The observation and analysis of lunar occultations of stars with an emphasis on improvements to data acquisition instrum...

MISSING IMAGE

Material Information

Title:
The observation and analysis of lunar occultations of stars with an emphasis on improvements to data acquisition instrumentation and reduction techniques
Physical Description:
xxvi, 504 leaves : ill. ; 28 cm.
Language:
English
Creator:
Schneider, Glenn H, 1955-
Publication Date:

Subjects

Subjects / Keywords:
Occultations   ( lcsh )
Astronomical instruments   ( lcsh )
Astronomical photometry   ( lcsh )
Information storage and retrieval systems -- Astronomy   ( lcsh )
Imaging systems in astronomy   ( lcsh )
Astronomy thesis Ph.D
Dissertations, Academic -- Astronomy -- UF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1985.
Bibliography:
Includes bibliographical references (leaves 497-502).
Additional Physical Form:
Also available online.
Statement of Responsibility:
by Glenn H Schneider.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030280403
oclc - 16396414
System ID:
AA00025825:00001

Full Text










THE OBSERVATION AND ANALYSIS OF LUNAR OCCULTATIONS OF
STARS WITH AN EMPHASIS ON IMPROVE11ENTS TO DATA
ACQUISITION INSTRUMENTATION AND REDUCTION TECHNIQUES










By

GLENN H SCHNEIDER
















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA 1985


















This work is dedicated to

Rose Wenig

and

Al ice Schneider,

my late, beloved (grandmothers.

















If the stars should appear one night in a thousand
years, how would men believe and adore; and preserve for many generations the remembrance of the city of God which had been shown!

'Emerson', Nature; Addresses, and Lectures
















ACKNOWLEDGEMENTS



The debts accrued at the successful completion of any endeavor which spans a number of days requiring four digits to count are many and varied. This work is no exception. Without the input and influence of many people the task which was set before me could never have come to fruition. On two levels, both personal and professional, my thanks go out to all those who helped make this project possible. The task of acknowledging all those who contributed to this study is formidable, and indeed if properly done would fill volumes. To those whom I may fail to personally acknowledge here, I offer my apology, for there are uncountable numbers who have contributed either directly, or indirectly to this study.

To my parents, my grandfather Max Wenig, and

grandmothers, Rose Wenig and Alice Schneider, I owe a debt which I can never hope fully to repay. With their love and encouragement, I was raised in an atmosphere which nurtured my intellectual curiosity. I was never denied the opportunity to explore, and to grow in any direction my inquisitiveness had taken me. To them go my most profound gratitude, respect and love.

To my committee chairman, John P. Oliver, go my warmest thanks for first suggesting this course of investigation.


iv











Without his assistance and guidance throughout my years at the University of Florida, this dissertation would never have been. I also thank my other committee members, Heinrich K. Eichhorn, Howard L. Cohen, Haywood C. Smith, and Ralph G. Selfridge, for their help, valuable suggestions, and at times much needed criticisms as this study progressed. To the latter, and to the University's Center for Intelligent Machines and Robotics I offer additional thanks for securing the use of the computational resources required for this investigation. I would also like to thank Ben Adenbaum and Warner Computer Systems, Inc., for granting the use of their facilities during the early stages of this project.

My personal thanks extend to Frank B. Wood, and Kwan-Yu Chen, both of whom often provided much needed information related to this study. In addition, through their efforts I was afforded the unique opportunity to hone my fledgling skills as an instrumentalist by actively working on the development of the South Po!e Observatory.

My fellow graduate students proved to be a source of a

wealth of information and ideas. To them I offer not only my thanks but the hope that there may indeed be life after graduate school. In particular, I would like to acknowledge the support I received from Roger Ball, Joseph T. Pollock (who have already realized this hope), and Gregory L. Fitzgibbons. To Elaine Reeves, who soon will enter into indentured servitude as a graduate student herself, go my




v










thanks for proof reading the manuscript of this dissertation, and for helping lbag" Pallas.

For their fabrication of the mechanical components required (often on a moment's notice) in support of the occultation program, I am grateful for the work done by Harvey Nachtrieb and the machinists in the Department of Physics. Though still new to the Astronomy Department, Donald McNeil has been an outstanding asset during the latter stages of this project. I offer him my personal thanks. and a tip of the hat for his help with the 1984 annular eclipse, and for Ceres as well.

I would like to thank Marie Lukac of the United States Naval Observatory, David Dunham and the International Occultation Timing Association, and Robert Millis at Lowell Observatory for providing the occultation predictions used in planning the observing programs.

On a number of personal notes I would like to express my gratitude to Alan Nathans for his support and encouragement during my childhood. Also extended are my thanks to the members of the Amateur Observers' Society of New York. for helping to lay the foundation for my future work. To Susan Howard goes my deepest appreciation for her unfailing encouragement during my early years in Florida. I must also thank Laura E. Kay, who is at this very moment spending a long, cold winter at the South Pole, for her prodding and nudging me to finish this thing and get out of Florida.




vi










To William Tilichock (whereever he may be) I say thank you for introducing me to APL. one of the major tributaries of my life. And I thank to Kenneth E. Iverson for inventing what very well may be the most beautiful, aesthetically pleasing, and incomparably useful symbolic notations ever conceived.

Finally, I thank -from the bottom of my heart Karla

Rahman for her love and understanding during the 25-hour days and 32-day months while this dissertation was being prepared. Though to her I no say "I will"M, by the time this work is shelved I intend to say "[ do".




































vii
















TABLE OF CONTENTS

ACKNOWLEDGEMENTS .......................................... iv

LIST OF TABLES .......................................... xii

LIST OF FIGURES ......................................... xvii

ABSTRACT ................................................. xxv

CHAPTERS

I INTRODUCTION ........................................ I

Lunar Occultations: A Historical Synopsis .......... I
Information Which May Be Learned From the
Analysis of Lunar Occultations .................... 3
Goals of the Program of Occultation Observation .... 5

II INSTRUMENTATION ..................................... 8

Optical Equipment .................................. 8
The Seventy-Six Centimeter Telescope ............. 8
Location and description ...................... 8
The telescope light baffle .................... 8
Photoelectric Photometers ....................... IS
The Astromechanics photometer ................ IS
Optical filters .............................. 19
The portable photometer ...................... 20
Data Acquisition Electronics ...................... 20
The SPICA-IV/LODAS System ....................... 20
Design criteria for a new SPICA .............. 20
The SPICA-IV digital electronics ............. 22
Ancillary equipment .......................... 27
Power supplies ............................... 28
SPICA-IV system configuration ................ 29
Portable use ................................. 29
Analog Electronics .............................. 40
The WWVB receiver ............................ 40
Radio antennas ............................... 49
The photometer amplifier ..................... 50
Limitations of the Occultation Photometric
System ......... ... ; ............... 52
The Lunar Occultation Data u
c isi ion System
Software ......................................... 55
Software Design Considerations .................. 55
System timing ................................ 56
Memory usage ................................. 57


viii










Suplementary program documentation ........... 58
Peripheral Input/Output ......................... 62
User (Observer) I/0 .......................... 62
The video "strip chart recorder .............. 71
Data archival ................................ 73

