shift due to the earth's rotational velocity and its motion around the sun. This program was developed at the NRAO by Drs. R. N. Manchester and R. A. Gordon and made available by Dr. Steve T. Gottesman Finally a summary of the times spent observing and reducing pulsar data and the total time of terminal use is presented in Table 2-1. Table 2-1, Summary of time spent with pulsar data. Total Time Total Time of Data Reduced Total Time Digitizing Data Pulsar Observing Pulsar PSR2045-16 PSR1133+16 3 h 12 h

PAGE 30

CHAPTER 3 DATA ANALYSIS 3.1 PSR1133+16 Data Analysis From the 8 nights of data reduced for pulsar PSR1133+16 (26.3 MHz, Florida), pulses were found with no ambiguity on 2 nights. Pulses were found on April 28 and May 25, 1979, over averages of 340 and 390 periods respectively; the results are plotted in figures 3-1 through 3-7. On April 28, after averaging over 340 periods ("^6^44^) a pulse appears in bin 33 with an amplitude of ^^.3 o above the mean value (Figure 3-1) The 2 data blocks P112EA3 and P112EB3 used to obtain this final block have been plotted separately in figures 3-2 and 3-3. It can be seen that a pulse appears in both blocks at the same location (bin 33) Even though the pulse appearing in each block would not alone be considered to be statistically significant, the fact that they appear in the same location for two separated sets of data provides an excellent test of the validity of the pulse. On May 25, after averaging over 390 periods, a prominent pulse was obtained in bin 25 with an amplitude of 3.4 a; this pulse is shown in figure 3-4. The same test for validity as before was made. Blocks P112WA6 and P112WB6 are 22

PAGE 31

23 N y3 o o o o a a Â— 1 1 < < II 1 I K HC O O < tT O O r I u. >Â• i/l o Â— LJ o Â— Z o U X T < D M fVJ ^ O o o 111 0. c. o o (/> CO u LJ I/) 1/1 l/l o irt o l> t mi/) ti o o s CD u IX o L. IT) O O if> a n D a a < < C a r, in w Q <> II 0" CO U.' N K O C u < -p Â•a Q) TI U O O 0) u cn m tn w 'O CM 0 .H Â•H iH u CM T3 C o 13 n n M w 0) CN > rH 0 iH CM (U w (0 o (U 0 > D < rH r+ n ,H n iH rH 00 CM CO CI4 rH Â•H w <; .H :3 c CM o o Â— Â— _'\jr Â— ft. Â— -.o-i Â— Oc.fV Â— r. 7 : 'V a (\j (\j ry CM ty l\j ry r\i ri rn rin n r*) nf) I n 0) CP Â•H

PAGE 32

24 O o O' iL Z < H X Z 7 K o m i-i < o Z z z < u.X Ji Â— in *^ Â• o C r. It r X U c in o II in c: n U p. < L> in Q K II a r, u X o cy r, U n -1 II z < u: > c o < u a tr. > u < < 4 u Ift o a o Â• o r. CJ o -J < u Ll > < ir Li O d Z c ^ -J Â— 2:'-.ooooooOooooC>QoococioooooooooO">ooooooo Â— '*rJ^J^^;^^l-v^^J(^jC^J^J^g^rvJ<\J(^J^Jf^^fyf^^(\,(^Jlyf^-^J^u^^J Â— fyr^^^jivjfy^ Z 5 a X ^r, Â— rjfo.jirnt/r^xO'O Â— rynifir, o^too'o Â— f^J^l*l^o^e^

PAGE 33

25 Q CO o < Â— uoooooooooooinriooooooooooooooooooooooooo aiÂ— <-OOOOOOOOOOr\jOOUooJOoOotJ*JOOOC)(.>C>rtOO'-'OQOO Lj>ry(V(\jfVKv,(\.fu!Mf\jryrj fVNfvjP. rg(\jry(\)(Njfyf\jfy(\i(\af\j(N(\jr\'rvi'>JfV'cyry'fVjry a < ti u'Oinooooo(noifii/ium~iDoL'^mo^-*ocjco'-j'"oinou'u". ooo 2 *Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•..Â•Â•Â•..Â•Â• Â•Â•Â•Â•Â•Â•Â• n pq w W CM 0 iH Â•H cH M 0) O O 0 O iH CM m !-) 0) VD > rH 0 + (1) n CP iH rH CO > M m rH 3 CTl n 1 n PAGE 34 26 N 0. u. c in ^ K (ft ft; o o > I z ? 2 hi n Â— X X o n ft U 4 Lj in Â• (/) (/> o oo r. o II y r tf ff c < 2 u f n 13 L y^ X i U) K C o o Â• O tt~ CÂ£ LL a Uj < < a. VI *n II K cc Id o Z o n o n II z < V > o < < 5 10 a: c in < ai < < X '* I/) Jc da o II \r. H a c -1 < tL a \L > < ;u u ^,l^r^rlnnnI'lr'^rnr>r]^nnnK.lrnr)rl'lr,l-)ln^rn'^rl^^nnnr)I^J < "Â£^'^lrÂ•t'X-^-uiuL"l^Juu^^^^a;u--i O O 0) u CQ S Â• CM m -H O o (N U iH (D iH > CU O W 0) u o ^1 rH m > < rH + n rH H H in ft >. w iH C o I n 0) Â•H PAGE 35 27 IT) >i m c o N s -P fd T) 0) u o u (U !-4 <^ i< W CM O A< a) M O O O O rH rsi 01 > -H O + n cn rH M OJ 10 > CM rt; ul UJ C KD ."3 a^oooo'^ooooo'juoo(jouc)ooooootjoo-'i)o(>oooooo{v a. ^ ^ ^. ry /V' ri^ l^J <\i fvt rw c\( t^' a.' ft> fv (V ry fvj ry fvj ft) ^ UJ ooou-oinooir.ooooinu.ooirooounjininoooL-o'-ointfoininni > fyf^Jf^J^J^'^J'^Ja^Ja;'yc^JfyPy^J^,t^l^JryrvJf^Jl^^fyf^J(\J^gf^vvJa,^^Gf^,^Oc\,^^ Â— -* Â— f\( n < M fO 01 (T\ rH rH in I 0) CP H Cm PAGE 36 28 g ^ ul J O < Â— acoonOoooooooooooooooOooooooor-noooooooo a< ---uj cjOinn^r-Nr. o>'f\j-(\jrjf^NO'iP*/ryr^ipfvinoo(r(^-ai(\j-o a NC^Xfva)^.(r^oaJ:*^.K^^a'bxcr'^<^KlPL^aoa3xa^^-^.^^K^cc'o^ u Â— Â— Â— .J^Jp!-.-J-J-JJJJj^J PAGE 37 29 PAGE 38 30 plotted separately in figure 3-5 and 3-6. A pulse appears in bin 25 in block P112WA6, with an intensity 2.6 a above the mean value; also in block P112WB6, the pulse is weaker, being only 2.3 o, but is at the same location (bin 25). Once again we have the fact that 2 pulses appear in the same location in two independent sets of data. From a precise knowledge of the pulsar period and the elapsed time between the April 28 and May 25 appearances of the pulse, it should be possible to preserve its phase so that it would reappear in the same bin on the second date as it occupied on the first. However, this could not be done because of the period uncertainty. This problem will be addressed later when analyzing PSR2045-16 data. The pulsar was detected in only one beam on each of the two nights, disappearing when the beam was switched to the next position. This phenomenon will be discussed later in the conclusions, when polarization effects are considered. Before other tests made on this pulse are discussed, it is interesting to note that in bins 9 and 21 of block P112WA6 (figure 3-5) two more small pulses can be distinguished. Although neither of them alone would be considered statistically significant, they both seem to appear again in block P112WB6 (figure 3-6) in the same position. The pulse in bin 21 leads the main pulse in bin 25 by 125 msec (resolution 0,31 msec/bin) or approximately 38 of phase. The pulse in bin 9 is separated by 4 96 msec from the main pulse, which is about 150 (or 0.42 P, P = 1.88023 sec). PAGE 39 31 The pulse in bin 21 suggests the existence of a subpulse leading the main pulse. Several authors have indeed reported for this pulsar a complex main pulse composed of 2 subpulses, the leading subpulse decreasing in intensity with respect to the trailing subpulse as the frequency decreases (6, 7, 8). The separation between the two subpulses has been measured at higher frequencies and extrapolation from that data suggests a much closer separation for low frequencies than the separation obtained at 26.3 MHz between the pulses in bins 21 and 25. In 1972 Bruck and Ustimenko (4) reported some observations made at 25 MHz with a linearly polarized antenna; they found that this pulsar shows a variety of pulse shapes with rapid transitions from one to another. In one plot, they show a pulse composed of 2 subpulses and separated by -^148 msec. Despite the poor resolution used (P/16 or msec) the 2 subpulses are clearly defined. The separation of the subpulses as a function of frequency has been plotted in figure 3-8, including the result by Bruck and Ustimenko and our result (i.e. for the bin 21 and 25 subpulses) The plot suggests that either the pulse separation increases very rapidly toward lower frequencies, or the subpulse detected at 26.3 and 25 MHz is a new one which appears only at low frequencies. In figure 3-9 the average pulse profile obtained in this thesis is compared with that obtained by Bruck and Ustimenko. The phase of the PAGE 40 32 Â• Â— o Â— (N 1 Â— 00 (NJ IS LA Â• Â• Â• Â• H Eh 03 ITJ fC CO O CO D u -P -P -P o H Q Q w <Â— < < <: w < > > 2: B O O >H Eh CO m H D > Q ISl N S pq U H O Â• < o o < o < ( r H Â• 1 1-1 LÂ„l LJ,.I 1 -J L 1 1 < 1 1 c c o in o o o in o o o o o in o o >H u w D O w o in o o H Eh Â— W CO O ro o (1> x: +J o M-l >1 u c 0) 3 cr M m o c o H o (0 CO (0 C o Â• Â•H >XI +J rH fO + Sh n fd n Q.rH QJ rH w 0) 03 iH 3 CO ex 00 I m 0) 3 H Pt4 PAGE 41 33 N K in O H CTi n W t< PQ U O D Eh a U CQ O O CVJ 1^ in -ro T" H ec: W Eh ^ S H W w m s u H cn >H Eh to Eh ^ S M W w m i-q S rti u H Â— CO + CO o (1( .H 4-1 -H o m o M 0) > -H M rC 3 3 ^ a 02 13 ^ >H O m -H O O ^ N o 0) C CM rH 0) D -P > 0) CO H 03 ax: -p 0) m o c U Jh m 0) x; >iEh X! 0 W c o m H u e -P rc o X! U O 4J I o -p u Q) TI o c H TJ OJ -p m x: 03 C OJ 0) Â•H > e x: QJ ^H 13 H PAGE 43 35 pulse profile at 26.3 MHz has been shifted, to align the pulses in bin 21 and 25 with the pulses at 25 MHz, Although the subpulse in bin 21 is much weaker than that in bin 25 for the 390-period average, the two are almost the same strength in the 140-period average shown in figure 3-10 (block P112EA6) In the latter case, the heights are 2.68 a and 2.62 a for the bin 25 and 21 pulses, respectively. A possible implication is that the bin 21 pulse varies in intensity much more than does that of bin 25, A small pulse appears in bin 9, when blocks P112WA6 and P112WB6 are averaged together. This pulse is also present in each block average separately (figure 3-5 and 3-6) suggesting the possible existence of an interpulse leading the main pulse by 0.42 P (or 150). In Bruck and Ustimenko (4) a broad and poorly defined pulse seems to appear at 0.5 P from the main pulse, in the same plot where the double main pulse appears (figure 3-9). So far, no interpulse has been reported for this pulsar. 3.1. a Peak Flux Density, Mean Flux Density and Energy for Pulsar PSR1133+16 Several continuous radio sources were observed in order to calibrate the array, to obtain the minimum detectable flux density, and then to obtain an estimate of the peak flux density for pulsar PSR1133+16. The radio sources observed and their main parameters are presented in Table 3-1. PAGE 44 36 Table 3-1. Radio sources used for calibration purposes Flux Antenna Density Beam R.A. Declination ( Jy) Used 3C310 15"02. 9"^ 2615' 6" 382 IW, 2W 3C315 15^11.7"^ 2616' 8" 133 IW 2W 3C409 20^12.2"^ 2325' 15" 381 IE 3C433 21^21. 4"^ 2454 6" 267 IW Positions and flux densities were taken from the Clark Lake Survey at 26.3 MHz (16). Of these four sources, only 3C409 and 3C433 were used in the calibrations; the 3C310 and 3C315 measurements were discarded because their difference in R.A. is only 8 s"^ (corresponding to 2.2). Since the HPBW of the radiotelescope is 6.0 east-to-west, these two sources cannot be resolved. The parameters of the radiotelescope for observing the calibration sources were: Bandwidth (Av) =2 50 kHz Time constant (t) = 1.5 s Frequency (v) = 26.3 MHz The minimum detectable flux density S deduced from calibrations using these sources, is presented in Table 3-2. PAGE 45 37 Table 3-2. S^^^ and Galactic background temperatures deduced from calibrations. Radio Source S mm (Jy) Galactic background temp. K 3C409 3C433 101 97 38,975 27,840 The galactic background temperature for pulsar PSR1133+16 is: '"backgr. = 18,514K All the galactic background temperatures given are actual brightness temperatures, corrected for losses in the antenna. A Hewlett Packard noise generator was used for the calibrations, and a pair of type 5722 noise diodes served as the standard of comparison for the noise generator. Details of the calibrations are given in Appendix E. The minimum detectable flux density has the following expression (17): AS oc XT o -1 V min ^^^^ T^yst (3-1) where n = number of averaged records Av = pre-detection bandwidth of the receiver T = post-detection time constant "^syst system temperature In this case, the temperature of the system is completely dominated by the galactic background temperature, and in PAGE 46 38 equation (3-1) it is possible to substitute T ^ by: syst ^ T = T = T syst galactic background Assigning subscripts p to the pulsar and r to a continuous radio source, it is possible by using equation (3-1), to write two similar expressions, one for the pulsar and the other for the radio source. By dividing one by the other, the following expression can be obtained: n Av T T AS = Â— ^ ^ ^ S K (3-2) min p n Av T T mm r \^ ^) P P P r The constant K takes into account that the pulsar signal, instead of being continuous, is being sampled. Using the values obtained for the pulsar on May 25, 1979 n^ = 390 periods (453 s) Av = 8 kHz P T =12.5 ms P T = 18,514K P and the values for the radio source of: n =1 r Av^ = 250 kHz = 1.5 s AS^ = 101 Jy = 38,975K the following value is obtained: ^min p (considering k = 1) To obtain the peak flux density of the pulsar, it is necessary to make several assumptions: PAGE 47 39 A. In order to be consistent it has been considered that the value S for both the continuous source and the mm pulsar is that which would cause a deflection of 3 a (i.e., 3 standard deviations) from the mean value of the galactic noise background. The 3 o criterion was used to determine whether or not a pulsar pulse was present at the end of an integration run. B. It is assumed that the sample rate is fast enough compared with the post-detection time constant that very little information is lost in this process; then K can be considered ^1. (Details about sampling rates can be found in Appendix B. ) C. It is assumed that the pulsar signal is unpolarized, as in the calibration source. This is probably not correct; a strong linearly polarized component has been reported (18) The effect of variable Faraday rotation in the terrestrial ionosphere would then be to cause the plane of the linearly polarized component to be oriented (at different times) at any angles from parallel to perpendicular with respect to the antenna dipoles. As a result, if the polarization is actually 100% linear, the value of S as given bv (3-2) min ^ J > / may range anywhere from too low by a factor of two to too high by a factor of infinity. Thus, no signal at all would be detected if the linear polarization is perpendicular to the dipoles, while according to (3-2) S would not be min p zero. PAGE 48 40 By making these assumptions, once the pulsar reaches a value slightly higher than 3 o at the end of an averaging process, the peak flux density can then be considered equal ^ ^min (calculated for the corresponding number of periods averaged). Therefore, %eak p = \in p = Pulse energy is usually defined as: E = P OJ S(t) dt (3_3) where P = pulsar period S(t) = flux density as a function of time For pulsars with no interpulse, the limits of the integral are restricted to the duration of the pulse (base time duration) In order to make the computations simpler, it is usually assumed that the pulse has a relatively simple geometrical shape. Izvekova et al. (7) assumed a triangular shape for most pulsars (including PSR1133+16) By looking at the pulse shape obtained on May 25, 1979, one can see that this is a reasonably good approximation, considering only the pulse centered at bin 25. Using this approximation, the pulse energy in equation (3-3) can be written as: ^P "T%eak p^base (3-4) where W^^^^^ = pulse width at base, corrected for dispersion and receiver time constant. PAGE 49 41 Substituting the values of: S = 150 Jy peak p ^ ^u^r.^ = 71 ms base the following value for E is obtained: P E = 5.3x10"^^ Jm'^Hz"-'It The mean flux density of the pulse is defined as: S = E /P f3-5) mean p \~> ->i Using the value obtained for E and the value for the P period P = 1.188 sec, the following value for the mean flux density is obtained: S = 4.4 5 Jy mean ^ It is interesting, now, to compare these values with values obtained by other authors. In 1973, Bruck and Ustimenko (4) observed the same pulsar at frequencies of 16.7, 20 and 25 MHz using a linearly polarized broadband array and reported peak flux densities in the range of 10 to 40 Jy. In 1968, Drake and Craft (19) observed the same pulsar, using the Arecibo 1000-foot radio telescope at frequencies of 111.5 and 195 Mhz and reported peak flux densities of 150 Jy at 111.5 MHz. Since these observations were made at a much higher frequency, they cannot be directly compared with the one at 26.3 MHz, nevertheless, the value obtained at the two frequencies gives an idea of the order of magnitude for the peak flux densities. A value of 38 Jy can be deduced from the data at 61 MHz given by Izvekova et al. (7) PAGE 50 42 Regarding the pulse energy and mean flux density, more recently, in 1979, Izvekova et al. (7), have given some results at 61 MHz. They obtained a value for the pulse "~ 2 6 ~ 2 ~ 1 energy of 0.89x10 Jm Hz and a mean flux density of ~ 2 6 0.75x10 Jy, using a linearly polarized antenna. In 1971, Slee and Hill (9), using a circularly polarized antenna, reported at 80 MHz a value for the pulse energy of ~ 2 6 ~ 2 ~ 1 2.2x10 Jm Hz for one night (with a median over 6 nights of l.Oxlo"^^ Jm"^Hz"-^) It is not easy to compare the values obtained in this thesis with values obtained by other authors, because of the polarization used and the different averaging times (not always given). Nevertheless, the value of pulse energy ~ 2 6 ""2 ~ 1 E = 5.3x10 Jm Hz obtained at 26.3 MHz, looks at least P ""2 6 ""2 1 comparable with the value of = 2.2xl0 Jm Hz obtained at 80 MHz (which also corresponds to a peak flux density of '\>110 Jy) According to Izvekova et al. (7) PSR1133+16 has a maximum in its average spectrum at v = 100+40 MHz. ^ max Kuz'minetal. (20), making simultaneous observations at several frequencies, found that the instantaneous spectrum changes from day to day. For several days it doesn't show a maximum in its spectrum but rather the intensity seems to increase towards lower frequencies. This can be considered as part of the explanation of why the peak flux density, the mean flux density and the pulse energy obtained at 2b. 3 MHz are higher than the values PAGE 51 43 10 ENERGY .-26 (Jm ^Hz xlO l.O .5 A .2 THIS THESIS Â• SIEBER (29) A SLEE AND HILL (9) O BRUCK AND USTIMENKO (4) o IZVEKOVA et al (8) 1 1 \ \ \ I'll I 10 20 30 40 50 100 200 FREQUENCY (MHz) Figure 3-11. Low frequency energy spectrum for pulsar PSR1133+16. PAGE 52 44 obtained at higher frequencies: the pulsar seems to show abnormally strong pulses at low frequencies during some particular nights, as it may have done in our case. Another interesting aspect of this has been pointed out by Ekers and Moffet (21) They observed PSR1133+16 at a wave length of 23 cm ("^1300 Mhz) From a histogram of energy and number of pulses with a given energy, they noticed a long tail in the distribution, towards the high energy end. The ratio of the mean value (0.020xl0~^^ Jm'^Hz""*") to the highest energy pulse ('^-0.090x10"^^ Jm'^Hz"-*") is about 4.5. Â— 9 Â— 1 If one takes the value for the mean energy of 5.3x10 Hz at 26.3 MHz as an average over a relative short period (^390 P) and compares it with the calculated mean energy of ~ 2 6 ~" 2 ~~ 1 1.6x10 Jm Hz obtained from the observations of Bruck and Ustimenko at 2 5 MHz (4) a value of 3.3 is obtained for the ratio. Such a ratio is not improbable according to the distribution function of Ekers and Moffet. Pulse energies obtained reported in this thesis and by several other authors are plotted as a function of frequency in Figure 3-11. 3.1.b PSR1133+16 Pulse Width Several factors can distort the actual pulse shape as observed; the factors that need to be considered are: A. Effect of the dispersion over the finite bandwidth of the receiver. B. Effect of the post detection time constant of the receiver PAGE 53 45 C. Polarization of the pulses and antenna. The first two factors will make the pulse wider than the actual pulse, but their effects can be corrected for if the pre-detection bandwidth, post-detection time constant of the telescope and the dispersion measure for the pulsar are known; a simple form needs also to be assumed for the pulse. In order to see the effect of dispersion let us consider the basic formula for the time of propagation through a plasma with a dispersion measure DM at a frequency v. The time of propagation is given by: t = kDM "2 (3-6) V where k is a constant depending on the plasma. Differentiating this equation with respect to v, taking finite At and Av (instead of dt and dv) and expressing the parameters in consistent units, the following expression can be obtained: At = 8. 3xl0"^DMAv (Vo ) ms (3-7) where DM = dispersion measure (parsec cm ) Av = bandwidth (kHz) = center frequency (hundreds of MHz) This relationship gives the time it takes for a dispersed pulse to sweep through a bandwidth Av centered at frequency v^. If a triangular shape for the pulse is assumed, (3-7) is transformed into (3-8) for the half intensity level: ^^0.5 4. 15xlO~-^DMAv (Vo)"^ ms (3-8) PAGE 54 46 The correction due to the time constant of the receiver (for the half intensity level and assuming a triangular pulse shape) is given by (8) : T = 0.693t (3-9) c where t is the post detection time constant of the receiver. No correction can be made due to the effect of polarization but some discussion about its effect will be given in the conclusions. Consequently, the corrected half intensity width of the pulse is: ^0.5 = ^0.5 Ob ^^0.5 ^c <^-10) where ^ is the observed half intensity width. Substituting in equations (3-8) and (3-9) the values DM = 4.8 pc cm""^ Av = 8 kHz Vo = 26. 3 MHz and T = 11. 25 ms, the following values are obtained for the corrections: At^ ^ = 8.8 ms T = 7.8 ms c Table 3-3 gives the values for the observed half intensity width and the half intensity corrected width for the pulses obtained on April 28 and May 25, 1979. The pulse width at the base is the width assuming a triangular shape, and is twice the corrected half intensity width. PAGE 55 47 rÂ— 1 00 1 ^ +J 1 0) 1 Q Â— 1 M K > E-i o >^ 1 ^ [m w w 1 H S o > CS] Eh u u H O <] o Â— o < Li 1 1 -i 1 \ L. o o in o o o lO o o ro in o o ro O CM CO (0 -p Â•H ? (U CO rH di >i N 4-* Â•H s CO 0) r \ H W m O w td u fa (U rH XT + 0) n m .H m 0 CO ft C 0 U -H ro +J to O M C n CP Â•H fa CO Eh Q D H PAGE 56 48 Table 3-3. Observed half intensity width, corrected half intensity width, and pulse width at base, for pulsar PSR1133+16 Date Wq ^ observed Wq ^ corrected base (ms) (ms) (ms) April 28, 1979 38 21 42 May 25, 1979 53 36 72 The graph of Figure 3-12 is a plot of the values of the corrected pulse half-intensity width at several frequencies (6, 8), including the values of this thesis obtained at 26.3 MHz on May 25, 1979; this is the more reliable of the two values presented in Table 3-3. 3.1.C Tests Performed with the Pulse Obtained on May 25, 1979 An independent and simple method for testing the pulses to see if they are indeed from a pulsar having the assumed period is to repeat the averaging process using both shorter and longer periods than that given by the DOPVEL program. If the observed pulses were spurious ones which happened to have approximately (but not exactly) the same period as that of the assumed pulsar, then the averaged intensity should either increase or decrease for the shorter period, and do the opposite for the longer period. If the averaged intensity is a maximum at the assumed pulsar period, then PAGE 57 49 it is almost a certainty that the pulses are indeed due to the pulsar. May 25, 1979, the date on which the best pulse was obtained, was chosen to make the test. Data were analyzed using periods longer and shorter than the assumed period by 40, 80, 160, +320, and 640 ys. Periods shorter or longer than the assumed period P by 80 ps should cause a shift of 1 bin at the end of the 390 period average. The result of this test is shown in Figure 3-13, where the relative averaged pulse intensities have been plotted as a function of the assumed period length, taking as the base the period P = 1.188023 s given by the DOPVEL program. It can be seen that the pulse nicely peaks with the assumed pulsar period, and decreases sharply for shorter and longer periods. The lack of symmetry of the plot might be due to the unsymetrical pulse shape. The computer plots of the results for the different periods used in this test are presented in Figures 3-14 to 3-24. The pulse intensity as a function of time has been plotted in figure 3-25 for the pulse obtained on May 25, 1979. This pulse was chosen because it was the better of the two and is well defined. This plot is useful not only to visualize the temporal behavior of the intensity of the pulse but also to check that no interference or isolated transient pulse is mistakenly taken as a pulsar pulse. PAGE 58 50 o o CO O U) CD ^ + o ^ O CO o o Â•H M 0) a Cn c H (0 > (N o QJ 00 o 00 o +J -H + 4-1 iH O II c o 0 Cn H o P OJ Q U Â• + O H d rK W ft Q. O w in H >i >i < -P fO H S w o c c o (U 0 CM -p C >X> Â•H iH + CO m 3 r-H CO Q) ft > H U -P (C tC w iH rH m rH I QJ U Â•H PL4 PAGE 59 51 (0 Â•H 0 (1) N >i r-t nJ s in (N >i n3 2 C 0 Â• TJ Q> 3. Â•X3 U O o U so PAGE 60 52 O o n. Q ^ < IV Â— o > y y z S o < z c I T o o Cl Q. ^(/ U". >o a Kl < > ll Cl ir. > u < < _) n Â• r\i 1 V) If 11 Ui o c < c u 'i. > < y L. u o J b; z t>~ ll n o o< a < ^ < xi -i ^'t. (\ fv rj rj {\j M n. f\i foi *y f\' r.ft. ft"M ^ (\j fti ry w <\( ru im rj oj c m rv ft' ft; ft""* (\i I IT PAGE 61 53 as Xi +J Â•H N >1 C cn in CN >i ni s o Â•a CO 0) M o 0 >X) U iH 1 ft +J to m 00 + 00 n Â• >-i II CO ft -d o CO Â•H iH u Q) I ro 0) Â•H Cm PAGE 62 54 cv o N P n II I < ^ n < *j tr, <( Â— o u. (J J y c o r. r-. o T I a o o r c c c (J K 1U' (/ t/* IT Co C II r y a < ^ r L IT. II C r. r r. r Â• c u < < r K V n. 4 r. f in I* r 7 < > > C < o <" 5 I (Cl u D < r < < J -1 c C r r 4. Â• A. \r (1 L' o _l < Q L' u. > (J < c: u U IT c Z 5 ,riu<\.cc ouc vr ~ c ttc u")'CicrO(jo>c (Mjo U'j-o ^^.^o ^0 f ff* o r^ r r / N Â— Â— Â• Â— r. f. r. fy r (** -P CO IX) r~ + CO ro Â• rH II tn 0 H rH PAGE 63 55 N >1 r\ O oc Q ft !Â• > 2 C If. in *^ k< C A. Â— Â— O u. Â— o 7" T 7 7 I'' UJ n r, -s s O T I C O r a r o 1>*tr Li^v 1/ I/) r cC r II N ? J r 5 r r Â• x r t( N r Or. r IT c fT tr t < < k I' *L 2 C r* ^ N. u'. IT I,: f p< s 5 c < !/ c > u < c < < <\ fj -J 0 p C f c Â• ft,' 1 f II u o < Â£ u L u. > u u er o T J o Z c ~ ~ r o c a < i S c o Â• (U M U O O U ^ dJ I (a p m CO (Ti + 00 II CO to O Â•H d 0) I r o Â— f\j f <Â• in Cn PAGE 64 56 A rt T X rt ^ K ^* Â— O 5 X a (D u I X o o Q. d o o 2 O II T I 4 i C 4 2 u rIT O u. Â• J) X i c o o n o >in u u c Uj < < tn ^ flC UJ o o fm o n H z < > > o j: < < 5 or tl UJ c o > u < < i> C a Q o Â• M ^ II UJ c -1 < X a Ui u. > 4 iX Ul r o X -1 D Oi Z o -^iioooooooooooooooooooooooooooooooooooooo oj Â— ryi or, irncj*;ai*.'opn f^ao IT. KoÂ— rviOo^xotJinNtMnfr-, Â— -ninir UJ pj (N. fM r. r. A, r r :y fvj f f\i rj (%j rj f\ (\( fv 'vj r. rj f\, *\i n. r\t (\j t\' > r\( r\. p, ojTj (\ (\. IV (V' IN ca r, t i\. i\, (\i f\' f\j I ^ r\. fv f,, -y <\j i.i n. rg (\, t> r c\. o. nj ry fui^Nc c7'^""r>i'^*L*o^(Oct o Â— ftjr^'ffC.^ST'j CO (N O 00 + 00 n Â• iH rH II o. (U 0 w Â•H Q) P.

PAGE 65

57 cc o o Q Q T y r sO u o II o u. y 1 > z o cr '3 n r, < 0 S I r oo It Or -0 (J tU J? -", in o n oc o II cr < ff C o r r t Â• in I X In ^o c n o f a u tt o N o t* C r. Tw 7 > > 5 O m < o d > < c < ^ % Q c Â•Â£ r\j fy (U -J < a a IL > < Li r c <Â• Â• Â•Â•Â•Â•Â•Â• Â• Â• a t a o I m u C7>

PAGE 66

58 O O 0 Q I > O t/. < ^ (/) \n U (V II Â— o u. > k> 7 O 111 nr) < -^ C X T a o o It a c MO oc C U 1~ V' UJ IT I/1/ o n Oo II a: 3 r <Â• (C 2 u fo r U. I r IT II e C O rr. u t fr tf L < < r II tt K o I tT O 0 1r i? n fUf. 1' c >> 7 o < < n > < c 4 _) -> t> 0 Q n o ft t' u c < tiu. c ui^oa-cc c^oc*cc'Occt*'*c C < IC C J (7f (Tp ^ o ft c a" a Â• Â• Â•Â•-Â•>--. > Mf\jf\Jf\.(\((\iP-r\j(\l(\'fy(\j(\j(\(f>{\jfyl\r ( Oo o rj ^ o o t> n r K. n o o o r o f rt Â• n O o o f ac a a J r. ^ r O O O r> Â— y !> r a n w' tf o 10 ru Â„ iTir i K ~0 "iiC c <3 a s iT ^ tOk' \r o r r > f\; fy ft' f\J (M ** L' iP N a t\, f\. r (\i f\j T' r. ^ Â— f\J K

PAGE 67

59 !? < o -S Â— f> > Â•* T M r". T I a; o o P CL C O y I to i/" (/) fr O (C t( (T } I X <^ 5 L' P rf lA r (T c f o r p. f L" L' u c: tf r < 1 f ^u ttUc 7 t c^ ^ N n if> O r r. II < >> s u < < c < i/> L > L < c < < p,i (\J c^ a o 'C { ii II L o C < ff u X' u. < or o -i r z II or< o o o Of > o oc (-1 o ro o r o o lij>fr'^f' r r>'"r;p" r. r rr I' n f Z J a. 2

PAGE 68

60 o N Â• < *>< o < .~ f > 7 y z r T li' p", fr O X I c. c ?^ (/", cv (A to I/* ^* Ofi r o 3 3 <* f 5 r r L Â• tr. I r o c o r If. a u a u. < < d K 1/ b1 a t ^ K o r l" ITc P (Vi 7 < > > o < 0* < Q > Q < < n C a r ft. {\. II L. c c u u. > c < CI c O T J c 2 Â•-aocooc.c.or coocoococ oooooooc rorr oorr^ooo^rir o Â„ c ^ in f <Â•-Â• ^ ff Â— Â— ? Â£ l' ^ < L' ^ ^ rj rj w r, ftj p. r(%j p (\j r c. (\j ru ^ fvj fy (V ftj f\' ty ft. p, c 2 Â— p,; n /-. Â— C. Js Â— Â— p rvf ^ (^r\ '\it\, f

PAGE 69

61 H o M ft. o o o Q Q. II T T I Z n 1< V *^ c Â— o L. V ( ) 3 3 7 O nr. < rr 1 1 L Â— ~ C c o L. n tt cc ftu C ^ IT ir, ^ r r a o 5 : c f -c y r r. o L y L'. II (M n oo f L n Lecu t r < r tr. 0 s o O n M 7 > >L' < < \J J 5 b. Of ^ <0 C < > t fy IN* 0 cc P H V II c < i a u u. > c tl y u ir. O } 2 1/^ L Olc a cf^opcnc-t'oc r ocoOcocr^rtoooOc-ooroOr ooecoOoo ** ^. if I' ir>f>^u'>tiru- (yr pyfv.'^Ajwr-. n,rj(\.rj(\jr'.t\j<\t(\(Pjft t\jf\JfVifyf*'(\ify<\jryf, f\;(v,fvj*vf\,jvjrgf,,fy -a N c (0 CTv IT) (N >i S Â§^ 0) 3 -a ^ o o ^ U X) 0) + u (0 Â•4-) ra >x) in iH 00 + CO n rH ro II PL. CL, u -o rtJ o CO -H ^3 (1) I n Â— Â— r. a f\, (\i r f\j i^j r r. n p; r ri m r, n QJ Â•H

PAGE 70

62 CO o o M w ^ c O nS 0) -a o Â•H 0) p c 0) CO (1) a. 0) c Â•H o 04 e Â•H 4-1 o c o H +J u c in rH u o 00 Â• 00 (Ti r-t W IT) CU >1 +J -H 05 C 0) -p c -H rH + cn n r-l rH > 01 no O Â•H U 0) a, H 4J r-l QJ > o H (U P > a, m in I n (1) H

PAGE 71

63 It can be regarded as a double check of the test previously described. Each point on the graph represents the intensity with respect to its corresponding average value of 10 consecutive periods, for bin 25 (peak of the pulse at the end of the averaging process) The intensity scale is in arbitrary units and the time scale is in periods P given by the DOPVEL program (P = 1.188023s). The plot shows that the intensity in bin 25 is almost always positive with respect to its corresponding 10-periods average value, and no clear periodic variations appear over the 390 periods plotted. 3.2 PSR2045-16 Data Analysis Of the 10 nights of data reduced for pulsar PSR2045-16, pulses were found on 4 nights: May 15, 16, 18 and 19, 1979. A summary of the dates, averaged pulse deflection in terms of standard deviations o, and number of periods averaged is presented in Table 3-4. Table 3-4. Summary of the date, pulse deflection and number of periods averaged each night for pulsar PSR2045-16 Pulse deflection Date in terms of a Periods averaged May 15, 1979 3.16 540 May 16, 1979 3.48 440 May 18, 1979 3.28 610 May 19, 1979 3.30 415

PAGE 72

64 Plots of the data for these dates are presented in Figures 3-26 to 3-30. Figure 3-31 corresponds to a plot of a typical day when only noise was present and no pulse appeared. In general the pulses were of different shapes, except for the last two days, for which the shapes were similar. Some have steep leading and trailing edges (as on May 16, Figure 3-28); others have only a steep trailing edge and a not so well defined leading edge (as on May 15, Figure 3-26). On May 18 and 19 (Figures 3-29 and 3-30) both pulses show a steep leading edge and much less steep trailing edge. On May 16 the pulse was clearly present with an intensity of 3.4 8 a above the mean value over an average of only 180 periods (Figure 3-27). On the other 3 days, pulses were not seen clearly until the end of the averaging process for that night. For partial averages in these cases, some indication of a pulse can be seen at the location where the pulse appears after completing the average, but its intensity never rises significantly above the noise level. No indication of the existence of a complex pulse or interpulse can be deduced from the plotted data. Pulsar PSR2045-16 presents at higher frequencies a complex pulse, composed of 3 subpulses, with the trailing subpulse decreasing dramatically in intensity with decreasing frequency (22); at 610 MHz the trailing pulse

PAGE 73

N S in in >i (13 s c 0 T3 Q) Â• 'O w U 0 0 o Â•H !-( dJ VD rH o 1 in in o U > to 0 0) u Cr> fO M 3 > a, I n Â•H

PAGE 74

66 ft.' cy fti ft.' N (\J o o o o o o Q. i c. a 1 Qz I T y 3 3 I c Â— <<<<<< z "1 (/) (/I KtH H h< (/) U) to U) in c ^ L'^ o >n tr. ''"i i/i fvi u z z X Z X X 7 7 >f O t\t iT CD < ti > o o o z o r r T X r r H u. c> c o o o X o o o o Â— Â— a c c a. 3. a X *- ^ c C J '-^ 3 N u 1K rH y1 IT t/1 1/* Vi l/> o 1*) i/icn i/i i/i wi I o n c^o o o n o o o o Â— Â— p. n u t_ o a. a '-^ u 0. < r < < < < o11 tn i/> i/^ Li u". ^ 1tt1^ 7a tTO o h1 N f., N. S fs. in O in IT IT, in Ul II z o < C4 o C >>>>>.> < 1/" <<<<<< LI Q. 1 I i 5 S I y > cr r '-> c u; u. Li .1 ;n ^ c Q < u UJ Ui O Â— ioooooooonoooooooooooooooooooooh-oooooooooooooooooo Â— OooNoOfwOOOO^^'>i,TU'inu" t-^iTir 4 ^ <3 -I uT''inif. if,inininjitnuiJinJiinir. inin* fs. (M iM '\j f\. cj oj ry cj i>i

PAGE 75

N K S in +> >1 S c o Ti 0) Â• T3 W 0 O U -r-i (U M o, rH O 1 CO in iH O M (N 0) > ft 01 w u <-{ 0) :3 > I

PAGE 76

68 o o o o tL 0. X K ^ < (J' II O J X O UJ at 1 1 2. ri U lU (/) ry l/) n n I/) H r y < u 1 O tu tn r I o o r) s K V Z O IX < < < in z z a a > a < s < < iT tn u> oo (\* a C. 0. eo Â• ti ^ H J u o o _) < D jc Ui li. > < o Â£ QO *-frr>oooooot')noooorjoooooo<^oor)OooooO'^oOooc^ooCOooo.->oooOoor>or)orvOovoooN u > Â— Â— Â— Â— Â— Â— --. Â— Â— Â— Â— Â— Â— -* Â— Â— Â— Â— Â— Â— Â— Â— -._ Â— Â— -. Â— Â— Â„ Â— Â— Â— Â—Â— Â— Â— a < O X Â— T N Â£! O iTl -* C Â— .j> o a, u *Â• T >r o X n o r r: J n : f^Jry^J^(^l^Jl\^Jr\Jt^J^^JfVJ^JfWf^JCyr,,^^J^J^J^J'J^y(^J^J(\;ry^J(\J^lf^.rJ'^Jf^l^ -jp. fMfj'Vj.-Si' if\ir\(Oi(\jr\j<\rjpgrinr>nm(irinnr.^

PAGE 77

N s in (Ti WD iH >1 fO S c o -a Â•o O 0 O 0 u u 0) iH o 1 in o CN > 0 0) u cn n3 rO U) rH Q) > a, : 00 CM I m (1) M Â•H

PAGE 78

70 o (\j cNj rj w o o o o o o Â—Â• Q C Q D. 0. ll z N < -r < < < c 1/1 '.1 */) i/> U". Â— Z H K K tt/l ifl yj 10 (/I u J' 't "J Â— u I n 1 1 z z N Â— ITJ O o o o o n Â• r. (K H r11 :r Lir c < < < < < t-h1i/l i/i i/itn ttI1o 0 O ff' (T J' N K o 1 o o >->>> I o a. < < < < < o < .1 y 5 I y I a > c; < u 3U; < 4 Li lOin *n if. UJ a z f 2 o< Â— U00000oCi00O0OOC>OOoC>O0NO0<^OO00o0S000K a < iz Â•Â•Â•Â•**c. Â•*Â•Â•.Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•Â•, Â•Â•Â•Â•Â•Â•Â•Â•Â•Â• ro n c ^~ i IT' ^ '"

PAGE 79

N in +J to CTi n3 C O Q) Â• u xs o o U -H Â•-i O I O U CM (1) > cn o 0) ^ CT> W !h 3 > A4 < CN I CO 0) -H Cm

PAGE 80

72 CO n o o o o o o o c G c. a It < < < < < z in m i/n/^ i/i o r X Ul i/> l/l l/l (/) < 0* o o o Ul lO o o u. > z a 3 z > ** V ^ CD > z z ri f 0> o o o o o o 03 o D ttK K (\d C < < < < U d (/) ^/i '/) 0'. yi t^^' tII in u 7> CT* O o I Â•0 o ) 3) IS (T iQ Â£ tn 15 It Z o >>->->> 4 (VJ < < < < < a I S I 5 S O < (/) 0: 0 < < < < n 1*) n rn > a < tr u c u' < (fuT^ in iT in (/I 4 IT n II ^ ft U UJ G -1 < -U > 'J Q :^ UJ OOr>nonOO-*OoOO'^OorOoOOOr)ri^>f>OO^OOOOOoOOOOoOOOOO O oo OOoOOOOofOciT. a < < f, r, Â— cj (M r -n ri r-, ri fy f^ Â— ^_ r Â— r :j r r., ry r'. rj n r.j r~ r-, r o, r r> 'Â•i -vi ri Â— ^ f, f, r. (\ r r > 'Â•I -n *^ ") ^ i*; n n *) 'Â•i n n r. n M n 1^ f*! n n n ^1 ^1 ^ < Â— Â— Â— Â— Â— Â— Â— Â— Â— Â— Â— Â— Â— Â— Â— Â— ^ Â— Â— Â— Â— Â— Â— ^Â— Â— Â— r

PAGE 81

N -p (T3 ro S c o Ti 0) Â• CO 0 o U -H 0) a. rH in 1 -H in o M > W O CM Q) u tn in u r^ 0) 3 > o 00 I n (U Â•H fa

PAGE 82

74 u o fy M ftj c\j o o o O o a c: a a II 3 Z Z X z < < < < o I I < u z 2 o o n UJ in in < z o X r I X c?h y o o o o u a a 1 a o u :n Â• J^J^ifi if) o X O ON O '0 r rn n u r 5 Z 3 c >> jz o a < < < < o < n Z Z 3 Z a a r -J c L) < OO O o fy fS: fv< ftj a a a c Q > n o f^, II J) (J o -J < c u. > < a 3 c 2 z o (/I u e o Â— Â— Â— Â— Â— Â— Â— ruff\;(\*(\,rgt\.r\f\i(\j^.<^rjp t0.rKtt>o Â— Pn^iT^Nff-o-r Â— p. f^^j^f^ Of. xc*oÂ— py

PAGE 83

s in -P (0 tfi O ^^ CP (13 W M iH OJ 3 > o U n I ro (U M CP Â•H

PAGE 84

76 o c co o o c r Q Q r y T I M 4 f N (/^> V : tn K K T < U" o c c o u O r O II u. > any u If ff r. If. 2 2 L X x r I G croc II O C.GO r n c t t^-ff* ('J < ^ < t ^hc r in IT I/-. 1c II C u ff O T o 3 > ^ N f.. n 1 N N in IS <* o >> > > < r. <.-<< 3 o 3 > J r C <,'. t: L. < < It < C C L. Â•r ir> in in J ^ < a Z g o (/) Lj r.o c < Â— a oor> u 'J o cr UJ > wc^>'o-o;''l^cc^oa(^:^(^cÂ•tl^uc^oc^i^^t7rÂ•oc^u'''.'o(^OJO'^_Â•cIC^r<^U'CoÂ•(>oooc^-'co(>c.o<>o u^c
PAGE 85

77 is the most intense, but the intensity decreases rapidly, and at 151 MHz is about 2/3 of the central subpulse. The intensity of the central and leading subpulses seems to remain almost constant with frequency. No indication of resolved subpulses can be obtained from the data at 45 MHz. Nevertheless, the difference in steepness of the leading and trailing edges might suggest the existence of an unresolved much less intense subpulse accompanying the main subpulse. This aspect will be discussed later when disI cussing the pulse width and the effect of polarization. 3. 2. a Peak Flux Density, Mean Flux Density and Energy for Pulsar PSR2045-16 In order to estimate the peak flux density of this pulsar, several radio sources were observed for calibration purposes. Radio source PKS2034-17, one of the closest to the pulsar, was clearly recorded despite the steepness of the galactic background (23). This radio source has apparently not been observed before at any frequency lower than 80 MHz. However, an extrapolated flux density at 45 MHz was obtained from the spectral index derived from flux density data at higher frequencies, as reported in the literature Spectral index can be defined (17) by the relation: (3-11)

PAGE 86

78 where S = flux density V = frequency n = spectral index For any two frequencies on the linear part of the relationship between S and V n ^"i^l (3-12) The best available values of the flux density for this source (24, 25) are at frequencies of 160 and 80 MHz: ^160 = ^80 = 15 Jy Substituting these values into equation (3-12) a value of n = 0,76 is obtained. Assuming a linear relationship for S as a function of ^ down to 45 MHz, the flux density at this frequency is found to be: S 5 = 23 Jy Applying the previously used criteria, the following value for the minimum detectable flux density at 45 MHz was obtained for the region near source PKS2045-17: S = 6 Jy min The radiotelescope parameters for observing the radio source PKS2045-17 were: Av = 3 MHz T = 3 S

PAGE 87

79 where Av = bandwidth, t = post detection time constant and n^ = niamber of records averaged. It can be shown that the brightness temperature T and the frequency v are related in a similar way as S and V are related in equation (3-12) : vn (3-13, Using the galactic background temperature at 38 MHz given by Blythe (26) : T = 12,000K IT = 7,000K where = galactic background temperature near the pulsar \ = galactic background temperature near the radio source PKS2034-17 and the value for the galactic background spectral index given by Findlay (27) valid for the range 20 to 400 MHz: n = 2.65 the galactic background temperatures T and T at P r 45 MHz were found to be: Tp = 7,630K = 4,452K Using equation (3-2) and the appropriate parameters of the radio telescope, the minimum detectable flux density can be obtained for the pulsar observations.

PAGE 88

80 The radio telescope parameters when observing PKS2034-17 are: Av =3 MHz r ^r = ^ ^ ^ = 1 T = 4,452K r S = 8 Jy min ^ The parameters when observing the pulsar are: Av = 10 kHz P T = 11.25 ms n^ = variable, depends on the day Tp = 7,630K The peak flux density (which is also the minimum detectable flux density) for the different days is shown in Table 3-5; this table also contains the values of the mean pulse energy and mean flux density computed using equations (3-4), (3-5) and the values for the pulse width given in Table 3-6. No known measurements of flux density and energy at low radiofrequencies exist for pulsar PSR2045-16. The lowest frequency data on this pulsar are the 80 MHz observations made by Slee and Hill (9). They attempted to observe the pulsar on 3 nights, failing to detect it and giving an upper limit for the pulse energy of Â— 2 6 Â— 2 Â— 1 <0. 17x10 Jm Hz The failure to detect this pulsar is not surprising since it has been reported to have large

PAGE 89

81 0 >1 m CO N Â•H S Cn U in irH S-l 4-1 (C3 B m 3 (U c/) S in I n o XJ ra Eh r-l >i in m in Cm -P cn 00 -H C CO rH f-H CN (0 c OJ Q iH 1 N (U CO CN iH 1 e 00 in Hi >-i m in 0) Â•H CO c Q X rH X (0 ft u 0) c w -a 0) u > CO T3 O Â•W S-l (1) LT) in 00 m CM m rH iH rH rH rH o o o in rH rH in CTi o^ rH rH rH rH -p Â• fC3 in VO 00 Q rH rH rH rH >1 >1 >1 >i (0 s tt) p Â•r) (U c C 0)

PAGE 90

82 AVERAGE PULSE ENERGY (Jm'^Hz""'-) xlO -26 .5 .4 .3 Z 20 I J L THIS THESIS O BACKER (30) A SIEBER (29) # SLEE AND HILL (9) JI 30 40 50 1 00 200 FREQUENCY (MHz) 300 400 Figure 3-32. Pulsar PSR2045-16 low frequency energy spectrum.

PAGE 91

83 intensity variations, being often undetectable (28, 29). Our experiences at 45 MHz confirm this behavior also since it was possible to detect this pulsar only in the first nights of the observing period; it faded out for the rest of the nights. -2 1 An unconfirmed pulse energy of '^'1.9 Jm Hz has been given by Backer at 34 MHz (30). The lowest frequency at which published data on pulse energy is available is 410 MHz; the energy is 0.120x10"'^^ Jm~^Hz~"'" (29, 31). Our measurement at 45 MHz seems rather high (perhaps suggesting an undetected systematic error) although no direct comparison is possible due to the lack of published information at low frequencies on this pulsar. Pulse energy as a function of the frequency has been plotted in Figure 3-32, including the data mentioned in the last paragraphs. 3.2.b Pulse Width for Pulsar PSR2045-16 In order to obtain the true half intensity pulse width of PSR2045-16 at 45 MHz, the measured half intensity pulse width has been corrected for pulse dispersion and post detection time constant using equations (3-8) and (3-9) respectively. In order to correct for pulse dispersion, the following values for the parameters were used: Av = 10 kHz Vo = 45 MHz DM = 11.5 pc cm"^

PAGE 92

84 The correction for pulse dispersion Atg ^, using the above values, is: At_ = 10.5 ms U b The correction due to the post detection time constant of T = 11 25 ms is: c T = 7. 8 ms c The true (or corrected) half intensity pulse width Wq ^ is as before (equation 3-10) : Wn c; = W_ c At. T 0.5 0.5 obs 0.5 c The result for the 4 days is shown in Table 3-6. Table 3-6. Observed and corrected half intensity pulse width and pulse width at base (assuming triangular shape). Date W 0.5 obs (ms) W 0.5 (ms) Width at base (ms) May 15, 1979 44 31 62 May 16, 1979 38 25 50 May 18, 1979 44 31 62 May 19, 1979 44 31 62 The lowest frequencies at which pulse shape and width measurements have been given for this pulsar are 150 and 151 MHz (10, 22). Lyne et al. (22) have given the separation between subpulses at a number of frequencies; the values are plotted in Figure 3-33. The separation between

PAGE 93

85 P CN to <; Â— -P o QJ CM H Q U ?H iH W < +J Â• 2 Oh 0) W W Eh W W W 2 W CO U 2 Bh 1-1 EH U >^ KM 1-M CQ a, CQ Q m 2 H 2 <; Â• to O CO O CO H H H o w Eh 2 CO EH < m Eh <: H di H 2 111 < eq w W w Pi P CO U CO Eh CO Â• 0 o o o o o I I I I o o J L o o in o o o o ro o o OJ o o o in o o O CM O in 2 o H Eh <: w CO i-^l a. m P W E Oi CO Â— o O ro o Tl ni u> c; 0 (0 4j C _Q 0 0 -H u cn C (0 14-1 (11 (d ^ 0} (0 -P (0 n Â£ Â•H -M Â•P fl3 -H N Pi q; W 1 Â— I >H fl) CL CJ W) 2 iH (U t/3 W 3 Â£ -H J3 Cli -M CO w 3 cri,c t/1 C 4-) Â•H U3 CO rH 3 -H 1 r-l Xi ld U -P c O -H C CM -H CO N U E Li (11 (Tl M 1 A CO tji r-H 0 3 M -P Oi 4-1 (ti n 1 (^ (U M Cn H fa

PAGE 94

86 the leading and central subpulses and the trailing and central subpulses is plotted separately. As in a previously mentioned case, the trailing pulse seems to decrease very rapidly in intensity toward lower frequencies (22) and it is likely that at 45 MHz it no longer exists or at least it can be very weak. The mean value at 45 MHz for the pulse width at its base (obtained by measuring the width at half intensity and assuming a triangular shape) of 61 ms has also been plotted in Figure 3-33. The plotted value for the pulse width at its base, lies close to the extrapolation of a straight line passing through the points representing the separation between the leading and central subpulses. A possible conclusion of this is that the pulse observed at 45 MHz is composed of two unresolved pulses which might correspond to the leading and central subpulses at higher frequencies; the pulse width at its base would correspond to the separation between the two subpulses. No polarization measurements have been made for this pulsar at low frequency. The lowest frequency at which polarization measurements are available is 409 MHz (32, 33). The mean percentage of linear polarization at 409 MHz is 40%, and it changes during the course of the pulse, reaching about 70% for the first subpulse. The stability of the linear polarization position angle is low for the wings of the integrated pulse profile, due to the occasional

PAGE 95

87 occurrence of pulses with orthogonal polarization with respect to the mean, but is high at the center of the pulse (32) If the degree of linear polarization of the pulse is higher at lower frequencies, there will probably be a distortion in the pulse shape when it is received with a linearly polarized antenna. This distortion will depend on the variation of the polarization angle within the pulse. Assuming that the linearly polarized component in the central part of the pulse is parallel (or nearly parallel) to the antenna, then the integrated pulse profile will appear strong in the center; however the slopes of the leading and trailing edges will be weaker and the averaged pulse profile will be narrow. More details about the effect of the degree of linearly polarized component on the pulse profile and apparent intensity variations are given in the conclusions. 3.2.C Period of Pulsar PSR2045-16 from Observations on Two Consecutive Days The apparent mean pulse period of a pulsar can be calculated from the elapsed time between two pulses; in order to increase the accuracy it is desirable that the elapsed time between the two pulses be much longer than the pulsar period. An attempt was made to obtain the apparent mean pulse period from observations made 24^ apart (elapsed time %24 ) for pulsar PSR2045-16. On two occasions the pulsar was detected on two consecutive days;

PAGE 96

88 the first occasion was May 15 and 16, 1979, and the second May 18 and 19, 1979. The pulse period can be obtained by subtracting the epoch of apparition of the pulse E on two consecutive days and dividing by an integer number N. The epoch of the pulse can be calculated from the relationship: where ; E = T ^ 4. + nxR (3-14) start T = time start digitizing s car t n = bin number where the pulse occurs R = plotting time resolution The information related to the two pairs of days has been summarized in Table 3-7. The mean apparent period can be calculated from the relationship: P = f (3-15) where AE: difference in epoch of pulse apparition on two consecutive days (elapsed time between pulse apparitions) The mean apparent periods obtained with this method are compared with the DOPVEL periods in Table 3-8.