III NUMERICAL MODELING OF LUNAR OCCULTATIONS ........... 76

The Physical Characterization of an Occultation
Intensity Curve .................................. 76
The Generation of a Model Occultation Intensity
Curve ...................................... ..... 78
Monochromatic Point Source Approximation ........ 78 Lunar Limb Effects .............................. 79
A Monochromatic Extended Source and Limb
Darkening ..... ................................. 81
Polychromatic Intensity Curve................... 86
The Effects of Discrete Modeling ................ 90
The Effects of Instrumental Optical Response .... 91
The Polychromatic Extended Source Intensity
Curve .......................................... 93
Models of Double Stars .......................... 94
The Differential Corrections (DC) Fitting
Procedure ........................................ 96
A Note on Time .................................. 96
Choosing Initial Parameters for the DC
Procedure ...................................... 97
Parametric Adjustment...........................100
The DC Fitting Procedure for Close Double
Stars ........................................ 103
Validation of the Fitting Procedures............106
Numerical Experiments to Improve the Fitting
Procedure ....................................... 109
Application of Partial Parametric Adjustments..113
Parametric Grouping into Computational
Subsets ....................................... 115
Uniqueness of the Solution ..................... 121
Smoothing of the Observational Data ............ 122
N-point unweighted smoothing ................ 123
N-point weighted, exponential smoothing......124
Smoothing by forward and inverse Fourier
transformtion .............................. 125

IV COMPUTATIONAL DATA REDUCTION PROCEDURES ........... 134

A Choice of Programming Languages: The APL
Decision ........................................ 134
Downloading and Uploading of Observational
Data ............................................ 136
The Occultation Reduction Workspace (OCCRED) ..... 138
Global Variables Used by OCCRED ................ 139
The Computational Differential Corrections
Procedure ..................................... 141



ix










The Two-Star Differential Corrections
Procedure (DC2) ................................ 154
Global parameters for DC2.................... 154
The APL function DC2.......................... 155
Preprocessing of the Observational Data .........158 Presentation of the Results of the DC Run .......158

V THE OCCULTATION OBSERVATIONS AND RESULTS OF
THEIR ANALYSIS..................................... 162

Presentation Format................................ 162
Format and Content of the Tables............... 162
The occultation summnary table................ 162
Variance-covariance, correlation, and
dI/dPi . . . . . . . . . 167
Observed and computed intensity values .......168
Format and Content of the Graphs................ 169
Graph of the entire event, RAWAPLOT ...........169
The integration plot! INTPLOT................ 170
Graphic depiction of the best fit, FITPLOT..171
Noise statistics of the observation,
NOI SEPLOT.................................... 172
Power spectra, POWERRPLOT.................... 174
Sensitivity of solution to variation of
parameters, PDPLOT........................... 176
Discussion of Individual Occultation Events .......177
ZC0916 (1 Geminorum)............................. 177
Historical notes.............................. 177
The observat ion............................... 179
Reduction and analysis of the
I Geminorum A observation................... 181
Reduction and analysis of the
I Geminorum Bobservation................... 186
A discussion of the 1 Geminorum results ......188
Concluding remarks on 1 Geminorum ......... 9
ZC1221 (9 Cancri)................................ 204
ZC1222.. . ........................................................ 215
X07589............................ ............... 223
X07598........................................... 235
X13534........................................... 245
X13607........................................... 256
ZC1462........................................... 265
X 18067........................................... 277
ZC2209 (32 Librae)............................... 285
ZC3214........................................... 296
X31590............................................ 307
ZC3458 (334B. Aquarii).......................... 316
X01217.......................................................... 319
X01246........................................... 329
ZC0126........................................... 337
X01309........................................... 343
ZC3158 (37 Capricorni)........................... 358
ZC0835........................................... 369
X07145........................................... 378


x










X07202 .. . . . . . . . . . 387
X07247............................................ 396
X09514............................................ 401
ZC1030 (Epsilon Geminorum)....................... 413
Summary of the Occultation Observations ...........416
Future Directions for the Occultation Proqram .... 424 APPEND ICES

A LODAS/EO7 ASSEMBLY LISTING.......................... 434

B OCCTRIS ASSEMBLY LISTING........................... 470

C LISTING OF THE APL WORKSPACE OCCPREP............... 475

D LISTING OF THE APL WORKSPACE OCCRED................ 477

E LISTING OF THE APL WORKSPACE OCCPLOTS.............. 489

LIST OF REFERENCES......................................... 497

B IOGRA~PHI CAL SKETCH........................................ 503




































xi

















LIST OF TABLES

2-1 SPECIFICATIONS FOR THE 76-CM. OPTICAL BAFFLE TUBE... .16 2-2 DETERMINATION OF THE PRIMARY MIRROR FOCAL LENGTH... .17 2-3 DIAPHRAGM DESIGNATIONS, LINEAR, AND ANGULAR FIELD
SIZES................................................. 19

2-4 DESIGN CRITERIA FOR SPICA-IV/LODAS................... 21

2-5 STELLAR INTENSITY READINGS WITH THE RHO
SPICA-IV/LODAS....................................... 53

2-6 LODAS/E07 SUBROUTINES................................. 63

2-7 LODAS/EO7 DATA TABLES..............................