PAGE 97

89 m o n n u (d CO U 0 =1 u o 0 ON n tl t iH a> (0 Â• c W 0 <] 00 Â•H iH + VO Â•H 00 00 nj & Id 0) ID 00 r-H to rH o 00 H n 0^ =1 in in CO VD 00 (N rH m Â• m cn Â• m o o o 0 w CN m o n 6 6 S E o 00 0 ro Ch 0) a o o o o Â•H 0) 0 u vo ,Â— I ,Â—1 rH rH m n CO ro n m o o o o Â•H rri o o o o Â•XS rH r~* 1 M 1 nJ in '!J< Xi o O c 00 00 0 Pi (N Cum Â• 1 n (Tl CTl (D en CTl CTl (Tl iH rH rH 1 >1 >1 >1 ro (C S s

PAGE 98

90 Table 3-8. Comparison of the mean apparent pulsar periods for PSR2045-16 obtained from observations with the DOPVEL periods. Mean apparent period DOPVEL period AP Date (s) (s) (ps) 1.96137547 1.961383 -8 May 15, 1979 May 16, 1979 May 18, 1979 1.96135847 1.961386 -28 May 19, 1979 or 1.96140311 +14 It can be seen that the periods do not agree. Unfortunately, a jump in the epoch was noticed in the quartz time standard of Maipu, after the first pair of days; that makes the figures for the last pair of days (May 18 and May 19, 1979) very unreliable. In any case, this discrepancy, at least for May 15 and May 16, demonstrates that the DOPVEL period cannot be used to bring in phase two consecutive days of observations. In order to bring in phase pulses detected on two consecutive days (within 0.5 bin), the apparent mean period for each day needs to be known with an accuracy of at least \ = (f ) /N (3-16) where N = Integer number of periods between two apparitions. Expressed in terms of the elapsed time T and the period P, (3-16) becomes:

PAGE 99

91 T where N = Â— h Assuming an elapsed time of 24 (86400 s) a resolution of 0.0315 s and a period of 1.961386 s, the following accuracy for the period is obtained: A 0.36 ys c This accuracy should bring the two pulses in phase within +0.5 bin. It is not clear what caused the discrepancies between the DOPVEL period and the period deduced from the observations; it could be an instrumental problem or a precision problem in the calculations in the DOPVEL program. Both possibilities were checked but the result was inconclusive; further investigations need to be made. 3.2.d Pulsar PSR2045-16 Pulse Intensity as Function of Time Figure 3-34 is a plot of the relative pulse intensity as a function of time for the pulse obtained on May 16, 1979; this pulse is the strongest and best defined. Each point plotted represents the relative intensity with respect to the noise average, averaged over 10 periods; the time axis scale is in pulsar periods. It can be seen that the pulse has large intensity variations. Most of the time it has a positive value with respect to the noise average, and it seems to show an almost periodic intensity modulation of '^^140 periods (peaks at periods 20 160 and 320).

PAGE 100

92 w H < > H H C/2 H H 2 Pi u J Â— I Â— \ I i (D -46 -4d CM d J Â— I Â— I I L 8 o o ro o o CVJ O o t fO M CO (U 1 P U vj lU CD to ftj (J C Â•ri to I M d) CO CD to E (D -I-' -i ^ Â•H (C II s tfl C -H 0 QJ 1 0 > in -H H U -P O QJ fO CM ex OJ w o K (li -H d d I I 00 n (U d

PAGE 101

CHAPTER 4 CONCLUSIONS Two pulsars have been successfully detected using narrow bandwidth and long averaging times. Pulsar PSR1133+16 was detected with the University of Florida large array at 26.3 MHz and pulsar PSR2045-16 was detected using the Maipu Radio Observatory large array at 45 MHz. Pulsar PSR1133+16 was strong enough to be considered statistically significant (pulse deflection larger than 3 standard deviations) in the data of 2 nights (out of 10 nights reduced) Pulses of PSR2045-16 were found on 4 nights (out of 10 reduced) with enough intensity to reach the 3 standard deviation level. Typical averaging times for both pulsars were in the range 400 to 600 periods. Both pulsars show large intensity variations on a time scale of days. Typically, one would appear after an averaging time as short as '^^200 periods, and then would fade, being undetectable for averaging times as long as '\/1500 periods over succeeding days. The peak pulse flux density obtained for PSR1133+16 is about 150 Jy; for PSR2045-16 the peak flux density is in the range of 120 to 140 Jy. These values (which have relatively large probable error, in the order of 50%), are higher than the values of the peak flux density measured by others at 93

PAGE 102

94 nearby frequencies. They are also larger than the extrapolated flux densities from published data at higher frequencies Also, the pulse shape and half intensity width seem to differ from extrapolated values based on data at higher frequencies. In order to find a satisfactory explanation of the intensity variations of the two pulsars, it is necessary to consider both previously reported intensity variations and the data available on polarization at higher frequencies. Pulsar PSR1133+16 and several other pulsars have been reported to show extremely variable flux densities both from pulse to pulse and on a longer time scale, when observed at 81.5 MHz (34). These intensity variations are not believed to be caused by interplanetary scintillation, and therefore must be attributed to the source. Pulsar PSR2045-16 has been reported as often being undetectable for several days at 40 8 MHz, but when it is active it can produce pulses stronger than any other of the first pulsars discovered (28) Besides these intensity variations, the intensity of the received signal will show larger and more frequent variations when received by a linearly polarized antenna, if the pulses are strongly linearly polarized. Manchester et al. have found that most pulsars present a relatively high degree of linear polarization (32). PSR1133+16 was found to have a degree of linear polarization of 33% at

PAGE 103

95 403 MHz, 50% at 86 MHz and 60% at 63 MHz, which is the lowest frequency at which polarization has been measured for this pulsar (35, 36). There is a clear dependence of percentage of linear polarization upon wavelength, and the tendency is for the percentage of linear polarization to increase with increasing wavelength. Pulsar PSR1133+16 has a complex pulse, composed of two subpulses; the percentage of linear polarization is variable throughout the pulse and reaches about 70% in the middle of it at 403 MHz. For individual pulses, the position angle of the plane of linear polarization is highly stable for all but the first component, which usually shows two preferred positions perpendicular to each other. The result of this is that the degree of linear polarization of the integrated pulse is low for the first subpulse and high for the rest (32) Pulsar PSR2045-16 has a complex pulse composed of 3 subpulses. Manchester et al. (32) measured the degree of linear polarization to be 40% at 403 MHz, with the polarization being even larger for the first subpulse ('^-70%) The position angle of the plane of linear polarization varies smoothly throughout the pulse; a change of about 30 occurs between the first and third subpulses. This pattern of position angle variation within a pulse is stable for many consecutive pulses. No known polarization measurements are available at lower frequencies for this pulsar. Even if the linear polarization position angle were stable when measured at a point above the terrestrial

PAGE 104

96 ionosphere, the ionosphere can introduce relatively large variations in the angle as measured at ground level. This is due to variations in the amount of Faraday rotation in the ionosphere. This effect is particularly strong at longer wavelengths because the angle of rotation 6 is proportional to the square of the wavelength A: e RM-A^ (4-1) where RM = rotation measure of the ionospheric path. The RM depends linearly on the electron density, N, the component of the earth' s magnetic field intensity which is parallel to the direction of propagation, B,, and the thickness of the ionosphere, L: RM cc \ B||NdL (4-2) o-^ It can be shown that the percent variation of 9 is equal to the percent variation of the electron density N, if the thickness L and the magnetic field BÂ„ of the ionosphere remain constant; that is: AO Â„ AN "F ^ = IT ^ (4-3) It can also be shown that ^ equals when N and BÂ„ are constant, and equals when N and L are constant. Manchester has obtained values of ionospheric RM ranging from 0.1 to 6 rad m"^, depending on the time of day and the declination of the pulsars (37). Taking these two extreme values, it is possible to obtain the change in rotation angle, AG, for let us say, a 1% change in AN at

PAGE 105

97 both 26.3 and 45 MHz. The result of this calculation is shown in Table 4-1. Table 4-1. Variation of the position angle AO due to variations of 1% in AN, ABÂ„or AL in the ionosphere. .Q AN AB|| .Â„ AN AB|, AO for Â— Â— A9 for Â— Â— AL II AL I" Frequency ^ T = ^ 17 ^^^'^ (MHz) RM = 0.1 rad m RM = 6 rad m ^ 26.3 7.4 447 45 2.5 150 From the values obtained in Table 4-1, one should expect large random (or nearly random) variations of the position angle for large values of RM at the two frequencies considered. This effect could produce a modulation which distorts the pulse shape, in addition to the long term intensity variations. For low and medium values of RM, the variations of the rotation angle A8 are not so important, since they are small. There is another effect that can be important for large values of RM; that is the change in position angle when observing the pulsar in different regions of the sky (i.e. when observing the pulsar with different beams) or when observing the pulsar on different nights. For example, at 45 MHz, with a RM = 1 rad m"^, a change of 4% of N, L or BÂ„ (or a combination of them) will cause the position angle to be orthogonal to its previous position. If the pulsar

PAGE 106

98 radiation is relatively highly linearly polarized, this should produce a significant drop in the intensity of the received signal when using a linearly polarized antenna. This last effect can provide a possible explanation of the disappearance of the pulsar signal when switching to a different beam of the 26.3 MHz array. Regarding the effect of the polarization throughout the pulse, the pulse shape can be modified when using a linearly polarized antenna; the signal will appear more intense for the part of the pulse with polarization that matches the polarization of the antenna, and abnormally weak for the part of the pulse with position angle unfavorable for the antenna. Also if a given part of the pulse presents random changes in the position angle from normal to orthogonal (which led to a low integrated polarization for that part of the pulse) the intensity of the averaged pulse will be abnormally weak in that part. This seems to be the case with PSR1133+16, in which the observed intensity at 26.3 MHz of the second subpulse is usually stronger than the intensity of the first subpulse; the latter is often undetectable or appears only occasionally. The separation of these two subpulses at 26.3 MHz is much larger {^-125 ms) than the predicted separation for the two subpulses (which compose the main pulse) obtained from data at higher frequencies; however the separation at 26.3 MHz does agree with the results obtained at 25 MHz (4)

PAGE 107

99 The values for the peak flux density, the mean flux density and the mean energy are higher than the values obtained from extrapolation at higher frequencies, with the exception of a value of peak flux density of 150 Jy obtained at the relatively high frequency of 111,5 MHz for PSR1133+16. Izvekova et al. (7) and Kuz'minetal. (20) made simultaneous observations of this pulsar at several frequencies in order to obtain the spectrum. They found that the average spectrum (average over several nights) presents a turn-over at about 60 MHz. Nevertheless they also found that the spectrum shows large variations from day to day in the range 61 151 MHz, often with a change in the sign of the spectral index. It is interesting to note that the spectrum of some individual nights does not show a turn-over but rather an increasing intensity with decreasing frequencies. This might explain (at least in part) the unusually large intensity of the pulses at 26.3 MHz, when compared with average values obtained by other authors at higher frequencies. Averaging of pulses for two different nights was not possible. For pulsar PSR2045-16, the period obtained from the DOPVEL program differs from the period deduced from observations on two consecutive nights by amounts between 8 to 28 ys; this discrepancy is large compared with the required precision for averaging over two consecutive days (i.e., ^ tenths of jjs) No attempt was made for PSR1133+16 to average data of two different nights since the only two

PAGE 108

100 nights with data are 27 days apart. It is not clear up to the present whether the discrepancy of the periods is due to instrumental problems or insufficient precision in the calculations of the predicted period. From the results obtained in this thesis, it has been shown that it is possible to detect pulsars at both 26.3 MHz and 45 MHz, with certain limitations. In the light of the results obtained, some improvements and modifications in the equipment and data reduction can be suggested which will make it possible to obtain even more information about the received emission. One such modification is already being put into effect. A greatly improved transistorized 26.3 MHz receiver which will replace the old receiver is under construction and is nearly completed. Recording at a much wider bandwidth (perhaps 50-100 kHz) and utilizing a de-dispersion technique, would considerably increase the signal-to-noise ratio. The wide band signal would initially be recorded using an instrumental tape recorder of suitable bandwidth. it could then be played back into the Saicor Real Time Spectrum Analyzer, the output of which could be digitized and further analyzed by means of a microprocessor or a minicomputer (like the PDP 1134) After the data has been converted to a power spectrum appearing in the 200 output frequency channels of the Saicor spectrum analyzer, the frequency dispersion can then be removed by means of the computer. With a 100 kHz-bandwidth

PAGE 109

101 system of this type, the signal-to-noise ratio would be increased by a factor of |/-^ or 3.5, over what it was with the 8 kHz-bandwidth system. Another advantage of a wide bandwidth system, and perhaps an even more important one, is its ability to determine the degree of linear polarization of the pulsar signals by making use of Faraday rotation in the terrestrial ionosphere. As stated previously, the rotation of the plane of linear polarization varies, due both to ionospheric fluctuations and to a continual change in the ray path as the earth rotates, so that the polarization plane is frequently parallel and frequently perpendicular to the antenna dipoles. The degree of linear polarization is readily calculable from the measured intensities at times of parallel and perpendicular orientations. This appears to be an excellent method for determining linear polarization; it has apparently not previously been used for pulsars. Also, if the rotation measure (RM) of the ionosphere above the observatory can be determined independently, the position angle of the plane of linear polarization of the pulsar signal at the top of the ionosphere can be calculated. It may also be possible to obtain still more information about polarization. Since the dipoles of the 26.3 MHz array in the Dixie County radio observatory have the capability to be rotated to any position, half of the array can be rotated in 90 respect to the remaining half, making possible to obtain information about the circularly

PAGE 110

102 polarized component of the emission. If for some pulsar there is enough signal-to-noise ratio, it would also be possible to set the antenna in such a way that one third of the array is oriented at 0, one third at 45 and the last third at 90; from the information obtained in this configuration it is possible to deduce the Stokes parameters of the incident wave. Any pulsar polarization measurement made at 26.3 MHz and 4 5 MHz would be useful since none have been made at frequencies lower than 63 MHz (36) Improvements need to be made in the calibrations in order to obtain more accurate values for the peak flux density. This can be accomplished by simulating the pulsar pulse using a precise time standard and a frequency synthesizer, in order to obtain periods within some y seconds of the observed period. It should be possible to obtain estimations of peak flux densities by digitizing and processing this signal in a similar manner to the actual pulsar. It will be also necessary to find out the reason for the discrepancy between the DOPVEL period and the period obtained from the observations; simulations with an accurate period should help in determining if the discrepancy is due to instrumental problems. A more careful analysis of the accuracy in the calculations of the DOPVEL program needs to be made. The solution of this problem will make it possible to average several nights and obtain the long-term average pulse characteristics.

PAGE 111

APPENDIX A LIST OF PULSARS Tables A-1 and A2 contain the pulsar candidates to be observed using the University of Florida 26.3 MHz array and the 45 MHz array at the Maipu Radio Observatory. Table A-3 gives the references from which the data at the different frequencies were obtained. 103

PAGE 112

104 -p Â•H c Â•H H in ^ IJi 4J C o ^ (0 fa u Â•H -P O (0 rH C 3 O S-i 0) o -p H 0) c 0) rO 0) g >1 Â•H VD W (N C I OJ O Q rH X 3 M I 0) C (N I <1 1^ g 00 ^ 00 I Q O o Â— Â•H W (U (d n rH 04 in 0\ m iH CN o O O 00 CO in in CM in en U3 CT\ o 00 on i-D CN CM N N N K N N N K EC K X S s: CN S s s: rvj (N O rH iH o o fN r-i in
PAGE 113

105 x: -P 0) Â•H .H C Cn-P C o (U C (0 (U Â•Â— -p Â•rH Eh U -H -P I U ^ U O OJ o >i Â•P >H T) rH C W ^3 -H (0 (u e ft S K -P Â— c H i-H CN I O M OJ N (N I I I 3 C ft W tM I K O rH I Id N K -P to < o g rH Â— I Q -a o -H W iH Â— 0) ft u Id (0 3 in IT) o C7> in (N n O O o o o o O O O o o o lO o o in ? V c' N N N N N N E K S S s s (N 00 00 rH O rH o rH o in CM rH in CSJ N N N N X s S o rH < n n iH rH Â• o rH rH C2J (p CSJ A N N N N N X X K o o o 00 o O m I-H 'S' rH o o rH in o U3 ro ^ n 00 < in n C-l CN n o VD 00 CO r-> o o o o o rH CN >1 u c OJ cr 0) M m 4-> x: -p -p rH X5 (0 > -P O C c o H -p g >H o c M

PAGE 114

106 Table A-3. List of references from which the data at the different frequencies was obtained. Frequency References MHz 34 Backer dO) 53 Sieber (29) 61 Izvekova et al. (8) 80 Slee and Hill (9) 102 Izvekova et al. (7) 151 Lyne et al. (22) 408 Komesaroff et al. (10) and Manchester et al. (38)

PAGE 115

APPENDIX B SAMPLING RATE, TIMING PULSE FREQUENCY AND DATA POINTS PER SECOND For the configuration of the data reduction system used (Figure 2-2), the sampled values are sent to the Amdahl 470/V6 computer in real time. In that condition the rate of transmission of data into the Amdahl is the limiting factor for the sampling rate. The absolute maximum transmission rate to the Amdahl 470/V6 is 1200 Band (120 characters/second) The maximum useful rate is 60 characters/second; 40 characters/second is a more conservative (but reliable) rate. Since the A/D converter has 12 bits, which means 3 characters per data point, the transmission rate will be 40/3 s 13 data points per second. The number of data points per second (DPS) obtained from the KIM I is given by the relationship (39) : DPS = 1 TPPxNAVExTSAMP (B-1) where TPP KIM I timing pulse period (ms) NAVE Number of sampled values averaged before sending to the terminal TSAMP = Parameter for the KIM I program The TPP for the KIM I must not be shorter than 4 ms Hz) ; the value used in the digitizing process for the 107

PAGE 116

10 8 pulsar was 8 ms (125 Hz) The niomber of sampled values averaged (NAVE) before sending through the terminal to the Amdahl is fixed at 8. Table B-1 gives the relationship between the TPP, TSAMP and the DPS ( 3 9) Table B-1. Values of DPS as TSAMP. a function of TPP and TPP (ms) FREQ (Hz) 01 02 TSAMP 03 04 05 4 250 31.25 15.625 10.4167 7. 8125 6. 250 8 125 15.075 7.8125 5. 2083 3. 906 3.125 Data Points Per Second (DPS) In order to meet the transmission rate requirements of 13 data points per second using 8 ms for the TPP, a safe value of TSAMP = 02 must be used; that gives a value of 7.8125 DPS. If the tapes are played back at the same speed that they are recorded, then the resolution would be also 7.815 DPS, which is a poor resolution. However, if the tapes are played back with a slow-down of 4:1 the time resolution will be increased by a factor of 4. Table B-2 gives the number of DPS obtained in real time and in actual or original pulsar time (39)

PAGE 117

109 Table B-2. Number of data points per second in real and original pulsar time. Samples per second Real time 15.675 10.4167 7.8125 6.250 5.2083 Original pulsar time 62.7 41.667 31.25 25.00 20.8332 Since NAVE is fixed at 8, that means that 8 sampled values are averaged together before producing one data point; in that way, the tape is being sampled 8 times faster than the data point rate. As an example, if a TPP = 8 ms and a TSAM = 02 are being used, 7.8125 DPS are obtained, but the tape is being sampled at the rate of 62.5 samples/second in real time, which in original pulsar time is 250 samples/second.

PAGE 118

APPENDIX C SCHEMATICS 110

PAGE 119

Ill o -P e u J a. o o 13 X O O to I u (U Cn

PAGE 120

112 in EH O AAA O T _l O O o u H -P e u

PAGE 121

APPENDIX D COMPUTER PROGRAMS 113

PAGE 122

MCAPGM Progr

PAGE 123

115 o o o u < a. n \ K ftj o N. o OIU e> ZOZ < Â• Â— o \ o o-c. o O 7 u. o -1 ~tc< 11 z Â— s lu hc o < z a -ITr O 3 OE S -< 'U u. ZJ -l/>< Â• < Z -N -t~*-t--l-CIQIÂ— Â— occoooo -iLjir Â— w7 ^ III (L < LJ < C Z K I }liu. r u. u. tr o o Â• Â• o o O II o Â• II o Â• o o ir o II u.. II o L K c a Ni Tir C Â— L KJ Â— -^ca Â• t Â— cc X Â— Â• O Â• ID O (\J CD CM M I rj Â— Â• Â• I -tÂ• Â• 13 _| K (Â• Â• U J 00 Â• Â• a u a o iil'.iC t ^4^4UJUJ a Q K Â— do. Â• Â• u C C Â• Â• Ul o Z Z C C I/I Â• < < 7 Z Â• Â• < < _J + Â— .a wr: -OJE *^ -4 0) c z o < Â• Â• Â• Â• OlO or; o o z -IUJuiUJUj"* \ Â• Â• Â• Â• a tr H U UlUi Â— III a a 01 r" ci ID 0 IhÂ— Â— Â— Â— H II II w Â— w ^ ax ti. U. L. k UI d C Â•----Olz CO O II NZ* o u Â• L U. Â— to _IO c c z c Â•nj UJ K a Â•0 I in a I otII O It Â— o << ai"< Bo Â• OS MM Â• o Io < < Â— O 0 c" c z o ^ .o a N Â—o <-Â• i-~ << Ol-< Â— Q Â• on MM o I *-in II II in w tti-oo Â— r Ul "t II II Â— < a o CD c u UJ zi----a 0 f\j fy m n IIin H < D< I tQ T a < a OOU o u. u. o ry Â— ul 4^ hIn II z II < u. oo Ul UJ < f OK Â• Â• C O J I IIII O fy U Â— u, Â— II < + CI (T (T Â• a DCC-Â— oo UiH Â— Â— -<|a II --U. z u. oiu. u. o I Â• o c Â— -o o I< o o < o Â— Q O o LJ c Â• o ZO 1u. < z :? < Â• 4r Â— r a: oz oo Ul Â— Â•< rj Â— Â— Â— a o IB Â— z -I Â— c z z Â• Â• < X Â— q: ''<5I< o Â• Â— Z C UJ UJUt< X 2 ?0(r: lU-Z --HON*Â— 0'0*0 Â•a^ Â— Â— Â— L. Â— c Â— o Â•oil -7n core* wKtotioii-' -no ^Â•--Â•DDliJtoilCH>-0 ) o o o < K OOOOOOOOO Â— Â— Â— Â— Â— Â„f\;f\ag Q. oooooooooooooocoooocooooooooo o ooooooooooooooooooooooooooooo

PAGE 124

116 K O o o < < M o Â• V Â• z m -z z CUJ z > < U 1 < -2 > Â— _o 1 Â— z z -2 o Â— --Z 2(D > Â— w "X 7cr + Â• zoz r z --Z+ z II (C H tn Â— Z T ., Â— Â— 3 r 2f T o Â— **:3(/>in II ZD n w It (I ^ ^ I/; rjzit-zZ(~it Z T ? 5 T Z liJ O(DZt0(/l(/^i/)(j< Z S O T a: w DZ o (/I Â•Olrt V -tor II T ::) y i/i u :^ t/i II r; (/itT ti 7 II n o n MO tl K u T y z >o D o < c v u in tr < z o L. T cy Â• V D z ui ec,, T *..L UJ D tmet I Â• < Â— c It t/l Â• UJ UJ c V o> O** Uf c < in> a N z J I ' < Â• *f It ty X (M o Â— o: <\. M y I a o -ca a ff iL c o J O Z w H ^ Z O 3 o ^ Â— ^( C d) +. O H II Â— ^ Â— Â— Â— c J II _( in OO Â— 'ijL-Jiiu.jj_i'r< Â— o Â— o*Â• Â•^cczrj a ito o Â— zuj~ujKj-i--ii )iinKii-.C Â— ^f\jirit-.->fy II Â— It Â— _j Â— Â•-'w^*-wH--wrjirw^^.2r - Â— Hyt-t-t-l-t-l-t-lÂ— ZHHQ -Z*-Zw Â— w Â— c < IE ^ tr C O OOCJOO C Z z o 1-U--U. U. lLU-I^DaOJ_J_i-IJJ_l_|-tECOt-D Â— Q ^^-acnz>L.ar. caacacrzuow <0 (A II Q CK UJ < > U < z Â• o Q. a X ~ -* < z tn i/i c O K o .-^ Â• < w u lU U Â— LJ O X in UJ z cr y o CCN < Â• or o tU X Â• UJ D w-* > V) bl Odiio -J L. f' Â— ta M y o i/i (J X D*^a Q < << Wt-z U. in UI^Q 0O ZOCliiU z U) a UJ o a; Hcor in < O UUJH 2 in Cin < t_J o o z O U.' D O DC] 1/1 a Â— u. o w u ^ u uiin ? -J ^C 0O Q. lio -i Z U< D liicr Ofy coy U. y ii z 0 1 < y r c: in H UJ o I y y _ico I u r Â— r: 3 -1 _j Â— ztz a a a O iP (vj o r o o Â• -o -o c o c> c^^^oÂ•cÂ•oÂ•0' a o o o o o o o o o o o o o o o o o o oooooo o o o o o o o oo o o o o o o o o o o oooooo OOOoOOoO o ^ Â— Â— Â— Â— Â— fy(\jrj(\j(\jfypgfy
PAGE 125

MCAPGMl Program

PAGE 126

118 N o \ o e za M A Ox Zor o Â— \ M o-a O ozk. N M Z-s lij Â— 0 (C 0 < z a Â— u. <-o 00 _( NO. Â• <_ .1 uu Â— a>ir X a Â•a< o-j Â• u-Oa s n Â• XÂ— <o o o<-zo oo-0; ZN -O a Â— fflÂ— t ffi U. M o --(--- -Â•Z-t--( a}< uj < Â• -Zl-T a u.ao ->ii. D M O Â• n o Â• o 3ao It X UJ 11 o M a s ui --Na .a Â— Â• BÂ— Â— 0 Â• V O Â• lO M I Â• I IÂ• Â• 3 _llIÂ• tU J DO Â• Â• t ff U I^Z IT laNN -aa 30 Â• lU o zoom Â• II II II II a X a H 00 Â— Â— Â—tLJ I lij< 11 II iLkiLUJacaoBJo <0 o Â• T8 << ll< Â•>o o << 01 TES OK Â•< -0 z Z Â• oa tvj n K Â• N H Â• U t 2S 0 <Â— a 0 >-< < < u K 0 -c m Â• 00 0 0 Â• 0 HO 0 zo > 0 < z 0 II II a Â• <-o oa datii Â— .QO -zo z le 10 Â— NÂ— Â— NUJ Â•* lij Â— 0 Â— Â• Â• t ^ Â£ Â£ < X Â• on HON Jin o Â• Â•tvJN H--Â— Â— H 03*' < O -01 K w< a U.UOO Â•>a L. o M rvi u in >o Â• < 2 HOI4-1 I OU. IÂ— win Â— Â— HIU* "'< -Q Â• X B Â— Â• oa (UN Â• o Â— Â• < Â— H < < H o< o w Q o o Â• o zo H < z Â• < Â— o Â• 00 a o Q Z> hi (y -* O COON OOH< O U lij< Z Â— X o t-t-||O0J,Â— otoo Â— i-> Â— COUJooooooo ooOo OooooooooooooooooooooooooooooooooooOoooooo ooooooooooooooooooooooooooooo 0000 oooooooooooooooooooooooooooooooooooooooooo I

PAGE 127

119 N m S o Z2 3I Â— ~ Â— o-zzÂ• z oz z II 0) II CD Â•> II Â— Â• Â— _z n Â— Â— t-T > < U I >-. < Â— I w -z Â• in Â• rg Â• Â• in Â• Â• N o IS Â• (>l M Â• Â• N CD e o s -U. N ni o n Â• S d o Â• Â• N LI IZ X o .. o N UJ z b. < a O a M Â• I Ul ct a -z O N 111 1 X w a N rg T < U) 1 a Â• CD< X 1 2a o z *IL z Â• z o o Â• c z z a -JZ Â• It. a z Â• z u. z: N Â• z ID ^3 z < 1CD O .. 1o < Z z ^ HU, UJ < >-_l o o < Â— L. -J -s Â•ZD :-z Â• a* ozz Z I II mio II z 11 II -7ZII I I Z DD3t> I/I i/) to u > u > Z B lO-ZD i O Ii/I Â•O t/1 3 V -m I inu X 3 s I >-3 inu. 13 w H D 3 wim II Â— z I/In no o Â— II ^ N in in kUJU I I z >> 05 30 < < oin tnu m oa. Iin -u u a N -o > O Â— UJ< < inÂ— > ct N Z < imi o 3 V o: ^ u in -uiSOIuOZ X O 1/1 > _1I m I < o o 13-2 3lfto Â• in Â— nx Â— Â— Im o n ^a 'n Â— a o w Qin Â— o m u ii.< Â• It T X M Â— m -
o Â— o Â— Â— in< >UJh*iiJ>t t II t< Â•< h-tÂ— -j-j Â• (Dl.(D|l(L(I Â— Â— "-Â•Â•"Â•Â•-Â•"III --DU. U.U. JQ. inz0(DZ IZZ Â— CD Â£ o*Â— Moaw O < H n Â— II Â— o-ia _iJ Â• t--UJf-K Â• Â•<Â•<_! OZffiK Ol-O Â• II Â— II JÂ— Jo Â— a Â— '^'O Â— ^(vi Â— 1Â— Â— _Nin Â— Â— ruK--II-(-l-l-K>->1 Â— irooooooo )o:o_i jjjj J J :iiuaa.aaaaa a z o in o s<^4 Oo uc c -U. zn o z ucr lu I c Ot U3 -z zÂ• Â— jin I -M I -t-a bj UJ Â• I J >-r> : P Â— X<-UJ Â• I # I 1^ Â•> CD K < c c in < u. Â— o z~ Â• <0 M Â— OOC4 Â• O OXIIu.c0 o u ujo Â•Â— aU. II DNSÂ— S N Â— --Z Â— l-O-l-wi. Â— Â— OUI < -UJ -wik.Ti-ia z Â— z Â—X Â—or Â— so Â— oooaootroao>aizuo3iu.Oku.iBU.cn o oonoÂ— inoin OOO Â— Â— Â— (MIM in u u o inuj I o M < Â• q: Â• Ul > UJ 1/1 m*^ loti -oo o q; o ouo r -CiUJ o a: in Â— UJ U Â— Ia in < -JO u UJ 30 in a Â— Iu. J au3 u. ujin z 0. oa u. o oo:o:!bi ZUJ< Â— minKj1 1 -lOO -33-1 J i-ziaa oooooooooooooooooo OOOOqO oooooooooooooooooo OoOOOO oooooooo o Â— Â— Â— Â— Â— Â— fy(y
PAGE 128

P2045 (or P1133) Program

PAGE 129

121 in O X 3 1 (\l y, O Â— V Â• O D S _j 3 O Â— Â— Â— O < O Â— O Â— w Â— > Â— Â— Â— w a 1/1 ti ^L. Li Â• N UJ z > -J .a < W 2 Q I c ^ c o o Â• tH e o i I Â— Â• a. O o ul Â— 'J */ Z O C L. Â• Li J. Â— O U. O > C Â— Â•*-C-< ^ : -\-. -I7 73 I Ju. z < Â•Z>x Â— I C 1^ < T n Â— _iJ^:i> Â• Â— -rr zijL' Z 'JO w < r u.J-JUJU.<03:mi X c a u t 0 u. w IL -> G Z II Q Z :3 Â• Â— .n > 2 00 < 0 n ^ Z Â• ^ > t~ 1 Â— Â— ^ > w c ^ n C ^ < z > Â— > Â£r Â• 0 ^ Â— z Â— Â— a -.<(/; Â— r_J _| UJ CilÂ£T, y U. i X I" 0> 2 Q z -J < r, cL lA -2 ^ Z Â• li 2 Z 2 :i II Â• 1fL eg Â• "i. Â• I 2 S < Â— Â— 0 Â— < li(\ tÂ— 11 7 3 UjI E li. U' 2 > a Â• < 2 -'*U.lU "l Â— Z2Â• 3V 5 o > < 'd? -C I t < J < y IX III Cj O I_( O L C >I T I LC ^ Li ^ Â— 0 o :> Â• Â• o loot' ^' ti II 1 z n c: z o 3 II 2 o 2 u Â• 3 ui -1 Â— o 2 O ^-j II II ti li. J 2 u. tii Z (-) z o-T 2 : Â• '3 X. ; o < -IÂ• II o M 2 5 2 ^ 00 2 I/) in c O ^ Z N a X it c 2 03 Â• tfi It LJ Â— wQ N ^ C ^ Â• O f\j w V) -i ii. a. Z u J U. Li r a r ^^ I ^ r ^ 2 l-" C /, L. n c > L.. cr ILT Â— < 11 Li ^ :d to Â• r\j jj Â— II I/. K fr ^ o w ^ii 11 r D o 3 > O'Xit-l: Â— l~< ui a ? 10 ta 1/1 ^.M tOK o^Z & .2 Â— N 2 U _J z -a .J -C -52 Â• C D Wl ::; 2 r -c zÂ• Â• z Â— > J Â— J r Â• X < -Â• II UJ c Â— < oo Â— a Â— II z_i; Â•H^klJiiC O fc-.* t, II T II a O J Â— 5aOa ItÂ— Â— _) -Â•LjnujUjna'UQ. a X UJ Ui f C c ^ < lu 0 UJ M K u IL 0 i/> 1*1 ,* EC Â• H X Ui Ul in X n tN 0 Â• X u. II < < (X II z u 0 0 > Â— < t/) Â• !t K t". Â• Â• 4 0 On < Â• Â• 0 Z CO Â•L. H 0 Â• (/) ^ -^ II l/^II Â— (J 1 X U' 0 IT. -0J < Â• < Â• c 0 ta Â• ry OL.' It U. u: t> 2 0 (/J 0 < OUJ a -la U". ^ h Â— 14J c n: Â• U < 1X 2 >c:i-< Â• Â— in Uj Â— < _i 3 n:c ^ cr jz _i< ^ Â• Â— fT Â• c: w 0 Z 2 Â• 2-(Nj -*oo 0 0 Â• Â• Â•0 Â• < 0 0 00.S -tf'. n I II n eg -1 w Â— lÂ„ Uj < K ? II Â— tl >Â• J -I o u -J O Â— Iw O U 1 00 uJ < Â• Ui ^ y r T Â— 2 a: u c N L c u i' u. t/ o o

PAGE 130

122 o in Q O II 3 a UJ (L z u< _(0 a o o u < < Ul o 3 Z -01 4 Â• < 1Â£ LI UJ o Â— Â— o o U Â• > O ,1. u o o UJ > < O o < < II C O II o II z Â• i 1 1 II ri < *" Â— < II 11 Q. 0 H Â— D< Â— Â— 1 Â— nL.a Â— iL Â— J > Â— i/'V.OU Ll>Ol/) wu.< Â— ir Â— Â— Â— > Â• Z K m yl J < (0 IL Â• Â—I _JÂ— OU.* ("0<0Â— QO < IZ iT < or-'>o Â— Â— Oi-^ Â— c Z >I> Â• r > 1Z^>* Â•!<< Â•Â•OÂ— Â•Â• Â— 1^ Â— < H II II Â—X Jl o 7) Â•* K Â•-**ujf>~'-a.''-J Â• Â•-Li Â• Z I _j 11 cv Â•rZvuj-*u;ti". wC(r Â• Â•--Â• Â•jO Â• Â•iij_j Â— ttaiO^-n c Â— Â• S -U.^< Â•-"U' C Â— -^-ill*Â— Â— 33' '^Ts^^ia D> Â— Â— Â— Diw r-(-OC II Â— II l-0 0l:.l-(.3OC->-Â— ^^.i.w|UXUtf iJll OL-uJ< < I' Â— Â— >>Â— y". ^ Â— Â— UJ<--uJ< in Â— 'X;33> >J.Q.3 t/. cn<'*^t-l-CK-(----l--3^-Hia Â— U. Â— J.li.w -J-J-*Â— ^ >C Â— Â— Â— Â— Â— w QUUUOOL.Â— a^ Â— aUC UOirCu.ti.O<" Â— L.U.U.l.i.U.O_J_>-Jj_jJJj:OLj[r3k-Z > Q JIO o li PJO o I<0 f>. Of a ^ L'l O

PAGE 131

APPENDIX E 26.3 MHz ARRAY CALIBRATIONS In order to obtain the minimum detectable flux density and then estimate the peak flux density for pulsar PSR1133+16, several calibrations were performed with the array. The standard noise source was provided by 2 Sylvania 5722 noise diodes; as a secondary noise source a Hewlett Packard amplifier was used. Several radio sources of known flux density and the galactic background temperatures were also used in the calibrations. Figure E-1 is a block diagram of the 26.3 MHz array signal transmission and calibration system. Usually the calibration signals are introduced at point (l) (Figure E-1). In order to improve the calibrations, the signal was later introduced at point (2) (Figure E-1); by doing this, any uncertainties in the gain (or losses) of the 10 preamplifiers and the 2 N-S Butler matrices are avoided. Two 1 inputto6 outputs 50-ohm power splitters were used to split the calibration signal into the 10 preamplifier inputs. To properly match all the outputs, two power splitters output were loaded with 50-ohm resistors. As a result of introducing the calibration signal at @ and then at (^ a net gain of =6 dB was determined for the 10 preamplifiers plus the 2 N-S Butler matrices; several Â• 123

PAGE 132

124 PQ piII ps o N PC PS W H o J II M 1-1 m Â— J rH < w II DS PL, u PS o m N TE J M Pm P3 II ^ O z l-l o M p m CO fcj <: Q M CM O M Q m Pd 1 SY CM IS NO 1 PiJ CO M Z O 2 O I <: z pd H H < V a^ pq c/2 Pi w w H H o M l-J PM CO CO PJ Z II p: o x> Â•H cd u CO n CM 0) 4J u o 00 n) Â•H Â•a u o .-I pq I Ul (U u a 00 Â•r(

PAGE 133

125 other gains and losses were also determined for the preamplifiers, 250 KHz filters and aluminum coaxial lines; all the losses and gains determined are shown in Figure E-1. E.l Losses Between the Dipoles and Point (2) The procedure to obtain the losses between the dipoles and point (2) which includes the 10 E-W Butler matrices, the hybrid rings, the matching networks and the parallel transmission lines, is not as straightforward as it is to obtain gain and losses for the rest of the system. An indirect method was therefore used in this thesis. First, the recorder deflection due to the galactic background was measured. Then a signal from the noise generator was introduced at point (2) (using the indicated switch) and was adjusted to give the same recorder deflection as that from the galactic noise. From the ratio of the noise temperature of the introduced signal to that of the part of the sky from which the galactic signal was obtained, which is available in the literature, the losses can be obtained. A summary of the results is shown in Table E-1. The noise temperature for the noise diodes has been calculated using the following relationship (17): where T,, = ambient temperature {= 290 K) e = electron charge (= LSxlO'"*"^ C) I = diode's plate direct current R = plate load resistance k = Plank constant (= 1.38xl0~^-^ j k)

PAGE 134

126 (0 u N K n CO m o Â— 1-^ o -H -M I -a -Â— c u :3 ^ (13 o o a;^^ H (13 M g o (0 CQ CP 0) 4-1 o u Q) O w Â— O Â£1, C o Q) O Eh (C ^4 to Â— ~ X! 4J CQ H C C t3 >H O -H ^ n3 -H o u 4J a, (0 4J^ 4-> m c c rH O -H Â— (0 -H O U 4-> X! -H I (C Â•H G H O U -H 0 4-1 Q I 4J C C 0) o U -H Â•H U) W in LD IT) o n CTv 00 o in o 00 o in (N (N in in o o o r-\ VD o o cn in rH 00 O o 00 r~ (^ rH fN n CO CM (N un CM in o 0 0 Â•H VD O o (0 Â• Â• 4-1 Xi o Â£ e e in o CN in in ^: CTl O rH <-\

PAGE 135

127 Substituting in equation (E-1) the values for the constants, using the value R = 50 ohm and expressing I in milliamperes the following relationship can be deduced: T = 290 + 289.85 I (E-2) Equation E-2 gives the temperature in K and it is valid for R = 50 ohm. Galactic background temperatures for the first two points in the sky have been obtained from Turtle et al. (40) at 26.3 MHz. The value for the third point was obtained from measurements made by Blythe (26) at 38 MHz and corrected to 26.3 MHz using the galactic background spectral index given by Findlay (27) of n 2.65, which is valid for the range 20-400 MHz. It is interesting to compare the losses obtained in Table E-1 with the losses obtained by summing the estimated dB attenuations along the path between the dipoles and point d) The list of estimated or measured losses of the individual components is given in Table E-2 (41) The average of the measured losses given in Table E-1 is 7.2 dB which agrees well with the losses computed from Table E-2.

PAGE 136

128 Table E-2. Losses between dipoles and point (2) considering the partial losses. Minimnm rlR ^ 1 ^ i 1 Xill L4JII \_i JL> riaxiiTiuin urS Full wave twin line 30 .30 Hybrid ring .45 .45 Phase cable 80 1.10 Coaxial cable line 1. 90 1. 90 E-W Butler matrix 1.44 1. 98 Line connecting E-W to N-S Butler Matrices .5 1. 10 Total 5.39 6.73 The corrected noise temperatures at point (g) given in Table E-1, is the noise temperature of the noise source corrected for the 5 dB losses in the aluminum coaxial line and the losses in the power splitter (factor 0.76); losses in the power splitter consider the losses due to the two 50-ohm resistors and the losses at 26.3 MHz of the power splitter itself. E.2 Antenna Effective Area By using the losses between the dipoles and point (2) already obtained and calibrations done with radio source 3C433, the antenna effective area can be calculated; the method consists of obtaining the change in antenna temperature

PAGE 137

129 AT for a radio source of known flux density S. The a. relationship that gives the effective area is: 2k AT Ae (E-3) AT^ is the change in antenna temperature referred to the dipoles, corrected for losses e between point (g) and the dipoles. If AT' is the change in temperature measured at point (2) then the actual change in antenna temperature is: AT' AT^~ (E-4) where e = transmission line efficiency, which in this case considers the losses between point (g) and the dipoles. From the calibrations, the following values were obtained for 3C433: S = 267 jy AT' = 295K a T = 25,000K e = 0.19 Using equations E-3 and E-4, the following value is obtained: = 16,050 m^ This value needs to be corrected by a factor p^^ = 0.69, given in the Array program (11) which considers the fact that the source is off the center of the beam. The corrected effective area of the antenna is then: A^ = 23,260 m^

PAGE 138

130 The value of the effective area can be also expressed in terms of the change in antenna temperature in K, divided by the flux density in Jy (K/Jy); this quantity is usually denoted as A* ( 12) e AT X p a "^1 Substituting in E-5 the values AT = AT' /e = 1553K a a S = 267 Jy p^ = 0.69 the following value for A* is obtained e A* = 8.3 K/Jy

PAGE 139

LIST OF REFERENCES A. Hewish, S. J. Bell, J. D. H. Pilkington, P. F. Scott, R. A. Collins, "Observations of a Rapidly Pulsating Radio Source," Nature, 217 709 (1968). Yu. M. Bruck, J. G. Davies, A. D. Kuz'min, A. G. Lyne V. M. Molofeev, B. Rowson, B. Yu. Ustimenko and Yu. P. Shitov, "Radio-emission Spectra of Five Pulsars in the 17-1420 MHz Range," Soviet Astronomy, 2_2, 588 (1979). F. N. Bash, F. A. Bozyan and G. W. Torrence, "Observations of CP1919 and NP0532 at 38 MHz," Astrophy sical Letters, 7, 39 (1970) Yu. M. Bruck and B. Yu. Ustimenko, "Decaraetric Pulse Radio Emission from PSR0809, PSR1133, and PSR1919," Nature Physical Science, 242 58 (1973). Yu. M. Bruck and B. Yu. Ustimenko, "Decametric Radio Emission from Four Pulsars," Nature, 260 766 (1976). H. D. Craft Jr. and J. M. Comella, "Frequency Dependent Pulse Width for CP1133," Nature, 220, 676 (1968) V. A. Izvekova, A. D. Kuz'min, V. M. Molofeev and Yu. P. Shitov, "Energy Spectra and Pulse Shapes of Pulsars at Metre Wavelengths ," Australian Journal of Physics, 32. 25 (1979) Â— V. A. Izvekova, A. D. Kuz'min, V. M. Molofeev and Yu. P. Shitov, "Radio Flux Density and Pulse Profile of Pulsars at 102.5 and 61 MHz," Soviet Astronomy, 23, 179 (1979) ^ Â— O. B. Slee and E. R. Hill, "Pulsar Energy Fluxes at 80 MHz," Australian Journal of Physics, 2A, 441 (1971). M. M. Komesaroff D. Morris and D. J. Cooke, "Linear Polarization and Pulse Shape Measurements on Nine Pulsars," Astrophysical Letters, 5, 37 (1970).

PAGE 140

132 11. M. D. Desch, "26.3 MHz Array Reference Manual," unpublished 12. M. D. Desch, Ground Based and Spacecraft Studies of Jupiter at Dacameter and Hectometer Wavelengths (Ph.D. Dissertation, University of Florida, 1976) 13. M. D. Desch, T. D. Carr and J. Levy, "Observations of Jupiter at 26.3 MHz Using a Large Array," Icarus, 25 12 (1975). 14. F. Reyes, Radiotelescopio en 45 MHz para Fuentes Extragalacticas Memoria para optar al titulo de Ingeniero Civil Electricista (Thesis, University of Chile, 1977) 15. J. May, F. Reyes and J. Aparici, "A 45 MHz Radiotelescope for Galactic and Extragalactic Research," First LatinAmerican Regional Astronomy Meeting, Santiago, Chile (1978) 16. M. R. Viner and W. C. Erickson, "26.3 MHz Radio Source Survey. II. Radio Sources Positions and Fluxes," The Astronomical Journal, 80^, 931 (1975). 17. J. D. Kraus, Radio Astronomy (Mac Graw Hill Inc., New York, 1966). 18. G. R. Huguenin, R. N. Manchester and J. H. Taylor, "Properties of Pulsars," The As trophysical Journal, 169, 97 (1971). 19. F. D. Drake and H. D. Craft, "Pulse Structure of Four Pulsars," Science, 160, 758 (1968). 20. A. D. Kuz'min, V. M. Molofeev, Yu P. Shitov, J. G. Davies, A. G. Lyne and B. Rowson, "Spectra of Nine Pulsars at 61-1420 MHz," Monthly Notices of the Royal Astronomical Society, 185, 441 (1978) 21. R. D. Ekers, A. T. Moffet, "Further Observations of Pulsating Radio Sources at 13 cm," Nature, 220, 756 (1968). 22. A. G. Lyne, F. G. Smith and D. A. Graham, "Characteristics of the Radio Pulses from the Pulsars," Monthly Notices of the Royal Astronomical Society, 153, 337 (1971). 23. J. May and J. Aparici, Personal Communication (1980).

PAGE 141

133 24. B. Y. Mills, 0. B. Slee and E. R. Hill, "A Catalogue of Radio Sources Between Declinations +10 and -20," Australian Journal of Physics, 11, 300 (1958). 25. 0. B. Slee, "Culgoora-3 List of Radio Source Measurements," Australian Journal of Physics, As trophysical Supplements, Number 43, (Nov., 1977). 26. J. H. Blythe, "Results of a Survey of Galactic Radiation at 38 MHz," Monthly Notices of the Royal Astronomical Society, IT?, 652 (1957). 27. J. W. Findlay, "Absolute Intensity Calibrations in Radio Astronomy, Annual Review of Astronomy and Astrophysics 4^, (1966) 28. A. J. Turtle, A. E. Vaughan, "Discovery of Two Southern Pulsars," Nature, 219, 689 (1968). 29. W. Sieber, "Pulsar Spectra: A Summary," Astronomy and Astrophysics, 2Q, 237 (1973). 30. D. C. Backer, Personal Communication to S. T. Gottesman (1975) 31. R. N. Manchester and J. H. Taylor, Pulsars (W. H. Freema and Co, San Francisco, 1977) 32. R. N. Manchester, J. H. Taylor, G. R. Huguenin, "Observations of Pulsars Radio Emission II. Polarization of Individual Pulses," Astrophy sical Journal, 196 83 (1975). 33. G. R. Huguenin, R. N. Manchester and J. H. Taylor, "Properties of Pulsars," Astrophysical Journal, 169, 97 (1971). 34. J. D. H. Pilkington, A. Hewish, S. J. Bell and T. W. Cole, "Observations of Some Further Pulsed Radio Sources," Nature, 218, 126 (1968). 35. Yu. I. Alekseev and S. A. Suleimanova, "Polarimetry of Pulsars at 3.5 m Wavelength," Soviet Astronomy, 21, No. 2, 180 (1977) 36. Yu. P. Shitov, "Fine Structure of Pulsar Radio Spectra," Soviet Astronomy, _16, 383 (1972). 37. R. N. Manchester, "Pulsar Rotation and Dispersion Measures and the Galactic Magnetic Field," The Astrophysical Journal, 172, 43 (1972).

PAGE 142

134 38. R. N. Manchester, A. G. Lyne J. H. Taylor, J. M. Durdin, M. I. Large and A. G. Little, "The Second Molonglo Pulsar Survey Discovery of 155 Pulsars," Monthly Notices of the Royal Astronomical Society, 185, 409 (1978) 39. J. P. Oliver, Personal communication (1978). 40. A. J. Turtle, J. F. Pugh, S. Kenderdine and I. I. K. Pauliny-Toth, "The Spectrum of the Galactic Radio Emission," Monthly Notices of the Royal Astronomical Society, 124, 297 (1962). 41. J. Levy, Personal communication (1980)

PAGE 143

. BIOGRAPHICAL SKETCH Francisco Reyes was born in Santiago, Chile, on December 8, 1941. After finishing high school and with a strong desire to study electrical engineering (in which he was encouraged by his parents, sisters and brother), he entered the School of Engineering of the University of Chile. It was there where he became interested in astronomy while taking a course with Professor Claudio Anguita. His interest grew when he was allowed to use the old Heide refractor at Cerro Calan. He started his amateur astronomical activities by building his own small Newtonian reflector and spending hours and hours watching the interesting southern sky (in which he was enthusiastically supported by his two sisters Lucia and Guadalupe and his brother Juan) His professional interest turned to radio astronomy when in 197 4 Jorge May asked him to work on the 4 5 MHz array in Maipu as a thesis project. In 1975 he married Maria Angelica, and they now have two daughters: Carolina and Daniela (who just joined the family) In 1977 he obtained a degree in electrical engineering from the University of Chile, and in September of the same year 135

PAGE 144

136 he came to the University of Florida to continue his studies in astronomy. While studying at the University of Florida he has been teaching elementary physics laboratories, elementary astronomy laboratories and digital electronics laboratories. His other interests are mountain climbing, Andean music and photography.

PAGE 145

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree of Master of Science. Thomas D. Carr, Chairman Professor of Astronomy and Physics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree of Master of Science. Alex G. Smith' Professor of Astronomy and Physics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree/of Kast^^ of 3

125
other gains and losses were also determined for the
preamplifiers, 250 KHz filters and aluminum coaxial lines;
all the losses and gains determined are shown in Figure E-l.
E.l Losses Between the Dipoles and Point (j)
The procedure to obtain the losses between the dipoles
and point (2), which includes the 10 E-W Butler matrices,
the hybrid rings, the matching networks and the parallel
transmission lines, is not as straightforward as it is to
obtain gain and losses for the rest of the system. An
indirect method was therefore used in this thesis. First,
the recorder deflection due to the galactic background was
measured. Then a signal from the noise generator was intro
duced at point (2) (using the indicated switch), and was
adjusted to give the same recorder deflection as that from
the galactic noise. From the ratio of the noise temperature
of the introduced signal to that of the part of the sky from
which the galactic signal was obtained, which is available
in the literature, the losses can be obtained. A summary of
the results is shown in Table E-l.
The noise temperature for the noise diodes has been
calculated using the following relationship (17):
T = T + inf where T0 = ambient temperature (= 290K)
_ I Q
e = electron charge (= 1.6x10 C)
I = diode's plate direct current
k = Plank constant (= 1.38xl0~23 J K)

ACKNOWLEDGMENTS
I wish to express my deepest gratitude to my research
advisor, Dr. Thomas D. Carr, for his guidance and encourage
ment, and his patience in reading and correcting the manu
script. It has been a privilege to work under him in this
project, and I have benefited in many ways as his student.
I am also grateful to Dr. Alex G. Smith for providing
me with financial support and for serving as a committee
member. I give my appreciation to Dr. John P. Oliver for
the development of the necessary hardware and software, for
his advice and for serving as a committee member.
My deep appreciation goes to Dean Claudio Anguita and
Professor Hugo Moreno of Facultad de Ciencias Fisicas y
Matemticas of University of Chile for their encouragement
and interest in my career and for providing me with finan
cial support. I am also grateful to Jorge May and Juan
Aparici for their encouragement and optimism and for pro
viding me with the data at 45 MHz.
I wish to thank Jorge Levy for his help in the operation
and calibration of the 26.3 MHz array, Chris St. Cyr for his
encouragement and help in the computational part and Wesley
Greenman and Richard Flagg for their help with the electronics.
in

Figure 3-29. Pulsar PSR2045-16 recorded on May 18, 1979, at 45 MHz.
Average over 610 periods.

43
10
ENERGY
(Jm"2Hz_1)
*10
-26
1.0
.5
.4
.2

this thesis
SIEBER (29)
A SLEE AND HILL (9)
BRUCK AND USTIMENKO (4)
o IZVEKOVA et al (8)
30 40 50
FREQUENCY (MHz)
100
200
Figure 3-11.
Low frequency energy spectrum for
pulsar PSR1133+16.

4
period of a pulsar is known to be a good accuracy, it is
possible to average many periods (N typically in the range
of a thousand periods), greatly improving the S/N ratio.
The University of Florida Radio Observatory, located
in Dixie County about 50 miles west of Gainesville, and the
Maip Radioastronomical Observatory of the University of
Chile, located 20 miles southwest of Santiago, have long
been engaged in a cooperative program of low frequency radio
astronomy. Although most of the effort over the years has
gone into the study of Jupiter's radio emission, other
decameter-wavelength sources are also investigated. Each
of the two observatories has a very large antenna array, in
Florida array consists of 640 dipoles operating at a center
frequency of 26.3 MHz, while the Chile array has 528 dipoles
and operates at a center frequency of 45 MHz. Both arrays
were believed to be suitable for observing pulsars, although
previous attempts to use them for this purpose had not been
successful.
The goal of the research on which this thesis is based
was to develop a low frequency pulsar data acquisition and
reduction system, and to use it to detect and measure the
important parameters of at least one pulsar with each of the
two large arrays (i.e. the one in Florida and the one in
Chile). This goal was achieved.
<

OBSERVATION OF PULSARS AND OTHER SOURCES
BY
FRANCISCO REYES
A THESIS PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
1980

to Maria Angelica
Carolina
and Daniela

ACKNOWLEDGMENTS
I wish to express my deepest gratitude to my research
advisor, Dr. Thomas D. Carr, for his guidance and encourage
ment, and his patience in reading and correcting the manu
script. It has been a privilege to work under him in this
project, and I have benefited in many ways as his student.
I am also grateful to Dr. Alex G. Smith for providing
me with financial support and for serving as a committee
member. I give my appreciation to Dr. John P. Oliver for
the development of the necessary hardware and software, for
his advice and for serving as a committee member.
My deep appreciation goes to Dean Claudio Anguita and
Professor Hugo Moreno of Facultad de Ciencias Fisicas y
Matemticas of University of Chile for their encouragement
and interest in my career and for providing me with finan
cial support. I am also grateful to Jorge May and Juan
Aparici for their encouragement and optimism and for pro
viding me with the data at 45 MHz.
I wish to thank Jorge Levy for his help in the operation
and calibration of the 26.3 MHz array, Chris St. Cyr for his
encouragement and help in the computational part and Wesley
Greenman and Richard Flagg for their help with the electronics.
in

My thanks goes to Hector Alvarez, Mauricio Bitrn and
Fernando Olmos of Maip Radio Observatory for their help
with the 45 MHz data.
I am grateful to the Astronomy Department and the
Physics Department and in particular to Drs. Heinrich
Eichhorn, Robert J. Leacock, F. Eugene Dunnam, Charles F.
Hooper and Richard E. Garrett for their financial support
in the form of teaching assistantships and fee waivers.
My gratitude goes to Mrs. Brenda Dobson for her
patience in correcting my English and for her devotion in
typing and editing the manuscript.
The realization of this project has been possible
thanks to the continuous support provided to the radio
astronomy program of the Department of Astronomy of the
University of Florida and to the Maip Radio Observatory
through a grant from National Science Foundation (principal
investigator Dr. Alex G. Smith).
IV

PAGE
ACKNOWLEDGMENTS iii
ABSTRACT vii
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 DATA ACQUISITION AND REDUCTION 5
2.1 Data Acquisition of Pulsar PSR1133+16... 6
2.2 Data Acquisition of Pulsar PSR2045-16... 10
2.3 Data Reduction 12
2.4 Block Designation for Each Pulsar 16
2.5 Computer Programs for7 Data Reduction.... 19
CHAPTER 3 DATA ANALYSIS 2 2
3.1 PSR1133+16 Data Analysis 22
3.1.a Peak Flux Density, Mean Flux Density
and Energy for Pulsar PSR1133+16 35
3.1. b PSR1133+16 Pulse Width 44
3.1.c Tests Performed with the Pulse
Obtained on May 25, 1979 48
3.2 PSR2045-16 Data Analysis 63
3.2.a Peak Flux Density, Mean Flux Density
and Energy for Pulsar PSR2045-16 77
3.2.b Pulse Width for Pulsar PSR2045-16 83
3.2.c Period of Pulsar PSR2045-16 from
Observations on Two Consecutive
Days 87
3.2.d PSR2045-16 Pulse Intensity as
Function of Time 91
CHAPTER 4 CONCLUSIONS 93
APPENDIX A LIST OF PULSARS 10 3
APPENDIX B SAMPLING RATE, TIMING PULSE FREQUENCY
AND DATA POINTS PER SECOND 10 7
APPENDIX C SCHEMATICS 110
v

APPENDIX D COMPUTER PROGRAMS
113
APPENDIX E 26.3 MHz ARRAY CALIBRATIONS 123
LIST OF REFERENCES 131
BIOGRAPHICAL SKETCH 135
vi

Abstract of Thesis Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
OBSERVATION OF PULSARS AND OTHER SOURCES
By
Francisco Reyes
December 1980
Chairman: Thomas D. Carr
Major Department: Astronomy
The investigation of the radio emission spectra of
pulsars at low frequencies is important since it helps to
clarify the mechanism responsible for this emission. At the
present, the low frequency end of the spectra is poorly
determined compared with more precise measurements avail
able at higher frequencies. Also, measurements of pulsar
intensity variations at low frequencies due to scintilla
tion are of value in the study of the interstellar medium.
Detection of pulsars at decametric wavelengths is
difficult because of the deterioration of the signal-to-
noise ratio caused by the high galactic background tempera
tures and the turn-over of the spectra of most pulsars.
Also the dispersion of the pulses imposes the use of narrow
bandwidth, reducing the sensitivity of the system.
Vll

Two pulsars have been observed at low radio frequencies
using narrow bandwidth to avoid pulse dispersion problems.
Pulsar PSR1133+16 was observed at 26.3 MHz using the
University of Florida large array; pulsar PSR2045-16 was
observed at 45 MHz using the Maipu Radiobservatory large
array (Chile). Long term integration techniques were used
to reduce the data.
Both pulsars were successfully detected and the average
pulse intensity and integrated pulse profile were obtained.
Time resolutions of P/38 were obtained for PSR1133+16 and
P/62 for PSR2045-16, where P is the pulse repetition period.
The two observed pulsars were selected because of their low
dispersion measure, long period (>1 sec), high pulse inten
sity and their location with respect to the galactic back
ground .
Several other radio sources were also observed and were
used to calibrate the effective area of the 26.3 MHz antenna.
Chairman
vm

CHAPTER 1
INTRODUCTION
Pulsars were discovered in 1967 by S. Jocelyn Bell and
Anthony Hewish. The first pulsar paper was published in
Nature in February 1968 by Hewish et al. (1) telling about
the discovery of pulsar PSR1919+21 and giving some of its
parameters. Since then, about 321 pulsars have been detec
ted and cataloged.
According to currently accepted models, pulsars are
rapidly spinning neutron stars originating from supernova
explosions. They are believed to have strong magnetic
12
fields on the order of 10 gauss, probably the strongest
known magnetic fields in the Universe. The neutron star is
believed to emit continuously a narrow beam of radio waves
which corotates with the star, producing the observed pulse
each time the beam sweeps past the earth. The periods of
the pulses which have been observed range from about 0.03
to 4.3 seconds. So far, for only 2 pulsars has it been
possible to find the optical counterpart of a pulsar.
Several mechanisms have been proposed to explain the
origin of the radio emission and how the rotational energy
of the neutron star is converted into electromagnetic wave
pulses; this subject still remains one of the least under
stood aspects of pulsars. Some of the important parameters
1

2
of a pulsar that need to be measured before realistic
models of the radio emission mechanism can be developed
are the spectrum of the radiation, the pulse shape, peak
intensity, and the time variations of these parameters. A
large number of measurements have been made in the range of
frequencies between 100 and 2,000 MHz, but only a few below
100 MHz (2). The low frequency part of the spectrum is
particularly interesting because most pulsars present a
turn-over of the spectrum at metric and decametric wave
lengths, and also because pulse shape and widths show some
peculiarities in this region. The spectral maximum for
most pulsars is in the range 60 to 250 MHz; this peak is
nearly symmetrical with respect to the frequency at which
the maximum occurs (2). Measurements of the low frequency
part of the pulsar spectra have been made by Bash et al. at
38 MHz (3), Bruck and Ustimenko at 10, 16.7, 20 and 25
MHz (4, 5), Craft and Cornelia at 40 MHz (6) and more
recently by Izvekova et al. at 61 and 102.5 MHz (7, 8).
Most pulsars measured at low frequencies are in that
part of the sky which is observable from northern hemis
phere observatories. Observations at the lowest frequen
cies have been carried out in the southern hemisphere only
by Australian observatories, at 80 and 150 MHz (9, 10).
Detection of pulsars at low radio frequencies is difficult
because of the deterioration of the signal-to-noise ratio
caused by the high galactic background temperatures, and
the low frequency spectral turn-over exhibited by most

3
pulsars. Also pulse dispersion in the interstellar medium
becomes more severe the lower the frequency. At higher
frequencies, compensation can be made for dispersion, per
mitting the use of relatively high bandwidths. However,
at the lower frequencies, it is generally necessary to
restrict the bandwidth because of dispersion, further
reducing the sensitivity of the detection system. Due to
the intrinsic short pulse durations of most pulsars (< 150
msec), short time constants must be used, imposing still
another limitation on the sensitivity of the system.
Nevertheless, the sensitivity of the system can be
improved by averaging many independent records of the same
source. If N records are averaged, the signal-to-noise
ratio (S/N) is improved by a factor of / For pulsars,
this can be accomplished by averaging groups of consecutive
periods. Digital sampling is generally used in such aver
aging. Due to the relatively short pulsar periods, the
signal needs to be sampled at a fast rate in order to have
enough time resolution to meet the requirements of the
sampling theorem. Since it is not feasible to manipulate
the large amounts of data generated in the digitizing proc
ess by hand, digital computing techniques must be used,
providing an efficient way to average, process, store and
plot that data.
In the past few years, the development of small and
inexpensive microprocessors has made possible improved
methods of pulsar data handling and processing. If the

4
period of a pulsar is known to be a good accuracy, it is
possible to average many periods (N typically in the range
of a thousand periods), greatly improving the S/N ratio.
The University of Florida Radio Observatory, located
in Dixie County about 50 miles west of Gainesville, and the
Maip Radioastronomical Observatory of the University of
Chile, located 20 miles southwest of Santiago, have long
been engaged in a cooperative program of low frequency radio
astronomy. Although most of the effort over the years has
gone into the study of Jupiter's radio emission, other
decameter-wavelength sources are also investigated. Each
of the two observatories has a very large antenna array, in
Florida array consists of 640 dipoles operating at a center
frequency of 26.3 MHz, while the Chile array has 528 dipoles
and operates at a center frequency of 45 MHz. Both arrays
were believed to be suitable for observing pulsars, although
previous attempts to use them for this purpose had not been
successful.
The goal of the research on which this thesis is based
was to develop a low frequency pulsar data acquisition and
reduction system, and to use it to detect and measure the
important parameters of at least one pulsar with each of the
two large arrays (i.e. the one in Florida and the one in
Chile). This goal was achieved.
<

CHAPTER 2
DATA ACQUISITION AND REDUCTION
Two pulsars, PSR1133+16 and PSR2045-16, were selected
(from the lists in Appendix A) for observation because of
their low dispersion measure, relative long period
(>1 second), high pulse intensity and favorable location
with respect to the galactic background. Pulsar PSR1133+16
was observed at 26.3 MHz using a bandwidth of 8 kHz, and
pulsar PSR2045-16 was observed at 45 MHz with a bandwidth
of 10 kHz.
The detected IF signals of the pulsar were recorded on
one channel of a magnetic tape, and a time code generator
signal on the other channel. The tapes were played back
(making a 4:1 slow down), the pulsar signal sampled, and
the resulting values stored in the NERDC Amdahl 470 computer.
The time code generator signal provides the timing pulses
for the sampling and also indicates Universal Time (U.T.).
The sampling of the signal was accomplished by a 12 bit
analog/digital converter (A/D) and a KIM I microprocessor.
Pulsar pulse averaging is carried out by a computer program
which averages consecutive periods of the sampled pulsar
signal, previously stored in the Amdahl computer.
5

6
2.1 Data Acquisition of Pulsar PSR1133+16
The pulsar PSR1133+16 was observed for 20 nights
between April 25, 1979, and June 8, 1979, using the Uni
versity of Florida 26.3 MHz large array, located at the
University of Florida Radio Observatory in Dixie County.
The 26.3 MHz array is a nearly filled rectangular array of
2
640 half-wave dipoles that covers about 30,000 m The
half power beamwidth (HPBW) is 2.5 north-to-south by 6.0
east-to-west. It is a phase steered array with full N-S
directional control of the beam in quarter-beamwidth incre
ments, and a selection of 8 E-W beam directions spaced one
beamwidth apart. Phasing is accomplished by means of a
network of Butler matrices, hybrid rings and plug-in phasing
cables.
North-south phasing of the array for a source having a
given declination is accomplished in two ways, a coarse
adjustment in large increments of declination and a fine
adjustment providing small steps. A change in the coarse
phasing requires the replacement of many plug-in cables
located throughout the array, and requires about 2 or 3 man
hours of labor. Fine phasing for the desired declination is
quickly accomplished by means of switches at a single loca
tion. The beam pointing can be altered in small increments
to any declination within a strip roughly 10 wide by means
of the fine phasing controls. Positioning the beam to a new
declination outside this strip requires a coarse phasing
readjustment. The right ascension of the beam is of course

7
controlled by the rotation of the earth. Scans by any or
all of the 8 selectable east-west beams (designated 4E, 3E,
2E, IE, 1W, 2W, 3W, 4W) are made by starting each scan at
the appropriate sideral time. A computer program called
ARRAY developed by Dr. M. D. Desch, is used to obtain infor
mation on cable lengths required for coarse phasing, switch
positions for fine phasing, and beam transit time across the
desired source (11) A more detailed description of the
antenna and its adjustment has been given by Desch,(12) and
Desch et al. (13).
In order to use the maximum gain of the array, only the
central E-W beam positions 2E, IE, 1W and 2W were used, and
within each beam only the central portion of it was recorded
(v8 minutes each beam). A maximum of 32 minutes was re
corded each night. Pulsar PSR1133+16 was selected because
2 6 2 1
of its relatively high pulse energy (^0.75x10 Wm Hz at
61 MHz), long period (1.188 s), and low dispersion (DM = 5
-3
parsec cm ). It was also easy to reach this pulsar using
the phasing the antenna had at that time (phased to make
Jupiter observations), without having to completely rephase
the array. Only the fine phasing needed readjustment. During
the period in which the pulsar was observed, it was transiting
Only 8 nights (out of the 20 observed) gave reliable
data, free from static and stations. The end of the observa
tion period coincided with the beginning of the thunderstorm

8
season, at which time it was necessary to shut the array
down for the summer.
A block diagram for the interconnections of the equip
ment used to observe the pulsar is shown in Figure 2-1. The
equipment basically included a Collins R 390 receiver, a
Magnecord 1022 magnetic tape recorder, a Tracor 5D fre
quency standard, and a Systron Donner 8154 time code gener
The time code generator (TCG) was synchronized against
WWV; this was accomplished by triggering an oscilloscope with
the 1 pulse per second (1 PPS) output of the TCG and measur
ing the delay of WWV 1 PPS. The 1 PPS of the TCG was brought
as close as possible to the 1 PPS of WWV (usually between
5 to 20 ms) by making short interruptions of the 1 MHz signal
from the Tracor frequency standard. The delay between the
two 1-PPS pulse trains was measured to an accuracy of about
0.5 ms, and from that information the U.T. was obtained,
except for the propagation delay between Boulder, Colorado,
and Dixie County Radio Observatory. No corrections have
relate the epoch of this observation with any other observa
tions.
The measured drift of the Tracor time standard is about
80 ys/h; this causes a drift in the epoch of about 100 ys
between pulsar transits by beams 2E and 2W (^lh16m); this
drift is small enough to be noticed, considering the time

9
Figure 2-1.
Block diagram of the equipment
interconnections used to record
pulsar PSR1133+16.

10
resolution of ^31.5 ms (P/38; P = period) obtained for
this pulsar.
The detected IF output of the receiver was recorded
on channel 1 of the Magnecord at 15 IPS and the recording
level for the galactic background was set to about 7 VU on
the recording level monitor, to prevent any saturation during
signal peaks.
The time code generator was recorded on channel 2 of the
Magnecord. This signal serves a double purpose: it provides
a time reference for the pulsar signal, and later on, in the
data reduction process, provides the timing signal for the
KIM I microprocessor.
As a back-up to obtain the correct time, at the begin
ning of each observing session the WWV signals were recorded
on channel 1 (the TCG also on channel 2) for about 2.5
minutes. The correct time can be obtained by measuring the
delay between the 1 PPS of WWV and the 1 PPS of the TCG.
The receiver was tuned to a center frequency of 26.3 MHz
and the bandwidth set to 8 kHz. Using data for higher fre
quencies an extrapolated pulse width (at the base) of about
80 ms was obtained. Pulse broadening due to dispersion
(DM = 5 parsec cm ) is about 18 ms (for a bandwidth
Av = 8 kHz), which was considered acceptable.
2.2 Data Acquisition of Pulsar PSR2045-16
Pulsar PSR2045-16 was observed for 11 nights between
May 15 and May 30, 1979, using the Maip Radio Astronomical

11
Observatory 45 MHz large array, located about 20 miles
south-west of Santiago, Chile. The 45 MHz array is a filled
rectangular array of 528 full-wave dipoles with an effective
2 ...
area of 12,000 m It is a transit instrument with 6 inde
pendently phased subsections in the N-S direction and can
be oriented to a zenith angle of about 45 in the meridian
plane. The HPBW is 2.1 north-to-south by 4.6 east-to-west.
A multiple beam system is generated by a Butler matrix. A
more detailed description of the array has been given by
Reyes (14) and May et al. (15).
Pulsar PSR2045-16 was recorded for 20 minutes each night,
which is the time it takes to drift between the E-W 3 dB
points of the beam. This pulsar was selected because of its
relatively high pulse intensity (^1.0*10 2^Wm 2Hz 1 at 34 MHz),
long period (1.96 ms) and its low dispersion measure (DM = 12
-3
parsec cm ). A zenith angle of only 18 at transit was also
another favorable factor considered.
The equipment used for the data acquisition of this
pulsar consisted of a modified General Dynamics telemetry
receiver, a Magnecord 1022 magnetic tape recorder, a Systron
Donner 8220 TCG and a Rhode and Schwartz quartz time standard.
The interconnections are similar to those shown in Figure 2-1.
During the observing sessions the pulsar was transiting at
dawn, and no interference was noticed. Nine of the eleven
nights gave reliable data; the other two nights were discarded
because of equipment malfunctions.

12
A bandwidth of 10 kHz was used; this is the minimum
bandwidth of the receiver. Using data obtained at higher
frequencies, an extrapolated pulsar pulse width at the
base of about 130 ms was obtained at 45 MHz. The pulse
broadening due to the 10 kHz bandwidth is 11 ms. This was
considered acceptable.
The quartz time standard is checked daily against the
Chilean National Observatory cesium time standard; the epoch
is believed to be known with an accuracy of about +10 ys.
The drift in time is about 3.6 ys/h (or about 0.00190 ys/P,
where P = 1.96 s, the pulsar period). The drift is 1.2 ys
in the 20 minutes between the east-to-west HPBW. For all
practical purposes considered in this research, this is
negligible.
2.3 Data Reduction
Data reduction for both pulsars was performed using the
facilities of the Department of Astronomy of the University of
Florida in Gainesville. The basic configuration and inter
connections of the equipment for data reduction are shown in
Figure 2-2. The schematics for the most relevant blocks are
presented in Appendix C.
As stated earlier, the IF signal in the receiver is
band limited to 8 or 10 kHz and appropriately smoothed. The
resulting audio frequency signal is recorded in analog form
on channel 1 of the Magnecord tape recorder at a tape speed
of 15 IPS. The time code generator is recorded on channel 2.

13
HAZELT1NE AMDAHL 370
TERMINAL COMPUTER
Figure 2-2
Block diagram of the data reduction equipment

14
In order to achieve the desired time resolution and to be
able to sample the data at an appropriate rate, the magnetic
tapes were played back at 3 3/4 IPS, providing a slow down
of 4:1.
Pulsar signals from channel 1 of the Magnecord are sent
through a diode detector and smoothing filter to the 12 bits
A/D converter. The filter time constant was adjusted to 45
ms for both the rising and falling edges of the noise pulses.
The TCG signal from channel 2 of the Magnecord is con
nected to the input of the TCG (working now as a reader) to
provide both a visual display of the time and a 1 PPS pulse
(taken from a modified output). The latter gives the syn
chronizing signal to start the sampling of the data at a
precisely known time. It is particularly important to know
the time at which the sampling is started in order to be able
to combine in phase 2 or more blocks of data.
In order to obtain the timing signal for the KIM I, the
TCG signal from channel 2 of the Magnecord is connected also
to a phase locked loop (PLL) circuit. The PLL circuit locks
at a particular frequency of the signal, filtering out the
rest of the frequencies present in the complex wave form of
the TCG signal. The 1000 Hz signal of the TCG is transformed
to 250 Hz due to the 4:1 slow down. The PLL was carefully
tuned to 250 Hz, and the output then divided by 2. The resul
tant 125 Hz constitutes the timing signal for the KIM I.
A gated DC output from the PLL circuit was monitored on
a Brush Mark II chart recorder. This provides a good means

15
to detect any loss of synchronism or dropout of the TCG
playback signal. If there is any loss of synchronism
between the input and PLL signals, then the gated DC output
falls to zero, giving a visual indication to stop the
sampling process. If synchronization were lost, the
pulsar signals would be averaged out of phase, wiping out
any pulse at the end of the integration process. A means
was provided for starting the sampling at the next 1 PPS
pulse from the TCG occurring after a switch was thrown.
This made it possible to know the exact time at which sampling
started. This was accomplished by connecting the start switch
to the D input and the 1-PPS signal to the clock (c) input of
a D-type flip-flop. The output of the flip flop gates the
125 Hz signal in a NAND gate (see schematic in Appendix C).
When the start switch is turned on, the next 1 PPS of the
TCG enables the 125 Hz signal to go into the KIM I and the
sampling is started.
The sampled values (coded in hexadecimal) are sent
through the Hazeltine terminal to the Amdahl 470 computer
and stored there for further processing and analysis.
Since the maximum number of cards that can be stored
under a new data file name is 493, it was necessary to
break up the data into several "blocks" to meet this require
ment. The 8 minutes of data obtained in one beam for pulsar
PSR1133+16 were broken up in 2 blocks of 4 minutes each,
generating about 312 cards for each block (for a TSAMP= 02
and timing signal frequency = 125 Hz; see Appendix B). The

16
20 minutes of data obtained each night for pulsar PSR2045-16
was broken up into 4 blocks of 5 minutes each, generating
about 390 cards for each block.
The method of breaking up the data into blocks proved to
be useful also because one can inspect the results of the
average of each block separately, making it possible to search
by eye for a possible pulse that might appear at the same loca
tion in the different blocks. It is highly unlikely that
interference transients would appear in the same location in
each of several blocks.
Monitoring of the DC gated output of the PLL circuit is
very useful since synchronism is sometimes lost when a tape
is played back. If that happens, then that portion of the
tape is skipped and a new block started.
The KIM I program which is necessary to digitize the
data is loaded into the microprocessor from a cassette tape
recorder.
2.4 Block Designation for Each Pulsar
Many blocks of data were generated in the data reduction
process. In order to avoid any confusion with the data and
to be able to distinguish quickly between blocks, a designa
tion scheme was developed for referring to a specific data
block. The designation format contains in abbreviated form
the pertinent information for that block. For pulsar
PSR1133+16, the following block designation was adopted:

17
P
1 X X Y Z
Number that designates the night
when the pulsar was recorded
Letter that designates the block
within the beam
Beam in which the pulsar was recorded
Designates pulsar PSR1133+16
Antenna beam designations are as follows:
2E
2nd
beam
east
of
transit
IE
1st
beam
east
of
transit
1W
1st
beam
west
of
transit
2W
2nd
beam
west
of
transit
In order to distinguish between the different blocks
within a given beam, a capital letter was used, usually A and
B. In case it became necessary to break up the data into
more than two blocks (i.e. due to loss of synchronism of the
TCG signal) C was also used.
The indicator for the date of the night on which the
pulsar was observed was an integer, as follows:
Night Date
1
2
3
4
5
6
7
April 25, 1979
April 27, 1979
April 28, 1979
April 30, 1979
May 20, 1979
May 25, 1979
May 26, 1979
May 30, 1979
8

18
The designation scheme for the blocks of data from
pulsar PSR2045-16 is slightly different, as follows:
P 2 0 4 5 X Y
Number that designates the night when
the pulsar was recorded
Letter that designates the block
within that night
Designates pulsar PSR2045-16
The block designation is usually one of the letters A,
B, C, D, E or F.
The nights were numbered as follows:
Night Date

May
15,
1979
1
May
16,
1979
2
May
17,
1979
3
May
18,
1979
4
May
19,
1979
5
May
22,
1979
6
May
23,
1979
7
May
24,
1979
8
May
29,
1979
9
May
30,
1979
The reduced data for each block was stored on a disk
called UF.BO030601.S5.XXXXX, where XXXXX is the name of the
block. The data have been recorded also on a digital magnetic
tape called PULSAR.

19
2.5 Computer Programs for Data Reduction
In order to analyze the data and to plot the values, a
set of two computer programs called PULSAR and MCAPGM are
used. The basic parameters necessary to analyze the data must
be given to the PULSAR program; they are:
N = Number of bits digitized (=12)
TPP = Timing pulse period (ms)
NPA = Number of pulse intervals averaged
P = Pulsar period (s)
PLOFF = Plot offset
MAG = Plot magnification factor
The MCAPGM program (using the PULSAR program parameters)
takes the first group of pulse intervals to be averaged (NPA)
from the data of a particular block and calculates the over
all average and the standard deviation. These two quantities,
and the average for each bin, are plotted. The program next
takes the second set of intervals from the data and does the
same as before. Also the first set of intervals is averaged
with the second set, computing a new average (cumulative) for
each bin, an overall average, and a new standard deviation
and plots the values for the bins. This process continues
including each time the next group of intervals (NPA).
The first set of programs is run once only to check the
quality of the data,to detect any problem and to obtain the total
overall average for a block which will be used in a second run.
The second set of programs called PULSITO and MCAPGM1
is similar to the already described PULSAR and MCAPGM, but

20
this time, the data is taken from disk (where it has been
previously stored), produces a magnified plot and punches
cards with the final values for each bin, the total block
average and the final standard deviation. This set of cards
is used later on, to stack 2 or more blocks.
Usually after the process already described, the digi
tized data is recorded on the PULSAR magnetic tape and
released from the data files.
In order to stack 2 or more blocks, the sets of cards
of each block (except the first, which is used as reference)
is shifted the proper number of bins so that the bins can be
averaged in phase. The number of bins to be shifted in each
block is given by the following relationship:
N
bins
(2-1)
where = Number of bins the block i needs to be shifted with
respect to block 1
T^ = Time start sampling block 1
T^ = Time start sampling block i
P = Pulsar period
P
R = Time resolution (= r r-. )
Number bins
Frac = Fractional part
T. -T
The fractional part of ^ corresponds to the fraction
of the period between the time of beginning of block i and
the time of beginning of the first block. The relationship
given by 2-1 was evaluated using an HP-29C calculator.

21
A last program called P2045 (or P1133) reads the cards
of the blocks to be averaged (previously shifted), calculates
the average for each bin (weighting by the number of periods
in each bin), computes the final average of the combined
blocks, the standard deviation o, and plots the data. Also
plotted is the 3 o reference line, which is used as a reference
to aid in deciding whether or not a pulse was present at the
end of the averaging.
The KIM I program to digitize the data and the MCAPGM
program were developed by Dr. John P. Oliver; MCAPGMl is a
A program called DOPVEL was used to obtain the apparent
pulsar period for the night and time when the pulsar was
observed. This program gives the pulsar period at a particu
lar site and at a given time, correcting for Doppler shift
due to the earth's rotational velocity and its motion around
the sun. This program was developed at the NRAO by Drs. R. N.
Manchester and R. A. Gordon and made available by Dr. Steve T.
Gottesman.
Finally a summary of the times spent observing and re
ducing pulsar data and the total time of terminal use is
presented in Table 2-1.
Table 2-1. Summary of time spent with pulsar data.
Pulsar
Total Time
Observing
Pulsar
Total Time
of
Data Reduced
Total Time
Digitizing
Data
PSR1133+16
10h 40m
4
h
10
m
16h 04m
3
h
40
m
3
h
h
12
PSR20 4 5-16

CHAPTER 3
DATA ANALYSIS
3.1 PSR1133+16 Data Analysis
From the 8 nights of data reduced for pulsar
PSR1133+16 (26.3 MHz, Florida), pulses were found with
no ambiguity on 2 nights. Pulses were found on April 28
and May 25, 1979, over averages of 340 and 390 periods
respectively; the results are plotted in figures 3-1 through
3-7.
On April 28, after averaging over 340 periods (^6 44 )
a pulse appears in bin 33 with an amplitude of ^3 o above
the mean value (Figure 3-1). The 2 data blocks P112EA3 and
P112EB3 used to obtain this final block have been plotted
separately in figures 3-2 and 3-3. It can be seen that a
pulse appears in both blocks at the same location (bin 33).
Even though the pulse appearing in each block would not
alone be considered to be statistically significant, the
fact that they appear in the same location for two sepa
rated sets of data provides an excellent test of the validity
of the pulse.
On May 25, after averaging over 390 periods, a promi
nent pulse was obtained in bin 25 with an amplitude of
3.4 a; this pulse is shown in figure 3-4. The same test for
validity as before was made. Blocks P112WA6 and P112WB6 are
22

PULSAR PSR1133AI6 p ER I 00 1 1 879 9 9 SEC FRE CUENCY*26.3 MHZ
P112EA3 APRIL 23 7} T S T ART 02H 03M 03S STOP O2H OSM SOS TSANP 02
PI12LB3 APRIL 28 79 T ST AP T 02H O 5M 55S STOP 02M 1 OM OOS TSAMP 02
BLOCK AVERAGEs 120*161 SIGMAs 0* 06 9 3 SIGMA* 0*208 MAGNIFICATIONS 100
NUM'TEK r.F PLPCKSs 2 AVERAGING T IMF.* 39A.6 SEC
n!N AVFRAGF PERI OOS
NT" BER AVERAGES
1
120024
340
I
2
120.156
340
1
3
12 0.12?
340
1
4
123.197
340
1
5
120.197
340
1
6
120 .20 6
340
1
7
12053
340
1
8
1 20.1 74
340
1
o
120.221
340
1
1 0
1 23.1 29
340
1
1 1
1 20.1 26
340
1
12
120 .032
l ( 5
I
1 3
120.J 1
340
1
1 4
120.107
340
1
15
120.06?
340
1
16
i 2: .oo i
340
1
1 7
120.226
3 40
| TIME
19
120.1 74
340
1
1 9
120.200
340
1
? 3
1 2 0 1 3 8
3 40
1
PI
l 19.99 7
3 40
1
22
120.138
340
1
23
i20.29 1
340
1
24
120259
1 4 0
1
25
120.050
340
1
26
120.121
340
1
27
1 20 .094
340
I
20
1 20.224
340
I
29
12 0.050
340
I
30
1 20 1 4 i
340
1
31
120.100
340
1
2?
120.091
3 40
1
3 3
l 20 .4?7
340
1
34
120.273
340
1
35
i2C.ieo
340
26
120.173
340
1
37
120.097
340
1
29
1 2 0. 050
2 1 8
1
INTENSITY
I
I
I
I
I
I
I
33 3.0 SIGMA
I
I
I
I
I
t
I
I
I
I
I
I
I
I
I
NO
U>
Figure 3-1.
Pulsar PSR1133+16 (Blocks P112EA3 and P112EB3)
on April 28, 1979. Average over 340 periods.
recorded at 26.3 MHz

PULSAR PSP I 133+II P>riun = l Sf.C f HF UUtJNC V&263 MHZ
ll?CAl APRiL 28 70 T ST ART 02M 03M OOS STOP 0 ?H 0Â£M 50S TSAMP 02
R_CCK
A VCRAGE = l?C. 160
SI CM A r
0. 13 7
3 Sir,MA =
0.411
MACN IF Â¡CAT ION =
NU M JFR
OF MLCCKS= 1
AVIR AC.1NC
T I MC =
lfrP.5 SCC
=>I N
AVEPAGF
nF P l
A Vf W.
I
119,^16
1 4 0
2
120.107
1 40
3
120.193
1 40
4
120.021
1 40
5
12 C. 150
1 40
6
120.2?9
1 40
7
1 20. 3 1 4
1 40
3
120.150
1 40
9
120.307
1 40
10
120.093
1 40
1 1
120.064
1 40
12
12 0.100
1 40
1 3
120.193
1 40
14
l 20.0 36
l 40
15
120029
1 40
1 6
l20.143
1 40
1 7
120.257
1 40
18
120 .242
1 40
19
120.207
1 40
20
120.01b
1 40
21
120.029
1 40
22
120.357
1 4 0
23
120.326
1 40
24
1 20 .24 1
l 40
25
1 19.979
1 40
26
120.2 79
1 40
2 7
l 20 29
1 40
2 8
1 20 .329
1 40
29
1 2C .0 79
1 4 0
30
l 19.971
1 4 J
31
120.014
1 40
22
120.214
1 4 C
33
1 20 .4 79
1 40
24
120. 3 1 4
1 4 0
25
120.086
1 40
36
1 20 .007
1 40
27
120.107
l 40
38
l 20 .0 56
1 8
TIME
INTENSITY
NO
Figure 3-2.
Pulsar PSR1133+16 (Block P112EA3) recorded at 26.3 MHz on April 28,
1979. Average over 140 periods.

PUL SAP PSR I 1 33* 16 PERIf30=l 107999 SEC F RE OUENCV=26. J MHZ
p 1 1 2ED3 APRIL 28 79 T ST ART 02 H 05M 55S STOP 02H | OM OOS TSAMP 02
HLCCK AvEkAGE= l 1 9.ft?0 S1GMA = 0. 1 1 A 3 SIGMA= 0.342 MAGNIFICATIONS 100
NUMBER OF BLOCKS= l AVERAGING TI 232. 1 SEC
PIN
NUMBER
REfIUOS
AVERAGED
1
1 1 9 1 6
2 ro
2
1 IR.50
200
3
1 19 545
2 00
A
1 19. 7P C
200
5
l 19 69 0
200
6
1 19.65 0
200
7
1 l9 .6 7C
2 CC
6
1 19.650
2 CO
9
119.620
2 00
1 0
1 19.6 15
200
1 1
1 19.630
2 GO
12
1 19. 120
25
1 3
1l9.665
2 CO
1 A
119.770
2 00
15
119.5AS
2 CO
1 t
1 15.5 15
2 00
1 7
119.635
2 00
13
119.586
2 CO
19
119.655
200
?0
119.670
200
2 l
119.435
200
22
119.445
200
23
1 19. 720
2 CC
24
1 19.73 0
2 CO
25
l 19.560
200
26
1 1 9.4 7 C
2 00
27
119.390
2 00
28
119.610
200
2V
1l9.490
2 CO
33
1 l9.7? 0
2 00
31
1 l 9.6 ? 0
2 00
22
1 lv.465
200
33
l 19,650
200
24
115.705
2 CC
35
1 19.7 05
2 CO
36
119.7 6 C
2 00
37
119.550
2 00
23
1 19.5 10
200
TIME
INTENSITY
NJ
OI
Figure 3-3.
Pulsar PSR1133+16 (Block P112EB3) recorded at 26.3 MHz on April 28,
1979. Average over 200 periods.

PULSAP PSR1133M6 PEWIOP=l 188023 SEC FRE QUENC Y = 26 3 m HZ
P 1 1 2* A 6
MAY
25
79
TSTAP1
0 I H
"AOM
1 OS
STOP
0 I H
3 AM
10S
TS AMP
02
P112*B6
MAY
25
79
TS7 ART
0IH
3AM
1 OS
STOP
0 IH
38M
0AS
T SAMP
02
etOCK AVERAGE* 122.609
SIGMA
r
0.
08 1
3 SIGMA =
0.
2 A 3
MAGNIF [CAT ION*
1 00
NUMBER (JF BLC K S =
0 IN
NJMHE R
AVERAGE
1
122.681
P
122.68?
T
1 2 2 7 2 3
A
122. 6 1 0
5
l23.603
6
l22.595
7
122.613
n
l ?>. 592
9
l 2 ? 7 6 6
10
122. tA e
1 1
I 22.6 .3 C
1 2
l2?.530
1 3
1225A 3
1 A
122.5 1 0
IS
12?.589
16
1 22. 620
1 7
l22.505
l 8
122.607
IP
122.546
20
l22.ABA
2 1
122.651
22
i22.se?
23
12?.A90
2A
122.700
2 5
l2.BOA
26
122.728
27
122. 6 0 C
28
l2? 52C
1 2 2.5 1 5
30
1 22.597
31
122638
32
122.57?
33
122. 6 A3
7 4
122 .533
35
122.651
36
l 22.55 J
7 7
122.595
38
1 22.555
PENI00 Â£
AVERAGED
??A |
300 I
390 I
390 |
390 I
390 I
390 |
390 I
3 9 0 I
390 I
3 90 I
3 90 I
390 |
390 I
3 90 I
390 |
390 I
390 I
390 I TIME
390 |
390 |
390 I
390 I
3 90 1
390 I
390 I
390 |
390 I
3 90 I
390 I
390 I
390 |
390 I
3 90 I
390 I
390 I
390 I
2 16 I
2
AVEWAGING T1 ME =
AS?.7 S^C
INTENSITY
71
25 3.4 SIGMA
I
I
I
I
I
I
I
I
I
I
I
ro
03
Figure 3-4. Pulsar PSR1133+16 (Blocks P112WA6 and P112WB6) recorded at 26.3 MHz
on May 25, 1979. Average over 390 periods.

PULSAR PSR1 133M6 PER IODs 1 19O023 S EC FREOUENC Ys 26. 3 MHZ
P112WAO MAY 25 79 TSTART O IH 3DM IOS STCP 01H 34M IOS TSAMP 02
EL CCK
AVERAGE- 12?.CIO
SI GMA =
0. 105
3 SI GMA =
0. 315
MAGN IF I CATION =
NUMBER
O' BLCC KS= 1
AV ERAC1NG
T 1 ME =
232.2 SEC
C3 I N
A VE R AGE
PER 1OS
N ) M(JÂ£R
4V E^AGED
l
l 22.6<0
200
2
1 22.64 0
2 00
3
122.740
2 CO
A
122 .655
200
5
122. Et>0
2 CO
ft
i2? .e 15
2 CO
7
122.590
2 00
3
l22.6 i C
2 CO
i
l22.7?5
200
1 C
1 22 .630
2 00
1 1
122.530
2 CO
1 2
122.510
2C0
1 3
122.510
2 00
1 4
122.505
2 CO
15
122.495
2C0
i e
122.590
2 00
1 7
1 22 *56C
2 CO
13
122.565
2 CO
1 9
122.5 20
200
20
122.520
2 CO
21
l22.750
200
22
122.7Â£E
2 CO
23
1 22.57C
2 CO
24
l22.665
2 00
25
122.915
2 CO
25
122.770
200
2 7
1 22 .530
200
23
122.430
2 CO
2 i
122.475
2 CO
30
I 22.590
2 00
3 1
1 22.76 0
2 C 0
3 2
122.630
2 00
33
1 227 05
2 00
34
122 .61 *
2 CO
35
122.670
2 CO
36
l22.625
200
?7
1 22.455
2 CO
39
1 22.692
26
1
I
i
I
I
I
I
I
I
I TIME
I
I
I
I
I
I
I
I
I
I
I
I
I
I
INTENSITY
I
I
I
Â¡
i
i
i
i
i
i
i
i
i
fO
Figure 3-5.
Pulsar PSR1133+16 (Block P112WA6) recorded at 26.3 MHz on May 25,
1979. Average over 200 periods.

ELOCK AVERAGE- 121
NUM9CP CF BLOCKS *
PULSAR PSW 1 1 J 3 1 fe CRJOr)s 1 1 8 8 0 2 3 SEC F WE Q UCNC V= 26 3 MHZ
P112WB6 may 25 79 T ST AR T 01H 3AM JOS STOP 01H 3BM 04S TSAMP
800 SIGMAs 0* 105 3 SIGMA* 0.315 MAGNIFICATION*
1 AVERAGING TIME= 220.5 SEC
02
1 00
RI N
A VC c AGE
PFRIOOS
NJMQER
AV EuAGlO
I
121,792
24
1
2
121,916
1 90
1
3
121.895
1 90
1
4
121 .753
l 90
1
5
12 1 .637
1 SO
1
6
121.76?
1 90
1
7
121.337
1 90
1
3
121.763
l 90
1
9
1 2?.000
l to
1
10
12 ieoe
1 90
1 1
121 .92 t
1 90
J
12
12 1 .74 2
1 90
|
1 3
121.768
190
1
l 4
1 2 1 705
1 90
1
1 5
121.379
1 90
16
121 .642
l 90
l 7
121.621
1 90
1 3
121.842
1 90
1 0
l 2 1 .76 2
l 90
?C
121 .637
l 90
c 1
1 21 .737
1 90
2
121.589
1 90
23
121.595
1 90
24
121.926
1 90
25
1 22.042
l 90
26
121.874
1 50
27
121.863
1 90
23
121.305
1 90
2 i
121.747
1 90
30
l21.795
l 90
3 1
121.700
1 SC
3 ">
121.700
l 90
33
121 .763
1 90
24
121.637
1 90
35
1 21.3? 1
1 50
3b
121.668
1 90
37
12 1 .932
1 90
33
l 2l 726
1 90
TIME
INTENSITY
ro
CD
Figure 3-6.
Pulsar PSR1133+16 (Block P112WB6) recorded at 26.3 MHz on May 25,
1979. Average over 190 periods.

HL ZC K AV f ^ A jfr
r**M
r A ^ P
1
1J.G
1 <
* fi; jfu-| .
1 J7J i.
LC f
F! OUT NCY = 26
J MHZ
n 1 !
v.r a 7
4 AV
??
76
1 f. T 4 f* T
0 Oh
1 AM
*0f
STuP
0 0 H
1 7M
40S
TS-'MP
02
r i \
:e'7
'1 A V
?o
7
1 ST A-' 1
00M
1 7M
A Ob
rTi_p
00 H
?0M
OOS
TSAMP
0?
p l l
ItAT
AY
?
7c
TSTAtT
0 OH
A 4 M
Oi-S
S TCiP
00H
36M
0 6 S
T b A M P
0?
CM 1
1 tP7
AY
2 6
7 V
T ST AKT
0 0 H
7 4M
oe.s
3 I .ip
0 0 H
4 PM
ocs
TSAMP
0?
1 1
| h A7
* A Y
?l
7f>
T ST At- 1
3 IH
00 M
03!
STOP
Cl M
0 1 v
bOc
TSAMP
0?
PI 1
1 *0 7
AY
o *
7 v
T ST AC T
0 1 H
T/>V
30^
*T iJ>
01 H
OS M
OOS
TSAMP
02
n 1 1
l C 7
AY
?<
7 <>
7 ST AS T
0 l H
OSM
0 OS
ST UP
0 1 H
0 7M
tbS
T S AMP
0?
07
c
1 CM A =
3.
T4,?
1 si
GM A =
0
1 2<.
MACN
IF IC A T I ON =
10
NU y < p
(F LJC<
FIN
A VF_ A G?
PfK!70S
NU'* C
AV rc \ GI N
1
i 2b i ? r
'* 06
?
12: i 2o
0 3?
1 2r> 30
10 10
4
i 2 5.2 o e
10 10
j
l PS >^7
1 0 1 J
r
12r>. s*
1 0 10
7
1?S,'7
1010
1
12S.? 1 2
13 10
4
1 2b :S 4
IMS
I 0
1 *S. ? 3 1
10 10
11
123,19?
I 0 10
1 .2
1 PS ,75 0
1 0 10
l 3
12b. *A3
10 10
1 A
12b.726
1 3 10
I >
l ?* 1 Y?
ton
1 r.
1 2S. mo
10 10
1 7
1 2 b 5 S
10 1)
1
12b.157
*>70
I *
12b 150
6 14
0
126.1Jc
376
T 1
1 ?*; .10 7
MIC
22
l 2S ,77 7
1 J 1 0
3
12 b* 1 c 7
1010
?
l2 *4 C
1 1 0
12*. 1 M M
10 10

1 25 .222
10 10
7
I2i .*0-
1010
2
1 2 .
1 J l C
?S
12' ,7 0 7
1)10
2?
12i>. i o e
1010
7 1
12 b 1SS
1 0 10
3
l 25 75 3
13 10
3 7
1 2 5 1 f ?
10 10
*4
12b.; ?7
l 3 1 3
75
l ?r 1 So
10 1 0
J6
1 2 S 1 *> M
13 10
27
1 2 b 1 7 v
1 Cl 0
7m
12b.74 7
4 79
TIME
T
AVE'- AGINO T I ''c
il??.a scr
INTENSITY *
N3
Figure 3-7. Pulsar PSR1133+16 recorded at 26.3 MHz on May 26, 1979. No pulse
present.

30
plotted separately in figure 3-5 and 3-6. A pulse appears
in bin 25 in block P112WA6, with an intensity 2.6 a above
the mean value; also in block P112WB6, the pulse is weaker,
being only 2.3 a, but is at the same location (bin 25).
Once again we have the fact that 2 pulses appear in the
same location in two independent sets of data.
From a precise knowledge of the pulsar period and the
elapsed time between the April 28 and May 25 appearances of
the pulse, it should be possible to preserve its phase so
that it would reappear in the same bin on the second date
as it occupied on the first. However, this could not be
done because of the period uncertainty. This problem will
be addressed later when analyzing PSR2045-16 data.
The pulsar was detected in only one beam on each of the
two nights, disappearing when the beam was switched to the
next position. This phenomenon will be discussed later in
the conclusions, when polarization effects are considered.
Before other tests made on this pulse are discussed, it
is interesting to note that in bins 9 and 21 of block
P112WA6 (figure 3-5) two more small pulses can be distin
guished. Although neither of them alone would be considered
statistically significant, they both seem to appear again in
block P112WB6 (figure 3-6) in the same position. The pulse
in bin 21 leads the main pulse in bin 25 by 125 msec
(resolution ^31 msec/bin) or approximately 38 of phase.
The pulse in bin 9 is separated by 496 msec from the
main pulse, which is about 150 (or 0.42 P, P = 1.88023 sec).

31
The pulse in bin 21 suggests the existence of a
subpulse leading the main pulse. Several authors have
indeed reported for this pulsar a complex main pulse com
posed of 2 subpulses, the leading subpulse decreasing in
intensity with respect to the trailing subpulse as the
freguency decreases (6, 7, 8). The separation between the
two subpulses has been measured at higher frequencies and
extrapolation from that data suggests a much closer separa
tion for low frequencies than the separation obtained at
26.3 MHz between the pulses in bins 21 and 25. In 1972 Bruck
and Ustimenko (4) reported some observations made at 25 MHz
with a linearly polarized antenna; they found that this
pulsar shows a variety of pulse shapes with rapid transi
tions from one to another. In one plot, they show a pulse
composed of 2 subpulses and separated by ^148 msec. Despite
the poor resolution used (P/16 or ^74 msec), the 2 subpulses
are clearly defined.
The separation of the subpulses as a function of fre
quency has been plotted in figure 3-8, including the result
by Bruck and Ustimenko and our result (i.e. for the bin 21
and 25 subpulses). The plot suggests that either the pulse
separation increases very rapidly toward lower frequencies,
or the subpulse detected at 26.3 and 25 MHz is a new one
which appears only at low frequencies. In figure 3-9 the
average pulse profile obtained in this thesis is compared
with that obtained by Bruck and Ustimenko. The phase of the

2 00
-
150

THIS THESIS

BRUCK AND USTIMENKO (4)
too

CRAFT AND C0MELLA (6)
SEPARATION
A
IZVEKOVA et al. (7)
(ms)
-
o
IZVEKOVA et al. (8)
50
AND LYNE et al. (22)
40

A 0
30
A
o
0
20
o
1
1
.... i
1 1 1
lJ
L_
i i 1 1 1 i i
10
20
30 40 50
00
500
1000
u>
to
FREQUENCY
(MHz)
Subpulse separation as a function of frequency for the
pulsar PSR1133+16.
Figure 3-8.

20 -
BRUCK AND USTIMENKO (4), 25 MHz
OCTOBER 11, 1972
INTENSITY
(ARBITRARY
SCALE)
INTENSITY
(ARBITRARY
SCALE)
Figure 3-9. Comparison of the pulse profile for pulsar PSR1133+16
obtained in this thesis (average of 340 periods) with
that obtained by Bruck and Ustimenko (4) (average of
v212 periods). The bins of the 26.3 MHz pulse profile
have been shifted in order to align the pulses.

PULSA PSH1J1 It t~9 IQ'Js | i 8*J?3 SEC r P FO UE NC Y = 26 3 MM2
PI I 2 A6 MAY 25 75 T ST APT OlH 30M IOS STG OlH 3 AM IOS TSAMP 02
BLOCK AVERAGES 122.690
S I GMA =
0. 1 1 3
3 SIGMAr
0. J39
MAGNIFICATIONS 100
NUMBER OF BLUCKS= 1
AVFPAGING
TI ME =
162.5 SEC
t IN AVERAGE PrPJCOS
NUMBER AVERAGED
l
12?.714
1 4 0
1
2
l22.629
1 40
1
3
l22.657
1 40
1
4
1 22. 700
1 40
1
5
12? .57 1
l 40
1
122.593
1 40
1
7
122.636
1 40
)
B
12^65 D
l 40
1
9
l 22 .2 l
1 40
1
10
l22. 72 l
l 40
1
11
l 22.629
140
1
12
122 .57 1
1 40
1
1 3
1 2266 4
1 40
1
14
1 22.61 A
l 40
1
15
122.614
l 40
1
16
122.^79
1 40
1 I
1 7
122.693
1 40
1 TIME
1 Pj
122.05C
1 40

19
122.636
1 40
|
2 J
122.7a 3
l 40
1
21
l 22. 9 b 6
1 40
1
2?
122.914
1 40
1
23
122.664
1 40
|
24
l22.607
1 40
1
25
l22 .99 3
1 40
I
26
l22 .756
l 40
1
7
122.571
1 40
28
1 22 .5 36
1 40
1
29
122456
1 40
1
30
122.593
l 40
1
31
122.756
l 40
1
32
122. 736
1 40
1
*3
122.664
1 40
1
34
l22664
1 40
1
25
122.779
1 40
I
26
122.72 i
1 40
1
37
122.579
I 40
1 |
38
123. 00 0
i e
|
TNTFNSITÂ¥-
2.62 SIC^
I
I
I
2.68 sicm
OJ
Figure 3-10.
Pulsar PSR1133+16 (Block P112WA6) recorded at 26.3 MHz on May 25,
1979. Average over 140 periods.

35
pulse profile at 26.3 MHz has been shifted, to align the
pulses in bin 21 and 25 with the pulses at 25 MHz.
Although the subpulse in bin 21 is much weaker than
that in bin 25 for the 390-period average, the two are
almost the same strength in the 140-period average shown in
figure 3-10 (block P112EA6). In the latter case, the
heights are 2.68 a and 2.62 a for the bin 25 and 21 pulses,
respectively. A possible implication is that the bin 21
pulse varies in intensity much more than does that of bin 25.
A small pulse appears in bin 9, when blocks P112WA6 and
P112WB6 are averaged together. This pulse is also present
in each block average separately (figure 3-5 and 3-6), sug
gesting the possible existence of an interpulse leading the
main pulse by 0.42 P (or 150). In Bruck and Ustimenko (4)
a broad and poorly defined pulse seems to appear at 0.5 P
from the main pulse, in the same plot where the double main
pulse appears(figure 3-9). So far, no interpulse has been
reported for this pulsar.
3.1.a Peak Flux Density, Mean Flux
Density and Energy for Pulsar PSR1133+16
Several continuous radio sources were observed in order
to calibrate the array, to obtain the minimum detectable
flux density, and then to obtain an estimate of the peak
flux density for pulsar PSR1133+16.
The radio sources observed and their main parameters are
presented in Table 3-1.

36
Table 3-1. Radio sources used for calibration purposes.
R.A.
Declination
Flux
Density
(Jy)
Antenna
Beam
Used
3C310
15h02.9m
2 6 15'6 "
382
1W, 2W
3C315
15h11.7m
2616'8"
133
1W, 2W
3C409
20h12.2m
23 25115
381
IE
3C433
21h21.4m
24 54'6"
267
1W
Positions and flux densities were taken from the Clark
Lake Survey at 26.3 MHz (16). Of these four sources, only
3C409 and 3C433 were used in the calibrations; the 3C310 and
3C315 measurements were discarded because their difference in
R.A. is only 8.8 (corresponding to 2.2). Since the HPBW of
the radiotelescope is 6.0 east-to-west, these two sources
cannot be resolved.
The parameters of the radiotelescope for observing the
calibration sources were:
Bandwidth (Av) = 250 kHz
Time constant (t) = 1.5 s
Frequency (v) =26.3 MHz
The minimum detectable flux density S deduced from
min
calibrations using these sources, is presented in Table 3-2.