2-8 LODAS/EO7 KEYBOARD COMMANDS........................... 66

3-1 PARAMETERS CHARACTERIZING AN OCCULTATION
INT~ENISITY( CURVE...................................... 76

3-2 PARAMETERS AFFECTING THE OBSERVED INTENSITY CUVE... 77 3-3 BRIGHTNESS DISTRIBUTION FOR A MODEL STAR WITH
GRID PARAMETER=-6..................................... 83

3-4 BRIGHTNESS DISTRIBUTION FOR A MODEL STAR WITH
GRID PARAMETER=-6 AND A LIMB DARKENING
COEFFICIN OF 1.0................................................. 84

3-5 PARAM'ETERS FOR GENERATING THE INITIAL MODEL.......... 98

3-6 COMPARATIVE SAMPLE OF DC FITTING TO SYNTHETIC
CURVE................................................ 116

3-7 COMPARATIVE SAMPLE OF DC2 FITTING TO SYNTHETIC
CURVE................................................ 118

4-1 GLOBAL VARIABLES CREATED BY THE APL FUNCTION
INPUT................................................ 140

4-2 GLOBAL VARIABLES RESIDENT IN THE OCCRED WORKSPACE..141 4-3 GLOBAL VARIABLES CREATED BY THE APL FUNCTION DC .... 153 4-4 INFORMATION PRESENTED BY THE APL FUNCTION OUTPUT ... 160


xii









4-5 TWO-STAR QUANTITIES PRESENTED BY THE APL FUNCTION

DC2 ............................................... 161

5-1 THE OCCULTATION OBSERVATION OF ZC0916 (I GEM) ...... 179 5-2 TWO-STAR SOLUTION FOR 1 GEM A ....................... 183

5-3 TWO-STAR SOLUTION FOR 1 GEM B ....................... 188

5-4 RELATIVE BRIGHTNESSES AND MAGNITUDES OF THE
INDIVIDUAL COMPONENTS OF 1 GEM ..................... 192

5-5 TIMES OF GEOMETRICAL OCCULTATIONS OF THE
COMPONENTS OF I GEM ............... ................... 195

5-6 ZC0916: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 201 5-7 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE 1 GEM A SOLUTION .............................. 202

5-8 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE 1 GEM B SOLUTION .............................. 203

5-9 THE OCCULTATION SUMMARY TABLE FOR ZC1221 ........... 205

5-10 ZC1221: OBSERVED, COMPUTED AND RESIDUAL VALUES ..... 212

5-11 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE ZC1221 SOLUTION ............................... 213

5-12 THE OCCULTATION SLMMARY TABLE FOR ZC1222 ........... 216

5-13 ZC1222: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 220 5-14 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE ZC1222 SOLUTION ............................... 221

5-15 THE OCCULTATION SUMMARY TABLE FOR X07589 ........... 225

5-16 X07589: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 232 5-17 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X07589 SOLUTION ............................... 233

5-18 THE OCCULTATION SUMMARY TABLE FOR X07598 ........... 236

5-19 X07598: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 240 5-20 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X07598 SOLUTION ............................... 242

5-21 THE OCCULTATION SUMMARY TABLE FOR X13534 ........... 246

5-22 X13534: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 252 xiii










5-23 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR

THE X13534 SOLUTION ............................... 254

5-24 THE OCCULTATION SUMMARY TABLE FOR X13607 ........... 257

5-25 X13607: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 262 5-26 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X13607 SOLUTION ............................... 263

5-27 THE OCCULTATION SUMMARY TABLE FOR ZC1462 ........... 270

5-28 ZC1462: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 274

5-29 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE ZC1462 SOLUTION ............................... 275

5-30 THE OCCULTATION SUMMARY TABLE FOR X18067 ........... 279

5-31 X18067: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 284

5-32 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X18067 SOLUTION ............................... 287

5-33 THE OCCULTATION SUMMARY TABLE FOR ZC2209 ........... 289

5-34 ZC2209: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 297 5-35 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE ZC2209 SOLUTION ............................... 298

5-36 THE OCCULTATION SUMMARY TABLE FOR ZC3214 ........... 299

5-37 ZC3214: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 304 5-38 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE ZC3214 SOLUTION ............................... 305

5-39 THE OCCULTATION SLMiMARY TABLE FOR X31590 ........... 308

5-40 X31590: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 311 5-41 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X31590 SOLUTION ............................... 314

5-42 THE OCCULTATION SUMMARY TABLE FOR X01217 ........... 320

5-43 X01217: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 323 5-44 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X01217 SOLUTION ............................... 325

5-45 THE OCCULTATION SUMMARY TABLE FOR X01246 ........... 330


xiv












5-46 X01246: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 333 5-47 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X01246 SOLUTION ............................... 335

5-48 THE OCCULTATION SUMMARY TABLE FOR ZC0126 ........... 340

5-49 ZC0126: OBSERVED, COMPUTED, AND RESIDUAL VALUES
RESI DUALS ......................................... 345

5-50 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE ZC0126 SOLUTION ............................... 347

5-51 THE OCCULTATION SUMMARY TABLE FOR X01309 ........... 351

5-52 X01309: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 354 5-53 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE XO1309 SOLUTION ............................... 356

5-54 THE OCCULTATION SUMMARY TABLE FOR ZC3158 ........... 359

5-55 ZC3158: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 364

5-56 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE ZC3158 SOLUTION ............................... 366

5-57 THE OCCULTATION SUMMARY TABLE FOR ZC0835 ........... 370

5-58 ZC0835: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 372 5-59 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE ZC0835 SOLUTION ............................... 376

5-60 THE OCCULTATION SUMMARY TABLE FOR X07145 ........... 380

5-61 X07145: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 381 5-62 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X07145 SOLUTION ............................... 384

5-63 THE OCCULTATION SUMMARY TABLE FOR X07202 ........... 388

5-64 X07202: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 393 5-65 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X07202 SOLUTION ............................... 394

5-66 THE OCCULTATION SUMIMARY TABLE FOR X07247 ........... 397

5-67 X07247: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 400




xv










5-68 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X07247 SOLUTION................................. 403

5-69 THE OCCULTATION SUMMARY TABLE FOR X09514............ 407

5-710 X09514: OBSERVED, COMPUTED, AND RESIDUAL VALUES .... 409 5-71 VARIANCE-COVARIANCE AND CORRELATION MATRICES FOR
THE X95147 SOLUrTION'................................. 411

5-72 DERIVED QUANTITIES FOR THE OCCULTATION BINARIES .... 418 5-73 STELLAR ANGULAR DIAMETERS............................ 418

5-74 COORDINATED UNIVERSAL TIMES OF GEOMETRICAL
OCCULTATION'S.. . . .. . ........................................ 419










































xvi
















LIST OF FIGURES

2-1 The Rosemary Hill Observatory 76-cm. telescope ..... 10 2-2 Layout of the occultation baffle system ............ 12

2-3 The occultation light-baffle tube and ring ......... 15

2-4 Schematic diagram of SPICA-IV/LODAS clock .......... 26

2-5 Schematic diagram of A-to-D circuit ................ 26

2-6 The SPICA-IV/LODAS 8-inch disk drive package ....... 31 2-7 SPICA-IV/LODAS system configuration ................ 32

2-8 The A-to-D Converter/Clock board ................... 34

2-9 SPICA-IV/LODAS backplate and connector layout ...... 36 2-10 The SPICA-IV/LODAS rolling rack and photometric
equipment ......................................... 38

2-11 Photograph of SPICA-IV/LODAS at Macon, Georgia ..... 42 2-12 Photograph of SPICA-IV/LODAS in the Everglades ..... 44 2-13 The WWVB and WWV receivers, and HV power supply .... 48 2-14 LODAS foreground program logic flow chart .......... 59