37
Table 3-2. S and Galactic background temperatures
min ^
deduced from calibrations.
S .
min
(Jy)
Galactic background
temp. K
3C409
101
38,975
3C433
97
27,840
The galactic background temperature for pulsar
PSR1133+16 is:
T, = 18,514K
backgr.
All the galactic background temperatures given are actual
brightness temperatures, corrected for losses in the antenna.
A Hewlett Packard noise generator was used for the cali
brations, and a pair of type 5722 noise diodes served as the
standard of comparison for the noise generator. Details of
the calibrations are given in Appendix E.
The minimum detectable flux density has the following
expression (17) :
AS oc
min
x t
syst
(3-1)
nAvi
where n = number of averaged records
Av = pre-detection bandwidth of the receiver
T = post-detection time constant
Tsyst = sYstem temperature
In this case, the temperature of the system is completely
dominated by the galactic background temperature, and in

38
equation (3-1), it is possible to substitute TSySt by:
T = T =: T
syst galactic background
Assigning subscripts p to the pulsar and r to a con
tinuous radio source, it is possible by using equation
(3-1), to write two similar expressions, one for the pulsar
and the other for the radio source. By dividing one by
the other, the following expression can be obtained:
AS .
mm p
n Av
r
n Av
P
T
r r
T
P P
T
_R
T
r
S .
min r
K
(3-2)
The constant K takes into account that the pulsar sig
nal, instead of being continuous, is being sampled.
Using the values obtained for the pulsar on May 25, 1979
n^ = 390 periods (453 s)
Av = 8 kHz
P
T = 12.5 ms
P
T = 18,514 K
P
and the values for the radio source of:
n =1
r
Av = 250 kHz
r
Tr = 1.5 s
ASr = 101 Jy
Tr = 38,975K
the following value is obtained:
Smin p = 149 Jy (considering K = 1).
To obtain the peak flux density of the pulsar, it is
necessary to make several assumptions:

39
A. In order to be consistent it has been considered
that the value S for both the continuous source and the
mm
pulsar is that which would cause a deflection of 3 a (i.e.,
3 standard deviations) from the mean value of the galactic
noise background. The 3 a criterion was used to determine
whether or not a pulsar pulse was present at the end of an
integration run.
B. It is assumed that the sample rate is fast enough
compared with the post-detection time constant that very
little information is lost in this process; then K can be
considered 'll. (Details about sampling rates can be found
in Appendix B.)
C. It is assumed that the pulsar signal is unpolarized
as in the calibration source. This is probably not correct;
a strong linearly polarized component has been reported (18)
The effect of variable Faraday rotation in the terrestrial
ionosphere would then be to cause the plane of the linearly
polarized component to be oriented (at different times) at
any angles from parallel to perpendicular with respect to
the antenna dipoles. As a result, if the polarization is
actually 100% linear, the value of as given by (3-2)
may range anywhere from too low by a factor of two to too
high by a factor of infinity. Thus, no signal at all would
be detected if the linear polarization is perpendicular to
the dipoles, while according to (3-2) S would not be
min p
zero.

40
By making these assumptions, once the pulsar reaches
a value slightly higher than 3 o at the end of an averaging
process, the peak flux density can then be considered equal
to S^n (calculated for the corresponding number of periods
averaged). Therefore,
S = S = 150 Jy
peak p mm p 1
Pulse energy is usually defined as:
rp
E = S(t) dt (3-3)
P o-
where P = pulsar period
S(t) = flux density as a function of time
For pulsars with no interpulse, the limits of the integral
are restricted to the duration of the pulse (base time dura
tion) .
In order to make the computations simpler, it is usually
assumed that the pulse has a relatively simple geometrical
shape. Izvekova et al. (7) assumed a triangular shape for
most pulsars (including PSR1133+16). By looking at the pulse
shape obtained on May 25, 1979, one can see that this is a
reasonably good approximation, considering only the pulse
centered at bin 25.
Using this approximation, the pulse energy in equation
(3-3) can be written as:
E
P
where W,
base
S w,
peak p base
= pulse width at base,
(3-4)
corrected for dispersion

41
Substituting the values of:
S = 150 Jy
peak p 1
W, = 71 ms
base
the following value for E is obtained:
P
E = 5.3xl0-26 Jm-2Hz_1
P
The mean flux density of the pulse is defined as:
S = E /P (3-5)
mean p
Using the value obtained for E and the value for the
P
period P = 1.188 sec, the following value for the mean flux
density is obtained:
S = 4.45 Jy
mean 1
It is interesting, now, to compare these values with
values obtained by other authors.
In 1973, Bruck and Ustimenko (4) observed the same
pulsar at frequencies of 16.7, 20 and 25 MHz using a linearly
polarized broadband array and reported peak flux densities in
the range of 10 to 40 Jy.
In 1968, Drake and Craft (19) observed the same pulsar,
using the Arecibo 1000-foot radio telescope at frequencies
of 111.5 and 195 Mhz and reported peak flux densities of
150 Jy at 111.5 MHz. Since these observations were made at
a much higher frequency, they cannot be directly compared
with the one at 26.3 MHz, nevertheless, the value obtained
at the two frequencies gives an idea of the order of magnitude
for the peak flux densities. A value of 38 Jy can be deduced
from the data at 61 MHz given by Izvekova et al. (7).

42
Regarding the pulse energy and mean flux density, more
recently, in 1979, Izvekova et al. (7), have given some
results at 61 MHz. They obtained a value for the pulse
26 2 1
energy of 0.89x10 Jm Hz and a mean flux density of
~26
0.75x10 Jy, using a linearly polarized antenna.
In 1971, Slee and Hill (9), using a circularly polarized
antenna, reported at 80 MHz a value for the pulse energy of
26 2 1
2.2x10 Jm Hz for one night (with a median over 6 nights
of l.OxlO-26 Jm_2Hz-1).
It is not easy to compare the values obtained in this
thesis with values obtained by other authors, because of the
polarization used and the different averaging times (not
always given). Nevertheless, the value of pulse energy
o r __ o __ i
E =5.3x10 Jm Hz obtained at 26.3 MHz, looks at least
P
26 2 1
comparable with the value of E = 2.2x10 Jm ^Hz obtained
P
at 80 MHz (which also corresponds to a peak flux density of
^110 Jy).
According to Izvekova et al. (7), PSR1133+16 has a
maximum in its average spectrum at v = 10040 MHz.
Kuz'min etal. (20), making simultaneous observations at
several frequencies, found that the instantaneous spectrum
changes from day to day. For several days it doesn't show a
maximum in its spectrum but rather the intensity seems to
increase towards lower frequencies.
This can be considered as part of the explanation of
why the peak flux density, the mean flux density and the
pulse energy obtained at 26.3 MHz are higher than the values

43
10
ENERGY
(Jm"2Hz_1)
*10
-26
1.0
.5
.4
.2

this thesis
SIEBER (29)
A SLEE AND HILL (9)
BRUCK AND USTIMENKO (4)
o IZVEKOVA et al (8)
30 40 50
FREQUENCY (MHz)
100
200
Figure 3-11.
Low frequency energy spectrum for
pulsar PSR1133+16.

44
obtained at higher frequencies: the pulsar seems to show
abnormally strong pulses at low frequencies during some
particular nights, as it may have done in our case.
Another interesting aspect of this has been pointed
out by Ekers and Moffet (21). They observed PSR1133+16 at
a wave length of 23 cm (^1300 Mhz). From a histogram of
energy and number of pulses with a given energy, they noticed
a long tail in the distribution, towards the high energy end.
The ratio of the mean value (0.020x10 ^ jm lj tjie
2 6 ~2 1
highest energy pulse (^0.090x10 Jm Hz ) is about 4.5.
26 1
If one takes the value for the mean energy of 5.3x10 Hz
at 26.3 MHz as an average over a relative short period
('i390 P) and compares it with the calculated mean energy of
26 2 1
1.6x10 Jm Hz obtained from the observations of Bruck
and Ustimenko at 25 MHz (4) a value of 3.3 is obtained for
the ratio. Such a ratio is not improbable according to the
distribution function of Ekers and Moffet. Pulse energies
obtained reported in this thesis and by several other authors
are plotted as a function of frequency in Figure 3-11.
3.1.b PSR1133+16 Pulse Width
Several factors can distort the actual pulse shape as
observed; the factors that need to be considered are:
A. Effect of the dispersion over the finite bandwidth
B. Effect of the post detection time constant of the

45
C. Polarization of the pulses and antenna.
The first two factors will make the pulse wider than the
actual pulse, but their effects can be corrected for if the
pre-detection bandwidth, post-detection time constant of the
telescope and the dispersion measure for the pulsar are
known; a simple form needs also to be assumed for the pulse.
In order to see the effect of dispersion let us con
sider the basic formula for the time of propagation through
a plasma with a dispersion measure DM at a frequency v.
The time of propagation is given by:
t = kDM ~2 (3-6)
v
where k is a constant depending on the plasma.
Differentiating this equation with respect to v, taking
finite At and Av (instead of dt and dv), and expressing the
parameters in consistent units, the following expression can
be obtained:
At = 8.3x10_3DMAv(v0)~3 ms (3-7)
_ 3
where DM = dispersion measure (parsec cm )
Av = bandwidth (kHz)
vQ = center frequency (hundreds of MHz)
This relationship gives the time it takes for a
dispersed pulse to sweep through a bandwidth Av centered at
frequency vQ.
If a triangular shape for the pulse is assumed, (3-7)
is transformed into (3-8) for the half intensity level:
AtQ = 4.15x10-3DMAv(v0)
ms
(3-8)

46
T = O.6 93t
c
The correction due to the time constant of the receiver
(for the half intensity level and assuming a triangular pulse
shape) is given by (8):
(3-9)
where x is the post detection time constant of the receiver.
No correction can be made due to the effect of polariza
tion but some discussion about its effect will be given in
the conclusions.
Consequently, the corrected half intensity width of
the pulse is:
W. c = At r T (3-10)
0.5 0.5 ob 0.5 c
where WQ ob is the observed half intensity width.
Substituting in equations (3-8) and (3-9) the values
DM = 4.8 pc cm ^
Av = 8 kHz
vQ = 26.3 MHz
and T = 11.25 ms,
the following values are obtained for the corrections:
Atg = 8.8 ms
T =7.8 ms
c
Table 3-3 gives the values for the observed half
intensity width and the half intensity corrected width for
the pulses obtained on April 28 and May 25, 1979. The pulse
width at the base is the width assuming a triangular shape,
and is twice the corrected half intensity width.

I 00
PULSE
WIDTH
(ms)
50
A
O
THIS THESIS
O CRAFT AND
COMELLA (6)
A IZVEKOVA et al
(18)
40
O
30
20
_i I I I I I I 1 1 I I L
20 30 40 50 100 200
FREQUENCY (MHz)
500
Figure 3-12.
Pulsar PSR1133+16 half intensity pulse width as a
function of frequency.

48
Table 3-3. Observed half intensity width, corrected
half intensity width, and pulse width
at base, for pulsar PSR1133+16
Date
WA c observed
U D
Wq corrected
W,
base
(ms)
(ms)
(ms)
April 28, 1979
38
21
42
May 25, 1979
53
36
72
The graph of Figure 3-12 is a plot of the values of
the corrected pulse half-intensity width at several fre
quencies (6, 8), including the values of this thesis
obtained at 26.3 MHz on May 25, 1979; this is the more
reliable of the two values presented in Table 3-3.
3,l.c Tests Performed with the Pulse
Obtained on May 25, 1979
An independent and simple method for testing the pulses
to see if they are indeed from a pulsar having the assumed
period is to repeat the averaging process using both shorter
and longer periods than that given by the DOPVEL program.
If the observed pulses were spurious ones which happened to
have approximately (but not exactly) the same period as that
of the assumed pulsar, then the averaged intensity should
either increase or decrease for the shorter period, and do
the opposite for the longer period. If the averaged
intensity is a maximum at the assumed pulsar period, then

49
it is almost a certainty that the pulses are indeed due to
the pulsar.
May 25, 1979, the date on which the best pulse was
obtained, was chosen to make the test. Data were analyzed
using periods longer and shorter than the assumed period
by 40, +80, 160,1320, and 640 ys.
Periods shorter or longer than the assumed period P
by 80 ys should cause a shift of 1 bin at the end of the
390 period average.
The result of this test is shown in Figure 3-13, where
the relative averaged pulse intensities have been plotted
as a function of the assumed period length, taking as the
base the periodP = 1.188023 s given by the DOPVEL program.
It can be seen that the pulse nicely peaks with the assumed
pulsar period, and decreases sharply for shorter and longer
periods. The lack of symmetry of the plot might be due to
the unsymetrical pulse shape.
The computer plots of the results for the different
periods used in this test are presented in Figures 3-14 to
3-24.
The pulse intensity as a function of time has been
plotted in figure 3-25 for the pulse obtained on May 25,
1979. This pulse was chosen because it was the better of
the two and is well defined. This plot is useful not only
to visualize the temporal behavior of the intensity of the
pulse but also to check that no interference or isolated
transient pulse is mistakenly taken as a pulsar pulse.

300
200
RELATIVE
INTENSITY
OF
AVERAGED
PULSE
(ARBITRARY
SCALE)
Ul
o
100
_L
X
X
JL
X
X
X
X
X
X
-800 -600 -400 -200 p +200
(ys) AVERAGING PERIOD
+ 400
+ 600
(us)
+ 8 00
Figure 3-13 Relative pulse intensity as a function of the averaging period for
pulsar PSR1133+16 on May 25, 1979. P = 1.188023 s.

3 J.
/ AF 3 ; 5 l m
is =nn- i. im7
irt.A *.*-:* -
p: qu" Nr y* ? *.
3 *17
p 1
1 ? W A 6 mAY 25
79 7START 01H 3)M
1 OS STCP
01H 3AM IOS
T S A M P
02
p 1
i?*r)6 v'av
79 T ~ A? 0 1 H 3 A
1 OS STno
01 -1 38M 0A S
TSA MO
02
fl-DC< AV*r?AGE 1 2?.
609
ST v A r
C- ? 6 3 S J r.MA
= 0*?0*
MAGNiriCAT1HN=
1 00
NJH trc of MLOCKS-?
2
AVCFAGING
T | ME- A,2.2 r
N AVCBA SC HFOIOOS
SI !(|C5
4 VC RAG'S I*
1
1 ? so 7
3 o)
2
122 A 4 3
70)
1
122.55A
3 90
A
I 2 ? a 0 0
5
12? s a
2 ? 1
6
1 22.579
70)
7
l ?. 656
3 C)
0
12? S 9
30?
9
12?.68?
3 50
1 3
1 2?. A 7
3 9)
1 1
12?
3 90
1 2
1227?8
a 90
1 3
390
1 1
t 2 ? S 5 Q
*90
15
122.564
3 50
IS
12? .MS
3 9)
1 7
l ? .61 f
3 93
1 8
12? ,697
30)
IQ
1 22.58?
3 50
0
122.S4 1
39)
2 1
1 22.587
3 9)
r
12?.S4 |
390
23
\ 09 *,7"*
30)
2 A
122 600
1 QO
?5
1 22.6 a?
30)
1 ?> 5 |
30)
?7
12? s??
3 90
? J
122.579
390
n
1 7 q a
7 C)
30
i 22. A55
3 50
* 1
1?? .so?
3 on
V
12? 7 7 3
3on
3 3
\2? 650
3 f>)
"* A
122.607
300
7 7
122,515
70)
36
12? 636
3 50
7 f
1 ?? .63.3
390
3*3
l22 536
2 1?
TIME
INTENSITY
Figure 3-14. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187383 s (P-640 us).

*jlcat 3'ri t uih **3r-u i7702 see fptour.NCV?26 3 HZ
M 1 ? *46 v A Y
26
73
T5T Act
0 1 H 3 3 m
1 OS
STOP
01 H
TAM 13S TSAMp
02
* 11 y
25
70
7 r-T AO *
C IH 34
t OS
STOP
01 -
3M ORS T?AVD
0?
OL 0C< AVFFAGF= 122.609 SIGMA
o.
C53
3 s T G M A s
0
1 7?
MAGNIFICAT ION*
1 00
m 4rc r>* nt odes-
VF- AGING T 1 **F =
46?.6 SFC
* f N
A Vr C A F
P CC1OOS
N *ME3
A V F r A G r
1
| ?? ' 3 9
1 90
2
12? S *3
3C?
1
1 2? f>6*
7??
A
l 29 ? 71
3 93
S
12?
1C3
6
1 22.538
3 SO
7
l??.661
VO
9
12? *74
393
l?7,5in
ico
l 0
1 ??.' 3
109
11
12? 555
303
1 ?
122.677
3 CO
1 3
1 ? ^ r.
3 0
1 A
12? 554
303
1 S
l ?? 54 8
-C3
1 6
P?."31
3 03
1 7
12? 548
3 CO
1 5
l 2? *5.3 a
3 CO
!Q
l. 5 n 2
ico
2 '
12? 61
3 CO
"* 1
12?.5 1 7
700
?*>
122 559
300
23
12? .605
1 CO
?4
1 2? .5 7 A
3C3
T*c
12? 5 A 3
303
26
1 2?.565
3 CO
*7
1 2*.7?e
303
8
I? 5SO
303
?C
1 22.650
3 CO
?0
I * .7 0
3 93
1
12? 641
3C3
T *
1 22.6 87
303
3 1
1 2 ? 6 l c
3 00
A
12? *20
3 CO
35
1 22.685
3 CO
3*
127.606
3-0
17
122 S?5
ico
?**
12?.546
? 1 A
TIME
INTENSITY.
(J1
to
Figure 3-15. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187702 s (P-320 ys) .

9 JL r An f:ni 1
1 3 | 6 |. 117B63 0FC F R f OUE .*C Y = 2 6,3 HZ
P l 1 2 A 6 '* W
25
79 T5T A> T 0 |H 39 M
05 STOP 0 1 H ,3 AM 10S TSA*P
7 9
>112456 MY
2 r.
7 9 r-,TA" DIM 34 v
0C STOR 01H 3RM 04S TSAVP
0?
OLOrK A /ERAGE-
l 2?
609 sr *MA
-
? 0 7 7 1 6 1 *.M4s
0-209 MAGNIF1C1T 103 =
too
Mj'nrr
OF BLO?
K 5 c
9 AVE P AG I j G
T1 ME 45 7.6 Sr~c
A V r d A r, e
n nr ?
4 vn. AGE
r>
i
1 ?-* ,6 5ft
107
1
1
?
1 22 *' 1 A
7 7 i
1
< 1
1
3
1 2? 6.59
300
1
1 +- _

1
a
122 7?A
190
1

1
5
12? A|i
193
1

1
f,
122 565
9 90
1

1
7
1 ?** .**5 1
7 07
1

1
9
12? 60?
loo
1
1
9
1 22.649
3 00
1

I
t 7
1 22 54 i
|OQ
1
i
1
1 !
12? 554
1 o^
1
1
1 2
l22,574
1 90
1
1
1 3
l
3 00
1
1
1 4
122 554
303
1
1
1*3
1 22,600
300
1
1
1 6
1 22,4 70
lrO
1

l
1 7
12? 637
3 93
1
-v.

1
* ^
1 2 *.*'' ^ |
10?
1

1
1
122 5 76
300
I
*

1
27
122 554
16 0
1 TIME
7.
1
1
1 ?'',= 79
300
1
1
o 7
1 22 *6 7
390
1
> 1 1
1
23
122,635
7 07
1
*-
1
*>6
1 2? .5 >?
390
1
d.
1
2*3
1 2 * 666
193
1
f ~ -*
1
2*>
1 22.9 1 8
1 00
1
26
1
r>7
l ?*> .697
103
1
-T-
1
? 3
1 77 577
393
!
I
?-
122.50?
7 99
1
1
>
1 2 7.600
3 90
1
1
3
I?? 541
190
1
1

1
32
1 22.71 0
3 00
1
i
*
I
t
127,559
3 90
1
1
34
12? 3 3
100
1

1
35
127.564
390
1
> *
1
16
12?.653
390
1
/ Â¥
1
* 7
12? e 50
190
1 -

1
39
122.660
2 1 5
1

1
Figure 3-16.
Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187863 s (P-160 ys) .

PUL 3 AO P
C-W 1 l
1.1
1 6
>PP| nO = l
, 1 0 7 P A 1
S5C
FRrQJFNCY=2fc3 MHZ
"Mir* a*-
MAY
<;
5
- TAP*
o 1 H
? r.i i oc
S
01 H 1AM 1 OS TSAMfJ
02
ntl2wn6
M\Y
?5
7?
tgt apt
0 1 H
14M IOS
STOP
01 M 104 S TSAM3
02
BLOCK AVERAGE* 1 2? .6 09
1 GMA
r
IT >
7 r
I r,M a = c
.217
MAGNIFICATlONs
1 00
NUMBER CF BLOCKS? ? AVF5 AGING T: *tr: A3?- 6 s^c
P 1 N
AVFPAGf
3 p 51 o o r.
n j Mirn
AV EP AGED
i
1 22 6 5 A
224
7
l ?2 4 1
3 90
12? TP
10
A
122 339
190
122574
iso
'
1 ?? M 1
190
7
1 22 6 2 5
190
1 22 .' 1 2
3 90
9
1 2? *.oo
190
1?
122.646
1 OQ
1
122,>C1
190
1 ?
122 es
190
13
122.564
390
11
1 22 .S2 0
390
1 5
122 6 11
3 90
1 3
1 22 561
1 90
1 *
190
I 8
1 22 5 7
190
19
122.512
190
20
122.556
ion
r i
122.637
190
122554
190
23
P?|B
19?
2A
122.692
190
PI
1 ?? H 1
3 90
1 ?2 uo
ion
27
122.62*
1 90
.
l 22 -- 1 2
39 3
2 )
1 2 * 4)7
-on
3 0
122. 649
190
1
I 2* ,*-#
150
1 7
1 22 * 25
190
31
12?6*9
190
t t
122.525
390
3-
t 2.? 6 ?
190
33
122 587
1 90
v
1 22 .64 6
3 0
3H
122 6P3
2 15
TIME
INTENSITY
Ul
Figure 3-17. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187943 s (P-80 us).

"NIL
6 4 n PSW 1 1 33! 6
vnirnai.nmi >rr r u- ojescy-?6* 3 mhz
r> 11 ?M 46
MAY *r, 70
r
TA9- 01M 70M 106 STH9 01H 34M IOS T5AMD
0?
nl 1
2WH6
MAY ?5 70
T5
T A 9 7 0 1 H 7 AM 10S 6 T O7 01 H 3M 04 S T 54 w 9
A2
Ow oc*
4 vrc A r,**
12?
60 9
f
i gma* o.
? 7 M
3 6|5MA* 0.214 MAGNIFICATinM-
1 0 0
N)MHCP
OF BLOCK
6-
2
Avr
'AGING 7 I
Mr
46? 7 SrC
N M-*FP
AV cc AGED
1
12? 733
2?A
>*

1

1 >*.' 7 4
3 00

1
T
l
703
~
1
4
1 2? 50?
3 09
.
1
1 ? ? 6 9
390
. *
1
6
12? *0*
*90
1
7
1 22 618
* 90
y
1
12? .* 1 5
3 90
* *
1
o
12? 7 7 5
790
~~
1
l '
1 22.6?n
390
*
1
1 1
122.623
7 90
1
1 *>
122.579
3 90
*
1
1 7
l ?? .561
3 90
v
1
1 4
12?.^?3
390
* *-"tr~
!
1 5
122 6 | 7
10
1
1
1 22.5 79
790
, <^T
1
1 7
122.620
390
1
1 8
1 2? .*= >?
jco
,
1
1 9
1 2? MH
3 r0
1 TIME
V- *
I
"'
1?? .523
3 90
, *
1
1
12? 578
390
~ *
1
2?
122.620
3 90

1
" t
122.4 70
390

,
1
?4
122 71
J 9 0
--
1
25
1 22 0 31
3 90
.
25
i
2-
12% 7 33
3 c0
, ~
i
27
122 *73
390
1
12? 4 9
3 90
^
1
> ,
122.497
3 90

1
7 0
t ? 7 7 3 7
190
~~ <_
1
3 l
122 636
3 90
*

1
V
1 ?* ' 1 2
7o7
.
1
3 3
12? 643
390
,
\
3 a
122.53?
7 90
.
.
1
75
1 2? .*69
390
.
1
36
1 2? ,64 3
390
,
1
37
122 579
7 90

i
3&!
l22.539
2 15
1
ai
Ln
Figure 3-18. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187983 s (P-40 ys).

PULSAR PSRI
1 33
16
PERIOP-1
1 8 M 0 2 3
SEC F
RE QUENCY* 26.
3 MHZ
P 1 I 2* A h M AY
25
7 9
T ST AC 7
0 1 H
10M
I OS
STOP
0 1 H
34M lOS
T S AMP
02
o112*BO MAY
25
75
TST APT
0 IN
34 M
1 OS
STOP
0 1 H
38M 0 AS
T SAMP
02
PL OCK
AVERAGE* 122
60S
SI GMA*
0. OR 1
3 SIGMAs
0. 243
MAGNIF1C ATlONs
NUMBER
OF BLOC K S =
2
AVERAGING
IIMEs
452.7 SEC
IN
NJMRFR
AVERAGE NENIOOS
AV FRAGED
1
122.tBi
2 24
2
122.632
300
3
122.723
3 90
4
122.t 1 0
3 50
5
I 2?.6 02
390
6
122.RR5
3 90
7
12?A 13
390
n
1 ?*>. 592
3 90
9
122.766
.190
13
1 2 2. t 4 e
390
1 I
I 22.63C
3 90
1 2
122.630
J 90
1 3
122.543
390
1 4
122.5 | 0
39C
IS
122.560
3 90
16
1 22. (.2 0
390
1 7
l 22.595
190
1 8
1 22.60 7
390
1=
122.546
3 90
2D
122.484
390
2 I
I 22.5 1
390
22
122 .se?
390
23
122.490
390
24
122.700
3 90
25
1 22 R54
390
26
l22.728
350
27
12 2. 00
390
2 6
1 2252C
3 SC
20
122.515
3 50
30
1 22 567
390
2 1
122.636
390
32
122.672
350
33
122. 43
390
74
1 22 .533
360
35
122.651
3 50
2t
l22.653
350
7
122.595
350
33
l22.555
2 16
TIME
INTENSITY
OI
Figure 3-19. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188023 s (DOPVEL period).

PUL SAP P'jRI 1 3H! P FR I 03 = 1 1 * Of* 3 S GC F 9 E 3 Uc 9 C Y = 2 6 .0 ?
P112WA6 Y 7 5 79 TST Ac T 0 1H 0 9M | 0 STOP 01H 3 A M IOS TMD O?
P11PW06 MAY ?5 7? T5TACT O 1H UM ios ST3" O l H 3RM 04S TSA VP 02
BLOCK AVEPAGE= 122.609
HGM' r O- 079
3 SISMA* o. 2 36 MAGNIFICATIONS IDO
NJMF9 PF LOCKS* 2
Vi l AGI NC T I MF -
A OO.7 SrC
N
N A RES
1
2
A
5
I O
I 1
1
I 3
1 4
I *
1 A
l *
! 1
1 5
'I
?2
-
5
25
2
?-
3 O
2 1
30
**A
36
37
A VESA GE
12? *92
122 677
1 2 > A 3 3
122 636
12? 536
l 2? .60
12? 656
12? 700
1??.^4
12? 7?s
12? SA8
12?.6?0
12? A 8 A
12?.649
l2?.05O
122 60S
I 22 6 23
1 2? 09 7
l22 556
122.528
12? 551
1 2? 6 ng
122 .68 4
12? 755
122 792
1 2?. M *
12? 551
1 2?- 564
122.617
l 2* 351
122 633
1 2?* 4 1
12? 5?n
l2.56C
l2?.669
122 65a
122. 49 7
122.848
nr c1005
A VF5A GE 3
093
O 50
39?
093
3 90
350
93
3 50
3 0
3 90
O <50
390
093
7 50
.no
050
390
3 90
00o
3 90
3 o?
7 90
390
090
090
3 9->
OG0
0 90
.090
090
O 50
3 90
OG0
o 90
390
3 90
3 50
51
TIME
INTENSITY
Ln
>vj
Figure 3-20
Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188063 s (P+40 ys).

JLr 4 0 9*.9||3i|5 rRl?0= 1.
188103
5 e r r
* z QUz NC Y=26, 3 *-Z
M?vA6 MY 25 79 T ST An T 0I
3 0M
1 OS
STOP
01 M 34 4 l os T5AMO
02
PI 12*86 AAV 25 70 01 H
74 M
l 0r
STOP
01 H 38V 045 TSA vo
0?
R^0C< AvfTP* GE =
l 72
609 SI GMA-s 0, 079 3 5 1
',MA =
0
? 76
MAGN IF ICAT IO* =
1 00
NUV8PC
nf ^LorKSs
2 AVf PAGING TIMÂ£t 47 >.8 SPC
n i
a vp!agtg
i
1 2 ? 6 o o
7 00

1
"*
12 *70
390

1
3
1 22.589
790
1
4
12? *6*
> oo
*
1
s
12? 559
79 0
f
1
6
l 2? .58?
3 00
1
7
1 ?? > 7J
7^0
t
1
8
122 755
700
v "
1
9
122,674
3 90

1
l 1
12 677
7Q0
A
1
1 J
12? 510
700
1
1 2
12?.5M
700
1
1 3
t 2 ? 4 ^ ?
700

1
1 4
1 2? 666
3 90
>
1
1 22.e 7 5
700

.
1
1*
12?^3
7 OO
1
1 7
1 2? 0 7
750

1
1 J
1 ?? 5 1 ~
3 0
*
1
1 o
1 r, 38
7<-7

1
2
122 554
3 90
1
M
l 5 26
700
?2
12? *15
700
1
? 3
1 2? 7 1 8
390

1
?4
l?? .a |
7 90
*
1
25
12? 7*o
390
f
25
1
26
12?.67fc
3 90
+
1
1 2? .5 50
705
1
r p
1 2 "* 559
7C0
1
9',
122.60
790

+
1
i:
1 2.5 in
3 0
t

1
m
12 ? 0
7 9 0

1
32
12?t 37
7 90
1
1 7
!
350
-
1
3 4
12? 58?
790
.
1
7r,
1 2? .F 4 4
7 90
,
1
76
1 ?> * 6 \
700
* ->
1
3 7
1 2 5 36
300

=" x-T _
1
38
123.044
51
t
Figure 3-21. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188103 s (P+80 ys)

PULSAD PS*t 133M& lOn-sl 1B |fl.3 S^C FH!*OUENC Y= ?AO H?
P 1
1 ?* A* MAY ?*
7P
r AC ?1H ]
1 *S
-,Tnr>
01-1 34M
10S
TS AMP
0?
P 1
I2WH6 u A Y ?5
79
TSTAPT 01H Hm I
IOS
SI OP
0 1 H 3*M
OAS
TSAvir
?2
OL CTK
A/ EPAGFe 1 ? '' 60 9
r I r.v A =
J
0 7f 7 c 1 r, MA =
?7ft
MAGNIFI CAT ION-
1 00
N
i'F PL D f 5 2
AVrWAf.l NG
T I
F: 4 5 ? * R SC
P IN
AVC9AGE
' EP I no5
NJV7FT5
wrc AGrn
1
1 2? oQ">
700
">
P?, U
300
7
l?? AS1
70Q
4
12? 566
3 SO
3
1 2? ' ? 7
3 00
A
|?? 5RO
300
7
12? 700
300
*
122.649
390
O
l?* *61
7 c*0
1 '
122 633
390
1 1
1 ?? .*37
790
1 ?
12? A*6
.7 90
1 3
122,595
790
1 A
1??.*70
300
1 s
1?? 40?
790
16
122.646
390
1 7
12?,'OR
70-'
1
122 526
790
19
12?. 4 77
3 90
? o
1??.57c
700
? i
12?. 643
790
?>
I 2?r 05
390
?7
12? 7o->
700
?4
l 27,71 0
190
?r
12?. 7 4 7
390
?*
12? r- *o
707
27
1 22. 50?
790
77
1 2?.* ?3
3 CO
?Q
12? 57?
790
3?
1 2? *8 7
7 90
1
l ??.5 07
390
7 ?
12? '?
790
33
122 j 538
790
14
12?.*61
390
75
12? 570
30
7*>
1 22 *1 R
7 rjo
7*
1 2? * 36
390
38
122 009
5?
TIME
INTENSITY
Ln
VD
Figure 3-22.
Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188183 s (P+160 ys).

r>UL ^ap pjici | 31
1 6
FDinosi i 8*
a ? *; rC F
Rc OJF NC YI ?6.
3 'AHZ
pi i r* a+ may ?o
?Q
T CAC 0 1 H
105 STDO
01H 3AM 105
TSA mo
0?
'll 2Rf. M Y ? 5
7?
T5T A9 T 0| M "A A M
IDS STOP
0 1 H 3 MM 94S
TS*3
02
AvrCAGFs 1 ??.i
13 I r'* A?
0
0*.Q 3 ; 1 OMA r
CV 208
MAGMIF|CAT IDN =
o
o
N ) MHEP
of ni ncK3
? AVF* AGING
TJHF= 45? 9 S^C
n i s
Ai/FPAGP
I'fr 100 8
n;m*gr
av re AGEO
1
122 ftl ?
3 90
1 ?? .ft,? '
300
"A
1 ?? ft?0
3 90
4
12? 661
3 90
>
l 2?*44f)
390
ft
l ? ?l 7
3 QC
7
122 635
3 90
1 2? ft 1
IQft
9
12? ft A 1
Q0
1 3
122 5ft!
3 90
1 1
1 ??.8?5
3r0
1 2
1 2? '*1 3
3 90
I 3
122.589
3 90
1 4
1 22 .ft 3 >
3 90
1 5
l ?? ftO-*
3 *0
1 ft
!2?,5HT
3 90
* *
1 ?>.Sft4
\ Oft
1 8
1 2? -505
3 90
1 9
I 22.5?6
90
> 1
1 7. ? 8 s i
1 >0
2 1
122 774
3 90
5 p
122.^07
390
> 4
1 ?"*, fto>
A 90
24
122.684
3 90
?.
1 22 71 7
3 90
122 -l3
390
27
1 ??,ft 7 6
A 90
*
122.4 i4
3 90
pn
l ?* 57ft
0 o?
33
122 574
3 90
* !
1 22 .ft 3 3
390
1 >
1 ?? ft 1 T
3 40
33
1 2? -607
390
1
1 22 ft56
390
71;
I ?? 56Q
30Q
3ft
1 22 6 3 3
3 90
p,
122.525
226
38
122 86?
2 18
TIME
INTENSITY
03
O
Figure 3-23. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188342 s (P+320 ys).

ot)i c/v p oc p |
1 33*
1 6 '>e0 !n0*| 1 C13S63
S CC rp^(jrN'y = ?,, 1 ' + 7
91 1 2 WA*- M AY
2C
75 r ART 0 1 H I?'*
1 05
5TH9 0IH 3 105 TSAMP
02
l12*R6 wAY
25
70 TSTART 01' 34M
1 OS
5THP 01H 3m 045 T54MP
02
RLGC *
WPOAttE*
122
.613 F 1 <.M4 s
0-^76 3 5!GMA=
o
. ??7 MAGNIFICATIONS
100
NJ 9EK
nF BL H C< 5 x
9 AYE^AGING
' T r = 453. 2 cr"C
MJ 'HE 0
a \* ro a rpp
1
12?.746
390
+ Jf
1
;
!?*.! 7
un
1
12? >5 6
16 3
1
L
1 ?? .V
3 90
^CT7 +
l
5
1 2** ,'oo
360
f
I
6
12? 351
TOO
C
1
122,646
3?
1
i
l *60 5
3 5,C
X
I
5
12? 6?4
190
C*
1
! '
122.641
TO?
1
1 \
l 22.653
3 00
1
1 2
12.607
190

\
1 3
l22.459
390
+
1
1 4
l 2? 4
op
*
1
15
1 22. 6. 51
i 1
1-
1 ??.* '?
300
1
1 7
122 607
1O0
1
l
12? 666
3 90
1
1 >
1??.61
300
l
2 0
12? 707
<90
Â¥
1
21
1 22 751
300
21
1
??
122.65?
390
.
+ Jr
1
23
12? .6 71
390
Â¥ Jr
1
24
l 22. 653
i 50
Â¥ .
I
25
1 2
35p
1
?5
12? 6l
3 00
+
1
2"*
123.6H4
3 90
f 4
1
24
I2?.r 64
3 O')
t Â¥
1
P 9
IP? r,7g
70

1
30
12?.620
3 50
i
1
? 1
122.60?
3C?
I
4
1
3?
122 674
390
4
1
33
1 22 5 in
3 O0

1
22R
1
3C
1 2? 5?->
390
~~ Â¥
1
36
1 22.625
360
1
3 7
l 2?.*4 3
30Q
Â¥ \
1
3*
12? 635
2 20
r
1
Figure 3-24. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188553 s (P+640 ys).

J I I I L
Figure 3-25.
jii1iiiiiiiii!iiiii i i i i I i i i i i i i i i 1
100 200 300 400
TIME (PERIODS)
Relative pulse intensity as a function of time (periods) for
pulsar PSR1133+16 on May 15, 1979. Each point represents an
average over 10 periods. P = 1.188023 s.

63
It can be regarded as a double check of the test previously
described.
Each point on the graph represents the intensity with
respect to its corresponding average value of 10 consecutive
periods, for bin 25 (peak of the pulse at the end of the
averaging process). The intensity scale is in arbitrary
units and the time scale is in periods P given by the DOPVEL
program (P = 1.188023 s) The plot shows that the intensity
in bin 25 is almost always positive with respect to its
corresponding 10-periods average value, and no clear periodic
variations appear over the 390 periods plotted.
3.2 PSR2045-16 Data Analysis
Of the 10 nights of data reduced for pulsar PSR2045-16,
pulses were found on 4 nights: May 15, 16, 18 and 19, 1979.
A summary of the dates, averaged pulse deflection in
terms of standard deviations o, and number of periods
averaged is presented in Table 3-4.
Table 3-4. Summary of the date, pulse deflection and
number of periods averaged each night for
pulsar PSR2045-16
Date
Pulse deflection
in terms of a
Periods averaged
May
15,
1979
3.16
540
May
16,
1979
3.48
440
May
18,
1979
3.28
610
May
19,
1979
3.30
415

64
Plots of the data for these dates are presented in
Figures 3-26 to 3-30. Figure 3-31 corresponds to a plot
of a typical day when only noise was present and no
pulse appeared.
In general the pulses were of different shapes,
except for the last two days, for which the shapes were
similar. Some have steep leading and trailing edges (as
on May 16, Figure 3-28); others have only a steep trail
ing edge and a not so well defined leading edge (as on
May 15, Figure 3-26). On May 18 and 19 (Figures 3-29
and 3-30) both pulses show a steep leading edge and much
less steep trailing edge.
On May 16 the pulse was clearly present with an
intensity of 3.48 a above the mean value over an average
of only 180 periods (Figure 3-27). On the other 3 days,
pulses were not seen clearly until the end of the averag
ing process for that night. For partial averages in
these cases, some indication of a pulse can be seen at
the location where the pulse appears after completing the
average, but its intensity never rises significantly
above the noise level. No indication of the existence of
a complex pulse or interpulse can be deduced from the
plotted data.
Pulsar PSR2045-16 presents at higher frequencies a
complex pulse, composed of 3 subpulses, with the trailing
subpulse decreasing dramatically in intensity with
decreasing frequency (22); at 610 MHz the trailing pulse

Figure 3-26.
Pulsar PSR2045-16 recorded on May 15, 1979, at 45 MHz.
Average over 540 periods.

HLLCK AVERAGE-
PULSAR
PSR?
0A5
- 16
LR I UD
19613H3
SEC
EEC
OUf NCY=45
MHZ
P2045 A
MAY
1 5
79
1ST APT
0 9 H
A QM
?0S
STOP
0 9H
5PM
35S
T SAMP
02
P2045R
MAY
15
7^
TS7A t T
09 H
52M
A 3 S
STOP
09 H
5b M
SOS
T S AMP
02
P20 4 EC
MAY
I 5
7 9
tSTAR T
09 H
55*
5RS
STOP
09M
59M
1 AS
TS AMP
02
P?0450
MAY
1 5
79
TSTAKT
09H
59 M
1 6S
ST UP
0 9 H
02M
J9S
T S AMP
02
p ? 0 A 0 E
MA Y
1 5
79
TS1ACT
l OH
0 2M
AOS
S TCP
l OH
05M
55 S
T SAMP
0?
P2045F
MAY
15
7^
1 ST APT
1 OH
O^M
5 6 S
STUP
l OH
08 M
20S
TSAMP
02
D 0 7
S IGMA =
0.
J5 9
3
SI GMA
= 0
1 7 t
MA GN IFICA T ION =
l 00
NU'VJER CF D L CC K S 6 AVERAGING T 1 ME = 103?.5 SCC
pin
AVERAGE
PE R l CO S
NJVBE
A VCNAGED
l
12390A
5A0
2
1 2 a. 037
5 AO
l PA *50
5 AO
A
l?A.3 40
6 AO
5
123.995
5 1 0
f->
1 2 3. 1 7A
5 10
7
l 23.*>6A
5 1 0
9
12A.013
4 90
1
1 2.3.9P5
490
10
l23.977
U 90
1 1
123.901
A 90
l?
l 23 *66
490
1 3
12A.070
5 20
1 A
12A.0 52
540
15
1 2A. ORE
540
16
12A.092
17
123.987
5A0
1 8
1 2A l l l
540
1 9
l 24 .08 1
5 40
20
1 2 4. 0 7 A
540
2 1
1 2A. l92
5 40
22
l 2A .025
540
23
124.020
5 A 0
?A
124.000
5 40
?5
1 24.070
540
26
I 2a.000
5 40
?7
1?7.952
5 A 0
29
124.137
540
29
1 24. 0 A 8
540
o
127*963
540
31
lPA.08?
4 77
32
123. 5 39
540
1 7
123.94f
540
3A
l2a.025
543
25
1?4.046
54 3
127.979
540
37
123.9?A
540
38
1 23. Rye
540
39
123.954
540
AO
1 23 ,99 l
540
A 1
1 24.04 t
540
A?
123>39
540
A3
l 24 .055
540
A+
1 23. 97 A
5 40
A 5
l2A.004
540
46
123.989
540
A 7
1 24.021
530
A 8
124.302
530
A 9
123.966
5 30
50
1 2a. 0 23
4 El
5 1
124.007
530
5?
123.94 l
5 30
53
1 ?3877
530
5 A
123.957
397
55
1 23 .992
0 30
56
123. 967
530
7
1 PA. 007
4 67
58
121.983
540
59
1 2A 01 7
54 0
*0
l 23.970
540
61
1 24 .08 1
540
62
1 2 A. 025
4 77
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
time
!
I
I
I
I
I
|
I
I
I
I
I
I
I
I
I
I
INTENSITY
I
I
I
I
21 3.16 SIGMA
C7T
CJ\

Figure 3-27.
Pulsar PSR2045-16 recorded on May 16, 1979, at 45 MHz.
Average over 180 periods.

PULSAR PSR2045-16 PFRI00=l96l383 SEC FfiEQUENCY=45 MHZ
P204SA1 MAY 16 79 TST ART OVH 44M 34S STO 09 H A 7m 45S TSAVp 02
P 2 O 4 5EJ l MAY 16 79 TST AR T O 9H 4 7M 5 OS S 7 CP 09 H 51M OAS TSAMP 02
BLOCK AVERAGE = 123*90? SIGMA= 0*093 3 SIGMA= 0*279 MAGNIFICATIONS 100
NUMBER CF ULCKS= 2 AVERAGING TI ME = 349. 1 SEC
P I N
A V F r, A G F
PERIODS
N JM'JE R
AVERAGED
1
123*999
1 80
2
l 2A 1 1 1
i eo
J
123.949
l 60
A
I ?A .0 06
i eo
5
1 23 *P95
i eo
6
12 1.97 3
i eo
7
12 5* VAS
1 tO
e
123.461
i eo
9
123467
i eo
10
1 23 .56
i eo
1 1
124.022
i eo
12
123*912
i eo
1 3
123*811
1 t 0
l A
123.5 3 2
i ec
1 5
l24.006
I 80
le
124.079
1 60
1 7
1 2395 0
l 60
1 9
1 24.05 6
1 30
1 9
123.^17
iao
2 0
1 2- .039
i tO
2 1
123 .FtA
i eo
22
124.045
i eo
23
124.045
i eo
24
123*495
i a j
25
123.017
i ec
2 fc
124.006
1 60
27
124.306
i eo
28
1 2a i 56
i eo
29
l 23 */39
l 80
20
12 1.9 22
I 8 0
3 l
1 24.06 7
i eo
32
124.05?
1 60
33
124.017
1 60
34
124.061
1 90
35
124.1 17
i eo
36
1?4.0? 3
1 60
37
123.934
i eo
33
123.996
1 oO
39
1 ? .973
1 60
40
123.906
i eo
41
123.950
1 80
4 2
l24.900
i ec
A 3
123.967
i eo
44
123.964
1 80
45
124.lOC
i eo

123.923
l 80
47
1 ?3.*21
1 80
A 9
12.9c l
i eo
A 9
124.0 11
i eo
5 c
1 23.945
1 90
5 1
123. i 06
i eo
52
1 24 1 56
i eo
53
123.97P
1 dD
54
123.756
1 80
55
123.964
1 60
56
1 c 4.06 7
1 60
57
124.034
1 80
SH
I2J.HO!
l 1 7
59
124.0? t
1 80
60
12 1.945
1 80
61
123.884
l 80
62
1 2 4. 0 44
1 1 7
TIME
INTENSITY
09
00

Figure 3-28. Pulsar PSR2045-16 recorded on May 16, 1979, at 45 MHz.
Average over 440 periods.

PUL S AH
PSP?045-
1 6
Pfwior)*i
9b 1
in 3 SEC TRF UFNCY
-45
MHZ
P2045 A1
V AY
1 6
70
TSTAKT
0 9M
4AM
3 4 S
STOP
09 H
4 7M
49S
T SAMP
02
P?04blU
MAY
1 6
79
TSTAH T
09H
4 7 M
5 33
ST OP
09M
5 IM
C4S
TSAMP
02
P2045C1
MAY
1 6
79
T ST A9 T
09M
5 1 M
0 5S
STOP
09H
54 M
1 9 S
TSAMP
02
P2045U1
MAY
1 6
79
TST AKT
0 9 H
S4M
2 OS
STOP
09 H
57M
3 4 S
TSA MP
02
P20 4 5E l
MAY
l 6
79
T ST ART
09H
S7W
3SS
STOP
l OH
0 OM
i as
TSAMP
02
nt nCK
AVTPACE= 123
.98 1
S 1 GMA
=
0.
0 6 1
3 SIGMA
= 0
.183
MAGN IF 1 C A T ION =
100
NJ M0E8
OF LLOCK S=
5 AVLF-AGING
T I
ME = 8S3. 3
src
PIN
AVERAGE
PEP ICOS
Ku^MEP
AV EPAGTD
1
l23.950
4 40
2
1 24 .0 22
440
3
123.Q93
4 40
4
1 24 C2 2
440
5
l23.4?4
4 40
6
1 2.3.927
4 40
7
l 23.852
4 40
9
123.926
440
9
1 23.979
4 40
1 )
123.95
4 40
1 1
124.004
4 40
1 2
l23.866
384
1 3
123.9 34
4 40
1 4
123.886
4 40
15
123. 89 2
44 0
lb
l23995
4 40
1 7
124.052
440
18
124.047
4 4 0
1 9
123.893
4 40
2 0
1 2 3.94 3
4 40
2 1
1 23.91 1
44 0
2?
124.001
.3 7 7
23
1 23.882
4 40
24
1 24.0.3 1
4 4 0
25
1 2 3.9 29
4 4Q
26
124 ,02 3
4 4 0
27
124.192
440
28
l23.954
4 40
29
1 23.9 36
4 40
30
123.904
440
31
l24.06\
440
32
124.027
440
33
123.9 22
4 4 0
34
123.986
4 40
35
l24.022
4 40
36
124.013
4 40
37
123 .9 70
4 40
38
l23.995
4 40
39
123 .954
440
40
124.0 18
4 40
4 1
1 24.0 34
4 40
42
1 24.054
4 40
43
123.966
4 40
44
124.016
4 40
45
l 23 .993
440
46
124.004
440
47
123.931
4 40
48
124.002
377
49
1 24 .00 7
4 4 0
bC
123.993
440
5 1
123.986
440
52
124.061
4 40
53
123.982
4 40
54
1 2.3.804
4 40
55
1 24.0 l l
4 40
56
124.102
4 40
57
1 24.023
440
68
124.032
3 77
59
124.025
440
( 0
123*986
4 40
61
l 23 .969
4 40
t'
124.021
377
TIME
INTENSITY
I
I
I
I
I
I
I
27 3.48 SIGMA
I
I
I
I
I
1
I
o

Figure 3-29. Pulsar PSR2045-16 recorded on May 18, 1979, at 45 MHz.
Average over 610 periods.

PULSAR PSR2045-16 PGR 100:(9613 0 5 SEC FREQUENCY=45 MHZ
LOCK AVERAGES
1 3 I
P2 045 A3 A
MAY
18
7*
TSTART
09H
38 M
OOS
STOP
0 9 h
39 M
54 S
TSAMP
02
P?045H3A
MAY
1 8
79
T57ART
0 9H
4 0M
oos
STOP
0 9 H
4 4 M
5 9 S
TSAMP
02
P2045C3A
MA Y
1 8
79
TSTART
0 9H
45
OOS
S T C P
0 9H
48M
4 OS
TSAMP
02
2 04503 A
MA Y
1 8
79
TS TART
09H
49 M
oos
STOP
0 9H
* 4M
OOS
TSAMP
02
P2 045E3A
MA Y
1 b
79
T STAR T
09H
54M
0 1S
jTP
0 9 H
59 v
OOS
TSAMP
02
2*1 SIGMA= 0.057 3 SIGM4= 0.170 MAGNIFICATIONS 100
NUM3EP F QLOCKS= 5
AVfcRAGlNG TlMts 118P.8 SEC
BIN
AVE RAGE
Pt R I C DS
N'JMMCR
AVCRAGED
1
121.>57
b 1 2
2
1 31 .27 ft 10
3
1 31
ftl 0
4
131.324
6 1 0
0
1 Ji 32 6
o 1 0
f-
121 .263
fc 1 0
7
121.293
ft 10
8
111.478
5 33
)
1J1394
ft 1 0
1 c
131. 1 9 (i
ft 1 0
11
12 1 .27 7
Â£ C4
12
1 3 l ? b 3
ft 1 0
1 3
1 J1 J 34
ft 10
1 4
131.329
ft 10
15
13!.317
6 l 0
16
121.209
61 0
17
131.224
6 1 0
1 b
131 ? 7 5
ft 10
19
131.275
6 10
20
1 J1 1HH
c 1 0
21
1 31 .288
ft 10
22
121 .311
6 1 0
23
1 3 1 <24
ft 10
24
131.14 7
6 1 0
25
131 .18 6
t 10
26
131 39
6 1 0
27
121.270
6 1 0
23
131.230
ft 1 0
29
131.291
b 10
30
131.272
b 10
31
1 3 1 .T9
5 04
22
121.pe
6 1 C
3 3
131.337
6 1 0
34
131 .329
6 1 0
35
1 21 .293
b 1 0
3b
131.110
6 1 0
37
1 31 1^5
b 1 0
35
1 31 .13 C
1 0
29
131.281
ft 10
40
121.266
ft 1 0
4 l
1 ?l 1 i8
ft 10
42
1 31 3 ft 0
b 10
43
121 .289
61 0
44
131 I4
ft 1 0
45
131.402
ft 10
4 ft
1 31 28 8
ft 10
4 7
l 3 1 10 a
ft 1 0
48
1 J 1 44
6 1 C
49
121.20'
ft 10
50
131.294
6 1 0
51
131.216
bl 0
52
131.155
ft l 0
53
131.221
b l 0
54
131 .253
ft 1 0
55
l 3 1 9 1
ft 1 0
5b
131.301
ft 1 0
57
13 1 .244
6 10
58
131 .222
ft 1 0
59
121.216
6 1 C
60
131.32b
ft 1 0
6 1
131. J 4 5
5 04
62
12 1. ->64
575
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I i
I
* TIME
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
INTF.NSITY-
8 3.28 SIGMA
I
I
I
I
I
I
I
I
I

I
I
I
I
I
i
!
i
i
ro

Figure 3-30. Pulsar PSR2045-16 recorded on May 19, 1979, at 45 MHz
Average over 415 periods.