2-15 LODAS background program logic flow chart .......... 60

2-16 LODAS 20-character LED status display format ....... 71 2-17 The SPICA-IV/LODAS video "strip chart recorder'....75 3-1 Lunar occultation intensity curve for a
monochromatic point source ........................ 80

3-2 An example of a stellar quadrant grid .............. 80

3-3 Stellar limb darkening ............................. 85

3-4 Modeling an extended source from intensity
weighted monochromatic point sources .............. 87




xvii










3-5 Modeling a polychromatic source from intensity
weighted monochromatic extended sources ........... 89

3-6 The effect of discrete modeling on a 10millisecond-of-arc source ......................... 92
3-7 Sample two-star intensity curve used for
initial parameter selection ....................... 105

3-8 Iterative convergence of the DC fitting process...112 3-9 Iterative convergence of the DC2 fitting process..120 3-10 Eight-point unweighted smoothing of raw data ...... 126 3-11 Power spectra of selected synthetic curves ........ 129 3-12 Windowing effects of Fourier smoothing ............ 131

5-1 ZC0916 (I Gem) RAWPLOT .......................... 180

5-2 Detail of the 1 Gem double disappearance .......... 180

5-3 ZC0916 INTPLOT .................................. 182

5-4 ZC0916 INTPLOT detailing the double
disappearance .................................... 182

5-5 ZC0916-A FITPLOT ................................ 185

5-6 ZC0916-B FITPLOT ................................ 185

5-7 ZC0916-A PDPLOT ................................. 187

5-8 ZC0916-B PDPLOT ................................. 187

5-9 ZC0916 NOISEPLOT ................................ 189

5-10 ZC0916 POWERPLOT ................................ 190

5-11 The projected geometry of the ZC0916 stellar
components ....................................... 194

5-12 The P and 0 residuals for the visual
components of ZC0916 ............................. 198

5-13 The four-star component solution for ZC0916 ....... 200 5-14 ZC1221 (9 Cnc) RAWPLOT .......................... 206

5-15 ZC1221 INTPLOT .................................. 207

5-16 ZC1221 INTPLOT of the secondary event ........... 207



xviii










5-17 ZC1221 FITPLOT .................................. 208

5-18 ZC1221 PDPLOT ................................... 210

5-19 ZC1221 NOISEPLOT ................................ 210

5-20 ZC1221 POWERPLOT ................................ 214

5-21 ZC1222 RAWPLOT .................................. 217

5-22 ZC1222 INTPLOT .................................. 217

5-23 ZC1222 FITPLOT .................................. 219

5-24 ZC1222 PDPLOT ................................... 222

5-25 ZC1222 NOISEPLOT ................................ 222

5-26 ZC1222 POWERPLOT ................................ 224

5-27 X07589 RAWPLOT .................................. 226

5-28 X07589 INTPLOT .................................. 226

5-29 X07589 FITPLOT .................................. 229

5-30 X07589 Detailed FITPLOT ......................... 229

5-31 X07589 PDPLOT ................................... 231

5-32 X07589 NOISEPLOT ................................ 231

5-33 X07589 POWERPLOT ................................ 234

5-34 X07598 RAWPLOT .................................. 237

5-35 X07598 INTPLOT .................................. 237

5-36 X07598 FITPLOT .................................. 239

5-37 X07598 PDPLOT ................................... 243

5-38 X07598 NOISEPLOT ................................ 243

5-39 X07598 POWERPLOT ................................ 244

5-40 X13534 RAWPLOT .................................. 248

5-41 X13534 INTPLOT .................................. 248

5-42 X13534 FITPLOT .................................. 250

5-43 X13534 PDPLOT ................................... 253



xix









5-44 X13534 NOISEPLOT ................................ 253

5-45 X13534 POWERPLOT ................................ 255

5-46 X13607 RAWPLOT .................................. 259

5-47 X13607 INTPLOT .................................. 259

5-48 X13607 FITPLOT .................................. 260

5-49 X13607 PDPLOT ................................... 264

5-50 X13607 NOISEPLOT ................................ 264

5-51 X13607 POWERPLOT ................................ 266

5-52 ZC1462 RAWPLOT .................................. 267

5-53 ZC1462 INTPLOT .................................. 267

5-54 ZC1462 FITPLOT .................................. 269

5-55 ZC1462 FITPLOT of the occultation with Fourier
smoothed data .................................... 269

5-56 ZC1462 POWERPLOT ................................ 271

5-57 ZC1462 PDPLOT ................................... 276

5-58 ZC1462 NOISEPLOT ................................ 276

5-59 X18607 RAWPLOT .................................. 280

5-60 X18607 INTPLOT .................................. 280

5-61 X18067 Detailed INTPLOT with 5-point smoothing..281 5-62 X18067 FITPLOT .................................. 283

5-63 X18067 PDPLOT ................................... 286

5-64 X18067 NOISEPLOT ................................ 286

5-65 X18067 POWERPLOT ................................ 288

5-66 ZC2209 (32 Lib) RAWPLOT ......................... 291

5-67 ZC2209 INTPLOT .................................. 291

5-68 ZC2209 FITPLOT .................................. 292

5-69 ZC2209 PDPLOT ................................... 294

5-70 ZC2209 NOISEPLOT ................................ 294


xx









5-71 ZC2209 POWERPLOT ................................ 295

5-72 ZC3214 RAWPLOT .................................. 300

5-73 ZC3214 INTPLOT .................................. 300

5-74 ZC3214 FITPLOT .................................. 301

5-75 ZC3214 PDPLOT ................................... 303

5-76 ZC3214 NOISEPLOT ................................ 303

5-77 ZC3214 POWERPLOT ................................ 306

5-78 X31590 RAWPLOT .................................. 309

5-79 X31590 INTPLOT .................................. 309

5-80 X31590 FITPLOT .................................. 310

5-81 X31590 PDPLOT ................................... 313

5-82 X31590 NOISEPLOT ................................ 313

5-83 X31590 POWERPLOT ................................. 315

5-84 ZC3458 (334 B. Aquarii) RAWPLOT ................. 318

5-85 ZC3458 INTPLOT .................................. 318

5-86 X01217 RAWPLOT .................................. 321

5-87 X01217 INTPLOT .................................. 321

5-88 X01217 FITPLOT .................................. 322

5-89 X01217 PDPLOT ................................... 327

5-90 X01217 NOISEPLOT ................................ 327

5-91 X01217 POWERPLOT ................................ 328

5-92 X01246 RAWPLOT .................................. 331

5-93 X01246 INTPLOT .................................. 331

5-94 X01246 FITPLOT .................................. 332

5-95 X01246 Detailed FITPLOT with 5-point smoothing..332 5-96 X01246 PDPLOT ................................... 336