PULSAR PSW04S-
IF
I'tK 1 CK> = 1 Oft 1 3 78
see
rwto UENCY
= 45
MHZ
P20A5HA MAY
1 9
79
T f 7 A ( T
09M 39M
30S
ST CP
09 H
A AM
3 0 S
tsamp
02
P2 045C4 MAY
1 9
79
1ST APT
0 9H * A M
3 OS
STOP
09 M
A6M
J6S
T SA MP
02
P ? 0 A 50 A MAY
19
79
TST ART
09H 46M
3 7 S
STOP
O^H
5 0 m
35S
TSAMP
02
P 20 A 55 A MAY
I 9
7 9
TST AW T
09H SO M
AOS
STOP
09H
5 JM
32 S
TSAMP
02
PI UCK A V E w A GE 130.
3V2 SIGMA
-
0.
073
3 SIGMA
= 0
220
MAGNIFICATION=
1 00
NUMMGR OF DLOCK S=
A AVCP AGING
T I MF = 80A
7 SFC
n in
A VE W AG F
per ions
N.J'RER
A V c WAGE 9
1
130870
A 15
2
1 JO.A 32
A 15
3
1 30. 82 1
A 15
4
I 30 3 75
A 15
5
1 3 0 A 2 S
A 15
t>
120 .t 2 3
41 5
7
1 3 0 A 9 5
4 15
*J
130.800
4 15
V
1 3 0 A 2 S
A 1 5
l 0
l 30 2 78
A 15
1 1
130.341
A 15
12
130396
A 1 5
1 l
130.A 37
A 15
1 A
1 3 0 A 2 0
A 15
1 5
l30.37?
A 15
1 5
13 0 .A 3 S
A 1 5
1 7
l30.379
A 15
1
l JO .5 0 7
4 15
1 9
130.4 A 15
20
1 30.A 49
A 15
? I
130.44T
4 15
2?
130.2 1 1
4 15
23
130. 39A
4 1 5
? A
1 30A 35
4 15
25
1 30 .4 32
A 15
26
130.428
4 1 5
7 7
130.146
A 1 5
28
1 J 0 3 7 9
4 15
29
UO. 2 34
4 1 5
33
133.423
4 15
31
130.365
4 15
22
130.?P6
4 15
23
130.177
4 15
3A
130.309
3S5
35
130. 4 A 7
A 1 5
35
1 JO.A FA
A 15
37
1 3 0.37,?
A 15
28
l 30. 7 5 8
A 15
29
130.271
A 15
4 0
130 .298
A 15
4 I
130. A 21
331
4 2
1 8 8 A l 3
A 15
43
1 3 0 J 4 8
4 15
44
1 30.A 41
A 15
AS
1 3 0 A 1 8
4 15
45
130.464
4 15
47
1 30. A 7 1
4 1 5
48
133.363
A 15
A 9
1 JO 34 3
A 15
5 0
1 JC.290
4 1 8
S 1
120 .3 34
4 1 5
5?
130.355
A 15
53
1 JO .392
415
SA
130.375
A 15
5S
1 30.A I 8
4 15
5b
1 JO A l 2
3 73
57
1 30 A JO
4 15
58
130.440
4 15
59
1 JO AS8
3 C 9
t 0
l30.3MQ
4 15
61
l30365
4 15
62
1 JO.381
A 15
TIME
INTENSITY
I
I
I
I
I
6 3.3 SIGMA I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-J

Figure 3-31.
Pulsar PSR2045-16 recorded on May 17, 1979, at 45 MHz.
Average over 390 periods. No pulse present.

PULSAR rr>w? 045- I (> Pf. 9 | OP*1 7* | 7** a S FC r RF OJE NC Yr % s MHZ
?,'5A?A MAY 17 7 9 rr;TA-T OOh AIM
>?0A5n?A MAY 17 79 TSTAJT 09H 4' =>2045024 MAY 17 79 TSTA~T 09M ARM
P2 0 A *>F ? A MAY 17 M T S T ART 0H 5 ** 4
O^S STn'* 09H ASM 00S TSAMO 02
n IS STOP 0 9H A PM "*9 5 TSAM> 92
AOS STOP 09H 5? M OSS TS AMP 02
3f.S STOP "QH 5SM 30 S TSAMP 0?
LOCK AV F.R AGP = 12^.829
ricma- o.064
MAC.N IF I CAT IP Ms 100
NU'PFO OF rlOCKSs A
AVFCAGIMG 5 I MC 766.3 S PC
n I m AVFCAGE *Ecin)S
NMQEh AV F.R A OrD
1 I 28 c 30 3C0
? 128 37 790
* 128. 32 300
a 128 801 390
5 1 28, B?a 3 90
~ 128,7 0-, 3 00
y l28 858 318
p. 128 83 1 3 90
128.55 390
10 128895 39J
11 120 92 A 3 SO
12 128.733 390
13 128 "70 300
1 A 1 28,826 390
18 1 R t 1 7C0
IS 128,*Op 390
17 120.775 390
18 178,767 790
1 12b 837 39C
2^ 1 28. '3 390
21 I 20 79* 3^0
22 128 706 390
27 128.770 3C0
2A l28 7SO 390
25 123 011 -*qo
2' 128- 3 19 390
2 7 1 23,7 ) 7 .3 S?
28 128 796 3 90
29 1 28.3 37 3 00
30 128 77P 300
71 128773 390
*'* 1 2 7 6 5 390
? 3 123-733 3 CO
34 120 793 390
3 i 129, 7SA 390
36 128 779 300
37 1 28. -726 390
*7 1 23 798 300
39 128 7U 390
AO 128. 8 A9 7 90
A I I 3.020 3 00
A? 123 67 390
A3 120 866 39C
A 1 l ? 7 7 *> 390
AS 123 79| 390
AS 1 20, 855 390
A7 I 28,998 390
a 8 123 852 390
AO 128,*24 390
60 1 28.7P 3 700
51 128007 700
6*> 123,9 75 390
5 J 1 28 *>4 O 790
61 120.900 717
5 5 1 2 0 * 5 1 3 50
56 l279 790
57 128 ^67 795
60 l20. 96 390
63 IP.60P 700
7.5 120 0 37 7oS
61 120,13? 390
6? 120.307 3 06
TIME
INTENSITY
03

77
is the most intense, but the intensity decreases
rapidly, and at 151 MHz is about 2/3 of the central
subpulse. The intensity of the central and leading
subpulses seems to remain almost constant with fre
quency. No indication of resolved subpulses can be
obtained from the data at 45 MHz. Nevertheless, the
difference in steepness of the leading and trailing
edges might suggest the existence of an unresolved
much less intense subpulse accompanying the main sub
pulse. This aspect will be discussed later when dis
cussing the pulse width and the effect of polarization.
3.2.a Peak Flux Density, Mean Flux
Density and Energy for Pulsar PSR2045-16
In order to estimate the peak flux density of this
pulsar, several radio sources were observed for calibration
purposes. Radio source PKS2034-17, one of the closest
to the pulsar, was clearly recorded despite the steepness
of the galactic background (23). This radio source has
apparently not been observed before at any frequency lower
than 80 MHz. However, an extrapolated flux density at
45 MHz was obtained from the spectral index derived from
flux density data at higher frequencies, as reported in the
literature.
Spectral index can be defined (17) by the relation:
(3-11)

78
where S = flux density
v = frequency
n = spectral index
For any two frequencies on the linear part of the
relationship between S and ^ ,
The best available values of the flux density for this
source (24, 25) are at frequencies of 160 and 80 MHz:
sico =8-8 Jy
sso = 15 Jy
Substituting these values into equation (3-12), a value
of n = 0.76 is obtained.
Assuming a linear relationship for S as a function of
~ down to 45 MHz, the flux density at this frequency is
found to be:
S45 = 23 Jy
Applying the previously used criteria, the following
value for the minimum detectable flux density at 45 MHz
was obtained for the region near source PKS2045-17:
S = 6 Jy
mm J
source PKS2045-17 were:
Av = 3 MHz
T = 3 S
n
r
1

79
where Av = bandwidth, r = post detection time constant
and n = number of records averaged.
It can be shown that the brightness temperature T
and the frequency v are related in a similar way as S and
v are related in equation (3-12):
Using the galactic background temperature at 38 MHz
given by Blythe (26):
T = 12,000K
P
T = 7,000K
r
where T = galactic background temperature near the pulsar
IT
= galactic background temperature near the radio
source PKS2034-17
and the value for the galactic background spectral
index given by Findlay (27), valid for the range 20 to 400
MHz:
n = 2.65
the galactic background temperatures and at
45 MHz were found to be:
T = 7,630 K
P
T = 4 ,452K
r
Using equation (3-2) and the appropriate parameters
of the radio telescope, the minimum detectable flux
density can be obtained for the pulsar observations.

80
The radio telescope parameters when observing
PKS2034-17 are:
Av =3 MHz
r
T = 3 s
r
n =1
r
T = 4,452K
r
S = 8 Jy
min J
The parameters when observing the pulsar are:
Av = 10 kHz
P
n
P
T
P
= 11.25 ms
= variable, depends on the day
= 7,630 K
The peak flux density (which is also the minimum
detectable flux density) for the different days is shown
in Table 3-5; this table also contains the values of the
mean pulse energy and mean flux density computed using
equations (3-4), (3-5) and the values for the pulse width
given in Table 3-6.
No known measurements of flux density and energy at
low radiofrequencies exist for pulsar PSR2045-16. The
lowest frequency data on this pulsar are the 80 MHz obser
vations made by Slee and Hill (9). They attempted to
observe the pulsar on 3 nights, failing to detect it and
giving an upper limit for the pulse energy of
<0.17x10 Jm Hz The failure to detect this pulsar
is not surprising since it has been reported to have large

Table 3-5. Summary of Peak flux densities, Mean Pulse energies and
Mean flux densities for Pulsar PSR2045-16 at 45 MHz.
Date
Periods Averaged
Peak Flux Density
Jy
Mean Pulse
-2 -1
Energy Jm Hz
Mean Flux
Density Jy
May
15,
1979
540
125
3.8
1.95
May
16,
1979
440
139
3.4
1.73
May
18,
1979
610
118
3.5
1.8
May
19,
1979
415
143
4.4
2.25
Unweighted
Mean
3.775

82
4
m THIS THESIS
3
0 BACKER (30)
A SIEBER (29)
2
O SLEE AND HILL (9)
AVERAGE 1
PULSE
-
ENERGY
-
-2 -1
(Jm Hz )
~
xlO-26
.5
.4
.3
'
.2
T
\
A
.1
2(
3 30 40 50 100 200 300 400
FREQUENCY (MHz)
Figure 3-32. Pulsar PSR2045-16 low frequency energy
spectrum.

83
intensity variations, being often undetectable (28, 29).
Our experiences at 45 MHz confirm this behavior also since
it was possible to detect this pulsar only in the first
nights of the observing period; it faded out for the rest
of the nights.
-2 -1
An unconfirmed pulse energy of ^1.9 Jm Hz has been
given by Backer at 34 MHz (30). The lowest frequency at
which published data on pulse energy is available is 410
MHz; the energy is 0.120x10 ^ jm ^z 1 (29, 31). Our
measurement at 45 MHz seems rather high (perhaps suggesting
an undetected systematic error), although no direct
comparison is possible due to the lack of published informa
tion at low frequencies on this pulsar.
Pulse energy as a function of the frequency has been
plotted in Figure 3-32, including the data mentioned in
the last paragraphs.
3.2.b Pulse Width for Pulsar PSR2045-16
In order to obtain the true half intensity pulse width
of PSR2045-16 at 45 MHz, the measured half intensity pulse
width has been corrected for pulse dispersion and post
detection time constant using equations (3-8) and (3-9)
respectively.
In order to correct for pulse dispersion, the follow
ing values for the parameters were used:
Av = 10 kHz
Vo = 45 MHz
DM = 11.5 pc cm ^

84
the
of T
W
0.5
The correction for pulse dispersion AtQ 5 using
above values, is:
At0.5 = 10'5 mS
The correction due to the post detection time constant
= 11.25 ms is:
c
T =7.8 ms
The true (or corrected) half intensity pulse width
is as before (equation 3-10):
WA c = Wn At. c T
0.5 0.5 obs. 0.5 c
The result for the 4 days is shown in Table 3-6.
Table 3-6. Observed and corrected half intensity
pulse width and pulse width at base
(assuming triangular shape).
Date
W
0.5 obs
(ms)
W
0.5
(ms)
Width at base
(ms)
May 15,
1979
44
31
62
May 16,
1979
38
25
50
May 18,
1979
44
31
62
May 19,
1979
44
31
62
The lowest frequencies at which pulse shape and width
measurements have been given for this pulsar are 150 and
151 MHz (10, 22). Lyne et al. (22) have given the separa
tion between subpulses at a number of frequencies; the
values are plotted in Figure 3-33. The separation between

200
THIS THESIS
CENTRAL SUBPULSES. LYNE et al
O SEPARATION BETWEEN (22
TRAILING AND CENTRAL
SUBPULSES. LYNE et al (22)
PULSE
SEPARATION
(ms) 50
40
30
20
O
O
O
00
Ul
1 1 1 I I 1 1 I I I I I I 1 1
10 20 30 40 50 100 200 300 400 500 1000
FREQUENCY (MHz)
Figure 3-33. Pulsar PSR2045-16 subpulse separation as a function of
frequency, including the pulse width at the base obtained
at 45 MHz in this thesis.

86
the leading and central subpulses and the trailing and
central subpulses is plotted separately.
As in a previously mentioned case, the trailing pulse
seems to decrease very rapidly in intensity toward lower
frequencies (22), and it is likely that at 45 MHz it no
longer exists or at least it can be very weak. The mean
value at 45 MHz for the pulse width at its base (obtained
by measuring the width at half intensity and assuming a
triangular shape) of 61 ms has also been plotted in Figure
3-33.
The plotted value for the pulse width at its base, lies
close to the extrapolation of a straight line passing
through the points representing the separation between the
leading and central subpulses. A possible conclusion of
this is that the pulse observed at 45 MHz is composed of
two unresolved pulses which might correspond to the leading
and central subpulses at higher frequencies; the pulse
width at its base would correspond to the separation between
the two subpulses.
No polarization measurements have been made for this
pulsar at low frequency. The lowest frequency at which
polarization measurements are available is 409 MHz (32, 33).
The mean percentage of linear polarization at 409 MHz is
40%, and it changes during the course of the pulse, reaching
about 70% for the first subpulse. The stability of the
linear polarization position angle is low for the wings of
the integrated pulse profile, due to the occasional

87
occurrence of pulses with orthogonal polarization with
respect to the mean, but is high at the center of the
pulse (32). If the degree of linear polarization of the
pulse is higher at lower frequencies, there will probably
be a distortion in the pulse shape when it is received
with a linearly polarized antenna. This distortion will
depend on the variation of the polarization angle within
the pulse. Assuming that the linearly polarized component
in the central part of the pulse is parallel (or nearly
parallel) to the antenna, then the integrated pulse profile
will appear strong in the center; however the slopes of the
leading and trailing edges will be weaker and the averaged
pulse profile will be narrow. More details about the
effect of the degree of linearly polarized component on the
pulse profile and apparent intensity variations are given
in the conclusions.
3.2.c Period of Pulsar PSR2045-16
from Observations on Two Consecutive Days
The apparent mean pulse period of a pulsar can be
calculated from the elapsed time between two pulses; in
order to increase the accuracy it is desirable that the
elapsed time between the two pulses be much longer than
the pulsar period. An attempt was made to obtain the
apparent mean pulse period from observations made 24n
apart (elapsed time %24 ) for pulsar PSR2045-16. On two
occasions the pulsar was detected on two consecutive days;

88
the first occasion was May 15 and 16, 1979, and the second
May 18 and 19, 1979.
The pulse period can be obtained by subtracting the
epoch of apparition of the pulse E on two consecutive days
and dividing by an integer number N.
The epoch of the pulse can be calculated from the
relationship:
where:
E
T
start
+
nxR
(3-14)
T = time start digitizing
s tax t
n = bin number where the pulse occurs
R = plotting time resolution
The information related to the two pairs of days has
been summarized in Table 3-7.
The mean apparent period can be calculated from the
relationship:
P
AE
N
(3-15)
where AE: difference in epoch of pulse apparition on two
consecutive days (elapsed time between pulse
apparitions)
The mean apparent periods obtained with this method
are compared with the DOPVEL periods in Table 3-8.

Table 3-7. Epoch and difference in epoch of pulse apparition for Pulsar
PSR20 45-16.
R
Date
n
(s)
E
May
15,
1979
27
0.0316
0 9h
4 9m
20?66434
May
16,
1979
8
0.0316
0 9h
44m
34 f8 5415
May
18,
1979
8
0.0316
0 9h
3 8m
OOf25308
May
19,
1979
6
0.0316
09h
34m
30 fl 8 981
AE
(s)
86114?190
N
43905
86189?9367
43944
or 43943
00
VO

90
Table 3-8. Comparison of the mean apparent pulsar
periods for PSR2045-16 obtained from
observations with the DOPVEL periods.
Date
Mean apparent period
(s)
DOPVEL period
(s)
AP
(PS)
May
15,
1979
1.96137547
1.961383
-8
May
16,
1979
May
18,
1979
1.96135847
1.961386
-28
May
19,
1979
or 1.96140311
+ 14
It can be seen that the periods do not agree.
Unfortunately, a jump in the epoch was noticed in the quartz
time standard of Maipu, after the first pair of days; that
makes the figures for the last pair of days (May 18 and
May 19, 1979) very unreliable.
In any case, this discrepancy, at least for May 15 and
May 16, demonstrates that the DOPVEL period cannot be used
to bring in phase two consecutive days of observations.
In order to bring in phase pulses detected on two
consecutive days (within 0.5 bin), the apparent mean period
for each day needs to be known with an accuracy of at least
where N = Integer number of periods between two apparitions.
Expressed in terms of the elapsed time T and the
period P, (3-16) becomes:
A
c
?x-
2 T
(3-17)

91
T
where N = .
Assuming an elapsed time of 24 (86400 s), a resolu
tion of 0.0315 s and a period of 1.961386 s, the following
accuracy for the period is obtained:
A = 0.36 ys
c
This accuracy should bring the two pulses in phase
within 0.5 bin.
It is not clear what caused the discrepancies between
the DOPVEL period and the period deduced from the observa
tions; it could be an instrumental problem or a precision
problem in the calculations in the DOPVEL program. Both
possibilities were checked but the result was inconclusive;
further investigations need to be made.
3.2.d Pulsar PSR2045-16 Pulse Intensity
as Function of Time
Figure 3-34 is a plot of the relative pulse intensity
as a function of time for the pulse obtained on May 16,
1979; this pulse is the strongest and best defined. Each
point plotted represents the relative intensity with respect
to the noise average, averaged over 10 periods; the time
axis scale is in pulsar periods. It can be seen that the
pulse has large intensity variations. Most of the time it
has a positive value with respect to the noise average, and
it seems to show an almost periodic intensity modulation of
^140 periods (peaks at periods 20, 160 and 320).

100 200 300 400
TIME (PERIODS)
Figure 3-34. Relative pulse intensity as a function of time (periods) for pulsar
PSR2045-16 on May 16, 1979. Each point represents an average over
10 periods. P = 1.961383 s.

CHAPTER 4
CONCLUSIONS
Two pulsars have been successfully detected using
narrow bandwidth and long averaging times. Pulsar
PSR1133+16 was detected with the University of Florida large
array at 26.3 MHz and pulsar PSR2045-16 was detected using
the Maipu Radio Observatory large array at 45 MHz. Pulsar
PSR1133+16 was strong enough to be considered statistically
significant (pulse deflection larger than 3 standard devia
tions) in the data of 2 nights (out of 10 nights reduced).
Pulses of PSR2045-16 were found on 4 nights (out of 10
reduced) with enough intensity to reach the 3 standard devia
tion level. Typical averaging times for both pulsars were in
the range 400 to 600 periods.
Both pulsars show large intensity variations on a time
scale of days. Typically, one would appear after an aver
aging time as short as ^200 periods, and then would fade,
being undetectable for averaging times as long as ^1500
periods over succeeding days.
The peak pulse flux density obtained for PSR1133+16 is
about 150 Jy; for PSR2045-16 the peak flux density is in the
range of 120 to 140 Jy. These values (which have relatively
large probable error, in the order of 50%), are higher than
the values of the peak flux density measured by others at
93

94
nearby frequencies. They are also larger than the
extrapolated flux densities from published data at higher
frequencies.
Also, the pulse shape and half intensity width seem
to differ from extrapolated values based on data at higher
frequencies.
In order to find a satisfactory explanation of the
intensity variations of the two pulsars, it is necessary
to consider both previously reported intensity variations
and the data available on polarization at higher frequencies.
Pulsar PSR1133+16 and several other pulsars have been
reported to show extremely variable flux densities both from
pulse to pulse and on a longer time scale, when observed at
81.5 MHz (34). These intensity variations are not believed
to be caused by interplanetary scintillation, and therefore
must be attributed to the source.
Pulsar PSR2045-16 has been reported as often being
undetectable for several days at 408 MHz, but when it is
active it can produce pulses stronger than any other of the
first pulsars discovered (28) .
Besides these intensity variations, the intensity of
the received signal will show larger and more frequent
variations when received by a linearly polarized antenna,
if the pulses are strongly linearly polarized. Manchester
et al. have found that most pulsars present a relatively
high degree of linear polarization (32). PSR1133+16 was
found to have a degree of linear polarization of 33% at

95
403 MHz, 50% at 86 MHz and 60% at 63 MHz, which is the
lowest frequency at which polarization has been measured
for this pulsar (35, 36). There is a clear dependence of
percentage of linear polarization upon wavelength, and the
tendency is for the percentage of linear polarization to
increase with increasing wavelength. Pulsar PSR1133+16
has a complex pulse, composed of two subpulses; the per
centage of linear polarization is variable throughout the
pulse and reaches about 70% in the middle of it at 403 MHz.
For individual pulses, the position angle of the plane of
linear polarization is highly stable for all but the first
component, which usually shows two preferred positions
perpendicular to each other. The result of this is that
the degree of linear polarization of the integrated pulse is
low for the first subpulse and high for the rest (32).
Pulsar PSR2045-16 has a complex pulse composed of 3
subpulses. Manchester et al. (32) measured the degree of
linear polarization to be 40% at 403 MHz, with the polariza
tion being even larger for the first subpulse (^70%). The
position angle of the plane of linear polarization varies
smoothly throughout the pulse; a change of about 30 occurs
between the first and third subpulses. This pattern of
position angle variation within a pulse is stable for many
consecutive pulses. No known polarization measurements are
available at lower frequencies for this pulsar.
Even if the linear polarization position angle were
stable when measured at a point above the terrestrial

96
ionosphere, the ionosphere can introduce relatively large
variations in the angle as measured at ground level. This
is due to variations in the amount of Faraday rotation in
the ionosphere. This effect is particularly strong at
longer wavelengths because the angle of rotation 0 is
proportional to the square of the wavelength A:
9 RM-A2 (4-1)
where RM = rotation measure of the ionospheric path.
The RM depends linearly on the electron density, N,
the component of the earth' s magnetic field intensity which
is parallel to the direction of propagation, Bn and the
thickness of the ionosphere, L:
fL
RM B|, NdL (4-2)
o-1
It can be shown that the percent variation of 0 is
equal to the percent variation of the electron density N,
if the thickness L and the magnetic field B of the ionosphere
remain constant; that is:
A0 AN _
T % IT %
(4-3)
It can also be shown that equals when N and B(|
AB
are constant, and equals -5-" when N and L are constant.
Manchester has obtained values of ionospheric RM
-2
ranging from 0.1 to 6 rad m depending on the time of day
and the declination of the pulsars (37). Taking these two
extreme values, it is possible to obtain the change in
rotation angle, A0, for let us say, a 1% change in AN at

97
both 26.3 and 45 MHz. The result of this calculation is
shown in Table 4-1.
Table 4-1. Variation of the position angle A0 due
to variations of 1% in AN, AB(|Or AL in
the ionosphere.
Frequency
(MHz)
AQ AN ABi,
A0 for IT' R
AL N B|1
or = 1%, and
L -2
AQ an
A0 for -TT-,
AL
r = 1%,
AB
and
-2
m
26.3
7.4
447
45
2.5
150
From the values obtained in Table 4-1, one should expect
large random (or nearly random) variations of the position
angle for large values of RM at the two frequencies considered.
This effect could produce a modulation which distorts the
pulse shape, in addition to the long term intensity varia
tions. For low and medium values of RM, the variations of the
rotation angle A0 are not so important, since they are small.
There is another effect that can be important for large
values of RM; that is the change in position angle when
observing the pulsar in different regions of the sky (i.e.
when observing the pulsar with different beams), or when
observing the pulsar on different nights. For example, at
_ o
45 MHz, with a RM = 1 rad m a change of 4% of N, L or Bn
(or a combination of them), will cause the position angle
to be orthogonal to its previous position. If the pulsar

98
radiation is relatively highly linearly polarized, this
should produce a significant drop in the intensity of the
received signal when using a linearly polarized antenna.
This last effect can provide a possible explanation of the
disappearance of the pulsar signal when switching to a
different beam of the 26.3 MHz array.
Regarding the effect of the polarization throughout
the pulse, the pulse shape can be modified when using a
linearly polarized antenna; the signal will appear more
intense for the part of the pulse with polarization that
matches the polarization of the antenna, and abnormally
weak for the part of the pulse with position angle unfavor
able for the antenna. Also if a given part of the pulse
presents random changes in the position angle from normal
to orthogonal (which led to a low integrated polarization
for that part of the pulse), the intensity of the averaged
pulse will be abnormally weak in that part. This seems to
be the case with PSR1133+16, in which the observed intensity
at 26.3 MHz of the second subpulse is usually stronger than
the intensity of the first subpulse; the latter is often
undetectable or appears only occasionally. The separation
of these two subpulses at 26.3 MHz is much larger (^125 ms)
than the predicted separation for the two subpulses (which
compose the main pulse) obtained from data at higher fre
quencies; however the separation at 26.3 MHz does agree
with the results obtained at 25 MHz (4).

99
The values for the peak flux density, the mean flux
density and the mean energy are higher than the values
obtained from extrapolation at higher frequencies, with
the exception of a value of peak flux density of 150 Jy
obtained at the relatively high frequency of 111.5 MHz for
PSR1133+16. Izvekova et al. (7) and Kuz'min etal. (20)
made simultaneous observations of this pulsar at several
frequencies in order to obtain the spectrum. They found
that the average spectrum (average over several nights)
presents a turn-over at about 60 MHz. Nevertheless they
also found that the spectrum shows large variations from
day to day in the range 61 151 MHz, often with a change in
the sign of the spectral index. It is interesting to note
that the spectrum of some individual nights does not show a
turn-over but rather an increasing intensity with decreasing
frequencies. This might explain (at least in part) the
unusually large intensity of the pulses at 26.3 MHz, when
compared with average values obtained by other authors at
higher frequencies.
Averaging of pulses for two different nights was not
possible. For pulsar PSR2045-16, the period obtained from
the DOPVEL program differs from the period deduced from
observations on two consecutive nights by amounts between
8 to 28 ys; this discrepancy is large compared with the
required precision for averaging over two consecutive days
(i.e., ^ tenths of ys). No attempt was made for PSR1133+16
to average data of two different nights since the only two

100
nights with data are 27 days apart. It is not clear up to
the present whether the discrepancy of the periods is due
to instrumental problems or insufficient precision in the
calculations of the predicted period.
From the results obtained in this thesis, it has been
shown that it is possible to detect pulsars at both 26.3
MHz and 45 MHz, with certain limitations. In the light of
the results obtained, some improvements and modifications
in the equipment and data reduction can be suggested which
One such modification is already being put into effect.
A greatly improved transistorized 26.3 MHz receiver which
will replace the old receiver is under construction and is
nearly completed.
Recording at a much wider bandwidth (perhaps 50-100 kHz),
and utilizing a de-dispersion technique, would considerably
increase the signal-to-noise ratio. The wide band signal
would initially be recorded using an instrumental tape
recorder of suitable bandwidth. It could then be played back
into the Saicor Real Time Spectrum Analyzer, the output of
which could be digitized and further analyzed by means of a
i
microprocessor or a minicomputer (like the PDP 1134).
After the data has been converted to a power spectrum appear
ing in the 200 output frequency channels of the Saicor
spectrum analyzer, the frequency dispersion can then be
removed by means of the computer. With a 100 kHz-bandwidth

101
system of this type, the signal-to-noise ratio would be
or 3.5, over what it was
with the 8 kHz-bandwidth system.
Another advantage of a wide bandwidth system, and per
haps an even more important one, is its ability to deter
mine the degree of linear polarization of the pulsar signals
by making use of Faraday rotation in the terrestrial iono
sphere. As stated previously, the rotation of the plane of
linear polarization varies, due both to ionospheric fluctua
tions and to a continual change in the ray path as the earth
rotates, so that the polarization plane is frequently
parallel and frequently perpendicular to the antenna dipoles.
The degree of linear polarization is readily calculable from
the measured intensities at times of parallel and perpendicu
lar orientations. This appears to be an excellent method for
determining linear polarization; it has apparently not pre
viously been used for pulsars.
Also, if the rotation measure (RM) of the ionosphere
above the observatory can be determined independently, the
position angle of the plane of linear polarization of the
pulsar signal at the top of the ionosphere can be calculated.
tion about polarization. Since the dipoles of the 26.3 MHz
array in the Dixie County radio observatory have the
capability to be rotated to any position, half of the array
can be rotated in 90 respect to the remaining half, making
possible to obtain information about the circularly

polarized component of the emission. If for some pulsar
there is enough signal-to-noise ratio, it would also be
possible to set the antenna in such a way that one third
of the array is oriented at 0, one third at 45 and the last
third at 90; from the information obtained in this configu
ration it is possible to deduce the Stokes parameters of the
incident wave. Any pulsar polarization measurement made at
26.3 MHz and 45 MHz would be useful since none have been made
at frequencies lower than 63 MHz (36).
Improvements need to be made in the calibrations in order
to obtain more accurate values for the peak flux density.
This can be accomplished by simulating the pulsar pulse using
a precise time standard and a frequency synthesizer, in order
to obtain periods within some \i seconds of the observed period.
It should be possible to obtain estimations of peak flux
densities by digitizing and processing this signal in a similar
manner to the actual pulsar. It will be also necessary to
find out the reason for the discrepancy between the DOPVEL
period and the period obtained from the observations; simula
tions with an accurate period should help in determining if
the discrepancy is due to instrumental problems. A more care
ful analysis of the accuracy in the calculations of the
DOPVEL program needs to be made. The solution of this problem
will make it possible to average several nights and obtain the
long-term average pulse characteristics.

APPENDIX A
LIST OF PULSARS
Tables A-l and A-2 contain the pulsar candidates to be
observed using the University of Florida 26.3 MHz array and
the 45 MHz array at the Maipu Radio Observatory.
Table A-3 gives the references from which the data at
the different frequencies were obtained.
103

Table A-l. Pulsar candidates to be observed with the University of Florida
26.3 MHz array.
Pulsar
Period
(s)
DM -3
(pc cm )
AtD
@ 8 KHz
(ms)
Pulse
Energy
-29
*10
-2 -1
Jm Hz
Flux
Density
xlO-26 Jy
Pulse
Width
Half
Intensity
(ms)
Galactic
Back
ground
Temp.
(K)
Zenith
Angle
at
Transit
()
0031-07
0. 94 3
10
36.5
930
@ 61 MHz
1.0
@ 61 MHz
92
@102 MHz
32,550
37
0809+74
1.292
6
21.9
2,980
@ 61 MHz
2.3
@ 61 MHz
125
@ 61 MHz
66,855
44
0834+06
1.274
13
49.1
3,970
@ 61 MHz
3.1
@ 61 MHz
24
@ 61 MHz
24,660
24
1133+16
1.188
5
18.25
1,105
@ 53 MHz
0.75
@ 61 MHz
47
@ 61 MHz
27,620
14
1508+55
0.739
20
73
1,000
@ 61 MHz
1.35
@ 61 MHz
13
@ 102 MHz
20,713
25
1919+21
1.337
12
45.3
3,800
@ 61 MHz
2.8
@ 61 MHz
28
@ 102 MHz
78,908
9
104

Table A-2. List of pulsar candidates to be observed with Maipu Radio Observatory
45 MHz array.
Pulse Flux
Pulsar
Period
(s)
DM
(pc cm j
AtD
@ 10 KHz
(ms)
Energy
-29
xlO
t I
Jm Hz
Density
. -26
x 10
Wm_2Hz_1
Pulse
Width
Half
Intensity
(ms)
Galactic
Back
ground
Temp.
(K)
Zenith
Angle
at
Transit
()
0031-07
0.94295
10.89
9.9
930
@ 61 MHz
1.0
@ 61
MHz
92
@ 102 MHz
^6,500
26
0628-28
1.24441
34.36
31.3
2,340
@ 61 MHz
1.9
@ 61
MHz
105
@ 61 MHz
^6,400
5
0826-34
1.84896
52
47.4
300
@408 MHz
N. A.
(*)
775
@ 408 MHz
^6,400
0.5
0834+06
1.27376
12.85
11.7
3,970
@ 61 MHz
N. A.
(*)
24
@ 61 MHz
< 6,000
39
1749-28
0.56255
50.88
46.3
750
@ 80 MHz
>10
@ 34
MHz
vlO
@ 408 MHz
^34,000
5
2045-16
1.96157
11.51
10.47
N. A. (*)
^1
Â§ 34
MHz
v83
@ 151 MHz
^8,500
17
(*) Information not available at that frequency.
105

106
Table A-
-3. List of references from which the data at
the different frequencies was obtained.
Frequency
MHz
References
34
Backer (30)
53
Sieber (29)
61
Izvekova et al. (8)
80
Slee and Hill (9)
102
Izvekova et al. (7)
151
Lyne et al. (22)
408
Komesaroff et al. (10)
and Manchester et al. (38)

APPENDIX B
SAMPLING RATE, TIMING PULSE FREQUENCY
AND DATA POINTS PER SECOND
For the configuration of the data reduction system used
(Figure 2-2), the sampled values are sent to the Amdahl 470/V6
computer in real time. In that condition the rate of trans
mission of data into the Amdahl is the limiting factor for the
sampling rate.
The absolute maximum transmission rate to the Amdahl
470/V6 is 1200 Band (120 characters/second). The maximum use
ful rate is 60 characters/second; 40 characters/second is a
more conservative (but reliable) rate. Since the A/D con
verter has 12 bits, which means 3 characters per data point,
the transmission rate will be 40/3 = 13 data points per
second.
The number of data points per second (DPS) obtained from
the KIM I is given by the relationship (39):
DPS = TPPxNAVExTSAMP (B-l)
where TPP = KIM I timing pulse period (ms)
NAVE = Number of sampled values averaged before
sending to the terminal
TSAMP = Parameter for the KIM I program
The TPP for the KIM I must not be shorter than 4 ms
(250 Hz); the value used in the digitizing process for the
107

108
pulsar was 8 ms (125 Hz). The number of sampled values
averaged (NAVE) before sending through the terminal to the
Amdahl is fixed at 8.
Table B-l gives the relationship between the TPP, TSAMP
and the DPS (39).
Table B-l.
Values
TSAMP.
of DPS as
a function
of TPP
and
TPP
(ms)
FREQ
(Hz)
01
02
TSAMP
03
04
05
4
250
31.25
15.625
10.4167
7.8125
6.250
8
125
15.075
7.8125
5.2083
3.906
3.125
Data Points
Per Second
(DPS)
In order to meet the transmission rate requirements of
13 data points per second using 8 ms for the TPP, a safe value
of TSAMP = 02 must be used; that gives a value of 7.8125 DPS.
If the tapes are played back at the same speed that
they are recorded, then the resolution would be also 7.815 DPS,
which is a poor resolution. However, if the tapes are played
back with a slow-down of 4:1 the time resolution will be in
creased by a factor of 4.
Table B-2 gives the number of DPS obtained in real time
and in actual or original pulsar time (39).

109
Table B-2. Number of data points per second in real
and original pulsar time.
Samples per second
Real time
15.675
10.4167
7.8125 6.250
5.2083
Original
pulsar
time
62.7
41.667
31.25 25.00
20.8332
Since NAVE is fixed at 8, that means that 8 sampled values
are averaged together before producing one data point; in that
way, the tape is being sampled 8 times faster than the data
point rate. As an example, if a TPP = 8 ms and a TSAM = 02
are being used, 7.8125 DPS are obtained, but the tape is being
sampled at the rate of 62.5 samples/second in real time, which
in original pulsar time is 250 samples/second.

APPENDIX C
SCHEMATICS
110

Figure C-1. Phase locked loop (PLL) schematic.
Ill

180
VW'
+ 5 V
Figure C-2. Synchronizer and divider schematic.
112

APPENDIX D
COMPUTER PROGRAMS
113

MCAPGM Progr

MA I N
DATE
G0C04
12/27/37
PAGE 0001
POPTRAN IV G LEVEL 21
C NEW MCA PROGRAM FOR SPICA DATA
000 1
INTFGER44 DATA(36).TITLC(20) ,N0IN( 100 ) NI3INI ( 1001 .NURI N( 100)
0002
REAL *4 BIN( l00).PLOT(51) PLOT I (51 1 HUF( I 00 ) B INK 1 00 ) .BK AVE( 100)
0003
OATA ALF 0/4 H /AL F X/4 H /,ALFC/4H' /.ALFP/4H* /
0004
1
DO 40 1=1.100
0005
8 IN( I )=0 0
C006
niNl ( I )=0.0
000 7
NBIN(I)=0
0005
NOIN I ( I ) = 0
0009
40
CONT IN'JE
001 0
DO 60 1=1.51
001 1
PLOT I ( I )= ALFB
00 1 2
50
PLOT ( I ) = ALFO
00 1 3
PLOT( 1 ) = A LF C
00 1 4
PLOT (26 ) = ALFP
0C1 5
PLOT(51)=ALCC
00 1 6
PLOTI(1)=ALFC
001 7
PLOT I (26 ) = ALFP
00 1 5
LO TI (51 1=ALFC
001 9
0020
505
FORMAT(I5/cl00/IS/F10.0/F5.0/15)
002 1
0022
5000
FORM AT(20A4)
0023
IF( IRIT.EQ.O )IBIT = 12
0024
IF(IPSCAL.EQ.0)IPSCAL=1
0025
M A X D= 24
0026
IF(IRIT.LT.12)MAXD=36
0027
I F(NANA.EQ.O)NANA*20
0028
WRITE(6.606)TITLE.iniT,TSAMP,NANA.TPER.PZcR0.IPSCAL
0029
506
1
FORMAT( l 48TAOT'/QX.?0A4/10X,110.Ft2.4,Il0.Fi2.6.F10.0.
1 110)
0030
IF( ( I BIT E Q 12)ANO.(PZZRO.GT.2046. ) )PZrRO=0.0
0031
IF( ( I3IT.E3.12). AND. (PZEPO.LT.-2 048. ) ) P7F R0=0.0
0032
IF((I3IT.E0.8).AND.(PZERC.GT.128.))PZERO=0.0
0033
IF((IBIT.EQ.8).AND.(PZERO.LT.-128.))PZERO=0.0
C
GET TER IN MILL1SEC
0034
T PER= TPER*1000.
0035
NBMX = TPER/TSAMP4-1 .0
0036
NPER=0
0037
T DAT = TSAMP
0038
I 8= 0
0039
80
I D = 0
0040
IF( I 3IT.LT.12)GO TO 82
0041
0042
500
FORMAT(2473.18)
0043
GO TO 120
0044
82
RpAD(5.501 ,END = 4 0 0) (DAT A( I ). I = 1.36) .NTFS7
0045
50 1
FCRMAT( 36Z2. I 8 )
0046
120
IF(NTEST.NE.O )G0 TO 1
004 7
ID=I0*1
0 C 4 8
T DAT = TDAT 4-TSAMP
0049
TNPE R = F|_0 AT ( NPFP 1 )*TPER
0050
I F( TDAT-TNPER) 14 0. 160.160
0051
140
IB=IRFl
0052
IF( ( I 91 T E Q. 1 2 ) AND. (DAT A( IP) .GT *2048. ) )DATA( ID) = DATA! ID)-4096.
0053
IF((IBIT.EQ.8).AND.(DATA(ID).GT.128.))DATA(ID)=DATA(ID )-256.
0054
RUF(18)=D A T A(ID)
0055
IF(ID.EQ.MAXD)GO TO 80
0056
GO TO 120
0057
160
DO 180 I=1,IB
0058
BIN( I )=UIN( I )BUF( I)
0059
180
NRIN( I )=N9 I N( 1)41
0060
NPE R = NPF R1
006 1
IF(M00(NPER.NANA).EQ.O)GO TO 220
0062
200
I B= l
0063
IF(( I BIT.EJ. 12).AND. (DAT AC IP).GT .2040. ) )DATA( 1D) = 0ATA( ID) 4096.
0064
IF((IRIT.E0.8)-AND.(DATACID).GE.128.))DATA(ID)=DATA(ID)-256.
0065
RUF(IB)=DATA(ID)
0066
IF(I 0 .EQ.MAXO)GD TO 80
0067
GO TO 120
0068
220
WRITE(6.510)NPER
0069
51 0
FORMAT! 1 OX, l l 0 )
0070
SUM=0.0
007 1
SUMN=0.0
0072
SUM 5Q = 00
0073
SUMI=0.0
0074
SUMNI=0.0
0075
SUMSQI=0.0
115

007ft
0077
0070
0 0 79
0080
008 1
0082
0093
0004
0005
0086
0087
0080
0009
0090
009 1
0092
0093
0094
0095
0096
0097
0098
0099
0 100
0 101
0 102
0 1 03
0 104
0105
0 106
0107
0 108
0 109
0 110
0 111
0112
0 113
0 114
0 115
0116
0117
0 110
0 119
0 120
0 12 1
0 122
0 123
0 124
0125
0 126
0 127
0 128
0 129
0 1 30
OUUUUUUUU
IV G LEVEL 21
MAI N
DATE 80004 12/27/37
PAGE
on 230 I=1.N8MX
BINI ( I )=BINI ( I )+BJN( I I
NBINI( lUNQINKI ) 4NBIN( I )
SUM=SUMfBIM(I)
SUMI SUMI MU NM I )
SUMN=SUMN + N5 IN( I >
SUMNI = SUMNI-N01NI ( I )
230 CONTINUE
AV*sSUM/SUMN
AVEISUMI/SOMNI
DO 235 1=1.NflMX
SUM SO SUM S3* ( 3 N( 1 ) AVE*NBIN( I > > *42
SUMS 91= SUM SQ I < 8 IN1 ( I ) A VE I N3 I N I ( I ) )**2
235 CONTINUE
PMS= SORT{ SUM SO/SUMN)/SORT(SUMN/FLOAT(N0MX) )
RMS1 = SORT(SUMSOI/SUMNI )/SORT(SUMNI/FLOAT( NOMX) )
WRITE{6.530 > AVERMS.AVE I .RMS I
530 FORMAT(30XBLOCK AVERAGE! r 0,2 SIGMA! ,FA.3,
1 12 X INTEGRAL AVERAGE! P8.2. SIGMA! F8 3)
^0 240 1=1 .N8 MX
I P= 1
01NA=0.0
IF(NBIN BINA = BIN( I )/FLO AT(NO IN( I ) )
IF(IB IT.EQ.12)I=(1N( I )/FLOAT(NBIN( I I l-PZERC)/2 048.*2 5.
1 p L OAT ( IPSC AL ) 2 6
IF{ I BIT.E0.8)IP=(BINC I )/FLOAT(N9IN(I ) >-PZERO)/I2 8.*2 5.
1 FFLOAT ( I PSCAL ) *26
511 IF( IP .LT.1 ) IP=1
I F( IP.GT.51 ) IP=51
PLOT(IP)=ALFX
PI=1
O INIA = 0.0
IF(NRINI(I).EQ.O)GO TO 512
BINI A=B INI (I )/FLOAT(N0IN 1(1))
IF(IBIT.EQ.12)IPI = (0INI( I )/FLO AT (N9INI (I ) >-PZERO )/2O4 0.4 25.
1 F|_0 A T ( IPSC AL ) *26
IF(IBIT.EQ.8)IPI=(BINI(I )/FLO AT(NBINI(I) )-PZERO)/I28.*25.
1 rLOAT(IPSCAL)426
512 IF( IP I .LT.l ) IPI = 1
IF( ID IGT.51 ) IPI=51
PLOT I ( IPI )= A LF x
3KAVE( I )= BINI A
NUB IN( I )= N3INI (I )
W RITF(ft. 520) BINA.NBIN I) .BINIA.NBINI( I ).PLOT.PLOT I
52 0 FORMAT ( 1X.F9.3. I4.F10.3. 15. IX. 102A1)
PLOT(IP)=ALF0
PLOT IC IPX )= ALFB
PLOT( 1 ) = ALFC
PLOT(26) = AL F p
PLOT( 51 ) =A LF C
PLOT I (1 ) = ALFC
PLOT 1(26)=ALFP
PLOT 1(51 )= ALFC
nIN( I ) = 0.0
N BIN( I )=0
240 CONTINUE
GO TO 200
400 STOP
0
LSI 5 1
1 2
16.0000
1 0
1 .000000
0
1
THE FOLLOWING ARE SAMPLES OF THE CONTROL CARDS
(COMMENTS DO NOT NEED TO BE ON CARDS)
NUMBER OF BITS DIGITIZFD (8 OR 12)
TIMING PULSE PERIOD (MSEC ) .2MSEC*TS AMP 4NA VE(= 8)
NUMBER O'7 PULSE PERIODS AVERAGED
PULSAR PERIOD (SECONDS)
PLOT OFFSET
PLOT MAGNIFICATION FACTOR
0002
END
116

MCAPGM1 Progr

118
-<
-y v
ZO
CD
2
zox
<
--X
o- a
ozu.
Z-X
O
CD O
Z -

-IL X
o:*
OCD S
- U
L.
U
(T
<~o
OOJ
NO.
uu^
a*-
CD
a -o
o~-j
u.a
n
X
< cr - o
o<-
O-
a Z
a
CD
uw .
*o#
Ui
*h- <
JZ C
z-a
xo
o
x *o c
O tool
U. H O II H '
J--~*
*< ----Z
O

o
O
o
O

u.
1
i
-CM
-CM

O 1
O 1

cm
-O
O

<
<
IL
y-
y-
-J
<<
< <
< O
OH
O h-
u-
II <
II <

-o
-O
ft.
O
O H
C M
- O
o o
-O
-O
_J
Org
O II O
<
<
<
a -
n o o
L-
y-
u
UJ u.
o a o ii
y-
h-
< <
< <
CD
N
a uj ii o
CD
h
O H
a
a o
uj no cr
UJ
UJ
<
<
n a a uj
P
y-
-O
O

(X-
Q. UJN
Z
z

o
UJ
- NQ.

a
a x
- &-

4
<
UJ
h- o

O
o
N

<\J
n
(M pg
rvj pg
a
< X
* O

Z 4
O CsJ CM

h-
O
h-
or -
< <
OJ I OJ
N
H
Oh
cg
OID
ID X
Z o
1

O
(M
O
ft. o
OJ
->-

h-
Q.
O _Jh- h-

o-
O
O

X X
tO J

-o
h-
-o
<0
OO
w

Z X
X
a cr O o
<
< O
<
O
< -
< O
h- -
uj Uj & cr
y-
h- a o
o
h- <
z

h*
NNUIUJ
<
< UJ -
<-

o
N
h- a
a a n n
cuo
o cl
o < o
o
o < o
a -
_J
<
o. o.

y- o
c

o
X U.
M <
l* oCDh-
--u

-

o

< X
- U
II CM
OO UJ o
OO
O Q -
0,0 -
UJ
0*0
id
H cD
O 11
Z z O
HO
Oh- *
ZO h ~

ZO H
h- _l
h- ft
X < J
<<2Z -
*
* 4 0
< z u-

< z
Ah
< < -J 4
O II
ii cr a <
<~0 ?4
<
< o
cr
H O-
CD-
X < H
*j a
OO -
o o o. uj
OO ID
z
- iQO
UJ-
Q. U h-
O
-Z -
ftJM *-01
-z
z- xa -
OJ 4-
<
pg -
a o
ffl auuiinoo-

CM-h-
- X o <
OJLU
uj-o -0-0 CD
z
- o
z -
-lLIDUU.li.ll J jh-Oa
o o
o
OCD
CD UJ
, < x - z

< X

m Ju.k j jj<<^LOrf
UJ
OOOOZ
Q.
o n
eg uj H h- ft.
OOh<
-a
o Oh- <
o
o
hOO
UJ UJ UJ UI X
X
h- O N
O M Z 4 < Z
UJUJ < x Z CD 4 UJ
UJ U < 1
-X o
II < < M H II* o O UJ i
_jlD
a ex
<
Jlf) ^ o
iD hOh
o o ii z a c.
O O O o o*
y ii ii ii ---- *pg
J CM
<
h-
.
-<
H f-h- Y- a UJ UJ
O UJ ft. ft.
CD
y-
04 OJ
H- -
ni- i
id- oil. y-.
y- y- n a oj cd ii uj z
- LU U C.
h- h- II O CM
o o o
o O O u
O
< --z
H ZZ-
< O- -CD
O O CD CD Z
pgir- idn z h--
O- ~ IflKh^-Hhl-ht-C X OT n**QhZH
oooooooo oata.ao.aQQ:u.au.-i*iL
O ID O
D CDCD (DO. H H O I
.H H II M H
* axai-o
*uTirÂ¡ruj!sÂ¡!mSoj
.-HOhZ Zh- -
CD
OO
II w<
OU.UJ
-a
< o w <
IKOI
a < cr
OOU-O
u. oa u.
-UJ*-|i<4(DÂ£nCD* 0**ZO CD h-*oMOM-
h-C|ltrO Z h- UJ >- w h- Z a X h- - X h Z Zl/1
-ii
HOZ U.Hk.|i DlLOO CDQ.LlOU.U-DU.OCXO 5d3dd J
CD-0U-
h- h- CD OQCDZZ *
OJ O
CON
ID -
O
04
O o
pg -
cm
a cmh s o cMro n cr o pvjn < s g.
h- OOOOOOOOO rgrjNCJ(\jPg<\4(\|fU<\j
u o o o o o o o o o o o o o o o o o o o O o o o o o o O o O
o o O o OO o o o o o o o o o o o o o o o o o o o o o o o o
o m m mh-ao t'in^inoMiOo- Nngise^o- r\j'rwh-o,o nnnn nnnnnm*<*<*<*cniininDiD.oNNh'-NH-h-
O O O O OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
O O O O OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

FORTRAN IV G LEVEL 21
MAIN
DATE 79206
18/19/42
PAGE
007 6
0077
0078
0079
0080
008 1
0082
008 3
0084
0085
230
0086
ooe 7
0088
0089
0090
009 l
2 35
0092
009 3
5 30
0094
0095
0096
0097
0098
0099
0100
0 10 1
0102
0103
0 104
01 05
0106
0107
0 10 8
511
0 109
0 110
0111
0 112
512
0 113
0 114
0 115
0116
0 117
0 118
0119
0 120
0 12 1
0 122
0 12 3
0 124
0 125
0126
520
012 7
0 128
240
0129
400
0 130
4 06
0 13 1
408
0132
4 10
0 133
4 1 1
0 13 4
4 15
0135
420
0 136
4 25
C
C
COL S
C
c
c
c
c
c
DO 230 1 = 1 NO MX
SINK I ) = 6 IN I ( 1) +U IN ( 1 )
NBINI ( 1 ) SNUINI ( H NO IN( I )
SUM= SUM-f B 1 N(I)
SUM I =SUM I 4-0 IN I ( I)
SUMNsSUMN*KBlN( I)
SUMN 1 = SUMNINO I Ni C I )
CONT INUE
A VE = SUM/SUMN
A VE I S UM1/SUMNI
DO 235 1*1# NO MX
SUMSQ = SUMSOH01N< I ) A VE NBIN( I ) )*?
SUMSOI* SUMSQIC BIN I( IJ-AVE INB IN I ( I ) ) **2
CONTINUF
RMS= S0RT( SUMSQ/SUMN)/SORT(SUMN/FLOAT (NBMX) I
RMS I SORT (SUMSQ I/SUMNI ) /SORT( SUMNI/FLGAT(NBMX))
WRI Tf(6.530 >AVE .RMS .AVEl .RMS I
FORMAT(30X. 'BLOCK A VERAGE Z ,F8. 2. SIGMA:' ,F8.3.
12 X INTEGRAL AVERAGE: '.FB.2. SIGMA: .FQ.3I
DO 2 40 I = 1 .NBMX
I P= 1
BINA =0.0
IF(NOIN B IN* =B I N( I ) /F L O AT ( N B I N ( I ) )
IFUBIT.F0.I2) I Ps ( B INC I ) / FLO AT ( NB IN ( I ) )-PZ ERO )/ 2040. *25.
FLOATt IPSCAL) t26
IF( IOIT.FQ.8) IP = (BIN{ I )/FLOAT(NBIN( I ) )-PZERO)/I28.*25.
FLOAT( I PSC AL) 2 6
IFC IP.LT.l ) IP*l
IF( IP.GT.51 ) 1P*51
P LOT C IP)=ALFX
I PI* 1
BINI A = 0 0
IF BINIA=BINI(I)/FL3AT(NBINI( I))
IF(IB1T.EQ.12)IPI = * FLOAT ( I P SC A L ) 2 6
IF(IBIT.E0.8MPI=(BINI( I )/FLOAT ( NB I N I ( I ) ) -PZ ERO ) / 128. *25.
F LOAT( IPSC AL)26
IFC IP I .LT.1 ) IPI*1
IF( IPI .GT.51 ) IP I = 5l
PLOT I( I PI ) ALF X
BKAVEC I ) O INlA
NUBIN(I)=NBINI(I)
R I TE( 6.52 0 )B l NA, NB I N( I ) OI NI A NB I Ni ( I ) PL CJT PL OT I
FORMAT (1X.F9.3. I4.F10.3. 15. IX. 102A1 )
PLOT( IP)=ALF B
PLOT IC IP I ) *ALFB
PLOT CI ) = ALFC
PLOT(26)=ALFP
PLOT (51 )= ALFC
PLOT I (1 ) = ALFC
PLOT 1C 26)=ALFP
PLOT I (5 1 ) = ALFC
B INC I )=0.0
NOINC I ) = 0
CONT I NUE
GO TO 200
VRITE7. 406 ) T I TLE
FORM AT( 9X.20A4 )
DO 410 1=I.NBMX
* RITE(7,411 )DK AVE( I ).NUBIN( I ) I
FORMAT (F10.3.I 4.T76 13)
*PITE(7.420)AVEI.RMSI
FORMAT CF8.2.F8.3)
STOP
THE FOLLOWING ARE SAMPLES OF THE CONTROL CARDS
0
I 5
1 2
I 6.0000
1 0
I .000000
0
1
(COMMENTS DO NOT NEED TO OE ON CAROS)
NUMBER OF BITS OIGITIZFD (8 OR 12)
TIMING PULSE PERIOD (MSEC) .2MSEC4TS AMP 4N A VEC *8)
NUMBER OF PULSE PCRIODS AVERAGED
PULSAR PERIOD (SECONDS)
PLOT OF F SC T
PLOT MAGNIFICATION FACTOR
0002
0 137
END
119

P2045 (or P1133) Program

121
< .D3S I
XI T *2.3 4 = 3 WI i 9NI9VajA9 4X 9 49 I * SXDOU JO a3HwON. C 1 i 4 / ) i 9tNM3 01 oz
3N | 1* XieNt O I 94 >>311 6*
< 9I .a NOIj. X 3I3 I NO V * X 5eLJ1
9H9IS Gi*XSC*J* = VW01S.*XC t'BJ* .sdOVdilAv XDO 1U C I 1 /) i V *M I 00 fe*
9 V* 4 9 r SMi 101*3 IS 4 A vi DI < 00 9* ) 3 II ' Z*
(f02M2mVh*JOJ OVC <*
C 02* 1* I I ( 1 I 1 )1N3 alus *t = i ose oo >*
< / ZH* 9:A3Nl
anoaad Das 49#34 =ooia3d osi4 .vi-svozasd vs too *c ei )i 9*30.3 cce c*
oji die ( ce e *> an a .* 2*
< coft ) Ji i n oze i N
< 1* I)AVX-U = < T)3AV OIC C9
( T l )N!r.N = ( !> a3dN 6-..
NIbN M al O 1 C 00 65
( t n*i )NiunN4^3dns = a3cjns boc ss
N ItN* 1= IN 90C 30 SS
o#o=i3cjn ts
ioi9is*soi-i eb
< l >9IS =1019IS ZS
m avMi = Avioi soe is
02c ni no es
UiJlSt.sOlSHi 6 9
c ((I-NI9N ) iNiflN/Hins > /os3arb ) iaos*inioi s 99
Oi i]o(NiLN/Hi*nq )s j/,ii z 9
(xM3dN4nmn=Hi*ns 00 e 99
< x> 3d N < *1033 >1* es Jtir.S = 0S3fS s 9
2* \$ O ) S Jd = ( > )0S 33 9*
N ) 3 AV-AV 101= I N) 53*3 f
ni kjN i = x oce 00 cv
0*0-OS 3JOS 19
o *o =hi*ps c 9
IN30/1WON=AV101 6e
( I ) ladN+ IN 30 = IN3Q 052 bC
( I )U3dN9( I )3AV4 IhONsIrtPH Â£
N I GN *1 I O S 2 U O *C
o*c=in 30 se
o *c= Ikon 9e
iDa3dN/wnN = ir 3av ore v.e
N3G-(r ) jddN 2C
< r4i )ni triM n3o=n33 osi le
ir*1>niqon*iri)ava i4nN:hnN ce
mQNMsl OSI OC 62
O O =N 3CJ 92
o*c=wns lz
NI BN Iaf 002 OU 52
( !M I3JDQ* CN* IN) AVN1H* XlfeN *2=IN II 00 92
N!bN 41 =N 2 1 I OO e
19 ) A9X*l-< I ) AV X Jl= ) J3j 30 0 11 22
XlN C-ti C I I OC 12
50C OI 09 < l 3"1* >3QN> 3 I C 2
SON I1N0 3 02I 61
(G*3 Z^J) iVhdUJ 51 fe I
( ( I IAVXbl)St9 =( I )AVXi Zl
< I >9! S4 ( I ) AV>91< 5 l *5)0933 *t
< 9| Â£ C 13) 19*303 CI SI
< (r i ) av a i ) suv = c r i ) ax> ih coi 91
c r 1) n mnN ir 1) Avxiui o* 5 lux j 9 ei
MOKl=r COI QU 21
(9902 )19*933 6 l l
<02 4 l = 1 I 4< I 1 4 I > 1N3C! ) < 6 4 5)09 3 3 OI
X WN 1 =1 C2 I OO A
< *6 3 4 9 I C ) i vwnld 9 9
O01.33d 49 9* *NI UN 4X1UN< fe4 5>C9 33 Z
/ Hl/)(Nin*/ |M I/HI1'/W1 /IVA'/'Hl/'U H/*H l/XVAV 9190 *
99 4 I W9 I S4ld9 I S4 d.l AV 4 JO 19A d3931N| S
< oo i ) a3dN < oo i 4o i )MnrN mD93ini 9
WII4 1 330 4 <0 I >3 JU30* 13S J 3 4 I kON'kON TV3d L
<0 01 )f*S 3a <00 1 )S 3d4 < 00 l )3 A9 < C I >9 1 4 < O 1 ) AV x* X 4 < GO l 4 t l ) AVX 1u IV 3*J c
ucr *

7 1
KB ITC(6.42 5)
72
4?5
FOPMAT( // MIN'.TQ. 'AVrPAGET 17. PERIODS* 1
73
SITE!/ .430 1
74
430
FTI-MATI' N'JMMFr. * T l 7 AVERAGED )
76
OFF St T= T EK AV< 1 )
76
OFF 1 =0FFSET-0.5
7 7
DO <100 K = 1 Nil IN
79
VALUE= ( AVE(K 1-OFF 1 ) MA G
79
VALUE= 1 ADS < VALUE )
H "3
IF (VALUE.LE.O)VAL UE = 1
81
ir ( value c.r 100 iv alue= i oo
t?
3VLR< IfiTAV OFF 1 ) W AG
t
3 I C.PL= ( < 1CT AV 4T HS 1C.)-OFF l 1 *MAG
H 4
sk.i*( 96
IF (SI GMI.LF. .0) SIGMjsl
6
IF (SI M l.GC. 100 )c IGM 1=100
87
IF ( A VER.LF 0 1 A Vt P* 1
84
IF (AVER GC100 )AVfP= 100
89
IF (S 1GML.L 0 IS 1 GPL = 1
90
IF( 51 GPLGE 1001SIGPI =100
9 1
DO 660 J=1,100
92
6 50
LGT ( J) =MLNK
9 3
MLCT(l)=LIM
9%
PL CT ( 1U0)=L IM
95
PL C T (AVER)=AVAX
96
L0 T( SI GPL 1 = TP SiG
97
PLOT 98
PLTT ( VALUE 1 = VAL
99
R I T F. ( < 7?b )K AVE ( K ) .NPER ( K) ( PL CT (M J m= | | 00 I
1 00
72b
FORMAT (13.T8.F8.3.T17. 14.T25.10 0A1 )
1 01
800
C CM 1 PUF
1 0 2
*>R ITE ( t- .900 I
1 03
900
h U R M A T ( 1 )
104
STEP
105
ENO
F E NT RV
122

APPENDIX E
26.3 MHz ARRAY CALIBRATIONS
In order to obtain the minimum detectable flux density
and then estimate the peak flux density for pulsar PSR1133+16,
several calibrations were performed with the array.
The standard noise source was provided by 2 Sylvania
5722 noise diodes; as a secondary noise source a Hewlett
Packard amplifier was used. Several radio sources of known
flux density and the galactic background temperatures were
also used in the calibrations.
Figure E-l is a block diagram of the 26.3 MHz array
signal transmission and calibration system. Usually the
calibration signals are introduced at point (l) (Figure
E-l). In order to improve the calibrations, the signal was
later introduced at point (2) (Figure E-l); by doing this,
any uncertainties in the gain (or losses) of the 10 pre
amplifiers and the 2 N-S Butler matrices are avoided.
Two 1 input-to-6 outputs 50-ohm power splitters were
used to split the calibration signal into the 10 preampli
fier inputs. To properly match all the outputs, two power
splitters output were loaded with 50-ohm resistors.
As a result of introducing the calibration signal at (l)
and then at (2) a net gain of =6 dB was determined for the
10 preamplifiers plus the 2 N-S Butler matrices; several
123

Figure E-l. Block diagram for the 26.3 MHz array calibration.
124

125
other gains and losses were also determined for the
preamplifiers, 250 KHz filters and aluminum coaxial lines;
all the losses and gains determined are shown in Figure E-l.
E.l Losses Between the Dipoles and Point (j)
The procedure to obtain the losses between the dipoles
and point (2), which includes the 10 E-W Butler matrices,
the hybrid rings, the matching networks and the parallel
transmission lines, is not as straightforward as it is to
obtain gain and losses for the rest of the system. An
indirect method was therefore used in this thesis. First,
the recorder deflection due to the galactic background was
measured. Then a signal from the noise generator was intro
duced at point (2) (using the indicated switch), and was
adjusted to give the same recorder deflection as that from
the galactic noise. From the ratio of the noise temperature
of the introduced signal to that of the part of the sky from
which the galactic signal was obtained, which is available
in the literature, the losses can be obtained. A summary of
the results is shown in Table E-l.
The noise temperature for the noise diodes has been
calculated using the following relationship (17):
T = T + inf where T0 = ambient temperature (= 290K)
_ I Q
e = electron charge (= 1.6x10 C)
I = diode's plate direct current
k = Plank constant (= 1.38xl0~23 J K)

Table E-l. Summary of the readings and calibrations made in the 26.3 MHz Array.
Right
Ascen
sion
Declina
tion
Calibra
tion at
Point (l)
(dB)
Calibra
tion at
Point (2)
(dB)
Noise
Temp, at
Point (2)
(K)
Corrected
Noise
Tenfjb. at
Point (2)
(K)
Galactic
Back
ground
Temp. (*)
(K)
Ratio
Losses
(dB)
0 9h
4 5m
7.6
30.5
37
14,043
3,019
15,858
5.3
7.26
14h
m
50
26.0
29
34.7
23,881
5,063
25,000
4.93
6.94
19h
52m
26.0
27
32.7
37,704
7,994
44,000
5.5
7.4
(*)
Obtained
from references (26,
40)
Average
Value
5.25
7.2
126

127
Substituting in equation (E-l) the values for the
constants, using the value R = 50 ohm and expressing I in
milliamperes, the following relationship can be deduced:
T = 290 + 289.85 I (E-2)
Equation E-2 gives the temperature in K and it is valid
for R = 50 ohm.
Galactic background temperatures for the first two
points in the sky have been obtained from Turtle et al.(40)
at 26.3 MHz. The value for the third point was obtained from
measurements made by Blythe (26) at 38 MHz and corrected to
26.3 MHz using the galactic background spectral index given
by Findlay (27) of n = 2.65, which is valid for the range
20-400 MHz.
It is interesting to compare the losses obtained in
Table E-l with the losses obtained by summing the estimated
dB attenuations along the path between the dipoles and point
(2) The list of estimated or measured losses of the indi
vidual components is given in Table E-2 (41).
The average of the measured losses given in Table E-l is
7.2 dB which agrees well with the losses computed from
Table E-2.

128
Table E-2. Losses between dipoles and point (2)
considering the partial losses.
Minimum dB
Maximum dB
Full wave twin line
. 30
.30
Hybrid ring
.45
.45
Phase cable
. 80
1.10
Coaxial cable line
1.90
1.90
E-W Butler matrix
1.44
1.98
Line connecting E-W to
N-S Butler Matrices
.5
1.10
Total
5.39
6.73
The corrected noise temperatures at point (), given in
Table E-l, is the noise temperature of the noise source cor
rected for the 5 dB losses in the aluminum coaxial line and
the losses in the power splitter (factor 0.76); losses in the
power splitter consider the losses due to the two 50-ohm
resistors and the losses at 26.3 MHz of the power splitter
itself.
E.2 Antenna Effective Area
By using the losses between the dipoles and point (5)
3C433, the antenna effective area can be calculated; the
method consists of obtaining the change in antenna temperature

129
AT for a radio source of known flux density S. The
a
relationship that gives the effective area is:
2k AT
Ae = ^ (E-3)
AT^ is the change in antenna temperature referred to
the dipoles, corrected for losses e between point (2) and
the dipoles. If AT' is the change in temperature measured
ci
at point (2) then the actual change in antenna tempera
ture is:
AT'
AT = (E-4)
a e
where e = transmission line efficiency, which in this
case considers the losses between point (2)
and the dipoles.
From the calibrations, the following values were
obtained for 3C433:
S = 267 jy
AT' = 295K
a
T = 25,000 K
a
e = 0.19
Using equations E-3 and E-4, the following value is
obtained:
A = 16,050 m2
c
This value needs to be corrected by a factor p^ = 0.69,
given in the Array program (11), which considers the fact
that the source is off the center of the beam. The corrected
effective area of the antenna is then:
A = 23,260 m2
e

130
The value of the effective area can be also expressed
in terms of the change in antenna temperature in K, divided
by the flux density in Jy (K/Jy); this quantity is usually
denoted as A* (12).
e
AT x P-,
A| !;r (E-5)
Substituting in E-5 the values
AT = AT'/e = 1553K
a a
S = 267 Jy
P1 =0.69
the following value for A* is obtained
A* = 8.3 K/Jy

LIST OF REFERENCES
1.A. Hewish, S. J. Bell, J. D. H. Pilkington, P. F. Scott,
R. A. Collins, "Observations of a Rapidly Pulsating
Radio Source," Nature, 217, 709 (1968).
2. Yu. M. Bruck, J. G. Davies, A. D. Kuz'min, A. G. Lyne,
V. M. Molofeev, B. Rowson, B. Yu. Ustimenko and Yu. P.
Shitov, "Radio-emission Spectra of Five Pulsars in the
17-1420 MHz Range," Soviet Astronomy, 2_2, 588 (1979).
3. F. N. Bash, F. A. Bozyan and G. W. Torrence, "Observa
tions of CP1919 and NP0532 at 38 MHz," Astrophysical
Letters, 7, 39 (1970).
4. Yu. M. Bruck and B. Yu. Ustimenko, "Decametric Pulse
Radio Emission from PSR0809, PSR1133, and PSR1919,"
Nature Physical Science, 242, 58 (1973).
5.Yu. M. Bruck and B. Yu. Ustimenko, "Decametric Radio
Emission from Four Pulsars," Nature, 260, 766 (1976).
6. H. D. Craft Jr. and J. M. Cornelia, "Frequency
Dependent Pulse Width for CP1133," Nature, 220, 676
(1968).
7. V. A. Izvekova, A. D. Kuz'min, V. M. Molofeev and Yu.
P. Shitov, "Energy Spectra and Pulse Shapes of Pulsars
at Metre Wavelengths," Australian Journal of Physics, 32,
25 (1979).
8. V. A. Izvekova, A. D. Kuz'min, V. M. Molofeev and Yu.
P. Shitov, "Radio Flux Density and Pulse Profile of
Pulsars at 102.5 and 61 MHz," Soviet Astronomy, 23,
179 (1979).
9. O. B. Slee and E. R. Hill, "Pulsar Energy Fluxes at 80
MHz," Australian Journal of Physics, 2_4, 441 (1971).
10.M. M. Komesaroff, D. Morris and D. J. Cooke, "Linear
Polarization and Pulse Shape Measurements on Nine
Pulsars," Astrophysical Letters, 5, 37 (1970).
131

132
11. M. D. Desch, "26.3 MHz Array Reference Manual,"
unpublished.
12. M. D. Desch, Ground Based and Spacecraft Studies of
Jupiter at Dacameter and Hectometer Wavelengths (Ph.D.
Dissertation, University of Florida, 1976).
13. M. D. Desch, T. D. Carr and J. Levy, "Observations of
Jupiter at 26.3 MHz Using a Large Array," Icarus, 25,
12 (1975) .
14. F. Reyes, Radiotelescopio en 45 MHz para Fuentes
Extragalacticas, Memoria para optar al titulo de
Ingeniero Civil Electricista (Thesis, University of
Chile, 1977) .
15. J. May, F. Reyes and J. Aparici, "A 45 MHz Radiotelescope
for Galactic and Extragalactic Research," First Latin-
American Regional Astronomy Meeting, Santiago, Chile
(1978).
16. M. R. Viner and W. C. Erickson, "26.3 MHz Radio Source
Survey. II. Radio Sources Positions and Fluxes," The
Astronomical Journal, H), 931 (1975).
17. J. D. Kraus, Radio Astronomy (Mac Graw Hill Inc., New
York, 1966).
18. G. R. Huguenin, R. N. Manchester and J. H. Taylor,
"Properties of Pulsars," The Astrophysical Journal, 169,
97 (1971).
19. F. D. Drake and H. D. Craft, "Pulse Structure of Four
Pulsars," Science, 160, 758 (1968).
20.A. D. Kuz'min, V. M. Molofeev, Yu P. Shitov, J. G.
Davies, A. G. Lyne and B. Rowson, "Spectra of Nine Pulsars
at 61-1420 MHz," Monthly Notices of the Royal Astronomical
Society, 185, 441 (1978).
21. R. D. Ekers, A. T. Moffet, "Further Observations of
Pulsating Radio Sources at 13 cm," Nature, 220, 756
(1968).
22. A. G. Lyne, F. G. Smith and D. A. Graham, "Characteris
tics of the Radio Pulses from the Pulsars," Monthly
Notices of the Royal Astronomical Society, 153, 337
(1971).
23. J. May and J. Aparici, Personal Communication (1980).

133
24. B. Y. Mills, 0. B. Slee and E. R. Hill, "A Catalogue of
Radio Sources Between Declinations +10 and -20,"
Australian Journal of Physics, J^, 300 (1958).
25. O. B. Slee, "Culgoora-3 List of Radio Source Measure
ments," Australian Journal of Physics, Astrophysical
Supplements, Number 43, (Nov., 1977).
26. J. H. Blythe, "Results of a Survey of Galactic Radia
tion at 38 MHz," Monthly Notices of the Royal Astro
nomical Society, 117, 652 (1957).
27. J. W. Findlay, "Absolute Intensity Calibrations in
Radio Astronomy, Annual Review of Astronomy and Astro
physics 4, (1966).
28. A. J. Turtle, A. E. Vaughan, "Discovery of Two Southern
Pulsars," Nature, 219, 689 (1968).
29. W. Sieber, "Pulsar Spectra: A Summary," Astronomy and
Astrophysics, 28, 237 (1973).
30. D. C. Backer, Personal Communication to S. T. Gottesman
(1975).
31. R. N. Manchester and J. H. Taylor, Pulsars, (W. H. Free-
ma and Co, San Francisco, 1977).
32. R. N. Manchester, J. H. Taylor, G. R. Huguenin, "Observa
tions of Pulsars Radio Emission II. Polarization of
Individual Pulses," Astrophysical Journal, 196, 83 (1975).
33. G. R. Huguenin, R. N. Manchester and J. H. Taylor,
"Properties of Pulsars," Astrophysical Journal, 169, 97
(1971).
34. J. D. H. Pilkington, A. Hewish, S. J. Bell and T. W. Cole,
"Observations of Some Further Pulsed Radio Sources,"
Nature, 218, 126 (1968).
35. Yu. I. Alekseev and S. A. Suleimanova, "Polarimetry of
Pulsars at 3.5 m Wavelength," Soviet Astronomy, _21, No. 2,
180 (1977).
36. Yu. P. Shitov, "Fine Structure of Pulsar Radio Spectra,"
Soviet Astronomy, _16, 383 (1972).
37. R. N. Manchester, "Pulsar Rotation and Dispersion Measures
and the Galactic Magnetic Field," The Astrophysical Jour
nal, 172, 43 (1972).

134
38.R. N. Manchester, A. G. Lyne, J. H. Taylor, J. M.
Durdin, M. I. Large and A. G. Little, "The Second
Molonglo Pulsar Survey Discovery of 155 Pulsars,
Monthly Notices of the Royal Astronomical Society,
185, 409 (1978).
39. J. P. Oliver, Personal communication (1978).
40. A. J. Turtle, J. F. Pugh, S. Kenderdine and I. I. K.
Pauliny-Toth, "The Spectrum of the Galactic Radio
Emission," Monthly Notices of the Royal Astronomical
Society, 12, 297 (1962).
41.J. Levy, Personal communication (1980).

BIOGRAPHICAL SKETCH
Francisco Reyes was born in Santiago, Chile, on
December 8, 1941. After finishing high school and with
a strong desire to study electrical engineering (in which
he was encouraged by his parents, sisters and brother), he
entered the School of Engineering of the University of
Chile. It was there where he became interested in astronomy
while taking a course with Professor Claudio Anguita. His
interest grew when he was allowed to use the old Heide
refractor at Cerro Caln. He started his amateur astro
nomical activities by building his own small Newtonian
reflector and spending hours and hours watching the inter
esting southern sky (in which he was enthusiastically sup
ported by his two sisters Lucia and Guadalupe and his
brother Juan).
His professional interest turned to radio astronomy
when in 1974 Jorge May asked him to work on the 45 MHz
array in Maip as a thesis project.
In 1975 he married Maria Angelica, and they now have
two daughters: Carolina and Daniela (who just joined the
family).
In 1977 he obtained a degree in electrical engineering
from the University of Chile, and in September of the same year
135

136
he came to the University of Florida to continue his studies
in astronomy. While studying at the University of Florida
he has been teaching elementary physics laboratories, elemen
tary astronomy laboratories and digital electronics labora
tories.
His other interests are mountain climbing, Andean music
and photography.

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a thesis for the degree of Master of Science.
Thomas D. Carr, Chairman
Professor of Astronomy
and Physics
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a thesis for the degree of Master of Science.
Alex G. Smith
Professor of Astronomy
and Physics
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a thesis for the degree/of ifast^ of fkpience.
Jfyh P. Oliver
A'sociate Professor of Astronomy
This thesis was submitted to the Graduate Faculty of the
Department of Astronomy in the College of Liberal Arts
and Sciences and to the Graduate Council, and was accepted
as partial fulfillment of the requirements for the degree
of Master of Science.
December, 1980

HLLCK AVERAGE-
PULSAR
PSR?
0A5
- 16
LR I UD
19613H3
SEC
EEC
OUf NCY=45
MHZ
P2045 A
MAY
1 5
79
1ST APT
0 9 H
A QM
?0S
STOP
0 9H
5PM
35S
T SAMP
02
P2045R
MAY
15
7^
TS7A t T
09 H
52M
A 3 S
STOP
09 H
5b M
SOS
T S AMP
02
P20 4 EC
MAY
I 5
7 9
tSTAR T
09 H
55*
5RS
STOP
09M
59M
1 AS
TS AMP
02
P?0450
MAY
1 5
79
TSTAKT
09H
59 M
1 6S
ST UP
0 9 H
02M
J9S
T S AMP
02
p ? 0 A 0 E
MA Y
1 5
79
TS1ACT
l OH
0 2M
AOS
S TCP
l OH
05M
55 S
T SAMP
0?
P2045F
MAY
15
7^
1 ST APT
1 OH
O^M
5 6 S
STUP
l OH
08 M
20S
TSAMP
02
D 0 7
S IGMA =
0.
J5 9
3
SI GMA
= 0
1 7 t
MA GN IFICA T ION =
l 00
NU'VJER CF D L CC K S 6 AVERAGING T 1 ME = 103?.5 SCC
pin
AVERAGE
PE R l CO S
NJVBE
A VCNAGED
l
12390A
5A0
2
1 2 a. 037
5 AO
l PA *50
5 AO
A
l?A.3 40
6 AO
5
123.995
5 1 0
f->
1 2 3. 1 7A
5 10
7
l 23.*>6A
5 1 0
9
12A.013
4 90
1
1 2.3.9P5
490
10
l23.977
U 90
1 1
123.901
A 90
l?
l 23 *66
490
1 3
12A.070
5 20
1 A
12A.0 52
540
15
1 2A. ORE
540
16
12A.092
17
123.987
5A0
1 8
1 2A l l l
540
1 9
l 24 .08 1
5 40
20
1 2 4. 0 7 A
540
2 1
1 2A. l92
5 40
22
l 2A .025
540
23
124.020
5 A 0
?A
124.000
5 40
?5
1 24.070
540
26
I 2a.000
5 40
?7
1?7.952
5 A 0
29
124.137
540
29
1 24. 0 A 8
540
o
127*963
540
31
lPA.08?
4 77
32
123. 5 39
540
1 7
123.94f
540
3A
l2a.025
543
25
1?4.046
54 3
127.979
540
37
123.9?A
540
38
1 23. Rye
540
39
123.954
540
AO
1 23 ,99 l
540
A 1
1 24.04 t
540
A?
123>39
540
A3
l 24 .055
540
A+
1 23. 97 A
5 40
A 5
l2A.004
540
46
123.989
540
A 7
1 24.021
530
A 8
124.302
530
A 9
123.966
5 30
50
1 2a. 0 23
4 El
5 1
124.007
530
5?
123.94 l
5 30
53
1 ?3877
530
5 A
123.957
397
55
1 23 .992
0 30
56
123. 967
530
7
1 PA. 007
4 67
58
121.983
540
59
1 2A 01 7
54 0
*0
l 23.970
540
61
1 24 .08 1
540
62
1 2 A. 025
4 77
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
time
!
I
I
I
I
I
|
I
I
I
I
I
I
I
I
I
I
INTENSITY
I
I
I
I
21 3.16 SIGMA
C7T
CJ\

83
intensity variations, being often undetectable (28, 29).
Our experiences at 45 MHz confirm this behavior also since
it was possible to detect this pulsar only in the first
nights of the observing period; it faded out for the rest
of the nights.
-2 -1
An unconfirmed pulse energy of ^1.9 Jm Hz has been
given by Backer at 34 MHz (30). The lowest frequency at
which published data on pulse energy is available is 410
MHz; the energy is 0.120x10 ^ jm ^z 1 (29, 31). Our
measurement at 45 MHz seems rather high (perhaps suggesting
an undetected systematic error), although no direct
comparison is possible due to the lack of published informa
tion at low frequencies on this pulsar.
Pulse energy as a function of the frequency has been
plotted in Figure 3-32, including the data mentioned in
the last paragraphs.
3.2.b Pulse Width for Pulsar PSR2045-16
In order to obtain the true half intensity pulse width
of PSR2045-16 at 45 MHz, the measured half intensity pulse
width has been corrected for pulse dispersion and post
detection time constant using equations (3-8) and (3-9)
respectively.
In order to correct for pulse dispersion, the follow
ing values for the parameters were used:
Av = 10 kHz
Vo = 45 MHz
DM = 11.5 pc cm ^

Figure 3-28. Pulsar PSR2045-16 recorded on May 16, 1979, at 45 MHz.
Average over 440 periods.

OBSERVATION OF PULSARS AND OTHER SOURCES
BY
FRANCISCO REYES
A THESIS PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
1980

82
4
m THIS THESIS
3
0 BACKER (30)
A SIEBER (29)
2
O SLEE AND HILL (9)
AVERAGE 1
PULSE
-
ENERGY
-
-2 -1
(Jm Hz )
~
xlO-26
.5
.4
.3
'
.2
T
\
A
.1
2(
3 30 40 50 100 200 300 400
FREQUENCY (MHz)
Figure 3-32. Pulsar PSR2045-16 low frequency energy
spectrum.

PUL 3 AO P
C-W 1 l
1.1
1 6
>PP| nO = l
, 1 0 7 P A 1
S5C
FRrQJFNCY=2fc3 MHZ
"Mir* a*-
MAY
<;
5
- TAP*
o 1 H
? r.i i oc
S
01 H 1AM 1 OS TSAMfJ
02
ntl2wn6
M\Y
?5
7?
tgt apt
0 1 H
14M IOS
STOP
01 M 104 S TSAM3
02
BLOCK AVERAGE* 1 2? .6 09
1 GMA
r
IT >
7 r
I r,M a = c
.217
MAGNIFICATlONs
1 00
NUMBER CF BLOCKS? ? AVF5 AGING T: *tr: A3?- 6 s^c
P 1 N
AVFPAGf
3 p 51 o o r.
n j Mirn
AV EP AGED
i
1 22 6 5 A
224
7
l ?2 4 1
3 90
12? TP
10
A
122 339
190
122574
iso
'
1 ?? M 1
190
7
1 22 6 2 5
190
1 22 .' 1 2
3 90
9
1 2? *.oo
190
1?
122.646
1 OQ
1
122,>C1
190
1 ?
122 es
190
13
122.564
390
11
1 22 .S2 0
390
1 5
122 6 11
3 90
1 3
1 22 561
1 90
1 *
190
I 8
1 22 5 7
190
19
122.512
190
20
122.556
ion
r i
122.637
190
122554
190
23
P?|B
19?
2A
122.692
190
PI
1 ?? H 1
3 90
1 ?2 uo
ion
27
122.62*
1 90
.
l 22 -- 1 2
39 3
2 )
1 2 * 4)7
-on
3 0
122. 649
190
1
I 2* ,*-#
150
1 7
1 22 * 25
190
31
12?6*9
190
t t
122.525
390
3-
t 2.? 6 ?
190
33
122 587
1 90
v
1 22 .64 6
3 0
3H
122 6P3
2 15
TIME
INTENSITY
Ul
Figure 3-17. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187943 s (P-80 us).

35
pulse profile at 26.3 MHz has been shifted, to align the
pulses in bin 21 and 25 with the pulses at 25 MHz.
Although the subpulse in bin 21 is much weaker than
that in bin 25 for the 390-period average, the two are
almost the same strength in the 140-period average shown in
figure 3-10 (block P112EA6). In the latter case, the
heights are 2.68 a and 2.62 a for the bin 25 and 21 pulses,
respectively. A possible implication is that the bin 21
pulse varies in intensity much more than does that of bin 25.
A small pulse appears in bin 9, when blocks P112WA6 and
P112WB6 are averaged together. This pulse is also present
in each block average separately (figure 3-5 and 3-6), sug
gesting the possible existence of an interpulse leading the
main pulse by 0.42 P (or 150). In Bruck and Ustimenko (4)
a broad and poorly defined pulse seems to appear at 0.5 P
from the main pulse, in the same plot where the double main
pulse appears(figure 3-9). So far, no interpulse has been
reported for this pulsar.
3.1.a Peak Flux Density, Mean Flux
Density and Energy for Pulsar PSR1133+16
Several continuous radio sources were observed in order
to calibrate the array, to obtain the minimum detectable
flux density, and then to obtain an estimate of the peak
flux density for pulsar PSR1133+16.
The radio sources observed and their main parameters are
presented in Table 3-1.

127
Substituting in equation (E-l) the values for the
constants, using the value R = 50 ohm and expressing I in
milliamperes, the following relationship can be deduced:
T = 290 + 289.85 I (E-2)
Equation E-2 gives the temperature in K and it is valid
for R = 50 ohm.
Galactic background temperatures for the first two
points in the sky have been obtained from Turtle et al.(40)
at 26.3 MHz. The value for the third point was obtained from
measurements made by Blythe (26) at 38 MHz and corrected to
26.3 MHz using the galactic background spectral index given
by Findlay (27) of n = 2.65, which is valid for the range
20-400 MHz.
It is interesting to compare the losses obtained in
Table E-l with the losses obtained by summing the estimated
dB attenuations along the path between the dipoles and point
(2) The list of estimated or measured losses of the indi
vidual components is given in Table E-2 (41).
The average of the measured losses given in Table E-l is
7.2 dB which agrees well with the losses computed from
Table E-2.

PULSA PSH1J1 It t~9 IQ'Js | i 8*J?3 SEC r P FO UE NC Y = 26 3 MM2
PI I 2 A6 MAY 25 75 T ST APT OlH 30M IOS STG OlH 3 AM IOS TSAMP 02
BLOCK AVERAGES 122.690
S I GMA =
0. 1 1 3
3 SIGMAr
0. J39
MAGNIFICATIONS 100
NUMBER OF BLUCKS= 1
AVFPAGING
TI ME =
162.5 SEC
t IN AVERAGE PrPJCOS
NUMBER AVERAGED
l
12?.714
1 4 0
1
2
l22.629
1 40
1
3
l22.657
1 40
1
4
1 22. 700
1 40
1
5
12? .57 1
l 40
1
122.593
1 40
1
7
122.636
1 40
)
B
12^65 D
l 40
1
9
l 22 .2 l
1 40
1
10
l22. 72 l
l 40
1
11
l 22.629
140
1
12
122 .57 1
1 40
1
1 3
1 2266 4
1 40
1
14
1 22.61 A
l 40
1
15
122.614
l 40
1
16
122.^79
1 40
1 I
1 7
122.693
1 40
1 TIME
1 Pj
122.05C
1 40

19
122.636
1 40
|
2 J
122.7a 3
l 40
1
21
l 22. 9 b 6
1 40
1
2?
122.914
1 40
1
23
122.664
1 40
|
24
l22.607
1 40
1
25
l22 .99 3
1 40
I
26
l22 .756
l 40
1
7
122.571
1 40
28
1 22 .5 36
1 40
1
29
122456
1 40
1
30
122.593
l 40
1
31
122.756
l 40
1
32
122. 736
1 40
1
*3
122.664
1 40
1
34
l22664
1 40
1
25
122.779
1 40
I
26
122.72 i
1 40
1
37
122.579
I 40
1 |
38
123. 00 0
i e
|
TNTFNSITÂ¥-
2.62 SIC^
I
I
I
2.68 sicm
OJ
Figure 3-10.
Pulsar PSR1133+16 (Block P112WA6) recorded at 26.3 MHz on May 25,
1979. Average over 140 periods.

132
11. M. D. Desch, "26.3 MHz Array Reference Manual,"
unpublished.
12. M. D. Desch, Ground Based and Spacecraft Studies of
Jupiter at Dacameter and Hectometer Wavelengths (Ph.D.
Dissertation, University of Florida, 1976).
13. M. D. Desch, T. D. Carr and J. Levy, "Observations of
Jupiter at 26.3 MHz Using a Large Array," Icarus, 25,
12 (1975) .
14. F. Reyes, Radiotelescopio en 45 MHz para Fuentes
Extragalacticas, Memoria para optar al titulo de
Ingeniero Civil Electricista (Thesis, University of
Chile, 1977) .
15. J. May, F. Reyes and J. Aparici, "A 45 MHz Radiotelescope
for Galactic and Extragalactic Research," First Latin-
American Regional Astronomy Meeting, Santiago, Chile
(1978).
16. M. R. Viner and W. C. Erickson, "26.3 MHz Radio Source
Survey. II. Radio Sources Positions and Fluxes," The
Astronomical Journal, H), 931 (1975).
17. J. D. Kraus, Radio Astronomy (Mac Graw Hill Inc., New
York, 1966).
18. G. R. Huguenin, R. N. Manchester and J. H. Taylor,
"Properties of Pulsars," The Astrophysical Journal, 169,
97 (1971).
19. F. D. Drake and H. D. Craft, "Pulse Structure of Four
Pulsars," Science, 160, 758 (1968).
20.A. D. Kuz'min, V. M. Molofeev, Yu P. Shitov, J. G.
Davies, A. G. Lyne and B. Rowson, "Spectra of Nine Pulsars
at 61-1420 MHz," Monthly Notices of the Royal Astronomical
Society, 185, 441 (1978).
21. R. D. Ekers, A. T. Moffet, "Further Observations of
Pulsating Radio Sources at 13 cm," Nature, 220, 756
(1968).
22. A. G. Lyne, F. G. Smith and D. A. Graham, "Characteris
tics of the Radio Pulses from the Pulsars," Monthly
Notices of the Royal Astronomical Society, 153, 337
(1971).
23. J. May and J. Aparici, Personal Communication (1980).

49
it is almost a certainty that the pulses are indeed due to
the pulsar.
May 25, 1979, the date on which the best pulse was
obtained, was chosen to make the test. Data were analyzed
using periods longer and shorter than the assumed period
by 40, +80, 160,1320, and 640 ys.
Periods shorter or longer than the assumed period P
by 80 ys should cause a shift of 1 bin at the end of the
390 period average.
The result of this test is shown in Figure 3-13, where
the relative averaged pulse intensities have been plotted
as a function of the assumed period length, taking as the
base the periodP = 1.188023 s given by the DOPVEL program.
It can be seen that the pulse nicely peaks with the assumed
pulsar period, and decreases sharply for shorter and longer
periods. The lack of symmetry of the plot might be due to
the unsymetrical pulse shape.
The computer plots of the results for the different
periods used in this test are presented in Figures 3-14 to
3-24.
The pulse intensity as a function of time has been
plotted in figure 3-25 for the pulse obtained on May 25,
1979. This pulse was chosen because it was the better of
the two and is well defined. This plot is useful not only
to visualize the temporal behavior of the intensity of the
pulse but also to check that no interference or isolated
transient pulse is mistakenly taken as a pulsar pulse.

Figure 3-30. Pulsar PSR2045-16 recorded on May 19, 1979, at 45 MHz
Average over 415 periods.

BIOGRAPHICAL SKETCH
Francisco Reyes was born in Santiago, Chile, on
December 8, 1941. After finishing high school and with
a strong desire to study electrical engineering (in which
he was encouraged by his parents, sisters and brother), he
entered the School of Engineering of the University of
Chile. It was there where he became interested in astronomy
while taking a course with Professor Claudio Anguita. His
interest grew when he was allowed to use the old Heide
refractor at Cerro Caln. He started his amateur astro
nomical activities by building his own small Newtonian
reflector and spending hours and hours watching the inter
esting southern sky (in which he was enthusiastically sup
ported by his two sisters Lucia and Guadalupe and his
brother Juan).
His professional interest turned to radio astronomy
when in 1974 Jorge May asked him to work on the 45 MHz
array in Maip as a thesis project.
In 1975 he married Maria Angelica, and they now have
two daughters: Carolina and Daniela (who just joined the
family).
In 1977 he obtained a degree in electrical engineering
from the University of Chile, and in September of the same year
135

ot)i c/v p oc p |
1 33*
1 6 '>e0 !n0*| 1 C13S63
S CC rp^(jrN'y = ?,, 1 ' + 7
91 1 2 WA*- M AY
2C
75 r ART 0 1 H I?'*
1 05
5TH9 0IH 3 105 TSAMP
02
l12*R6 wAY
25
70 TSTART 01' 34M
1 OS
5THP 01H 3m 045 T54MP
02
RLGC *
WPOAttE*
122
.613 F 1 <.M4 s
0-^76 3 5!GMA=
o
. ??7 MAGNIFICATIONS
100
NJ 9EK
nF BL H C< 5 x
9 AYE^AGING
' T r = 453. 2 cr"C
MJ 'HE 0
a \* ro a rpp
1
12?.746
390
+ Jf
1
;
!?*.! 7
un
1
12? >5 6
16 3
1
L
1 ?? .V
3 90
^CT7 +
l
5
1 2** ,'oo
360
f
I
6
12? 351
TOO
C
1
122,646
3?
1
i
l *60 5
3 5,C
X
I
5
12? 6?4
190
C*
1
! '
122.641
TO?
1
1 \
l 22.653
3 00
1
1 2
12.607
190

\
1 3
l22.459
390
+
1
1 4
l 2? 4
op
*
1
15
1 22. 6. 51
i 1
1-
1 ??.* '?
300
1
1 7
122 607
1O0
1
l
12? 666
3 90
1
1 >
1??.61
300
l
2 0
12? 707
<90
Â¥
1
21
1 22 751
300
21
1
??
122.65?
390
.
+ Jr
1
23
12? .6 71
390
Â¥ Jr
1
24
l 22. 653
i 50
Â¥ .
I
25
1 2
35p
1
?5
12? 6l
3 00
+
1
2"*
123.6H4
3 90
f 4
1
24
I2?.r 64
3 O')
t Â¥
1
P 9
IP? r,7g
70

1
30
12?.620
3 50
i
1
? 1
122.60?
3C?
I
4
1
3?
122 674
390
4
1
33
1 22 5 in
3 O0

1
22R
1
3C
1 2? 5?->
390
~~ Â¥
1
36
1 22.625
360
1
3 7
l 2?.*4 3
30Q
Â¥ \
1
3*
12? 635
2 20
r
1
Figure 3-24. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188553 s (P+640 ys).

9
Figure 2-1.
Block diagram of the equipment
interconnections used to record
pulsar PSR1133+16.

Table E-l. Summary of the readings and calibrations made in the 26.3 MHz Array.
Right
Ascen
sion
Declina
tion
Calibra
tion at
Point (l)
(dB)
Calibra
tion at
Point (2)
(dB)
Noise
Temp, at
Point (2)
(K)
Corrected
Noise
Tenfjb. at
Point (2)
(K)
Galactic
Back
ground
Temp. (*)
(K)
Ratio
Losses
(dB)
0 9h
4 5m
7.6
30.5
37
14,043
3,019
15,858
5.3
7.26
14h
m
50
26.0
29
34.7
23,881
5,063
25,000
4.93
6.94
19h
52m
26.0
27
32.7
37,704
7,994
44,000
5.5
7.4
(*)
Obtained
from references (26,
40)
Average
Value
5.25
7.2
126

PULSAR rr>w? 045- I (> Pf. 9 | OP*1 7* | 7** a S FC r RF OJE NC Yr % s MHZ
?,'5A?A MAY 17 7 9 rr;TA-T OOh AIM
>?0A5n?A MAY 17 79 TSTAJT 09H 4' =>2045024 MAY 17 79 TSTA~T 09M ARM
P2 0 A *>F ? A MAY 17 M T S T ART 0H 5 ** 4
O^S STn'* 09H ASM 00S TSAMO 02
n IS STOP 0 9H A PM "*9 5 TSAM> 92
AOS STOP 09H 5? M OSS TS AMP 02
3f.S STOP "QH 5SM 30 S TSAMP 0?
LOCK AV F.R AGP = 12^.829
ricma- o.064
MAC.N IF I CAT IP Ms 100
NU'PFO OF rlOCKSs A
AVFCAGIMG 5 I MC 766.3 S PC
n I m AVFCAGE *Ecin)S
NMQEh AV F.R A OrD
1 I 28 c 30 3C0
? 128 37 790
* 128. 32 300
a 128 801 390
5 1 28, B?a 3 90
~ 128,7 0-, 3 00
y l28 858 318
p. 128 83 1 3 90
128.55 390
10 128895 39J
11 120 92 A 3 SO
12 128.733 390
13 128 "70 300
1 A 1 28,826 390
18 1 R t 1 7C0
IS 128,*Op 390
17 120.775 390
18 178,767 790
1 12b 837 39C
2^ 1 28. '3 390
21 I 20 79* 3^0
22 128 706 390
27 128.770 3C0
2A l28 7SO 390
25 123 011 -*qo
2' 128- 3 19 390
2 7 1 23,7 ) 7 .3 S?
28 128 796 3 90
29 1 28.3 37 3 00
30 128 77P 300
71 128773 390
*'* 1 2 7 6 5 390
? 3 123-733 3 CO
34 120 793 390
3 i 129, 7SA 390
36 128 779 300
37 1 28. -726 390
*7 1 23 798 300
39 128 7U 390
AO 128. 8 A9 7 90
A I I 3.020 3 00
A? 123 67 390
A3 120 866 39C
A 1 l ? 7 7 *> 390
AS 123 79| 390
AS 1 20, 855 390
A7 I 28,998 390
a 8 123 852 390
AO 128,*24 390
60 1 28.7P 3 700
51 128007 700
6*> 123,9 75 390
5 J 1 28 *>4 O 790
61 120.900 717
5 5 1 2 0 * 5 1 3 50
56 l279 790
57 128 ^67 795
60 l20. 96 390
63 IP.60P 700
7.5 120 0 37 7oS
61 120,13? 390
6? 120.307 3 06
TIME
INTENSITY
03

CHAPTER 2
DATA ACQUISITION AND REDUCTION
Two pulsars, PSR1133+16 and PSR2045-16, were selected
(from the lists in Appendix A) for observation because of
their low dispersion measure, relative long period
(>1 second), high pulse intensity and favorable location
with respect to the galactic background. Pulsar PSR1133+16
was observed at 26.3 MHz using a bandwidth of 8 kHz, and
pulsar PSR2045-16 was observed at 45 MHz with a bandwidth
of 10 kHz.
The detected IF signals of the pulsar were recorded on
one channel of a magnetic tape, and a time code generator
signal on the other channel. The tapes were played back
(making a 4:1 slow down), the pulsar signal sampled, and
the resulting values stored in the NERDC Amdahl 470 computer.
The time code generator signal provides the timing pulses
for the sampling and also indicates Universal Time (U.T.).
The sampling of the signal was accomplished by a 12 bit
analog/digital converter (A/D) and a KIM I microprocessor.
Pulsar pulse averaging is carried out by a computer program
which averages consecutive periods of the sampled pulsar
signal, previously stored in the Amdahl computer.
5

21
A last program called P2045 (or P1133) reads the cards
of the blocks to be averaged (previously shifted), calculates
the average for each bin (weighting by the number of periods
in each bin), computes the final average of the combined
blocks, the standard deviation o, and plots the data. Also
plotted is the 3 o reference line, which is used as a reference
to aid in deciding whether or not a pulse was present at the
end of the averaging.
The KIM I program to digitize the data and the MCAPGM
program were developed by Dr. John P. Oliver; MCAPGMl is a
A program called DOPVEL was used to obtain the apparent
pulsar period for the night and time when the pulsar was
observed. This program gives the pulsar period at a particu
lar site and at a given time, correcting for Doppler shift
due to the earth's rotational velocity and its motion around
the sun. This program was developed at the NRAO by Drs. R. N.
Manchester and R. A. Gordon and made available by Dr. Steve T.
Gottesman.
Finally a summary of the times spent observing and re
ducing pulsar data and the total time of terminal use is
presented in Table 2-1.
Table 2-1. Summary of time spent with pulsar data.
Pulsar
Total Time
Observing
Pulsar
Total Time
of
Data Reduced
Total Time
Digitizing
Data
PSR1133+16
10h 40m
4
h
10
m
16h 04m
3
h
40
m
3
h
h
12
PSR20 4 5-16

polarized component of the emission. If for some pulsar
there is enough signal-to-noise ratio, it would also be
possible to set the antenna in such a way that one third
of the array is oriented at 0, one third at 45 and the last
third at 90; from the information obtained in this configu
ration it is possible to deduce the Stokes parameters of the
incident wave. Any pulsar polarization measurement made at
26.3 MHz and 45 MHz would be useful since none have been made
at frequencies lower than 63 MHz (36).
Improvements need to be made in the calibrations in order
to obtain more accurate values for the peak flux density.
This can be accomplished by simulating the pulsar pulse using
a precise time standard and a frequency synthesizer, in order
to obtain periods within some \i seconds of the observed period.
It should be possible to obtain estimations of peak flux
densities by digitizing and processing this signal in a similar
manner to the actual pulsar. It will be also necessary to
find out the reason for the discrepancy between the DOPVEL
period and the period obtained from the observations; simula
tions with an accurate period should help in determining if
the discrepancy is due to instrumental problems. A more care
ful analysis of the accuracy in the calculations of the
DOPVEL program needs to be made. The solution of this problem
will make it possible to average several nights and obtain the
long-term average pulse characteristics.

I 00
PULSE
WIDTH
(ms)
50
A
O
THIS THESIS
O CRAFT AND
COMELLA (6)
A IZVEKOVA et al
(18)
40
O
30
20
_i I I I I I I 1 1 I I L
20 30 40 50 100 200
FREQUENCY (MHz)
500
Figure 3-12.
Pulsar PSR1133+16 half intensity pulse width as a
function of frequency.

180
VW'
+ 5 V
Figure C-2. Synchronizer and divider schematic.
112

APPENDIX C
SCHEMATICS
110

7
controlled by the rotation of the earth. Scans by any or
all of the 8 selectable east-west beams (designated 4E, 3E,
2E, IE, 1W, 2W, 3W, 4W) are made by starting each scan at
the appropriate sideral time. A computer program called
ARRAY developed by Dr. M. D. Desch, is used to obtain infor
mation on cable lengths required for coarse phasing, switch
positions for fine phasing, and beam transit time across the
desired source (11) A more detailed description of the
antenna and its adjustment has been given by Desch,(12) and
Desch et al. (13).
In order to use the maximum gain of the array, only the
central E-W beam positions 2E, IE, 1W and 2W were used, and
within each beam only the central portion of it was recorded
(v8 minutes each beam). A maximum of 32 minutes was re
corded each night. Pulsar PSR1133+16 was selected because
2 6 2 1
of its relatively high pulse energy (^0.75x10 Wm Hz at
61 MHz), long period (1.188 s), and low dispersion (DM = 5
-3
parsec cm ). It was also easy to reach this pulsar using
the phasing the antenna had at that time (phased to make
Jupiter observations), without having to completely rephase
the array. Only the fine phasing needed readjustment. During
the period in which the pulsar was observed, it was transiting
Only 8 nights (out of the 20 observed) gave reliable
data, free from static and stations. The end of the observa
tion period coincided with the beginning of the thunderstorm

HL ZC K AV f ^ A jfr
r**M
r A ^ P
1
1J.G
1 <
* fi; jfu-| .
1 J7J i.
LC f
F! OUT NCY = 26
J MHZ
n 1 !
v.r a 7
4 AV
??
76
1 f. T 4 f* T
0 Oh
1 AM
*0f
STuP
0 0 H
1 7M
40S
TS-'MP
02
r i \
:e'7
'1 A V
?o
7
1 ST A-' 1
00M
1 7M
A Ob
rTi_p
00 H
?0M
OOS
TSAMP
0?
p l l
ItAT
AY
?
7c
TSTAtT
0 OH
A 4 M
Oi-S
S TCiP
00H
36M
0 6 S
T b A M P
0?
CM 1
1 tP7
AY
2 6
7 V
T ST AKT
0 0 H
7 4M
oe.s
3 I .ip
0 0 H
4 PM
ocs
TSAMP
0?
1 1
| h A7
* A Y
?l
7f>
T ST At- 1
3 IH
00 M
03!
STOP
Cl M
0 1 v
bOc
TSAMP
0?
PI 1
1 *0 7
AY
o *
7 v
T ST AC T
0 1 H
T/>V
30^
*T iJ>
01 H
OS M
OOS
TSAMP
02
n 1 1
l C 7
AY
?<
7 <>
7 ST AS T
0 l H
OSM
0 OS
ST UP
0 1 H
0 7M
tbS
T S AMP
0?
07
c
1 CM A =
3.
T4,?
1 si
GM A =
0
1 2<.
MACN
IF IC A T I ON =
10
NU y < p
(F LJC<
FIN
A VF_ A G?
PfK!70S
NU'* C
AV rc \ GI N
1
i 2b i ? r
'* 06
?
12: i 2o
0 3?
1 2r> 30
10 10
4
i 2 5.2 o e
10 10
j
l PS >^7
1 0 1 J
r
12r>. s*
1 0 10
7
1?S,'7
1010
1
12S.? 1 2
13 10
4
1 2b :S 4
IMS
I 0
1 *S. ? 3 1
10 10
11
123,19?
I 0 10
1 .2
1 PS ,75 0
1 0 10
l 3
12b. *A3
10 10
1 A
12b.726
1 3 10
I >
l ?* 1 Y?
ton
1 r.
1 2S. mo
10 10
1 7
1 2 b 5 S
10 1)
1
12b.157
*>70
I *
12b 150
6 14
0
126.1Jc
376
T 1
1 ?*; .10 7
MIC
22
l 2S ,77 7
1 J 1 0
3
12 b* 1 c 7
1010
?
l2 *4 C
1 1 0
12*. 1 M M
10 10

1 25 .222
10 10
7
I2i .*0-
1010
2
1 2 .
1 J l C
?S
12' ,7 0 7
1)10
2?
12i>. i o e
1010
7 1
12 b 1SS
1 0 10
3
l 25 75 3
13 10
3 7
1 2 5 1 f ?
10 10
*4
12b.; ?7
l 3 1 3
75
l ?r 1 So
10 1 0
J6
1 2 S 1 *> M
13 10
27
1 2 b 1 7 v
1 Cl 0
7m
12b.74 7
4 79
TIME
T
AVE'- AGINO T I ''c
il??.a scr
INTENSITY *
N3
Figure 3-7. Pulsar PSR1133+16 recorded at 26.3 MHz on May 26, 1979. No pulse
present.

44
obtained at higher frequencies: the pulsar seems to show
abnormally strong pulses at low frequencies during some
particular nights, as it may have done in our case.
Another interesting aspect of this has been pointed
out by Ekers and Moffet (21). They observed PSR1133+16 at
a wave length of 23 cm (^1300 Mhz). From a histogram of
energy and number of pulses with a given energy, they noticed
a long tail in the distribution, towards the high energy end.
The ratio of the mean value (0.020x10 ^ jm lj tjie
2 6 ~2 1
highest energy pulse (^0.090x10 Jm Hz ) is about 4.5.
26 1
If one takes the value for the mean energy of 5.3x10 Hz
at 26.3 MHz as an average over a relative short period
('i390 P) and compares it with the calculated mean energy of
26 2 1
1.6x10 Jm Hz obtained from the observations of Bruck
and Ustimenko at 25 MHz (4) a value of 3.3 is obtained for
the ratio. Such a ratio is not improbable according to the
distribution function of Ekers and Moffet. Pulse energies
obtained reported in this thesis and by several other authors
are plotted as a function of frequency in Figure 3-11.
3.1.b PSR1133+16 Pulse Width
Several factors can distort the actual pulse shape as
observed; the factors that need to be considered are:
A. Effect of the dispersion over the finite bandwidth
B. Effect of the post detection time constant of the

200
THIS THESIS
CENTRAL SUBPULSES. LYNE et al
O SEPARATION BETWEEN (22
TRAILING AND CENTRAL
SUBPULSES. LYNE et al (22)
PULSE
SEPARATION
(ms) 50
40
30
20
O
O
O
00
Ul
1 1 1 I I 1 1 I I I I I I 1 1
10 20 30 40 50 100 200 300 400 500 1000
FREQUENCY (MHz)
Figure 3-33. Pulsar PSR2045-16 subpulse separation as a function of
frequency, including the pulse width at the base obtained
at 45 MHz in this thesis.

Figure 3-31.
Pulsar PSR2045-16 recorded on May 17, 1979, at 45 MHz.
Average over 390 periods. No pulse present.