5-97 X01246 NOISEPLOT ................................ 336



xxi










5-98 X01246 POWERPLOT ................................ 338

5-99 ZC0126 RAWPLOT .................................. 341

5-100 ZC0126 INTPLOT .................................. 342

5-101 ZC0126 INTPLOT of the secondary event ........... 342

5-102 ZC0126 FITPLOT .................................. 344

5-103 ZC0126 Detailed FITPLOT with 5-point smoothing..344 5-104 ZC0126 PDPLOT ................................... 348

5-105 ZC0126 NOISEPLOT ................................ 348

5-106 ZC0126 POWERPLOT ................................ 349

5-107 X01309 RAWPLOT .................................. 352

5-108 X01309 INTPLOT .................................. 352

5-109 X01309 FITPLOT .................................. 353

5-110 X01309 PDPLOT ................................... 355

5-111 X01309 NOISEPLOT ................................ 355

5-112 X01309 POWERPLOT ................................ 357

5-113 ZC3158 (37 Cap) RAWPLOT ......................... 360

5-114 ZC3158 INTPLOT .................................. 360

5-115 ZC3158 FITPLOT .................................. 362

5-116 ZC3158 PDPLOT ................................... 363

5-117 ZC3158 NOISEPLOT ................................ 363

5-118 ZC3158 POWERPLOT ................................ 368

5-119 ZC0835 RAWPLOT .................................. 371

5-120 ZC0835 INTPLOT .................................. 371

5-121 ZC0835 FITPLOT .................................. 374

5-122 ZC0835 PDPLOT ................................... 375

5-123 ZC0835 NOISEPLOT ................................ 375

5-124 ZC0835 POWERPLOT ................................ 377



xxii









5-125 X01745 RdWPLOT .................................. 379

5-126 X01745 INTPLOT .................................. 379

5-127 X07145 FITPLOT .................................. 383

5-128 X07145 PDPLOT ................................... 385

5-129 X07145 NOISEPLOT ................................ 385

5-130 X07145 POWERPLOT ................................ 386

5-131 X07202 RAWPLOT .................................. 389

5-132 X07202 INTPLOT .................................. 389

5-133 X07202 FITPLOT .................................. 390

5-134 X07202 Detailed FITPLOT with 5-point smoothinq..390 5-135 X07202 PDPLOT ................................... 392

5-136 X07202 NOISEPLOT ................................ 392

5-137 X07202 POWERPLOT ................................ 395

5-138 X07247 RAWaPLOT .................................. 398

5-139 X07247 INTPLOT .................................. 398

5-140 X07247 FITPLOT .................................. 399

5-141 X07247 PDPLOT ................................... 402

5-142 X07247 NOISEPLOT ................................ 402

5-143 X07247 POWERPLOT ................................ 404

5-144 X09514 RAWPLOT .................................. 406

5-145 X09514 INTPLOT .................................. 406

5-146 X09514 FITPLOT .................................. 408

5-147 X09514 PDPLOT ................................... 410

5-148 X09514 NOISEPLOT ................................ 410

5-149 X09514 POtJERPLOT ................................ 412

5-150 ZC1030 (Epsilon Gem) .............................. 415

5-151 ZC1030 INTPLOT .................................. 415



xxiii










5-152 Distribution function of observed lunar limb
sl1opes.............................................. 423


























































xxi v
















Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




THE OBSERVATION AND ANALYSIS OF LUNAR OCCULTATIONS OF
STARS WITH AN EMPHASIS ON IMPROVEMENTS TO DATA
ACQUISITION INSTRUMENTATION AND REDUCTION TECHNIQUES By

Glenn H Schneider

August 1985

Chairman: John P. Oliver
Major Department: Astronomy

A program of observation and analysis of lunar

occultations was conceived, developed, and carried out using the facilities of the Universit/ ji Florida's Rosemary Hill Observatory (RHO). The successful implementation of the program required investigation into several related areas. First, after an upgrade to the RHO 76-cm. reflecting telescope, a microprocessor controlled fast photoelectric data acquisition system was designed and built for the occultation data acquisition task. Second, the currently available model-fitting techniques used in the analysis of occultation observations were evaluated. A number of numerical experiments on synthesized and observational data were carried out to improve the performance of the numerical techniques. Among the numerical methods investigated were



xxv










solution schemes employing partial parametric ad~iustrnent, parametric grou~ping into computational subsets (randomly and on the basis the correlation coefficients), and reprocessing of the observational data by a number of smoothina techniques for a variety of noise conditions. Third. a turn-key computational software system, incorporating data transfer. reduction, graphics and disiplay, was developed to carry out all the necessary and related computational tasks in an interactive environment.

Twenty-four occultation observations were obtained during the period March 1983 to March 1984. The observational data and the solutions resulting from the subsequent reductions are presented graphically and tabularly for each of the occultation events. Several angular diameter determinations were made. Among those of particular interest were 32 Librae (12.1, +/- 1.9 milliseconds of arc).

1 Geminorum-Bi (5.9, +/- 0.8 milliseconds of arc), and X07598 (5.5, +/- 2.0 milliseconds of arc). The visual/spectroscopic binary 1 Geminorum was discovered to have a fourth, previously undetected, component. Two other stars, X13534 and X13-607, were found to be binary with compainions closer than 15 milliseconds of arc. Previously unknowjn faint companions were discovered for ZC1221 and ZC0126. Times of geometrical occultation for all the events (including the secondary components of the binary systems) were reported as part of a cooperative astrometric project to the International Lunar Occultation Center.



xxvi
















CHAPTER I
INTRODUCTION


Lunar Occultations: A Historical Synopsis

When the moon, as a result of its orbital motion, moves in front of a star as viewed by an Earth-based observer, the light from the star is obscured. Such an event is known as a lunar occultation. At the time of the star's disappearance (or reappearance) the moon's limb is seen to move across the stellar disc. MacMahon (1909) proposed that the angular diameter of the occulted star could be determined by measuring the time interval of lunar limb passage across the star. Immediately thereafter, Eddington (1909) correctly pointed out that this was not possible, as the star's light would be diffracted by the moon's limb, hence the problem could not be approached from the standpoint of simple geometrical optics. He noted that the time-scale of the disappearance phenomenon is essentially unchanged as a function of stellar diameter.

Williams (1939) showed that while the time-scale of the event is not affected by the angular diameter of the star the diffraction intensity as a function of time most certainly is. The time-scale of variation of the diffraction fringes projected on the Earth's surface resulting from a lunar occultation is on the order of tens of milliseconds, thus

I







2

requiring fast photometric observations. Whitford (1939). reported on observations of occultations of Nu Aquarii and Beta Capricorni using a photocell with a Cesium-Oxygen-Silver cathode on the 100-inch telescope. Neither of these events showed any deviation in the diffraction pattern from a point source, as his instrumental detection limit was approximately 5 milliseconds-of-arc. Yet the foundation for a powerful new technique for the acquisition of fundamental astronomical information, i.e. stellar diameters, was laid.

Over the ensuing three decades additional photoelectric occultation observations were carried out. The first angular diameter measurement was reported by Evans (1951) for the star Antares. This was followed, also by Evans (1959), with the determination of the angular diameter of Mu Geminorum. Over the next two decades other observations had been made, and additional theoretical work on the extraction of information from lunar occultation observations progressed. It was not until the advent of electronic computers and reliable fast photometric equipment that occultation observations could truly begin to be exploited. In a now classic series of papers by Nather and Evans (1970), Nather (1970), Evans (1970 and 1971), and Nather and McCants (1970) the theoretical and observational aspects of lunar occultations are discussed in detail. The methods reported in these papers serve as the foundation for modern investigation of lunar occultations.







3

Only in the last few years, with the revolution in both microcomputer and opto-electronic technology, have the tools essential to bringing the observation and analysis of lunar occultations come to fruition. The problems are still many, but the instrumental hurdle, at least, may now be cleared.



Information Which May Be Learned From the Analysis of Lunar Occultations

The analysis of the lunar occultation intensity curve, obtained from a fast photoelectric record of an occultation event can yield, in principle, a wealth of information. The degree to which any observation can be exploited depends upon a large number of variables. The geometry of the relative position of the moon and the star, the quality of the sky during the observation (seeing and scintillation), the physical nature of the source, and the response of the instrumental system, to mention only a few, can help or hinder the discovery of information hidden in the intensity curve.

In the case of an occultation of a single star the

angular diameter of the star can be determined. Coupled with either a knowledge of the stellar parallax, or the V and R stellar flux (Barnes and Evans, 197-6) this angular measurement can be transformed into an actual linear diameter. The observational techniques for direct measurement of stellar diameters are severely limited. Speckle interferometry (Lohman and Weigelt, 1980), Phase Correlation Interferometry (Brown, 1968), and Michelson







4

Interferometry (Brown, 1980), are the only other currently available techniques. All of these are restricted. instrumentally, to the measurement of diameters of very bright stars.

The size of extended non-stellar sources such as the emission shells of Be stars can be determined (White and Slettebak, 1980) as was investigated in the case of Zeta Tauri (Schmidtke and Africano, 1984). In ideal cases, with multiple observations the brightness distributions of such sources could be found.

The angular optical resolution achievable through occultation observations, on the order of milliseconds-of-arc, often leads to the discovery of stellar duplicity of stars previously thought to be single. An accurate determination of the separation of the components in a binary (Evans, 1971) or multiple system (Evans et al., 1977) can be found from simultaneous observations from more than one site (or a projected separation from a single observation). The individual brightnesses, of otherwise MunresolvedO binaries, or multiple systems, can be determined.

The field of chronometry is dependent upon, and enhanced enormously by, the precise measurement of the times of occultation events. These event timings lead to an accurate determination of the moon's longitude from which Ephemeris Time is derived. Time intervals are easily obtained with a precision of one part in 10 trillion through the use of







5

atomic clocks. The observation of dynamical phenomena must have a zero point reference to couple dynamical events to atomic time (Van Flandern, 1974). Hence the accumulation of a database of accurate timing measurements of occultation events (dynamical phenomena) is essential in helping provide corrections to Co-ordinated Universal Time (C. U. T.).

The determination of stellar positions, the fundamental concern of the field of astrometry can benefit from occultation observations. The predicted time of an occultation event for a given topocentric location depends upon a knowledge of the moon's actual longitude, the geometry of the point of contact of the lunar limb, and the position in a given co-ordinate system of the star under study. Prediction errors in the times of occultation can lead to an improved knowledge of the positions of stars (Van Flandern, 1975). In some cases, gross errors in the predicted times of occultation can be attributed to positional errors in the stellar catalogs.



Goals of the Program of Occultation Observation

The primary intent of this program was to concentrate the investigation in areas which could have lead to astrophysically significant, or interesting results. While it could not have been determined, a priori, which stars or stellar systems would Yield the most useful information, a judicious choice of program objects was called for. Many observational selection effects and instrumental limitations








6

dictated the types of stars for which the occultation method would be fruitful. The general nature of these constraints has been addressed by Taylor (1966) and Ridgway (1977). Hence, in order to be considered for the observing program, candidate stars had to meet a variety of selection criteria. Unfortunately this left only a handful of stars each year to be considered for this investigation.

As a result of the stringent limitations a great deal of attention was paid to improving observational techniques and the development of new microelectronic data acquisition instrumentation. Once implemented, improved instrumentation allowed relaxation, to some degree, of the observational constraints placed on candidate star selection, thereby increasing the number of stars which were available for study. In addition, several non-traditional methods of data reduction were tried, which in some cases proved to be advantageous over the more conventional procedures.

The primary emphasis of the program was to extract as much information as possible from lunar occultation observations. The determination of stellar angular diameters, the discovery of unsuspected stellar duplicity or multiplicity, and the elucidation of the parameters of double or multiple systems were paramount. Of secondary importance, but an item to which no concessions were made during evolution of the program, was the accurate measurement of the times of occultation events. These measurements were reported, as part of an international co-operative project.







7

to the United States Naval Observatory, and the International Lunar Occultation Center.

An outgrowth of the lunar occultation program has been the observation and analysis of occultations of stars by asteroids. Such observations typically yield information about the size and shape of the occulting body, as well as better astrometric positions of the occulted stars. Over two thousand asteroids are known and have been catalogued (Bender, 1979), and several hundred have orbits determined with a sufficient degree of precision to allow topocentric predictions of asteroidal occultations to be carried out well in advance of the anticipated event. Hence, the third aspect of this investigation of high speed occultation photometry extended the observational domain to asteroidal, as well as lunar occultations.















CHAPTER II
INSTRUMENTATION


Optical Equipment

The Seventy-Six Centimeter Telescope

Location and description. All observations, unless otherwise noted, were carried out at the University of Florida's Rosemary Hill Observatory (RHO) in Bronson, Florida, utilizing the seventy-six centimeter Tinsley reflecting telescope. This telescope is located at Latitude of 29 degrees 23 minutes 59.4 seconds North, Longitude 82 degrees 35 minutes 11.0 seconds West, at an altitude of 44 meters above mean sea level, based on the 1927 North American Datum. The primary mirror has a focal ratio nominally of f/4. In its Cassegrain configuration the telescope provides a nominal focal ratio of 4116. Figure 2-1 shows the seventy-six centimeter telescope and the associated optical and electronic equipment employed for the observations of lunar occultations.