APPENDIX D
COMPUTER PROGRAMS
113

to Maria Angelica
Carolina
and Daniela

109
Table B-2. Number of data points per second in real
and original pulsar time.
Samples per second
Real time
15.675
10.4167
7.8125 6.250
5.2083
Original
pulsar
time
62.7
41.667
31.25 25.00
20.8332
Since NAVE is fixed at 8, that means that 8 sampled values
are averaged together before producing one data point; in that
way, the tape is being sampled 8 times faster than the data
point rate. As an example, if a TPP = 8 ms and a TSAM = 02
are being used, 7.8125 DPS are obtained, but the tape is being
sampled at the rate of 62.5 samples/second in real time, which
in original pulsar time is 250 samples/second.

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EAR2QH07M_5JDNV0 INGEST_TIME 2014-12-11T21:10:59Z PACKAGE AA00026481_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES

38
equation (3-1), it is possible to substitute TSySt by:
T = T =: T
syst galactic background
Assigning subscripts p to the pulsar and r to a con
tinuous radio source, it is possible by using equation
(3-1), to write two similar expressions, one for the pulsar
and the other for the radio source. By dividing one by
the other, the following expression can be obtained:
AS .
mm p
n Av
r
n Av
P
T
r r
T
P P
T
_R
T
r
S .
min r
K
(3-2)
The constant K takes into account that the pulsar sig
nal, instead of being continuous, is being sampled.
Using the values obtained for the pulsar on May 25, 1979
n^ = 390 periods (453 s)
Av = 8 kHz
P
T = 12.5 ms
P
T = 18,514 K
P
and the values for the radio source of:
n =1
r
Av = 250 kHz
r
Tr = 1.5 s
ASr = 101 Jy
Tr = 38,975K
the following value is obtained:
Smin p = 149 Jy (considering K = 1).
To obtain the peak flux density of the pulsar, it is
necessary to make several assumptions:

Table A-2. List of pulsar candidates to be observed with Maipu Radio Observatory
45 MHz array.
Pulse Flux
Pulsar
Period
(s)
DM
(pc cm j
AtD
@ 10 KHz
(ms)
Energy
-29
xlO
t I
Jm Hz
Density
. -26
x 10
Wm_2Hz_1
Pulse
Width
Half
Intensity
(ms)
Galactic
Back
ground
Temp.
(K)
Zenith
Angle
at
Transit
()
0031-07
0.94295
10.89
9.9
930
@ 61 MHz
1.0
@ 61
MHz
92
@ 102 MHz
^6,500
26
0628-28
1.24441
34.36
31.3
2,340
@ 61 MHz
1.9
@ 61
MHz
105
@ 61 MHz
^6,400
5
0826-34
1.84896
52
47.4
300
@408 MHz
N. A.
(*)
775
@ 408 MHz
^6,400
0.5
0834+06
1.27376
12.85
11.7
3,970
@ 61 MHz
N. A.
(*)
24
@ 61 MHz
< 6,000
39
1749-28
0.56255
50.88
46.3
750
@ 80 MHz
>10
@ 34
MHz
vlO
@ 408 MHz
^34,000
5
2045-16
1.96157
11.51
10.47
N. A. (*)
^1
Â§ 34
MHz
v83
@ 151 MHz
^8,500
17
(*) Information not available at that frequency.
105

Figure 3-26.
Pulsar PSR2045-16 recorded on May 15, 1979, at 45 MHz.
Average over 540 periods.

PAGE
ACKNOWLEDGMENTS iii
ABSTRACT vii
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 DATA ACQUISITION AND REDUCTION 5
2.1 Data Acquisition of Pulsar PSR1133+16... 6
2.2 Data Acquisition of Pulsar PSR2045-16... 10
2.3 Data Reduction 12
2.4 Block Designation for Each Pulsar 16
2.5 Computer Programs for7 Data Reduction.... 19
CHAPTER 3 DATA ANALYSIS 2 2
3.1 PSR1133+16 Data Analysis 22
3.1.a Peak Flux Density, Mean Flux Density
and Energy for Pulsar PSR1133+16 35
3.1. b PSR1133+16 Pulse Width 44
3.1.c Tests Performed with the Pulse
Obtained on May 25, 1979 48
3.2 PSR2045-16 Data Analysis 63
3.2.a Peak Flux Density, Mean Flux Density
and Energy for Pulsar PSR2045-16 77
3.2.b Pulse Width for Pulsar PSR2045-16 83
3.2.c Period of Pulsar PSR2045-16 from
Observations on Two Consecutive
Days 87
3.2.d PSR2045-16 Pulse Intensity as
Function of Time 91
CHAPTER 4 CONCLUSIONS 93
APPENDIX A LIST OF PULSARS 10 3
APPENDIX B SAMPLING RATE, TIMING PULSE FREQUENCY
AND DATA POINTS PER SECOND 10 7
APPENDIX C SCHEMATICS 110
v

96
ionosphere, the ionosphere can introduce relatively large
variations in the angle as measured at ground level. This
is due to variations in the amount of Faraday rotation in
the ionosphere. This effect is particularly strong at
longer wavelengths because the angle of rotation 0 is
proportional to the square of the wavelength A:
9 RM-A2 (4-1)
where RM = rotation measure of the ionospheric path.
The RM depends linearly on the electron density, N,
the component of the earth' s magnetic field intensity which
is parallel to the direction of propagation, Bn and the
thickness of the ionosphere, L:
fL
RM B|, NdL (4-2)
o-1
It can be shown that the percent variation of 0 is
equal to the percent variation of the electron density N,
if the thickness L and the magnetic field B of the ionosphere
remain constant; that is:
A0 AN _
T % IT %
(4-3)
It can also be shown that equals when N and B(|
AB
are constant, and equals -5-" when N and L are constant.
Manchester has obtained values of ionospheric RM
-2
ranging from 0.1 to 6 rad m depending on the time of day
and the declination of the pulsars (37). Taking these two
extreme values, it is possible to obtain the change in
rotation angle, A0, for let us say, a 1% change in AN at

JLr 4 0 9*.9||3i|5 rRl?0= 1.
188103
5 e r r
* z QUz NC Y=26, 3 *-Z
M?vA6 MY 25 79 T ST An T 0I
3 0M
1 OS
STOP
01 M 34 4 l os T5AMO
02
PI 12*86 AAV 25 70 01 H
74 M
l 0r
STOP
01 H 38V 045 TSA vo
0?
R^0C< AvfTP* GE =
l 72
609 SI GMA-s 0, 079 3 5 1
',MA =
0
? 76
MAGN IF ICAT IO* =
1 00
NUV8PC
nf ^LorKSs
2 AVf PAGING TIMÂ£t 47 >.8 SPC
n i
a vp!agtg
i
1 2 ? 6 o o
7 00

1
"*
12 *70
390

1
3
1 22.589
790
1
4
12? *6*
> oo
*
1
s
12? 559
79 0
f
1
6
l 2? .58?
3 00
1
7
1 ?? > 7J
7^0
t
1
8
122 755
700
v "
1
9
122,674
3 90

1
l 1
12 677
7Q0
A
1
1 J
12? 510
700
1
1 2
12?.5M
700
1
1 3
t 2 ? 4 ^ ?
700

1
1 4
1 2? 666
3 90
>
1
1 22.e 7 5
700

.
1
1*
12?^3
7 OO
1
1 7
1 2? 0 7
750

1
1 J
1 ?? 5 1 ~
3 0
*
1
1 o
1 r, 38
7<-7

1
2
122 554
3 90
1
M
l 5 26
700
?2
12? *15
700
1
? 3
1 2? 7 1 8
390

1
?4
l?? .a |
7 90
*
1
25
12? 7*o
390
f
25
1
26
12?.67fc
3 90
+
1
1 2? .5 50
705
1
r p
1 2 "* 559
7C0
1
9',
122.60
790

+
1
i:
1 2.5 in
3 0
t

1
m
12 ? 0
7 9 0

1
32
12?t 37
7 90
1
1 7
!
350
-
1
3 4
12? 58?
790
.
1
7r,
1 2? .F 4 4
7 90
,
1
76
1 ?> * 6 \
700
* ->
1
3 7
1 2 5 36
300

=" x-T _
1
38
123.044
51
t
Figure 3-21. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188103 s (P+80 ys)

APPENDIX E
26.3 MHz ARRAY CALIBRATIONS
In order to obtain the minimum detectable flux density
and then estimate the peak flux density for pulsar PSR1133+16,
several calibrations were performed with the array.
The standard noise source was provided by 2 Sylvania
5722 noise diodes; as a secondary noise source a Hewlett
Packard amplifier was used. Several radio sources of known
flux density and the galactic background temperatures were
also used in the calibrations.
Figure E-l is a block diagram of the 26.3 MHz array
signal transmission and calibration system. Usually the
calibration signals are introduced at point (l) (Figure
E-l). In order to improve the calibrations, the signal was
later introduced at point (2) (Figure E-l); by doing this,
any uncertainties in the gain (or losses) of the 10 pre
amplifiers and the 2 N-S Butler matrices are avoided.
Two 1 input-to-6 outputs 50-ohm power splitters were
used to split the calibration signal into the 10 preampli
fier inputs. To properly match all the outputs, two power
splitters output were loaded with 50-ohm resistors.
As a result of introducing the calibration signal at (l)
and then at (2) a net gain of =6 dB was determined for the
10 preamplifiers plus the 2 N-S Butler matrices; several
123

Figure 3-27.
Pulsar PSR2045-16 recorded on May 16, 1979, at 45 MHz.
Average over 180 periods.

6
2.1 Data Acquisition of Pulsar PSR1133+16
The pulsar PSR1133+16 was observed for 20 nights
between April 25, 1979, and June 8, 1979, using the Uni
versity of Florida 26.3 MHz large array, located at the
University of Florida Radio Observatory in Dixie County.
The 26.3 MHz array is a nearly filled rectangular array of
2
640 half-wave dipoles that covers about 30,000 m The
half power beamwidth (HPBW) is 2.5 north-to-south by 6.0
east-to-west. It is a phase steered array with full N-S
directional control of the beam in quarter-beamwidth incre
ments, and a selection of 8 E-W beam directions spaced one
beamwidth apart. Phasing is accomplished by means of a
network of Butler matrices, hybrid rings and plug-in phasing
cables.
North-south phasing of the array for a source having a
given declination is accomplished in two ways, a coarse
adjustment in large increments of declination and a fine
adjustment providing small steps. A change in the coarse
phasing requires the replacement of many plug-in cables
located throughout the array, and requires about 2 or 3 man
hours of labor. Fine phasing for the desired declination is
quickly accomplished by means of switches at a single loca
tion. The beam pointing can be altered in small increments
to any declination within a strip roughly 10 wide by means
of the fine phasing controls. Positioning the beam to a new
declination outside this strip requires a coarse phasing
readjustment. The right ascension of the beam is of course

36
Table 3-1. Radio sources used for calibration purposes.
R.A.
Declination
Flux
Density
(Jy)
Antenna
Beam
Used
3C310
15h02.9m
2 6 15'6 "
382
1W, 2W
3C315
15h11.7m
2616'8"
133
1W, 2W
3C409
20h12.2m
23 25115
381
IE
3C433
21h21.4m
24 54'6"
267
1W
Positions and flux densities were taken from the Clark
Lake Survey at 26.3 MHz (16). Of these four sources, only
3C409 and 3C433 were used in the calibrations; the 3C310 and
3C315 measurements were discarded because their difference in
R.A. is only 8.8 (corresponding to 2.2). Since the HPBW of
the radiotelescope is 6.0 east-to-west, these two sources
cannot be resolved.
The parameters of the radiotelescope for observing the
calibration sources were:
Bandwidth (Av) = 250 kHz
Time constant (t) = 1.5 s
Frequency (v) =26.3 MHz
The minimum detectable flux density S deduced from
min
calibrations using these sources, is presented in Table 3-2.

16
20 minutes of data obtained each night for pulsar PSR2045-16
was broken up into 4 blocks of 5 minutes each, generating
about 390 cards for each block.
The method of breaking up the data into blocks proved to
be useful also because one can inspect the results of the
average of each block separately, making it possible to search
by eye for a possible pulse that might appear at the same loca
tion in the different blocks. It is highly unlikely that
interference transients would appear in the same location in
each of several blocks.
Monitoring of the DC gated output of the PLL circuit is
very useful since synchronism is sometimes lost when a tape
is played back. If that happens, then that portion of the
tape is skipped and a new block started.
The KIM I program which is necessary to digitize the
data is loaded into the microprocessor from a cassette tape
recorder.
2.4 Block Designation for Each Pulsar
Many blocks of data were generated in the data reduction
process. In order to avoid any confusion with the data and
to be able to distinguish quickly between blocks, a designa
tion scheme was developed for referring to a specific data
block. The designation format contains in abbreviated form
the pertinent information for that block. For pulsar
PSR1133+16, the following block designation was adopted:

40
By making these assumptions, once the pulsar reaches
a value slightly higher than 3 o at the end of an averaging
process, the peak flux density can then be considered equal
to S^n (calculated for the corresponding number of periods
averaged). Therefore,
S = S = 150 Jy
peak p mm p 1
Pulse energy is usually defined as:
rp
E = S(t) dt (3-3)
P o-
where P = pulsar period
S(t) = flux density as a function of time
For pulsars with no interpulse, the limits of the integral
are restricted to the duration of the pulse (base time dura
tion) .
In order to make the computations simpler, it is usually
assumed that the pulse has a relatively simple geometrical
shape. Izvekova et al. (7) assumed a triangular shape for
most pulsars (including PSR1133+16). By looking at the pulse
shape obtained on May 25, 1979, one can see that this is a
reasonably good approximation, considering only the pulse
centered at bin 25.
Using this approximation, the pulse energy in equation
(3-3) can be written as:
E
P
where W,
base
S w,
peak p base
= pulse width at base,
(3-4)
corrected for dispersion

PULSAR PSW04S-
IF
I'tK 1 CK> = 1 Oft 1 3 78
see
rwto UENCY
= 45
MHZ
P20A5HA MAY
1 9
79
T f 7 A ( T
09M 39M
30S
ST CP
09 H
A AM
3 0 S
tsamp
02
P2 045C4 MAY
1 9
79
1ST APT
0 9H * A M
3 OS
STOP
09 M
A6M
J6S
T SA MP
02
P ? 0 A 50 A MAY
19
79
TST ART
09H 46M
3 7 S
STOP
O^H
5 0 m
35S
TSAMP
02
P 20 A 55 A MAY
I 9
7 9
TST AW T
09H SO M
AOS
STOP
09H
5 JM
32 S
TSAMP
02
PI UCK A V E w A GE 130.
3V2 SIGMA
-
0.
073
3 SIGMA
= 0
220
MAGNIFICATION=
1 00
NUMMGR OF DLOCK S=
A AVCP AGING
T I MF = 80A
7 SFC
n in
A VE W AG F
per ions
N.J'RER
A V c WAGE 9
1
130870
A 15
2
1 JO.A 32
A 15
3
1 30. 82 1
A 15
4
I 30 3 75
A 15
5
1 3 0 A 2 S
A 15
t>
120 .t 2 3
41 5
7
1 3 0 A 9 5
4 15
*J
130.800
4 15
V
1 3 0 A 2 S
A 1 5
l 0
l 30 2 78
A 15
1 1
130.341
A 15
12
130396
A 1 5
1 l
130.A 37
A 15
1 A
1 3 0 A 2 0
A 15
1 5
l30.37?
A 15
1 5
13 0 .A 3 S
A 1 5
1 7
l30.379
A 15
1
l JO .5 0 7
4 15
1 9
130.4 A 15
20
1 30.A 49
A 15
? I
130.44T
4 15
2?
130.2 1 1
4 15
23
130. 39A
4 1 5
? A
1 30A 35
4 15
25
1 30 .4 32
A 15
26
130.428
4 1 5
7 7
130.146
A 1 5
28
1 J 0 3 7 9
4 15
29
UO. 2 34
4 1 5
33
133.423
4 15
31
130.365
4 15
22
130.?P6
4 15
23
130.177
4 15
3A
130.309
3S5
35
130. 4 A 7
A 1 5
35
1 JO.A FA
A 15
37
1 3 0.37,?
A 15
28
l 30. 7 5 8
A 15
29
130.271
A 15
4 0
130 .298
A 15
4 I
130. A 21
331
4 2
1 8 8 A l 3
A 15
43
1 3 0 J 4 8
4 15
44
1 30.A 41
A 15
AS
1 3 0 A 1 8
4 15
45
130.464
4 15
47
1 30. A 7 1
4 1 5
48
133.363
A 15
A 9
1 JO 34 3
A 15
5 0
1 JC.290
4 1 8
S 1
120 .3 34
4 1 5
5?
130.355
A 15
53
1 JO .392
415
SA
130.375
A 15
5S
1 30.A I 8
4 15
5b
1 JO A l 2
3 73
57
1 30 A JO
4 15
58
130.440
4 15
59
1 JO AS8
3 C 9
t 0
l30.3MQ
4 15
61
l30365
4 15
62
1 JO.381
A 15
TIME
INTENSITY
I
I
I
I
I
6 3.3 SIGMA I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-J

APPENDIX D COMPUTER PROGRAMS
113
APPENDIX E 26.3 MHz ARRAY CALIBRATIONS 123
LIST OF REFERENCES 131
BIOGRAPHICAL SKETCH 135
vi

87
occurrence of pulses with orthogonal polarization with
respect to the mean, but is high at the center of the
pulse (32). If the degree of linear polarization of the
pulse is higher at lower frequencies, there will probably
be a distortion in the pulse shape when it is received
with a linearly polarized antenna. This distortion will
depend on the variation of the polarization angle within
the pulse. Assuming that the linearly polarized component
in the central part of the pulse is parallel (or nearly
parallel) to the antenna, then the integrated pulse profile
will appear strong in the center; however the slopes of the
leading and trailing edges will be weaker and the averaged
pulse profile will be narrow. More details about the
effect of the degree of linearly polarized component on the
pulse profile and apparent intensity variations are given
in the conclusions.
3.2.c Period of Pulsar PSR2045-16
from Observations on Two Consecutive Days
The apparent mean pulse period of a pulsar can be
calculated from the elapsed time between two pulses; in
order to increase the accuracy it is desirable that the
elapsed time between the two pulses be much longer than
the pulsar period. An attempt was made to obtain the
apparent mean pulse period from observations made 24n
apart (elapsed time %24 ) for pulsar PSR2045-16. On two
occasions the pulsar was detected on two consecutive days;

Table A-l. Pulsar candidates to be observed with the University of Florida
26.3 MHz array.
Pulsar
Period
(s)
DM -3
(pc cm )
AtD
@ 8 KHz
(ms)
Pulse
Energy
-29
*10
-2 -1
Jm Hz
Flux
Density
xlO-26 Jy
Pulse
Width
Half
Intensity
(ms)
Galactic
Back
ground
Temp.
(K)
Zenith
Angle
at
Transit
()
0031-07
0. 94 3
10
36.5
930
@ 61 MHz
1.0
@ 61 MHz
92
@102 MHz
32,550
37
0809+74
1.292
6
21.9
2,980
@ 61 MHz
2.3
@ 61 MHz
125
@ 61 MHz
66,855
44
0834+06
1.274
13
49.1
3,970
@ 61 MHz
3.1
@ 61 MHz
24
@ 61 MHz
24,660
24
1133+16
1.188
5
18.25
1,105
@ 53 MHz
0.75
@ 61 MHz
47
@ 61 MHz
27,620
14
1508+55
0.739
20
73
1,000
@ 61 MHz
1.35
@ 61 MHz
13
@ 102 MHz
20,713
25
1919+21
1.337
12
45.3
3,800
@ 61 MHz
2.8
@ 61 MHz
28
@ 102 MHz
78,908
9
104

108
pulsar was 8 ms (125 Hz). The number of sampled values
averaged (NAVE) before sending through the terminal to the
Amdahl is fixed at 8.
Table B-l gives the relationship between the TPP, TSAMP
and the DPS (39).
Table B-l.
Values
TSAMP.
of DPS as
a function
of TPP
and
TPP
(ms)
FREQ
(Hz)
01
02
TSAMP
03
04
05
4
250
31.25
15.625
10.4167
7.8125
6.250
8
125
15.075
7.8125
5.2083
3.906
3.125
Data Points
Per Second
(DPS)
In order to meet the transmission rate requirements of
13 data points per second using 8 ms for the TPP, a safe value
of TSAMP = 02 must be used; that gives a value of 7.8125 DPS.
If the tapes are played back at the same speed that
they are recorded, then the resolution would be also 7.815 DPS,
which is a poor resolution. However, if the tapes are played
back with a slow-down of 4:1 the time resolution will be in
creased by a factor of 4.
Table B-2 gives the number of DPS obtained in real time
and in actual or original pulsar time (39).

128
Table E-2. Losses between dipoles and point (2)
considering the partial losses.
Minimum dB
Maximum dB
Full wave twin line
. 30
.30
Hybrid ring
.45
.45
Phase cable
. 80
1.10
Coaxial cable line
1.90
1.90
E-W Butler matrix
1.44
1.98
Line connecting E-W to
N-S Butler Matrices
.5
1.10
Total
5.39
6.73
The corrected noise temperatures at point (), given in
Table E-l, is the noise temperature of the noise source cor
rected for the 5 dB losses in the aluminum coaxial line and
the losses in the power splitter (factor 0.76); losses in the
power splitter consider the losses due to the two 50-ohm
resistors and the losses at 26.3 MHz of the power splitter
itself.
E.2 Antenna Effective Area
By using the losses between the dipoles and point (5)
3C433, the antenna effective area can be calculated; the
method consists of obtaining the change in antenna temperature

PULSAR PSP I 133+II P>riun = l Sf.C f HF UUtJNC V&263 MHZ
ll?CAl APRiL 28 70 T ST ART 02M 03M OOS STOP 0 ?H 0Â£M 50S TSAMP 02
R_CCK
A VCRAGE = l?C. 160
SI CM A r
0. 13 7
3 Sir,MA =
0.411
MACN IF Â¡CAT ION =
NU M JFR
OF MLCCKS= 1
AVIR AC.1NC
T I MC =
lfrP.5 SCC
=>I N
AVEPAGF
nF P l
A Vf W.
I
119,^16
1 4 0
2
120.107
1 40
3
120.193
1 40
4
120.021
1 40
5
12 C. 150
1 40
6
120.2?9
1 40
7
1 20. 3 1 4
1 40
3
120.150
1 40
9
120.307
1 40
10
120.093
1 40
1 1
120.064
1 40
12
12 0.100
1 40
1 3
120.193
1 40
14
l 20.0 36
l 40
15
120029
1 40
1 6
l20.143
1 40
1 7
120.257
1 40
18
120 .242
1 40
19
120.207
1 40
20
120.01b
1 40
21
120.029
1 40
22
120.357
1 4 0
23
120.326
1 40
24
1 20 .24 1
l 40
25
1 19.979
1 40
26
120.2 79
1 40
2 7
l 20 29
1 40
2 8
1 20 .329
1 40
29
1 2C .0 79
1 4 0
30
l 19.971
1 4 J
31
120.014
1 40
22
120.214
1 4 C
33
1 20 .4 79
1 40
24
120. 3 1 4
1 4 0
25
120.086
1 40
36
1 20 .007
1 40
27
120.107
l 40
38
l 20 .0 56
1 8
TIME
INTENSITY
NO
Figure 3-2.
Pulsar PSR1133+16 (Block P112EA3) recorded at 26.3 MHz on April 28,
1979. Average over 140 periods.

PULSAP PSR1133M6 PEWIOP=l 188023 SEC FRE QUENC Y = 26 3 m HZ
P 1 1 2* A 6
MAY
25
79
TSTAP1
0 I H
"AOM
1 OS
STOP
0 I H
3 AM
10S
TS AMP
02
P112*B6
MAY
25
79
TS7 ART
0IH
3AM
1 OS
STOP
0 IH
38M
0AS
T SAMP
02
etOCK AVERAGE* 122.609
SIGMA
r
0.
08 1
3 SIGMA =
0.
2 A 3
MAGNIF [CAT ION*
1 00
NUMBER (JF BLC K S =
0 IN
NJMHE R
AVERAGE
1
122.681
P
122.68?
T
1 2 2 7 2 3
A
122. 6 1 0
5
l23.603
6
l22.595
7
122.613
n
l ?>. 592
9
l 2 ? 7 6 6
10
122. tA e
1 1
I 22.6 .3 C
1 2
l2?.530
1 3
1225A 3
1 A
122.5 1 0
IS
12?.589
16
1 22. 620
1 7
l22.505
l 8
122.607
IP
122.546
20
l22.ABA
2 1
122.651
22
i22.se?
23
12?.A90
2A
122.700
2 5
l2.BOA
26
122.728
27
122. 6 0 C
28
l2? 52C
1 2 2.5 1 5
30
1 22.597
31
122638
32
122.57?
33
122. 6 A3
7 4
122 .533
35
122.651
36
l 22.55 J
7 7
122.595
38
1 22.555
PENI00 Â£
AVERAGED
??A |
300 I
390 I
390 |
390 I
390 I
390 |
390 I
3 9 0 I
390 I
3 90 I
3 90 I
390 |
390 I
3 90 I
390 |
390 I
390 I
390 I TIME
390 |
390 |
390 I
390 I
3 90 1
390 I
390 I
390 |
390 I
3 90 I
390 I
390 I
390 |
390 I
3 90 I
390 I
390 I
390 I
2 16 I
2
AVEWAGING T1 ME =
AS?.7 S^C
INTENSITY
71
25 3.4 SIGMA
I
I
I
I
I
I
I
I
I
I
I
ro
03
Figure 3-4. Pulsar PSR1133+16 (Blocks P112WA6 and P112WB6) recorded at 26.3 MHz
on May 25, 1979. Average over 390 periods.

Two pulsars have been observed at low radio frequencies
using narrow bandwidth to avoid pulse dispersion problems.
Pulsar PSR1133+16 was observed at 26.3 MHz using the
University of Florida large array; pulsar PSR2045-16 was
observed at 45 MHz using the Maipu Radiobservatory large
array (Chile). Long term integration techniques were used
to reduce the data.
Both pulsars were successfully detected and the average
pulse intensity and integrated pulse profile were obtained.
Time resolutions of P/38 were obtained for PSR1133+16 and
P/62 for PSR2045-16, where P is the pulse repetition period.
The two observed pulsars were selected because of their low
dispersion measure, long period (>1 sec), high pulse inten
sity and their location with respect to the galactic back
ground .
Several other radio sources were also observed and were
used to calibrate the effective area of the 26.3 MHz antenna.
Chairman
vm

P2045 (or P1133) Program

Figure C-1. Phase locked loop (PLL) schematic.
Ill

CHAPTER 1
INTRODUCTION
Pulsars were discovered in 1967 by S. Jocelyn Bell and
Anthony Hewish. The first pulsar paper was published in
Nature in February 1968 by Hewish et al. (1) telling about
the discovery of pulsar PSR1919+21 and giving some of its
parameters. Since then, about 321 pulsars have been detec
ted and cataloged.
According to currently accepted models, pulsars are
rapidly spinning neutron stars originating from supernova
explosions. They are believed to have strong magnetic
12
fields on the order of 10 gauss, probably the strongest
known magnetic fields in the Universe. The neutron star is
believed to emit continuously a narrow beam of radio waves
which corotates with the star, producing the observed pulse
each time the beam sweeps past the earth. The periods of
the pulses which have been observed range from about 0.03
to 4.3 seconds. So far, for only 2 pulsars has it been
possible to find the optical counterpart of a pulsar.
Several mechanisms have been proposed to explain the
origin of the radio emission and how the rotational energy
of the neutron star is converted into electromagnetic wave
pulses; this subject still remains one of the least under
stood aspects of pulsars. Some of the important parameters
1

12
A bandwidth of 10 kHz was used; this is the minimum
bandwidth of the receiver. Using data obtained at higher
frequencies, an extrapolated pulsar pulse width at the
base of about 130 ms was obtained at 45 MHz. The pulse
broadening due to the 10 kHz bandwidth is 11 ms. This was
considered acceptable.
The quartz time standard is checked daily against the
Chilean National Observatory cesium time standard; the epoch
is believed to be known with an accuracy of about +10 ys.
The drift in time is about 3.6 ys/h (or about 0.00190 ys/P,
where P = 1.96 s, the pulsar period). The drift is 1.2 ys
in the 20 minutes between the east-to-west HPBW. For all
practical purposes considered in this research, this is
negligible.
2.3 Data Reduction
Data reduction for both pulsars was performed using the
facilities of the Department of Astronomy of the University of
Florida in Gainesville. The basic configuration and inter
connections of the equipment for data reduction are shown in
Figure 2-2. The schematics for the most relevant blocks are
presented in Appendix C.
As stated earlier, the IF signal in the receiver is
band limited to 8 or 10 kHz and appropriately smoothed. The
resulting audio frequency signal is recorded in analog form
on channel 1 of the Magnecord tape recorder at a tape speed
of 15 IPS. The time code generator is recorded on channel 2.

Figure E-l. Block diagram for the 26.3 MHz array calibration.
124

CHAPTER 3
DATA ANALYSIS
3.1 PSR1133+16 Data Analysis
From the 8 nights of data reduced for pulsar
PSR1133+16 (26.3 MHz, Florida), pulses were found with
no ambiguity on 2 nights. Pulses were found on April 28
and May 25, 1979, over averages of 340 and 390 periods
respectively; the results are plotted in figures 3-1 through
3-7.
On April 28, after averaging over 340 periods (^6 44 )
a pulse appears in bin 33 with an amplitude of ^3 o above
the mean value (Figure 3-1). The 2 data blocks P112EA3 and
P112EB3 used to obtain this final block have been plotted
separately in figures 3-2 and 3-3. It can be seen that a
pulse appears in both blocks at the same location (bin 33).
Even though the pulse appearing in each block would not
alone be considered to be statistically significant, the
fact that they appear in the same location for two sepa
rated sets of data provides an excellent test of the validity
of the pulse.
On May 25, after averaging over 390 periods, a promi
nent pulse was obtained in bin 25 with an amplitude of
3.4 a; this pulse is shown in figure 3-4. The same test for
validity as before was made. Blocks P112WA6 and P112WB6 are
22

PULSAR PSR1 133M6 PER IODs 1 19O023 S EC FREOUENC Ys 26. 3 MHZ
P112WAO MAY 25 79 TSTART O IH 3DM IOS STCP 01H 34M IOS TSAMP 02
EL CCK
AVERAGE- 12?.CIO
SI GMA =
0. 105
3 SI GMA =
0. 315
MAGN IF I CATION =
NUMBER
O' BLCC KS= 1
AV ERAC1NG
T 1 ME =
232.2 SEC
C3 I N
A VE R AGE
PER 1OS
N ) M(JÂ£R
4V E^AGED
l
l 22.6<0
200
2
1 22.64 0
2 00
3
122.740
2 CO
A
122 .655
200
5
122. Et>0
2 CO
ft
i2? .e 15
2 CO
7
122.590
2 00
3
l22.6 i C
2 CO
i
l22.7?5
200
1 C
1 22 .630
2 00
1 1
122.530
2 CO
1 2
122.510
2C0
1 3
122.510
2 00
1 4
122.505
2 CO
15
122.495
2C0
i e
122.590
2 00
1 7
1 22 *56C
2 CO
13
122.565
2 CO
1 9
122.5 20
200
20
122.520
2 CO
21
l22.750
200
22
122.7Â£E
2 CO
23
1 22.57C
2 CO
24
l22.665
2 00
25
122.915
2 CO
25
122.770
200
2 7
1 22 .530
200
23
122.430
2 CO
2 i
122.475
2 CO
30
I 22.590
2 00
3 1
1 22.76 0
2 C 0
3 2
122.630
2 00
33
1 227 05
2 00
34
122 .61 *
2 CO
35
122.670
2 CO
36
l22.625
200
?7
1 22.455
2 CO
39
1 22.692
26
1
I
i
I
I
I
I
I
I
I TIME
I
I
I
I
I
I
I
I
I
I
I
I
I
I
INTENSITY
I
I
I
Â¡
i
i
i
i
i
i
i
i
i
fO
Figure 3-5.
Pulsar PSR1133+16 (Block P112WA6) recorded at 26.3 MHz on May 25,
1979. Average over 200 periods.

8
season, at which time it was necessary to shut the array
down for the summer.
A block diagram for the interconnections of the equip
ment used to observe the pulsar is shown in Figure 2-1. The
equipment basically included a Collins R 390 receiver, a
Magnecord 1022 magnetic tape recorder, a Tracor 5D fre
quency standard, and a Systron Donner 8154 time code gener
The time code generator (TCG) was synchronized against
WWV; this was accomplished by triggering an oscilloscope with
the 1 pulse per second (1 PPS) output of the TCG and measur
ing the delay of WWV 1 PPS. The 1 PPS of the TCG was brought
as close as possible to the 1 PPS of WWV (usually between
5 to 20 ms) by making short interruptions of the 1 MHz signal
from the Tracor frequency standard. The delay between the
two 1-PPS pulse trains was measured to an accuracy of about
0.5 ms, and from that information the U.T. was obtained,
except for the propagation delay between Boulder, Colorado,
and Dixie County Radio Observatory. No corrections have
relate the epoch of this observation with any other observa
tions.
The measured drift of the Tracor time standard is about
80 ys/h; this causes a drift in the epoch of about 100 ys
between pulsar transits by beams 2E and 2W (^lh16m); this
drift is small enough to be noticed, considering the time

78
where S = flux density
v = frequency
n = spectral index
For any two frequencies on the linear part of the
relationship between S and ^ ,
The best available values of the flux density for this
source (24, 25) are at frequencies of 160 and 80 MHz:
sico =8-8 Jy
sso = 15 Jy
Substituting these values into equation (3-12), a value
of n = 0.76 is obtained.
Assuming a linear relationship for S as a function of
~ down to 45 MHz, the flux density at this frequency is
found to be:
S45 = 23 Jy
Applying the previously used criteria, the following
value for the minimum detectable flux density at 45 MHz
was obtained for the region near source PKS2045-17:
S = 6 Jy
mm J
source PKS2045-17 were:
Av = 3 MHz
T = 3 S
n
r
1

J I I I L
Figure 3-25.
jii1iiiiiiiii!iiiii i i i i I i i i i i i i i i 1
100 200 300 400
TIME (PERIODS)
Relative pulse intensity as a function of time (periods) for
pulsar PSR1133+16 on May 15, 1979. Each point represents an
average over 10 periods. P = 1.188023 s.

118
-<
-y v
ZO
CD
2
zox
<
--X
o- a
ozu.
Z-X
O
CD O
Z -

-IL X
o:*
OCD S
- U
L.
U
(T
<~o
OOJ
NO.
uu^
a*-
CD
a -o
o~-j
u.a
n
X
< cr - o
o<-
O-
a Z
a
CD
uw .
*o#
Ui
*h- <
JZ C
z-a
xo
o
x *o c
O tool
U. H O II H '
J--~*
*< ----Z
O

o
O
o
O

u.
1
i
-CM
-CM

O 1
O 1

cm
-O
O

<
<
IL
y-
y-
-J
<<
< <
< O
OH
O h-
u-
II <
II <

-o
-O
ft.
O
O H
C M
- O
o o
-O
-O
_J
Org
O II O
<
<
<
a -
n o o
L-
y-
u
UJ u.
o a o ii
y-
h-
< <
< <
CD
N
a uj ii o
CD
h
O H
a
a o
uj no cr
UJ
UJ
<
<
n a a uj
P
y-
-O
O

(X-
Q. UJN
Z
z

o
UJ
- NQ.

a
a x
- &-

4
<
UJ
h- o

O
o
N

<\J
n
(M pg
rvj pg
a
< X
* O

Z 4
O CsJ CM

h-
O
h-
or -
< <
OJ I OJ
N
H
Oh
cg
OID
ID X
Z o
1

O
(M
O
ft. o
OJ
->-

h-
Q.
O _Jh- h-

o-
O
O

X X
tO J

-o
h-
-o
<0
OO
w

Z X
X
a cr O o
<
< O
<
O
< -
< O
h- -
uj Uj & cr
y-
h- a o
o
h- <
z

h*
NNUIUJ
<
< UJ -
<-

o
N
h- a
a a n n
cuo
o cl
o < o
o
o < o
a -
_J
<
o. o.

y- o
c

o
X U.
M <
l* oCDh-
--u

-

o

< X
- U
II CM
OO UJ o
OO
O Q -
0,0 -
UJ
0*0
id
H cD
O 11
Z z O
HO
Oh- *
ZO h ~

ZO H
h- _l
h- ft
X < J
<<2Z -
*
* 4 0
< z u-

< z
Ah
< < -J 4
O II
ii cr a <
<~0 ?4
<
< o
cr
H O-
CD-
X < H
*j a
OO -
o o o. uj
OO ID
z
- iQO
UJ-
Q. U h-
O
-Z -
ftJM *-01
-z
z- xa -
OJ 4-
<
pg -
a o
ffl auuiinoo-

CM-h-
- X o <
OJLU
uj-o -0-0 CD
z
- o
z -
-lLIDUU.li.ll J jh-Oa
o o
o
OCD
CD UJ
, < x - z

< X

m Ju.k j jj<<^LOrf
UJ
OOOOZ
Q.
o n
eg uj H h- ft.
OOh<
-a
o Oh- <
o
o
hOO
UJ UJ UJ UI X
X
h- O N
O M Z 4 < Z
UJUJ < x Z CD 4 UJ
UJ U < 1
-X o
II < < M H II* o O UJ i
_jlD
a ex
<
Jlf) ^ o
iD hOh
o o ii z a c.
O O O o o*
y ii ii ii ---- *pg
J CM
<
h-
.
-<
H f-h- Y- a UJ UJ
O UJ ft. ft.
CD
y-
04 OJ
H- -
ni- i
id- oil. y-.
y- y- n a oj cd ii uj z
- LU U C.
h- h- II O CM
o o o
o O O u
O
< --z
H ZZ-
< O- -CD
O O CD CD Z
pgir- idn z h--
O- ~ IflKh^-Hhl-ht-C X OT n**QhZH
oooooooo oata.ao.aQQ:u.au.-i*iL
O ID O
D CDCD (DO. H H O I
.H H II M H
* axai-o
*uTirÂ¡ruj!sÂ¡!mSoj
.-HOhZ Zh- -
CD
OO
II w<
OU.UJ
-a
< o w <
IKOI
a < cr
OOU-O
u. oa u.
-UJ*-|i<4(DÂ£nCD* 0**ZO CD h-*oMOM-
h-C|ltrO Z h- UJ >- w h- Z a X h- - X h Z Zl/1
-ii
HOZ U.Hk.|i DlLOO CDQ.LlOU.U-DU.OCXO 5d3dd J
CD-0U-
h- h- CD OQCDZZ *
OJ O
CON
ID -
O
04
O o
pg -
cm
a cmh s o cMro n cr o pvjn < s g.
h- OOOOOOOOO rgrjNCJ(\jPg<\4(\|fU<\j
u o o o o o o o o o o o o o o o o o o o O o o o o o o O o O
o o O o OO o o o o o o o o o o o o o o o o o o o o o o o o
o m m mh-ao t'in^inoMiOo- Nngise^o- r\j'rwh-o,o nnnn nnnnnm*<*<*<*cniininDiD.oNNh'-NH-h-
O O O O OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
O O O O OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

7 1
KB ITC(6.42 5)
72
4?5
FOPMAT( // MIN'.TQ. 'AVrPAGET 17. PERIODS* 1
73
SITE!/ .430 1
74
430
FTI-MATI' N'JMMFr. * T l 7 AVERAGED )
76
OFF St T= T EK AV< 1 )
76
OFF 1 =0FFSET-0.5
7 7
DO <100 K = 1 Nil IN
79
VALUE= ( AVE(K 1-OFF 1 ) MA G
79
VALUE= 1 ADS < VALUE )
H "3
IF (VALUE.LE.O)VAL UE = 1
81
ir ( value c.r 100 iv alue= i oo
t?
3VLR< IfiTAV OFF 1 ) W AG
t
3 I C.PL= ( < 1CT AV 4T HS 1C.)-OFF l 1 *MAG
H 4
sk.i*( 96
IF (SI GMI.LF. .0) SIGMjsl
6
IF (SI M l.GC. 100 )c IGM 1=100
87
IF ( A VER.LF 0 1 A Vt P* 1
84
IF (AVER GC100 )AVfP= 100
89
IF (S 1GML.L 0 IS 1 GPL = 1
90
IF( 51 GPLGE 1001SIGPI =100
9 1
DO 660 J=1,100
92
6 50
LGT ( J) =MLNK
9 3
MLCT(l)=LIM
9%
PL CT ( 1U0)=L IM
95
PL C T (AVER)=AVAX
96
L0 T( SI GPL 1 = TP SiG
97
PLOT 98
PLTT ( VALUE 1 = VAL
99
R I T F. ( < 7?b )K AVE ( K ) .NPER ( K) ( PL CT (M J m= | | 00 I
1 00
72b
FORMAT (13.T8.F8.3.T17. 14.T25.10 0A1 )
1 01
800
C CM 1 PUF
1 0 2
*>R ITE ( t- .900 I
1 03
900
h U R M A T ( 1 )
104
STEP
105
ENO
F E NT RV
122

15
to detect any loss of synchronism or dropout of the TCG
playback signal. If there is any loss of synchronism
between the input and PLL signals, then the gated DC output
falls to zero, giving a visual indication to stop the
sampling process. If synchronization were lost, the
pulsar signals would be averaged out of phase, wiping out
any pulse at the end of the integration process. A means
was provided for starting the sampling at the next 1 PPS
pulse from the TCG occurring after a switch was thrown.
This made it possible to know the exact time at which sampling
started. This was accomplished by connecting the start switch
to the D input and the 1-PPS signal to the clock (c) input of
a D-type flip-flop. The output of the flip flop gates the
125 Hz signal in a NAND gate (see schematic in Appendix C).
When the start switch is turned on, the next 1 PPS of the
TCG enables the 125 Hz signal to go into the KIM I and the
sampling is started.
The sampled values (coded in hexadecimal) are sent
through the Hazeltine terminal to the Amdahl 470 computer
and stored there for further processing and analysis.
Since the maximum number of cards that can be stored
under a new data file name is 493, it was necessary to
break up the data into several "blocks" to meet this require
ment. The 8 minutes of data obtained in one beam for pulsar
PSR1133+16 were broken up in 2 blocks of 4 minutes each,
generating about 312 cards for each block (for a TSAMP= 02
and timing signal frequency = 125 Hz; see Appendix B). The

007ft
0077
0070
0 0 79
0080
008 1
0082
0093
0004
0005
0086
0087
0080
0009
0090
009 1
0092
0093
0094
0095
0096
0097
0098
0099
0 100
0 101
0 102
0 1 03
0 104
0105
0 106
0107
0 108
0 109
0 110
0 111
0112
0 113
0 114
0 115
0116
0117
0 110
0 119
0 120
0 12 1
0 122
0 123
0 124
0125
0 126
0 127
0 128
0 129
0 1 30
OUUUUUUUU
IV G LEVEL 21
MAI N
DATE 80004 12/27/37
PAGE
on 230 I=1.N8MX
BINI ( I )=BINI ( I )+BJN( I I
NBINI( lUNQINKI ) 4NBIN( I )
SUM=SUMfBIM(I)
SUMI SUMI MU NM I )
SUMN=SUMN + N5 IN( I >
SUMNI = SUMNI-N01NI ( I )
230 CONTINUE
AV*sSUM/SUMN
AVEISUMI/SOMNI
DO 235 1=1.NflMX
SUM SO SUM S3* ( 3 N( 1 ) AVE*NBIN( I > > *42
SUMS 91= SUM SQ I < 8 IN1 ( I ) A VE I N3 I N I ( I ) )**2
235 CONTINUE
PMS= SORT{ SUM SO/SUMN)/SORT(SUMN/FLOAT(N0MX) )
RMS1 = SORT(SUMSOI/SUMNI )/SORT(SUMNI/FLOAT( NOMX) )
WRITE{6.530 > AVERMS.AVE I .RMS I
530 FORMAT(30XBLOCK AVERAGE! r 0,2 SIGMA! ,FA.3,
1 12 X INTEGRAL AVERAGE! P8.2. SIGMA! F8 3)
^0 240 1=1 .N8 MX
I P= 1
01NA=0.0
IF(NBIN BINA = BIN( I )/FLO AT(NO IN( I ) )
IF(IB IT.EQ.12)I=(1N( I )/FLOAT(NBIN( I I l-PZERC)/2 048.*2 5.
1 p L OAT ( IPSC AL ) 2 6
IF{ I BIT.E0.8)IP=(BINC I )/FLOAT(N9IN(I ) >-PZERO)/I2 8.*2 5.
1 FFLOAT ( I PSCAL ) *26
511 IF( IP .LT.1 ) IP=1
I F( IP.GT.51 ) IP=51
PLOT(IP)=ALFX
PI=1
O INIA = 0.0
IF(NRINI(I).EQ.O)GO TO 512
BINI A=B INI (I )/FLOAT(N0IN 1(1))
IF(IBIT.EQ.12)IPI = (0INI( I )/FLO AT (N9INI (I ) >-PZERO )/2O4 0.4 25.
1 F|_0 A T ( IPSC AL ) *26
IF(IBIT.EQ.8)IPI=(BINI(I )/FLO AT(NBINI(I) )-PZERO)/I28.*25.
1 rLOAT(IPSCAL)426
512 IF( IP I .LT.l ) IPI = 1
IF( ID IGT.51 ) IPI=51
PLOT I ( IPI )= A LF x
3KAVE( I )= BINI A
NUB IN( I )= N3INI (I )
W RITF(ft. 520) BINA.NBIN I) .BINIA.NBINI( I ).PLOT.PLOT I
52 0 FORMAT ( 1X.F9.3. I4.F10.3. 15. IX. 102A1)
PLOT(IP)=ALF0
PLOT IC IPX )= ALFB
PLOT( 1 ) = ALFC
PLOT(26) = AL F p
PLOT( 51 ) =A LF C
PLOT I (1 ) = ALFC
PLOT 1(26)=ALFP
PLOT 1(51 )= ALFC
nIN( I ) = 0.0
N BIN( I )=0
240 CONTINUE
GO TO 200
400 STOP
0
LSI 5 1
1 2
16.0000
1 0
1 .000000
0
1
THE FOLLOWING ARE SAMPLES OF THE CONTROL CARDS
(COMMENTS DO NOT NEED TO BE ON CARDS)
NUMBER OF BITS DIGITIZFD (8 OR 12)
TIMING PULSE PERIOD (MSEC ) .2MSEC*TS AMP 4NA VE(= 8)
NUMBER O'7 PULSE PERIODS AVERAGED
PULSAR PERIOD (SECONDS)
PLOT OFFSET
PLOT MAGNIFICATION FACTOR
0002
END
116

41
Substituting the values of:
S = 150 Jy
peak p 1
W, = 71 ms
base
the following value for E is obtained:
P
E = 5.3xl0-26 Jm-2Hz_1
P
The mean flux density of the pulse is defined as:
S = E /P (3-5)
mean p
Using the value obtained for E and the value for the
P
period P = 1.188 sec, the following value for the mean flux
density is obtained:
S = 4.45 Jy
mean 1
It is interesting, now, to compare these values with
values obtained by other authors.
In 1973, Bruck and Ustimenko (4) observed the same
pulsar at frequencies of 16.7, 20 and 25 MHz using a linearly
polarized broadband array and reported peak flux densities in
the range of 10 to 40 Jy.
In 1968, Drake and Craft (19) observed the same pulsar,
using the Arecibo 1000-foot radio telescope at frequencies
of 111.5 and 195 Mhz and reported peak flux densities of
150 Jy at 111.5 MHz. Since these observations were made at
a much higher frequency, they cannot be directly compared
with the one at 26.3 MHz, nevertheless, the value obtained
at the two frequencies gives an idea of the order of magnitude
for the peak flux densities. A value of 38 Jy can be deduced
from the data at 61 MHz given by Izvekova et al. (7).

101
system of this type, the signal-to-noise ratio would be
or 3.5, over what it was
with the 8 kHz-bandwidth system.
Another advantage of a wide bandwidth system, and per
haps an even more important one, is its ability to deter
mine the degree of linear polarization of the pulsar signals
by making use of Faraday rotation in the terrestrial iono
sphere. As stated previously, the rotation of the plane of
linear polarization varies, due both to ionospheric fluctua
tions and to a continual change in the ray path as the earth
rotates, so that the polarization plane is frequently
parallel and frequently perpendicular to the antenna dipoles.
The degree of linear polarization is readily calculable from
the measured intensities at times of parallel and perpendicu
lar orientations. This appears to be an excellent method for
determining linear polarization; it has apparently not pre
viously been used for pulsars.
Also, if the rotation measure (RM) of the ionosphere
above the observatory can be determined independently, the
position angle of the plane of linear polarization of the
pulsar signal at the top of the ionosphere can be calculated.
tion about polarization. Since the dipoles of the 26.3 MHz
array in the Dixie County radio observatory have the
capability to be rotated to any position, half of the array
can be rotated in 90 respect to the remaining half, making
possible to obtain information about the circularly

63
It can be regarded as a double check of the test previously
described.
Each point on the graph represents the intensity with
respect to its corresponding average value of 10 consecutive
periods, for bin 25 (peak of the pulse at the end of the
averaging process). The intensity scale is in arbitrary
units and the time scale is in periods P given by the DOPVEL
program (P = 1.188023 s) The plot shows that the intensity
in bin 25 is almost always positive with respect to its
corresponding 10-periods average value, and no clear periodic
variations appear over the 390 periods plotted.
3.2 PSR2045-16 Data Analysis
Of the 10 nights of data reduced for pulsar PSR2045-16,
pulses were found on 4 nights: May 15, 16, 18 and 19, 1979.
A summary of the dates, averaged pulse deflection in
terms of standard deviations o, and number of periods
averaged is presented in Table 3-4.
Table 3-4. Summary of the date, pulse deflection and
number of periods averaged each night for
pulsar PSR2045-16
Date
Pulse deflection
in terms of a
Periods averaged
May
15,
1979
3.16
540
May
16,
1979
3.48
440
May
18,
1979
3.28
610
May
19,
1979
3.30
415

97
both 26.3 and 45 MHz. The result of this calculation is
shown in Table 4-1.
Table 4-1. Variation of the position angle A0 due
to variations of 1% in AN, AB(|Or AL in
the ionosphere.
Frequency
(MHz)
AQ AN ABi,
A0 for IT' R
AL N B|1
or = 1%, and
L -2
AQ an
A0 for -TT-,
AL
r = 1%,
AB
and
-2
m
26.3
7.4
447
45
2.5
150
From the values obtained in Table 4-1, one should expect
large random (or nearly random) variations of the position
angle for large values of RM at the two frequencies considered.
This effect could produce a modulation which distorts the
pulse shape, in addition to the long term intensity varia
tions. For low and medium values of RM, the variations of the
rotation angle A0 are not so important, since they are small.
There is another effect that can be important for large
values of RM; that is the change in position angle when
observing the pulsar in different regions of the sky (i.e.
when observing the pulsar with different beams), or when
observing the pulsar on different nights. For example, at
_ o
45 MHz, with a RM = 1 rad m a change of 4% of N, L or Bn
(or a combination of them), will cause the position angle
to be orthogonal to its previous position. If the pulsar

Table 3-5. Summary of Peak flux densities, Mean Pulse energies and
Mean flux densities for Pulsar PSR2045-16 at 45 MHz.
Date
Periods Averaged
Peak Flux Density
Jy
Mean Pulse
-2 -1
Energy Jm Hz
Mean Flux
Density Jy
May
15,
1979
540
125
3.8
1.95
May
16,
1979
440
139
3.4
1.73
May
18,
1979
610
118
3.5
1.8
May
19,
1979
415
143
4.4
2.25
Unweighted
Mean
3.775

30
plotted separately in figure 3-5 and 3-6. A pulse appears
in bin 25 in block P112WA6, with an intensity 2.6 a above
the mean value; also in block P112WB6, the pulse is weaker,
being only 2.3 a, but is at the same location (bin 25).
Once again we have the fact that 2 pulses appear in the
same location in two independent sets of data.
From a precise knowledge of the pulsar period and the
elapsed time between the April 28 and May 25 appearances of
the pulse, it should be possible to preserve its phase so
that it would reappear in the same bin on the second date
as it occupied on the first. However, this could not be
done because of the period uncertainty. This problem will
be addressed later when analyzing PSR2045-16 data.
The pulsar was detected in only one beam on each of the
two nights, disappearing when the beam was switched to the
next position. This phenomenon will be discussed later in
the conclusions, when polarization effects are considered.
Before other tests made on this pulse are discussed, it
is interesting to note that in bins 9 and 21 of block
P112WA6 (figure 3-5) two more small pulses can be distin
guished. Although neither of them alone would be considered
statistically significant, they both seem to appear again in
block P112WB6 (figure 3-6) in the same position. The pulse
in bin 21 leads the main pulse in bin 25 by 125 msec
(resolution ^31 msec/bin) or approximately 38 of phase.
The pulse in bin 9 is separated by 496 msec from the
main pulse, which is about 150 (or 0.42 P, P = 1.88023 sec).