The telescope liQht baffle. Observing immediately

adjacent to the bright lunar disk increases the background sky illumination considerably. For a given spectral region there is nothing which can be done to enhance the signal to noise ratio considering noise due to atmospherically scattered light. It was found, however, that the Cassegrain

8



























Figure 2-1. The seventy-six centimeter Tinsley relecting telescope at the University of Florida's Rosemary Hill Observatory.







10










optical baffling for the seventy-six centimeter reflector could be improved upon greatly. A redesign of the optical baffle permitted a reduction in light scattered by the telescope optics and supporting structure. In addition.. a tighter baffle system enabled the rejection of off-axis rays. preventing them from reaching the focal plane.

The primary design considerations for a new baffle

system were two-fold. First, any baffles and/or field stops would have to be easily removable, so as not to impact on the telescope configuration required by concurrently running observing programs. Second, the unvignetted field of view at the focal plane had to marginally exceed the field obtained with the available diaphragm while minimizing extraneous light. An Astromechanics dual channel photometer was the primary instrument to be used for observing lunar occultations. This instrument has a maximum diaphragm opening corresponding to a field of view with a diameter of

thirty two seconds of arc.

Many designs were considered by constructing models of

the optical path, and ray tracing paraxial and marginal rays. It became quite obvious early in this investigation that no single tube/field stop design would be satisfactory for the task. The new baffle system would have to consist of two major elements: a tube with concentric annular field stops (rings)t and an annular ring placed around the Cassegrain secondary mirror. A schematic representation of the optical path is presented in Figure 2-2.










12

















0



L L





4.







0 M

-W


64



LD 0 CL.
40 40 L id



4-.








.-x

'U








13

To satisfy the first design objective a system of field stops was built into a baffle tube identical at the mounting base to the existing tube. This would allow easy change-over to either the new or old baffle tube. The Cassegrain secondary ring would be be installed on the already existing locking pins on the secondary mirror baffle, which serve to hold the mirror cover in place. The new baffle tube, with the inner concentric rings removed to show its construction, is shown in Figure 2-3 along with the secondary ring.

Since the position of the focal plane varies with

respect to the fixed position of the telescope superstructure and baffle assembly as a function of temperature, some degree of tolerance had to be allowed in the placement and size of field stops in the baffle tube. The system as designed would allow a one minute of arc unvignetted field of view at the focal plane at 15 degrees Celsius. This rather liberal tolerance was felt prudent considering the wide variation of climatic conditions experienced in northern Florida. Table 2-1 gives the specifications for the spacing and sizes of the inner ring stack. The secondary baffle ring has an outside diameter of 11.375 inches.

To actually perform the ray tracing, the optical path of the telescope had to be determined. Although the blueprints from Tinsley Laboratories claim the primary is an f/4 paraboloid this had to be tested as the engineering specifications did not necessarily represent the state of the completed telescope.




























































0





u
w
0
do






N
40
L CD
LL










15 77







16

TABLE 2-1
SPECIFICATIONS FOR THE OPTICAL BAFFLE TUBE


Ring Hole Diameter Ring Hole Diameter ---- ------------- ---- ------------1 3.56 5 2.24
2 3.23 6 1.91
3 2.90 7 1.67
4 2.57 8 1.24

Outside diameter or ring 1: 5.000 Outside diameter of rings 2-7: 4.812 Thickness of rings 1-8: 0.0625
Spacing between rings: 5.938
Bevel angle for all 8 holes: 45 degrees

Note: All linear measurements are in inches.


Using photographic plates of regions routinely monitored by the quasar research program, the plate scale of the primary mirror was found to be 67.065 (S.E. 0.06) seconds of arc per millimeter, corresponding to a focal length of 117.505 (S.E. 0.11) inches. The scale was determined by Pollock (1980) by measuring the positions of the quasar 0922+14 (denoted "GO) and the star SAO 098559 (denoted 'SO) for six plates taken over a period of seven years. Table 2-2 shows the plate measures used in determining the true scale of the primary mirror. The distance to the focal plane from the front surface of the Cassegrain secondary was measured at 15 degrees Celsius. From this the focal length of the Cassegrain system was found to be 1205 centimeters.

The assembled baffle tube is inserted through the

central hole in the primary mirror and is screwed into place. The old, wide-field baffle tube is removed by grabbing the outside end and twisting sharply counter-clockwise. This







17

will loosen its seating. When unscrewing the old baffle tube

one hand is kept below the lower end of the tube to prevent

it from hitting the primary mirror when it is completely

released. When screwing the occultation baffle tube into

TABLE 2-2
DETERMINATION OF THE PRIMARY MIRROR FOCAL LENGTH


PLATE DATE S to 0 0 to S MEAN DIFF SCALE
(Mm) (mm) (mm) '/mm
---------- ------ ------ --------- ----4/2-3/72 9.212 9.215
12.988 12.987 9.208 9.217
12.982 12.991 3.774 67.01

1/30-31/73 10.024 10.032
13.792 13.793 10.026 10.028
13.796 13.796 3.767 67.14

2/27-28/76 7.940 7.988
11.740 11.758 7.979 7.985
11.750 11.753 3.769 67.10

2/27-28/76 2.782 2.791
6.555 6.567 2.788 2.790
6.560 6.567 3.774 67.01

2/13-14/77 3.995 4.000
7.775 7.776
4.000 4.005
7.769 7.778 3.774 67.01

1/28-29/79 19.972 18.972
22.749 22.736 18.971 18.975
22.735 22.742 3.768 67.12

MEAN 67.065
(0.06)








18

position, care must also be taken to assure the tube is not being cross-threaded. To switch back to the old tube (which must be done if using the infra-red photometer available at RHO) the process is reversed. The Cassegrain secondary baffle ring slips over the end of the secondary containment cylinder after it has been aligned with the cylinder's three positioning pins. A small rotation will secure the ring position insets against the pins. Photoelectric Photometers

The Astromechanics Photometer. Unless otherwise stated, the photometer used throughout this investigation was a dual channel instrument manufactured by Astromechanics. This instrument splits the light path into two beams by means of dichroic filters so two wavelengths can be monitored simultaneously. Though the instrument can be used in dual channel mode, observations of lunar occultations obtained thus far at RHO have been observed only in one color. The photometer employs two dry ice cooled photomultiplier tubes (PtIT's). The PMIT in channel-i is a 6256S. while channel-2 uses a red sensitive 9684. In most cases 1200 Volts DC was applied to the PMT used. This photometer has a number of different available diaphragms. The linear diameters and angular field sizes of these diaphragms, along with their respective letter designations are listed in Table 2-3.