9 JL r An f:ni 1
1 3 | 6 |. 117B63 0FC F R f OUE .*C Y = 2 6,3 HZ
P l 1 2 A 6 '* W
25
79 T5T A> T 0 |H 39 M
05 STOP 0 1 H ,3 AM 10S TSA*P
7 9
>112456 MY
2 r.
7 9 r-,TA" DIM 34 v
0C STOR 01H 3RM 04S TSAVP
0?
OLOrK A /ERAGE-
l 2?
609 sr *MA
-
? 0 7 7 1 6 1 *.M4s
0-209 MAGNIF1C1T 103 =
too
Mj'nrr
OF BLO?
K 5 c
9 AVE P AG I j G
T1 ME 45 7.6 Sr~c
A V r d A r, e
n nr ?
4 vn. AGE
r>
i
1 ?-* ,6 5ft
107
1
1
?
1 22 *' 1 A
7 7 i
1
< 1
1
3
1 2? 6.59
300
1
1 +- _

1
a
122 7?A
190
1

1
5
12? A|i
193
1

1
f,
122 565
9 90
1

1
7
1 ?** .**5 1
7 07
1

1
9
12? 60?
loo
1
1
9
1 22.649
3 00
1

I
t 7
1 22 54 i
|OQ
1
i
1
1 !
12? 554
1 o^
1
1
1 2
l22,574
1 90
1
1
1 3
l
3 00
1
1
1 4
122 554
303
1
1
1*3
1 22,600
300
1
1
1 6
1 22,4 70
lrO
1

l
1 7
12? 637
3 93
1
-v.

1
* ^
1 2 *.*'' ^ |
10?
1

1
1
122 5 76
300
I
*

1
27
122 554
16 0
1 TIME
7.
1
1
1 ?'',= 79
300
1
1
o 7
1 22 *6 7
390
1
> 1 1
1
23
122,635
7 07
1
*-
1
*>6
1 2? .5 >?
390
1
d.
1
2*3
1 2 * 666
193
1
f ~ -*
1
2*>
1 22.9 1 8
1 00
1
26
1
r>7
l ?*> .697
103
1
-T-
1
? 3
1 77 577
393
!
I
?-
122.50?
7 99
1
1
>
1 2 7.600
3 90
1
1
3
I?? 541
190
1
1

1
32
1 22.71 0
3 00
1
i
*
I
t
127,559
3 90
1
1
34
12? 3 3
100
1

1
35
127.564
390
1
> *
1
16
12?.653
390
1
/ Â¥
1
* 7
12? e 50
190
1 -

1
39
122.660
2 1 5
1

1
Figure 3-16.
Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187863 s (P-160 ys) .

31
The pulse in bin 21 suggests the existence of a
subpulse leading the main pulse. Several authors have
indeed reported for this pulsar a complex main pulse com
posed of 2 subpulses, the leading subpulse decreasing in
intensity with respect to the trailing subpulse as the
freguency decreases (6, 7, 8). The separation between the
two subpulses has been measured at higher frequencies and
extrapolation from that data suggests a much closer separa
tion for low frequencies than the separation obtained at
26.3 MHz between the pulses in bins 21 and 25. In 1972 Bruck
and Ustimenko (4) reported some observations made at 25 MHz
with a linearly polarized antenna; they found that this
pulsar shows a variety of pulse shapes with rapid transi
tions from one to another. In one plot, they show a pulse
composed of 2 subpulses and separated by ^148 msec. Despite
the poor resolution used (P/16 or ^74 msec), the 2 subpulses
are clearly defined.
The separation of the subpulses as a function of fre
quency has been plotted in figure 3-8, including the result
by Bruck and Ustimenko and our result (i.e. for the bin 21
and 25 subpulses). The plot suggests that either the pulse
separation increases very rapidly toward lower frequencies,
or the subpulse detected at 26.3 and 25 MHz is a new one
which appears only at low frequencies. In figure 3-9 the
average pulse profile obtained in this thesis is compared
with that obtained by Bruck and Ustimenko. The phase of the

84
the
of T
W
0.5
The correction for pulse dispersion AtQ 5 using
above values, is:
At0.5 = 10'5 mS
The correction due to the post detection time constant
= 11.25 ms is:
c
T =7.8 ms
The true (or corrected) half intensity pulse width
is as before (equation 3-10):
WA c = Wn At. c T
0.5 0.5 obs. 0.5 c
The result for the 4 days is shown in Table 3-6.
Table 3-6. Observed and corrected half intensity
pulse width and pulse width at base
(assuming triangular shape).
Date
W
0.5 obs
(ms)
W
0.5
(ms)
Width at base
(ms)
May 15,
1979
44
31
62
May 16,
1979
38
25
50
May 18,
1979
44
31
62
May 19,
1979
44
31
62
The lowest frequencies at which pulse shape and width
measurements have been given for this pulsar are 150 and
151 MHz (10, 22). Lyne et al. (22) have given the separa
tion between subpulses at a number of frequencies; the
values are plotted in Figure 3-33. The separation between

77
is the most intense, but the intensity decreases
rapidly, and at 151 MHz is about 2/3 of the central
subpulse. The intensity of the central and leading
subpulses seems to remain almost constant with fre
quency. No indication of resolved subpulses can be
obtained from the data at 45 MHz. Nevertheless, the
difference in steepness of the leading and trailing
edges might suggest the existence of an unresolved
much less intense subpulse accompanying the main sub
pulse. This aspect will be discussed later when dis
cussing the pulse width and the effect of polarization.
3.2.a Peak Flux Density, Mean Flux
Density and Energy for Pulsar PSR2045-16
In order to estimate the peak flux density of this
pulsar, several radio sources were observed for calibration
purposes. Radio source PKS2034-17, one of the closest
to the pulsar, was clearly recorded despite the steepness
of the galactic background (23). This radio source has
apparently not been observed before at any frequency lower
than 80 MHz. However, an extrapolated flux density at
45 MHz was obtained from the spectral index derived from
flux density data at higher frequencies, as reported in the
literature.
Spectral index can be defined (17) by the relation:
(3-11)

64
Plots of the data for these dates are presented in
Figures 3-26 to 3-30. Figure 3-31 corresponds to a plot
of a typical day when only noise was present and no
pulse appeared.
In general the pulses were of different shapes,
except for the last two days, for which the shapes were
similar. Some have steep leading and trailing edges (as
on May 16, Figure 3-28); others have only a steep trail
ing edge and a not so well defined leading edge (as on
May 15, Figure 3-26). On May 18 and 19 (Figures 3-29
and 3-30) both pulses show a steep leading edge and much
less steep trailing edge.
On May 16 the pulse was clearly present with an
intensity of 3.48 a above the mean value over an average
of only 180 periods (Figure 3-27). On the other 3 days,
pulses were not seen clearly until the end of the averag
ing process for that night. For partial averages in
these cases, some indication of a pulse can be seen at
the location where the pulse appears after completing the
average, but its intensity never rises significantly
above the noise level. No indication of the existence of
a complex pulse or interpulse can be deduced from the
plotted data.
Pulsar PSR2045-16 presents at higher frequencies a
complex pulse, composed of 3 subpulses, with the trailing
subpulse decreasing dramatically in intensity with
decreasing frequency (22); at 610 MHz the trailing pulse

PUL SAP P'jRI 1 3H! P FR I 03 = 1 1 * Of* 3 S GC F 9 E 3 Uc 9 C Y = 2 6 .0 ?
P112WA6 Y 7 5 79 TST Ac T 0 1H 0 9M | 0 STOP 01H 3 A M IOS TMD O?
P11PW06 MAY ?5 7? T5TACT O 1H UM ios ST3" O l H 3RM 04S TSA VP 02
BLOCK AVEPAGE= 122.609
HGM' r O- 079
3 SISMA* o. 2 36 MAGNIFICATIONS IDO
NJMF9 PF LOCKS* 2
Vi l AGI NC T I MF -
A OO.7 SrC
N
N A RES
1
2
A
5
I O
I 1
1
I 3
1 4
I *
1 A
l *
! 1
1 5
'I
?2
-
5
25
2
?-
3 O
2 1
30
**A
36
37
A VESA GE
12? *92
122 677
1 2 > A 3 3
122 636
12? 536
l 2? .60
12? 656
12? 700
1??.^4
12? 7?s
12? SA8
12?.6?0
12? A 8 A
12?.649
l2?.05O
122 60S
I 22 6 23
1 2? 09 7
l22 556
122.528
12? 551
1 2? 6 ng
122 .68 4
12? 755
122 792
1 2?. M *
12? 551
1 2?- 564
122.617
l 2* 351
122 633
1 2?* 4 1
12? 5?n
l2.56C
l2?.669
122 65a
122. 49 7
122.848
nr c1005
A VF5A GE 3
093
O 50
39?
093
3 90
350
93
3 50
3 0
3 90
O <50
390
093
7 50
.no
050
390
3 90
00o
3 90
3 o?
7 90
390
090
090
3 9->
OG0
0 90
.090
090
O 50
3 90
OG0
o 90
390
3 90
3 50
51
TIME
INTENSITY
Ln
>vj
Figure 3-20
Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188063 s (P+40 ys).

136
he came to the University of Florida to continue his studies
in astronomy. While studying at the University of Florida
he has been teaching elementary physics laboratories, elemen
tary astronomy laboratories and digital electronics labora
tories.
His other interests are mountain climbing, Andean music
and photography.

86
the leading and central subpulses and the trailing and
central subpulses is plotted separately.
As in a previously mentioned case, the trailing pulse
seems to decrease very rapidly in intensity toward lower
frequencies (22), and it is likely that at 45 MHz it no
longer exists or at least it can be very weak. The mean
value at 45 MHz for the pulse width at its base (obtained
by measuring the width at half intensity and assuming a
triangular shape) of 61 ms has also been plotted in Figure
3-33.
The plotted value for the pulse width at its base, lies
close to the extrapolation of a straight line passing
through the points representing the separation between the
leading and central subpulses. A possible conclusion of
this is that the pulse observed at 45 MHz is composed of
two unresolved pulses which might correspond to the leading
and central subpulses at higher frequencies; the pulse
width at its base would correspond to the separation between
the two subpulses.
No polarization measurements have been made for this
pulsar at low frequency. The lowest frequency at which
polarization measurements are available is 409 MHz (32, 33).
The mean percentage of linear polarization at 409 MHz is
40%, and it changes during the course of the pulse, reaching
about 70% for the first subpulse. The stability of the
linear polarization position angle is low for the wings of
the integrated pulse profile, due to the occasional

134
38.R. N. Manchester, A. G. Lyne, J. H. Taylor, J. M.
Durdin, M. I. Large and A. G. Little, "The Second
Molonglo Pulsar Survey Discovery of 155 Pulsars,
Monthly Notices of the Royal Astronomical Society,
185, 409 (1978).
39. J. P. Oliver, Personal communication (1978).
40. A. J. Turtle, J. F. Pugh, S. Kenderdine and I. I. K.
Pauliny-Toth, "The Spectrum of the Galactic Radio
Emission," Monthly Notices of the Royal Astronomical
Society, 12, 297 (1962).
41.J. Levy, Personal communication (1980).

133
24. B. Y. Mills, 0. B. Slee and E. R. Hill, "A Catalogue of
Radio Sources Between Declinations +10 and -20,"
Australian Journal of Physics, J^, 300 (1958).
25. O. B. Slee, "Culgoora-3 List of Radio Source Measure
ments," Australian Journal of Physics, Astrophysical
Supplements, Number 43, (Nov., 1977).
26. J. H. Blythe, "Results of a Survey of Galactic Radia
tion at 38 MHz," Monthly Notices of the Royal Astro
nomical Society, 117, 652 (1957).
27. J. W. Findlay, "Absolute Intensity Calibrations in
Radio Astronomy, Annual Review of Astronomy and Astro
physics 4, (1966).
28. A. J. Turtle, A. E. Vaughan, "Discovery of Two Southern
Pulsars," Nature, 219, 689 (1968).
29. W. Sieber, "Pulsar Spectra: A Summary," Astronomy and
Astrophysics, 28, 237 (1973).
30. D. C. Backer, Personal Communication to S. T. Gottesman
(1975).
31. R. N. Manchester and J. H. Taylor, Pulsars, (W. H. Free-
ma and Co, San Francisco, 1977).
32. R. N. Manchester, J. H. Taylor, G. R. Huguenin, "Observa
tions of Pulsars Radio Emission II. Polarization of
Individual Pulses," Astrophysical Journal, 196, 83 (1975).
33. G. R. Huguenin, R. N. Manchester and J. H. Taylor,
"Properties of Pulsars," Astrophysical Journal, 169, 97
(1971).
34. J. D. H. Pilkington, A. Hewish, S. J. Bell and T. W. Cole,
"Observations of Some Further Pulsed Radio Sources,"
Nature, 218, 126 (1968).
35. Yu. I. Alekseev and S. A. Suleimanova, "Polarimetry of
Pulsars at 3.5 m Wavelength," Soviet Astronomy, _21, No. 2,
180 (1977).
36. Yu. P. Shitov, "Fine Structure of Pulsar Radio Spectra,"
Soviet Astronomy, _16, 383 (1972).
37. R. N. Manchester, "Pulsar Rotation and Dispersion Measures
and the Galactic Magnetic Field," The Astrophysical Jour
nal, 172, 43 (1972).

3 J.
/ AF 3 ; 5 l m
is =nn- i. im7
irt.A *.*-:* -
p: qu" Nr y* ? *.
3 *17
p 1
1 ? W A 6 mAY 25
79 7START 01H 3)M
1 OS STCP
01H 3AM IOS
T S A M P
02
p 1
i?*r)6 v'av
79 T ~ A? 0 1 H 3 A
1 OS STno
01 -1 38M 0A S
TSA MO
02
fl-DC< AV*r?AGE 1 2?.
609
ST v A r
C- ? 6 3 S J r.MA
= 0*?0*
MAGNiriCAT1HN=
1 00
NJH trc of MLOCKS-?
2
AVCFAGING
T | ME- A,2.2 r
N AVCBA SC HFOIOOS
SI !(|C5
4 VC RAG'S I*
1
1 ? so 7
3 o)
2
122 A 4 3
70)
1
122.55A
3 90
A
I 2 ? a 0 0
5
12? s a
2 ? 1
6
1 22.579
70)
7
l ?. 656
3 C)
0
12? S 9
30?
9
12?.68?
3 50
1 3
1 2?. A 7
3 9)
1 1
12?
3 90
1 2
1227?8
a 90
1 3
390
1 1
t 2 ? S 5 Q
*90
15
122.564
3 50
IS
12? .MS
3 9)
1 7
l ? .61 f
3 93
1 8
12? ,697
30)
IQ
1 22.58?
3 50
0
122.S4 1
39)
2 1
1 22.587
3 9)
r
12?.S4 |
390
23
\ 09 *,7"*
30)
2 A
122 600
1 QO
?5
1 22.6 a?
30)
1 ?> 5 |
30)
?7
12? s??
3 90
? J
122.579
390
n
1 7 q a
7 C)
30
i 22. A55
3 50
* 1
1?? .so?
3 on
V
12? 7 7 3
3on
3 3
\2? 650
3 f>)
"* A
122.607
300
7 7
122,515
70)
36
12? 636
3 50
7 f
1 ?? .63.3
390
3*3
l22 536
2 1?
TIME
INTENSITY
Figure 3-14. Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.187383 s (P-640 us).

APPENDIX B
SAMPLING RATE, TIMING PULSE FREQUENCY
AND DATA POINTS PER SECOND
For the configuration of the data reduction system used
(Figure 2-2), the sampled values are sent to the Amdahl 470/V6
computer in real time. In that condition the rate of trans
mission of data into the Amdahl is the limiting factor for the
sampling rate.
The absolute maximum transmission rate to the Amdahl
470/V6 is 1200 Band (120 characters/second). The maximum use
ful rate is 60 characters/second; 40 characters/second is a
more conservative (but reliable) rate. Since the A/D con
verter has 12 bits, which means 3 characters per data point,
the transmission rate will be 40/3 = 13 data points per
second.
The number of data points per second (DPS) obtained from
the KIM I is given by the relationship (39):
DPS = TPPxNAVExTSAMP (B-l)
where TPP = KIM I timing pulse period (ms)
NAVE = Number of sampled values averaged before
sending to the terminal
TSAMP = Parameter for the KIM I program
The TPP for the KIM I must not be shorter than 4 ms
(250 Hz); the value used in the digitizing process for the
107

LIST OF REFERENCES
1.A. Hewish, S. J. Bell, J. D. H. Pilkington, P. F. Scott,
R. A. Collins, "Observations of a Rapidly Pulsating
Radio Source," Nature, 217, 709 (1968).
2. Yu. M. Bruck, J. G. Davies, A. D. Kuz'min, A. G. Lyne,
V. M. Molofeev, B. Rowson, B. Yu. Ustimenko and Yu. P.
Shitov, "Radio-emission Spectra of Five Pulsars in the
17-1420 MHz Range," Soviet Astronomy, 2_2, 588 (1979).
3. F. N. Bash, F. A. Bozyan and G. W. Torrence, "Observa
tions of CP1919 and NP0532 at 38 MHz," Astrophysical
Letters, 7, 39 (1970).
4. Yu. M. Bruck and B. Yu. Ustimenko, "Decametric Pulse
Radio Emission from PSR0809, PSR1133, and PSR1919,"
Nature Physical Science, 242, 58 (1973).
5.Yu. M. Bruck and B. Yu. Ustimenko, "Decametric Radio
Emission from Four Pulsars," Nature, 260, 766 (1976).
6. H. D. Craft Jr. and J. M. Cornelia, "Frequency
Dependent Pulse Width for CP1133," Nature, 220, 676
(1968).
7. V. A. Izvekova, A. D. Kuz'min, V. M. Molofeev and Yu.
P. Shitov, "Energy Spectra and Pulse Shapes of Pulsars
at Metre Wavelengths," Australian Journal of Physics, 32,
25 (1979).
8. V. A. Izvekova, A. D. Kuz'min, V. M. Molofeev and Yu.
P. Shitov, "Radio Flux Density and Pulse Profile of
Pulsars at 102.5 and 61 MHz," Soviet Astronomy, 23,
179 (1979).
9. O. B. Slee and E. R. Hill, "Pulsar Energy Fluxes at 80
MHz," Australian Journal of Physics, 2_4, 441 (1971).
10.M. M. Komesaroff, D. Morris and D. J. Cooke, "Linear
Polarization and Pulse Shape Measurements on Nine
Pulsars," Astrophysical Letters, 5, 37 (1970).
131

MCAPGM1 Progr

APPENDIX A
LIST OF PULSARS
Tables A-l and A-2 contain the pulsar candidates to be
observed using the University of Florida 26.3 MHz array and
the 45 MHz array at the Maipu Radio Observatory.
Table A-3 gives the references from which the data at
the different frequencies were obtained.
103

10
resolution of ^31.5 ms (P/38; P = period) obtained for
this pulsar.
The detected IF output of the receiver was recorded
on channel 1 of the Magnecord at 15 IPS and the recording
level for the galactic background was set to about 7 VU on
the recording level monitor, to prevent any saturation during
signal peaks.
The time code generator was recorded on channel 2 of the
Magnecord. This signal serves a double purpose: it provides
a time reference for the pulsar signal, and later on, in the
data reduction process, provides the timing signal for the
KIM I microprocessor.
As a back-up to obtain the correct time, at the begin
ning of each observing session the WWV signals were recorded
on channel 1 (the TCG also on channel 2) for about 2.5
minutes. The correct time can be obtained by measuring the
delay between the 1 PPS of WWV and the 1 PPS of the TCG.
The receiver was tuned to a center frequency of 26.3 MHz
and the bandwidth set to 8 kHz. Using data for higher fre
quencies an extrapolated pulse width (at the base) of about
80 ms was obtained. Pulse broadening due to dispersion
(DM = 5 parsec cm ) is about 18 ms (for a bandwidth
Av = 8 kHz), which was considered acceptable.
2.2 Data Acquisition of Pulsar PSR2045-16
Pulsar PSR2045-16 was observed for 11 nights between
May 15 and May 30, 1979, using the Maip Radio Astronomical

PULSAR PSR2045-16 PFRI00=l96l383 SEC FfiEQUENCY=45 MHZ
P204SA1 MAY 16 79 TST ART OVH 44M 34S STO 09 H A 7m 45S TSAVp 02
P 2 O 4 5EJ l MAY 16 79 TST AR T O 9H 4 7M 5 OS S 7 CP 09 H 51M OAS TSAMP 02
BLOCK AVERAGE = 123*90? SIGMA= 0*093 3 SIGMA= 0*279 MAGNIFICATIONS 100
NUMBER CF ULCKS= 2 AVERAGING TI ME = 349. 1 SEC
P I N
A V F r, A G F
PERIODS
N JM'JE R
AVERAGED
1
123*999
1 80
2
l 2A 1 1 1
i eo
J
123.949
l 60
A
I ?A .0 06
i eo
5
1 23 *P95
i eo
6
12 1.97 3
i eo
7
12 5* VAS
1 tO
e
123.461
i eo
9
123467
i eo
10
1 23 .56
i eo
1 1
124.022
i eo
12
123*912
i eo
1 3
123*811
1 t 0
l A
123.5 3 2
i ec
1 5
l24.006
I 80
le
124.079
1 60
1 7
1 2395 0
l 60
1 9
1 24.05 6
1 30
1 9
123.^17
iao
2 0
1 2- .039
i tO
2 1
123 .FtA
i eo
22
124.045
i eo
23
124.045
i eo
24
123*495
i a j
25
123.017
i ec
2 fc
124.006
1 60
27
124.306
i eo
28
1 2a i 56
i eo
29
l 23 */39
l 80
20
12 1.9 22
I 8 0
3 l
1 24.06 7
i eo
32
124.05?
1 60
33
124.017
1 60
34
124.061
1 90
35
124.1 17
i eo
36
1?4.0? 3
1 60
37
123.934
i eo
33
123.996
1 oO
39
1 ? .973
1 60
40
123.906
i eo
41
123.950
1 80
4 2
l24.900
i ec
A 3
123.967
i eo
44
123.964
1 80
45
124.lOC
i eo

123.923
l 80
47
1 ?3.*21
1 80
A 9
12.9c l
i eo
A 9
124.0 11
i eo
5 c
1 23.945
1 90
5 1
123. i 06
i eo
52
1 24 1 56
i eo
53
123.97P
1 dD
54
123.756
1 80
55
123.964
1 60
56
1 c 4.06 7
1 60
57
124.034
1 80
SH
I2J.HO!
l 1 7
59
124.0? t
1 80
60
12 1.945
1 80
61
123.884
l 80
62
1 2 4. 0 44
1 1 7
TIME
INTENSITY
09
00

88
the first occasion was May 15 and 16, 1979, and the second
May 18 and 19, 1979.
The pulse period can be obtained by subtracting the
epoch of apparition of the pulse E on two consecutive days
and dividing by an integer number N.
The epoch of the pulse can be calculated from the
relationship:
where:
E
T
start
+
nxR
(3-14)
T = time start digitizing
s tax t
n = bin number where the pulse occurs
R = plotting time resolution
The information related to the two pairs of days has
been summarized in Table 3-7.
The mean apparent period can be calculated from the
relationship:
P
AE
N
(3-15)
where AE: difference in epoch of pulse apparition on two
consecutive days (elapsed time between pulse
apparitions)
The mean apparent periods obtained with this method
are compared with the DOPVEL periods in Table 3-8.

PULSAD PS*t 133M& lOn-sl 1B |fl.3 S^C FH!*OUENC Y= ?AO H?
P 1
1 ?* A* MAY ?*
7P
r AC ?1H ]
1 *S
-,Tnr>
01-1 34M
10S
TS AMP
0?
P 1
I2WH6 u A Y ?5
79
TSTAPT 01H Hm I
IOS
SI OP
0 1 H 3*M
OAS
TSAvir
?2
OL CTK
A/ EPAGFe 1 ? '' 60 9
r I r.v A =
J
0 7f 7 c 1 r, MA =
?7ft
MAGNIFI CAT ION-
1 00
N
i'F PL D f 5 2
AVrWAf.l NG
T I
F: 4 5 ? * R SC
P IN
AVC9AGE
' EP I no5
NJV7FT5
wrc AGrn
1
1 2? oQ">
700
">
P?, U
300
7
l?? AS1
70Q
4
12? 566
3 SO
3
1 2? ' ? 7
3 00
A
|?? 5RO
300
7
12? 700
300
*
122.649
390
O
l?* *61
7 c*0
1 '
122 633
390
1 1
1 ?? .*37
790
1 ?
12? A*6
.7 90
1 3
122,595
790
1 A
1??.*70
300
1 s
1?? 40?
790
16
122.646
390
1 7
12?,'OR
70-'
1
122 526
790
19
12?. 4 77
3 90
? o
1??.57c
700
? i
12?. 643
790
?>
I 2?r 05
390
?7
12? 7o->
700
?4
l 27,71 0
190
?r
12?. 7 4 7
390
?*
12? r- *o
707
27
1 22. 50?
790
77
1 2?.* ?3
3 CO
?Q
12? 57?
790
3?
1 2? *8 7
7 90
1
l ??.5 07
390
7 ?
12? '?
790
33
122 j 538
790
14
12?.*61
390
75
12? 570
30
7*>
1 22 *1 R
7 rjo
7*
1 2? * 36
390
38
122 009
5?
TIME
INTENSITY
Ln
VD
Figure 3-22.
Pulsar PSR1133+16 data recorded on May 25, 1979, analyzed with a
period P = 1.188183 s (P+160 ys).

CHAPTER 4
CONCLUSIONS
Two pulsars have been successfully detected using
narrow bandwidth and long averaging times. Pulsar
PSR1133+16 was detected with the University of Florida large
array at 26.3 MHz and pulsar PSR2045-16 was detected using
the Maipu Radio Observatory large array at 45 MHz. Pulsar
PSR1133+16 was strong enough to be considered statistically
significant (pulse deflection larger than 3 standard devia
tions) in the data of 2 nights (out of 10 nights reduced).
Pulses of PSR2045-16 were found on 4 nights (out of 10
reduced) with enough intensity to reach the 3 standard devia
tion level. Typical averaging times for both pulsars were in
the range 400 to 600 periods.
Both pulsars show large intensity variations on a time
scale of days. Typically, one would appear after an aver
aging time as short as ^200 periods, and then would fade,
being undetectable for averaging times as long as ^1500
periods over succeeding days.
The peak pulse flux density obtained for PSR1133+16 is
about 150 Jy; for PSR2045-16 the peak flux density is in the
range of 120 to 140 Jy. These values (which have relatively
large probable error, in the order of 50%), are higher than
the values of the peak flux density measured by others at
93

100
nights with data are 27 days apart. It is not clear up to
the present whether the discrepancy of the periods is due
to instrumental problems or insufficient precision in the
calculations of the predicted period.
From the results obtained in this thesis, it has been
shown that it is possible to detect pulsars at both 26.3
MHz and 45 MHz, with certain limitations. In the light of
the results obtained, some improvements and modifications
in the equipment and data reduction can be suggested which
One such modification is already being put into effect.
A greatly improved transistorized 26.3 MHz receiver which
will replace the old receiver is under construction and is
nearly completed.
Recording at a much wider bandwidth (perhaps 50-100 kHz),
and utilizing a de-dispersion technique, would considerably
increase the signal-to-noise ratio. The wide band signal
would initially be recorded using an instrumental tape
recorder of suitable bandwidth. It could then be played back
into the Saicor Real Time Spectrum Analyzer, the output of
which could be digitized and further analyzed by means of a
i
microprocessor or a minicomputer (like the PDP 1134).
After the data has been converted to a power spectrum appear
ing in the 200 output frequency channels of the Saicor
spectrum analyzer, the frequency dispersion can then be
removed by means of the computer. With a 100 kHz-bandwidth

PUL S AH
PSP?045-
1 6
Pfwior)*i
9b 1
in 3 SEC TRF UFNCY
-45
MHZ
P2045 A1
V AY
1 6
70
TSTAKT
0 9M
4AM
3 4 S
STOP
09 H
4 7M
49S
T SAMP
02
P?04blU
MAY
1 6
79
TSTAH T
09H
4 7 M
5 33
ST OP
09M
5 IM
C4S
TSAMP
02
P2045C1
MAY
1 6
79
T ST A9 T
09M
5 1 M
0 5S
STOP
09H
54 M
1 9 S
TSAMP
02
P2045U1
MAY
1 6
79
TST AKT
0 9 H
S4M
2 OS
STOP
09 H
57M
3 4 S
TSA MP
02
P20 4 5E l
MAY
l 6
79
T ST ART
09H
S7W
3SS
STOP
l OH
0 OM
i as
TSAMP
02
nt nCK
AVTPACE= 123
.98 1
S 1 GMA
=
0.
0 6 1
3 SIGMA
= 0
.183
MAGN IF 1 C A T ION =
100
NJ M0E8
OF LLOCK S=
5 AVLF-AGING
T I
ME = 8S3. 3
src
PIN
AVERAGE
PEP ICOS
Ku^MEP
AV EPAGTD
1
l23.950
4 40
2
1 24 .0 22
440
3
123.Q93
4 40
4
1 24 C2 2
440
5
l23.4?4
4 40
6
1 2.3.927
4 40
7
l 23.852
4 40
9
123.926
440
9
1 23.979
4 40
1 )
123.95
4 40
1 1
124.004
4 40
1 2
l23.866
384
1 3
123.9 34
4 40
1 4
123.886
4 40
15
123. 89 2
44 0
lb
l23995
4 40
1 7
124.052
440
18
124.047
4 4 0
1 9
123.893
4 40
2 0
1 2 3.94 3
4 40
2 1
1 23.91 1
44 0
2?
124.001
.3 7 7
23
1 23.882
4 40
24
1 24.0.3 1
4 4 0
25
1 2 3.9 29
4 4Q
26
124 ,02 3
4 4 0
27
124.192
440
28
l23.954
4 40
29
1 23.9 36
4 40
30
123.904
440
31
l24.06\
440
32
124.027
440
33
123.9 22
4 4 0
34
123.986
4 40
35
l24.022
4 40
36
124.013
4 40
37
123 .9 70
4 40
38
l23.995
4 40
39
123 .954
440
40
124.0 18
4 40
4 1
1 24.0 34
4 40
42
1 24.054
4 40
43
123.966
4 40
44
124.016
4 40
45
l 23 .993
440
46
124.004
440
47
123.931
4 40
48
124.002
377
49
1 24 .00 7
4 4 0
bC
123.993
440
5 1
123.986
440
52
124.061
4 40
53
123.982
4 40
54
1 2.3.804
4 40
55
1 24.0 l l
4 40
56
124.102
4 40
57
1 24.023
440
68
124.032
3 77
59
124.025
440
( 0
123*986
4 40
61
l 23 .969
4 40
t'
124.021
377
TIME
INTENSITY
I
I
I
I
I
I
I
27 3.48 SIGMA
I
I
I
I
I
1
I
o

129
AT for a radio source of known flux density S. The
a
relationship that gives the effective area is:
2k AT
Ae = ^ (E-3)
AT^ is the change in antenna temperature referred to
the dipoles, corrected for losses e between point (2) and
the dipoles. If AT' is the change in temperature measured
ci
at point (2) then the actual change in antenna tempera
ture is:
AT'
AT = (E-4)
a e
where e = transmission line efficiency, which in this
case considers the losses between point (2)
and the dipoles.
From the calibrations, the following values were
obtained for 3C433:
S = 267 jy
AT' = 295K
a
T = 25,000 K
a
e = 0.19
Using equations E-3 and E-4, the following value is
obtained:
A = 16,050 m2
c
This value needs to be corrected by a factor p^ = 0.69,
given in the Array program (11), which considers the fact
that the source is off the center of the beam. The corrected
effective area of the antenna is then:
A = 23,260 m2
e

99
The values for the peak flux density, the mean flux
density and the mean energy are higher than the values
obtained from extrapolation at higher frequencies, with
the exception of a value of peak flux density of 150 Jy
obtained at the relatively high frequency of 111.5 MHz for
PSR1133+16. Izvekova et al. (7) and Kuz'min etal. (20)
made simultaneous observations of this pulsar at several
frequencies in order to obtain the spectrum. They found
that the average spectrum (average over several nights)
presents a turn-over at about 60 MHz. Nevertheless they
also found that the spectrum shows large variations from
day to day in the range 61 151 MHz, often with a change in
the sign of the spectral index. It is interesting to note
that the spectrum of some individual nights does not show a
turn-over but rather an increasing intensity with decreasing
frequencies. This might explain (at least in part) the
unusually large intensity of the pulses at 26.3 MHz, when
compared with average values obtained by other authors at
higher frequencies.
Averaging of pulses for two different nights was not
possible. For pulsar PSR2045-16, the period obtained from
the DOPVEL program differs from the period deduced from
observations on two consecutive nights by amounts between
8 to 28 ys; this discrepancy is large compared with the
required precision for averaging over two consecutive days
(i.e., ^ tenths of ys). No attempt was made for PSR1133+16
to average data of two different nights since the only two

130
The value of the effective area can be also expressed
in terms of the change in antenna temperature in K, divided
by the flux density in Jy (K/Jy); this quantity is usually
denoted as A* (12).
e
AT x P-,
A| !;r (E-5)
Substituting in E-5 the values
AT = AT'/e = 1553K
a a
S = 267 Jy
P1 =0.69
the following value for A* is obtained
A* = 8.3 K/Jy

48
Table 3-3. Observed half intensity width, corrected
half intensity width, and pulse width
at base, for pulsar PSR1133+16
Date
WA c observed
U D
Wq corrected
W,
base
(ms)
(ms)
(ms)
April 28, 1979
38
21
42
May 25, 1979
53
36
72
The graph of Figure 3-12 is a plot of the values of
the corrected pulse half-intensity width at several fre
quencies (6, 8), including the values of this thesis
obtained at 26.3 MHz on May 25, 1979; this is the more
reliable of the two values presented in Table 3-3.
3,l.c Tests Performed with the Pulse
Obtained on May 25, 1979
An independent and simple method for testing the pulses
to see if they are indeed from a pulsar having the assumed
period is to repeat the averaging process using both shorter
and longer periods than that given by the DOPVEL program.
If the observed pulses were spurious ones which happened to
have approximately (but not exactly) the same period as that
of the assumed pulsar, then the averaged intensity should
either increase or decrease for the shorter period, and do
the opposite for the longer period. If the averaged
intensity is a maximum at the assumed pulsar period, then

14
In order to achieve the desired time resolution and to be
able to sample the data at an appropriate rate, the magnetic
tapes were played back at 3 3/4 IPS, providing a slow down
of 4:1.
Pulsar signals from channel 1 of the Magnecord are sent
through a diode detector and smoothing filter to the 12 bits
A/D converter. The filter time constant was adjusted to 45
ms for both the rising and falling edges of the noise pulses.
The TCG signal from channel 2 of the Magnecord is con
nected to the input of the TCG (working now as a reader) to
provide both a visual display of the time and a 1 PPS pulse
(taken from a modified output). The latter gives the syn
chronizing signal to start the sampling of the data at a
precisely known time. It is particularly important to know
the time at which the sampling is started in order to be able
to combine in phase 2 or more blocks of data.
In order to obtain the timing signal for the KIM I, the
TCG signal from channel 2 of the Magnecord is connected also
to a phase locked loop (PLL) circuit. The PLL circuit locks
at a particular frequency of the signal, filtering out the
rest of the frequencies present in the complex wave form of
the TCG signal. The 1000 Hz signal of the TCG is transformed
to 250 Hz due to the 4:1 slow down. The PLL was carefully
tuned to 250 Hz, and the output then divided by 2. The resul
tant 125 Hz constitutes the timing signal for the KIM I.
A gated DC output from the PLL circuit was monitored on
a Brush Mark II chart recorder. This provides a good means

MA I N
DATE
G0C04
12/27/37
PAGE 0001
POPTRAN IV G LEVEL 21
C NEW MCA PROGRAM FOR SPICA DATA
000 1
INTFGER44 DATA(36).TITLC(20) ,N0IN( 100 ) NI3INI ( 1001 .NURI N( 100)
0002
REAL *4 BIN( l00).PLOT(51) PLOT I (51 1 HUF( I 00 ) B INK 1 00 ) .BK AVE( 100)
0003
OATA ALF 0/4 H /AL F X/4 H /,ALFC/4H' /.ALFP/4H* /
0004
1
DO 40 1=1.100
0005
8 IN( I )=0 0
C006
niNl ( I )=0.0
000 7
NBIN(I)=0
0005
NOIN I ( I ) = 0
0009
40
CONT IN'JE
001 0
DO 60 1=1.51
001 1
PLOT I ( I )= ALFB
00 1 2
50
PLOT ( I ) = ALFO
00 1 3
PLOT( 1 ) = A LF C
00 1 4
PLOT (26 ) = ALFP
0C1 5
PLOT(51)=ALCC
00 1 6
PLOTI(1)=ALFC
001 7
PLOT I (26 ) = ALFP
00 1 5
LO TI (51 1=ALFC
001 9
0020
505
FORMAT(I5/cl00/IS/F10.0/F5.0/15)
002 1
0022
5000
FORM AT(20A4)
0023
IF( IRIT.EQ.O )IBIT = 12
0024
IF(IPSCAL.EQ.0)IPSCAL=1
0025
M A X D= 24
0026
IF(IRIT.LT.12)MAXD=36
0027
I F(NANA.EQ.O)NANA*20
0028
WRITE(6.606)TITLE.iniT,TSAMP,NANA.TPER.PZcR0.IPSCAL
0029
506
1
FORMAT( l 48TAOT'/QX.?0A4/10X,110.Ft2.4,Il0.Fi2.6.F10.0.
1 110)
0030
IF( ( I BIT E Q 12)ANO.(PZZRO.GT.2046. ) )PZrRO=0.0
0031
IF( ( I3IT.E3.12). AND. (PZEPO.LT.-2 048. ) ) P7F R0=0.0
0032
IF((I3IT.E0.8).AND.(PZERC.GT.128.))PZERO=0.0
0033
IF((IBIT.EQ.8).AND.(PZERO.LT.-128.))PZERO=0.0
C
GET TER IN MILL1SEC
0034
T PER= TPER*1000.
0035
NBMX = TPER/TSAMP4-1 .0
0036
NPER=0
0037
T DAT = TSAMP
0038
I 8= 0
0039
80
I D = 0
0040
IF( I 3IT.LT.12)GO TO 82
0041
0042
500
FORMAT(2473.18)
0043
GO TO 120
0044
82
RpAD(5.501 ,END = 4 0 0) (DAT A( I ). I = 1.36) .NTFS7
0045
50 1
FCRMAT( 36Z2. I 8 )
0046
120
IF(NTEST.NE.O )G0 TO 1
004 7
ID=I0*1
0 C 4 8
T DAT = TDAT 4-TSAMP
0049
TNPE R = F|_0 AT ( NPFP 1 )*TPER
0050
I F( TDAT-TNPER) 14 0. 160.160
0051
140
IB=IRFl
0052
IF( ( I 91 T E Q. 1 2 ) AND. (DAT A( IP) .GT *2048. ) )DATA( ID) = DATA! ID)-4096.
0053
IF((IBIT.EQ.8).AND.(DATA(ID).GT.128.))DATA(ID)=DATA(ID )-256.
0054
RUF(18)=D A T A(ID)
0055
IF(ID.EQ.MAXD)GO TO 80
0056
GO TO 120
0057
160
DO 180 I=1,IB
0058
BIN( I )=UIN( I )BUF( I)
0059
180
NRIN( I )=N9 I N( 1)41
0060
NPE R = NPF R1
006 1
IF(M00(NPER.NANA).EQ.O)GO TO 220
0062
200
I B= l
0063
IF(( I BIT.EJ. 12).AND. (DAT AC IP).GT .2040. ) )DATA( 1D) = 0ATA( ID) 4096.
0064
IF((IRIT.E0.8)-AND.(DATACID).GE.128.))DATA(ID)=DATA(ID)-256.
0065
RUF(IB)=DATA(ID)
0066
IF(I 0 .EQ.MAXO)GD TO 80
0067
GO TO 120
0068
220
WRITE(6.510)NPER
0069
51 0
FORMAT! 1 OX, l l 0 )
0070
SUM=0.0
007 1
SUMN=0.0
0072
SUM 5Q = 00
0073
SUMI=0.0
0074
SUMNI=0.0
0075
SUMSQI=0.0
115

20 -
BRUCK AND USTIMENKO (4), 25 MHz
OCTOBER 11, 1972
INTENSITY
(ARBITRARY
SCALE)
INTENSITY
(ARBITRARY
SCALE)
Figure 3-9. Comparison of the pulse profile for pulsar PSR1133+16
obtained in this thesis (average of 340 periods) with
that obtained by Bruck and Ustimenko (4) (average of
v212 periods). The bins of the 26.3 MHz pulse profile
have been shifted in order to align the pulses.

PULSAR PSR1133AI6 p ER I 00 1 1 879 9 9 SEC FRE CUENCY*26.3 MHZ
P112EA3 APRIL 23 7} T S T ART 02H 03M 03S STOP O2H OSM SOS TSANP 02
PI12LB3 APRIL 28 79 T ST AP T 02H O 5M 55S STOP 02M 1 OM OOS TSAMP 02
BLOCK AVERAGEs 120*161 SIGMAs 0* 06 9 3 SIGMA* 0*208 MAGNIFICATIONS 100
NUM'TEK r.F PLPCKSs 2 AVERAGING T IMF.* 39A.6 SEC
n!N AVFRAGF PERI OOS
NT" BER AVERAGES
1
120024
340
I
2
120.156
340
1
3
12 0.12?
340
1
4
123.197
340
1
5
120.197
340
1
6
120 .20 6
340
1
7
12053
340
1
8
1 20.1 74
340
1
o
120.221
340
1
1 0
1 23.1 29
340
1
1 1
1 20.1 26
340
1
12
120 .032
l ( 5
I
1 3
120.J 1
340
1
1 4
120.107
340
1
15
120.06?
340
1
16
i 2: .oo i
340
1
1 7
120.226
3 40
| TIME
19
120.1 74
340
1
1 9
120.200
340
1
? 3
1 2 0 1 3 8
3 40
1
PI
l 19.99 7
3 40
1
22
120.138
340
1
23
i20.29 1
340
1
24
120259
1 4 0
1
25
120.050
340
1
26
120.121
340
1
27
1 20 .094
340
I
20
1 20.224
340
I
29
12 0.050
340
I
30
1 20 1 4 i
340
1
31
120.100
340
1
2?
120.091
3 40
1
3 3
l 20 .4?7
340
1
34
120.273
340
1
35
i2C.ieo
340
26
120.173
340
1
37
120.097
340
1
29
1 2 0. 050
2 1 8
1
INTENSITY
I
I
I
I
I
I
I
33 3.0 SIGMA
I
I
I
I
I
t
I
I
I
I
I
I
I
I
I
NO
U>
Figure 3-1.
Pulsar PSR1133+16 (Blocks P112EA3 and P112EB3)
on April 28, 1979. Average over 340 periods.
recorded at 26.3 MHz

91
T
where N = .
Assuming an elapsed time of 24 (86400 s), a resolu
tion of 0.0315 s and a period of 1.961386 s, the following
accuracy for the period is obtained:
A = 0.36 ys
c
This accuracy should bring the two pulses in phase
within 0.5 bin.
It is not clear what caused the discrepancies between
the DOPVEL period and the period deduced from the observa
tions; it could be an instrumental problem or a precision
problem in the calculations in the DOPVEL program. Both
possibilities were checked but the result was inconclusive;
further investigations need to be made.
3.2.d Pulsar PSR2045-16 Pulse Intensity
as Function of Time
Figure 3-34 is a plot of the relative pulse intensity
as a function of time for the pulse obtained on May 16,
1979; this pulse is the strongest and best defined. Each
point plotted represents the relative intensity with respect
to the noise average, averaged over 10 periods; the time
axis scale is in pulsar periods. It can be seen that the
pulse has large intensity variations. Most of the time it
has a positive value with respect to the noise average, and
it seems to show an almost periodic intensity modulation of
^140 periods (peaks at periods 20, 160 and 320).

PULSAR PSR2045-16 PGR 100:(9613 0 5 SEC FREQUENCY=45 MHZ
LOCK AVERAGES
1 3 I
P2 045 A3 A
MAY
18
7*
TSTART
09H
38 M
OOS
STOP
0 9 h
39 M
54 S
TSAMP
02
P?045H3A
MAY
1 8
79
T57ART
0 9H
4 0M
oos
STOP
0 9 H
4 4 M
5 9 S
TSAMP
02
P2045C3A
MA Y
1 8
79
TSTART
0 9H
45
OOS
S T C P
0 9H
48M
4 OS
TSAMP
02
2 04503 A
MA Y
1 8
79
TS TART
09H
49 M
oos
STOP
0 9H
* 4M
OOS
TSAMP
02
P2 045E3A
MA Y
1 b
79
T STAR T
09H
54M
0 1S
jTP
0 9 H
59 v
OOS
TSAMP
02
2*1 SIGMA= 0.057 3 SIGM4= 0.170 MAGNIFICATIONS 100
NUM3EP F QLOCKS= 5
AVfcRAGlNG TlMts 118P.8 SEC
BIN
AVE RAGE
Pt R I C DS
N'JMMCR
AVCRAGED
1
121.>57
b 1 2
2
1 31 .27 ft 10
3
1 31
ftl 0
4
131.324
6 1 0
0
1 Ji 32 6
o 1 0
f-
121 .263
fc 1 0
7
121.293
ft 10
8
111.478
5 33
)
1J1394
ft 1 0
1 c
131. 1 9 (i
ft 1 0
11
12 1 .27 7
Â£ C4
12
1 3 l ? b 3
ft 1 0
1 3
1 J1 J 34
ft 10
1 4
131.329
ft 10
15
13!.317
6 l 0
16
121.209
61 0
17
131.224
6 1 0
1 b
131 ? 7 5
ft 10
19
131.275
6 10
20
1 J1 1HH
c 1 0
21
1 31 .288
ft 10
22
121 .311
6 1 0
23
1 3 1 <24
ft 10
24
131.14 7
6 1 0
25
131 .18 6
t 10
26
131 39
6 1 0
27
121.270
6 1 0
23
131.230
ft 1 0
29
131.291
b 10
30
131.272
b 10
31
1 3 1 .T9
5 04
22
121.pe
6 1 C
3 3
131.337
6 1 0
34
131 .329
6 1 0
35
1 21 .293
b 1 0
3b
131.110
6 1 0
37
1 31 1^5
b 1 0
35
1 31 .13 C
1 0
29
131.281
ft 10
40
121.266
ft 1 0
4 l
1 ?l 1 i8
ft 10
42
1 31 3 ft 0
b 10
43
121 .289
61 0
44
131 I4
ft 1 0
45
131.402
ft 10
4 ft
1 31 28 8
ft 10
4 7
l 3 1 10 a
ft 1 0
48
1 J 1 44
6 1 C
49
121.20'
ft 10
50
131.294
6 1 0
51
131.216
bl 0
52
131.155
ft l 0
53
131.221
b l 0
54
131 .253
ft 1 0
55
l 3 1 9 1
ft 1 0
5b
131.301
ft 1 0
57
13 1 .244
6 10
58
131 .222
ft 1 0
59
121.216
6 1 C
60
131.32b
ft 1 0
6 1
131. J 4 5
5 04
62
12 1. ->64
575
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I i
I
* TIME
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
INTF.NSITY-
8 3.28 SIGMA
I
I
I
I
I
I
I
I
I

I
I
I
I
I
i
!
i
i
ro

MCAPGM Progr

18
The designation scheme for the blocks of data from
pulsar PSR2045-16 is slightly different, as follows:
P 2 0 4 5 X Y
Number that designates the night when
the pulsar was recorded
Letter that designates the block
within that night
Designates pulsar PSR2045-16
The block designation is usually one of the letters A,
B, C, D, E or F.
The nights were numbered as follows:
Night Date

May
15,
1979
1
May
16,
1979
2
May
17,
1979
3
May
18,
1979
4
May
19,
1979
5
May
22,
1979
6
May
23,
1979
7
May
24,
1979
8
May
29,
1979
9
May
30,
1979
The reduced data for each block was stored on a disk
called UF.BO030601.S5.XXXXX, where XXXXX is the name of the
block. The data have been recorded also on a digital magnetic
tape called PULSAR.

90
Table 3-8. Comparison of the mean apparent pulsar
periods for PSR2045-16 obtained from
observations with the DOPVEL periods.
Date
Mean apparent period
(s)
DOPVEL period
(s)
AP
(PS)
May
15,
1979
1.96137547
1.961383
-8
May
16,
1979
May
18,
1979
1.96135847
1.961386
-28
May
19,
1979
or 1.96140311
+ 14
It can be seen that the periods do not agree.
Unfortunately, a jump in the epoch was noticed in the quartz
time standard of Maipu, after the first pair of days; that
makes the figures for the last pair of days (May 18 and
May 19, 1979) very unreliable.
In any case, this discrepancy, at least for May 15 and
May 16, demonstrates that the DOPVEL period cannot be used
to bring in phase two consecutive days of observations.
In order to bring in phase pulses detected on two
consecutive days (within 0.5 bin), the apparent mean period
for each day needs to be known with an accuracy of at least
where N = Integer number of periods between two apparitions.
Expressed in terms of the elapsed time T and the
period P, (3-16) becomes:
A
c
?x-
2 T
(3-17)

Abstract of Thesis Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
OBSERVATION OF PULSARS AND OTHER SOURCES
By
Francisco Reyes
December 1980
Chairman: Thomas D. Carr
Major Department: Astronomy
The investigation of the radio emission spectra of
pulsars at low frequencies is important since it helps to
clarify the mechanism responsible for this emission. At the
present, the low frequency end of the spectra is poorly
determined compared with more precise measurements avail
able at higher frequencies. Also, measurements of pulsar
intensity variations at low frequencies due to scintilla
tion are of value in the study of the interstellar medium.
Detection of pulsars at decametric wavelengths is
difficult because of the deterioration of the signal-to-
noise ratio caused by the high galactic background tempera
tures and the turn-over of the spectra of most pulsars.
Also the dispersion of the pulses imposes the use of narrow
bandwidth, reducing the sensitivity of the system.
Vll

20
this time, the data is taken from disk (where it has been
previously stored), produces a magnified plot and punches
cards with the final values for each bin, the total block
average and the final standard deviation. This set of cards
is used later on, to stack 2 or more blocks.
Usually after the process already described, the digi
tized data is recorded on the PULSAR magnetic tape and
released from the data files.
In order to stack 2 or more blocks, the sets of cards
of each block (except the first, which is used as reference)
is shifted the proper number of bins so that the bins can be
averaged in phase. The number of bins to be shifted in each
block is given by the following relationship:
N
bins
(2-1)
where = Number of bins the block i needs to be shifted with
respect to block 1
T^ = Time start sampling block 1
T^ = Time start sampling block i
P = Pulsar period
P
R = Time resolution (= r r-. )
Number bins
Frac = Fractional part
T. -T
The fractional part of ^ corresponds to the fraction
of the period between the time of beginning of block i and
the time of beginning of the first block. The relationship
given by 2-1 was evaluated using an HP-29C calculator.

300
200
RELATIVE
INTENSITY
OF
AVERAGED
PULSE
(ARBITRARY
SCALE)
Ul
o
100
_L
X
X
JL
X
X
X
X
X
X
-800 -600 -400 -200 p +200
(ys) AVERAGING PERIOD
+ 400
+ 600
(us)
+ 8 00
Figure 3-13 Relative pulse intensity as a function of the averaging period for
pulsar PSR1133+16 on May 25, 1979. P = 1.188023 s.

My thanks goes to Hector Alvarez, Mauricio Bitrn and
Fernando Olmos of Maip Radio Observatory for their help
with the 45 MHz data.
I am grateful to the Astronomy Department and the
Physics Department and in particular to Drs. Heinrich
Eichhorn, Robert J. Leacock, F. Eugene Dunnam, Charles F.
Hooper and Richard E. Garrett for their financial support
in the form of teaching assistantships and fee waivers.
My gratitude goes to Mrs. Brenda Dobson for her
patience in correcting my English and for her devotion in
typing and editing the manuscript.
The realization of this project has been possible
thanks to the continuous support provided to the radio
astronomy program of the Department of Astronomy of the
University of Florida and to the Maip Radio Observatory
through a grant from National Science Foundation (principal
investigator Dr. Alex G. Smith).
IV

ELOCK AVERAGE- 121
NUM9CP CF BLOCKS *
PULSAR PSW 1 1 J 3 1 fe CRJOr)s 1 1 8 8 0 2 3 SEC F WE Q UCNC V= 26 3 MHZ
P112WB6 may 25 79 T ST AR T 01H 3AM JOS STOP 01H 3BM 04S TSAMP
800 SIGMAs 0* 105 3 SIGMA* 0.315 MAGNIFICATION*
1 AVERAGING TIME= 220.5 SEC
02
1 00
RI N
A VC c AGE
PFRIOOS
NJMQER
AV EuAGlO
I
121,792
24
1
2
121,916
1 90
1
3
121.895
1 90
1
4
121 .753
l 90
1
5
12 1 .637
1 SO
1
6
121.76?
1 90
1
7
121.337
1 90
1
3
121.763
l 90
1
9
1 2?.000
l to
1
10
12 ieoe
1 90
1 1
121 .92 t
1 90
J
12
12 1 .74 2
1 90
|
1 3
121.768
190
1
l 4
1 2 1 705
1 90
1
1 5
121.379
1 90
16
121 .642
l 90
l 7
121.621
1 90
1 3
121.842
1 90
1 0
l 2 1 .76 2
l 90
?C
121 .637
l 90
c 1
1 21 .737
1 90
2
121.589
1 90
23
121.595
1 90
24
121.926
1 90
25
1 22.042
l 90
26
121.874
1 50
27
121.863
1 90
23
121.305
1 90
2 i
121.747
1 90
30
l21.795
l 90
3 1
121.700
1 SC
3 ">
121.700
l 90
33
121 .763
1 90
24
121.637
1 90
35
1 21.3? 1
1 50
3b
121.668
1 90
37
12 1 .932
1 90
33
l 2l 726
1 90
TIME
INTENSITY
ro
CD
Figure 3-6.
Pulsar PSR1133+16 (Block P112WB6) recorded at 26.3 MHz on May 25,
1979. Average over 190 periods.

106
Table A-
-3. List of references from which the data at
the different frequencies was obtained.
Frequency
MHz
References
34
Backer (30)
53
Sieber (29)
61
Izvekova et al. (8)
80
Slee and Hill (9)
102
Izvekova et al. (7)
151
Lyne et al. (22)
408
Komesaroff et al. (10)
and Manchester et al. (38)