1?9

TABLE 2-3

DIAPHRAGM1 DESIGNATIONS, LINEAR AND ANGULAR FIELD SIZES
------------------Letter Designation Diameter (mm) Field (arc-secs.) -------------------- ------------- ----------------G 1.98 32.5
H 1.52 25.0
1 0.93 15.2
J 0.51 8.3


Optical -filters. Occultation observations made with the Astromechanics photometer employed Johnson V and B filters, as well as intermediate bandwidth yellow and blue interference filters. One inch diameter interference filters were obtained from Pomfret Research Optics, and are designated '* and *b respectively. The *yt filter, Pomfret part number 20-5400-1, has a peak spectral transmission at 5400 Angstroms and a Full Width at Half Maximum (FWHt1) of 100 Angstroms. The "b filter, Pomfret part number 20-4700-1 has a peak spectral transmission at 4700 Angstroms and a FWHM also of 100 Angstroms. These filters were selected in spectral regions for which M and K stars are relatively free of major absorption features. Of course, late type stars are riddled with a myriad of spectral lines. Hence the choice of these particular filters was somewhat of a compromise. Spectra typical G, K, and M, stars presented by Keenan and McNeil (1976) were examined, and on average were found least plagued with absorption lines at wavelengths of 4716 and 5408 Angstroms. These would have been the ideal central wavelengths for selected filters, but the cost of custom made filters was prohibitive. The filters procured were selected







20

to be as close to these wavelengths as possible from a stock list. The H-Beta line at 4861 Angstroms is outside of the "b" filter passband. While other lines are found at 4716 and 4670 Angstroms, the integrated passband is less subject to absorption losses than adjacent regions. The TiO band at 5448 Angstroms enters into the "y* filter passband, but it is centered close to the lower half-power point.

While an actual set of narrow band filters was not

available, a digital spectrum scanner employing Ebert-Fastie optics (Parise, 1978) is part of the standard equipment at RHO. The scanner can be used in a non-scanning mode as a variable-passband tunable filter. This in fact was done with great success in observing the occultation of Zeta Tauri in the passband of its H-Alpha emission.

The portable photometer. A small, lightweight

photometer employing a IP21 photomultiplier tube and built-in Johnson U, B, and V filters was used exclusively for events observed from remote sites. This instrument is discussed in detail by Chen and Rekenthaler (1966).




Data Acquisition Electronics

The SPICA-IV/LODAS System

Design criteria for a new SPICA. The concept of a Small Portable Interactive Computer for Astronomy (SPICA) was first conceived by Dr. John P. Oliver. The first SPICA system was implemented on a KIM-I computer, and is the precursor to the three generations of SPICAs which have followed. The common








21

thread linking the first SPICA to the latest version, the SPICA-IV, is the use of a 6502 microprocessor. Though each major upgrade to the SPICA systems has involved more hardware an effort has been made in SPICA-IV to retain portability, or at the very least transportabil1i ty.

The Lunar Occultation Data Acquisition System (LODAS) is the software control program designed to carry out the task of fast photometric data acquisition on SPICA-IV. It is, in actuality, inaccurate to say that LODAS was designed for SPICA-IV, or SPICA-IV for LODAS. The system requirements were such that the hardware and software grew together in a complementary fashion. The major design criteria for SPICA-IV/LODAS are listed in Table 2-4.

TABLE 2-4
DESIGN CRITERIA FOR SPICA-IV/LODAS

1. Data acquisition rates up to and including 1 kiloHertz must be supported.

2. The system must support a minimum of two simultaneous data acquisition channels.

3. Memory space must be provided to hold a minimum of two seconds of data in each channel,! at a rate of 1kHz.

4. 12-bit sample resolution should be used, to give a large dynamic range and eliminate last minute gain switching.

5. The system should retain easy transportability.

6. The system must function in the abscence of a disk drive, or a disk operating system.

7. A user friendly command structure and display must be
impl emented.

8. The control program must reside in Read Only Memory.










In all cases these criteria were met. and in the first three cases they were exceeded.

The SPICA-IV digital electronics. The heart of the SPICA-IV/LODAS system is an Advanced Interactive Microcomputer, model AIM-65, manufactured by Rockwell International. The AIM-65 has proven to be an invaluable design and development tool for the LODAS system as well as for several other astronomical data acquisition systems implemented at the Rosemary Hill Observatory. The AIM-65 is an 8-bit microcomputer with sixteen address lines incorporating a 6502 microprocessor chip. Up to 4-kilobytes of Random Access Memory (RAM), in the form of paired IK-by-4 bit chips (i.e. 2114's) can be accommodated on the AIM board. A 6522 Versatile Interface Adaptor (VIA), which is a programmable chip holding 16 bidirectional I/0 lines, four control lines, and two timers, serves as an interface to the "outside world* through an expansion connector on back of the AIM board. A standard ASCII keyboard, a 20 character alphanumeric LED display, and a thermal printer are provided for user I/0. The AIM-65 accomodates 24-kilobytes of ROM space in five 4-kilobyte Read Only Memory (ROM) sockets. memory mapped in the areas of $8000 to $FFFF. The AIM-65 operating system is normally resident in two ROM's occupying the uppermost 8-kilobytes of address space.

The SPICA-IV/LODAS system uses three boards manufactured by Micro-Technology Unlimited (MTU) to expand its RAM memory and to support peripheral devices. The first of these is a








23

16-kilobyte dynamic RAM board, model number K-1016 (MTU. 1979). This board is address assignable only on 8-kilobyte boundaries. Since the AIM-65 board is designed to hold 4-kilobytes of RAM, addressing the RAM board on 8-kilobyte boundaries would leave a 4-kilobyte hole in the system address space. While this in itself is not a problem, it will be seen that all available 64-kilobytes of AIM-65 address space must be used in configuring the system to meet the design criteria. This 4-kilobyte hole would then require a 68-kilobyte address space which the AIM-65 does not support. With this in mind the 2114 RAM chips were removed from the AIM board and the lower 16 kilobytes (address range $0000 to $3FFF) of system RAM reside contiguously on the dynamic RAM board.

A major design consideration was for LODAS to be able to function even if there were a hardware failure of either the disk drive or its controller board. A Shugart model 801, 8-inch floppy disk drive is used for primary data storage. The second MTU board is a Double Density Disk Controller (DDDC), model number K-1013 (MTU, 1980a), used in conjunction with the disk drive. The Channel Oriented Disk Operating System (CODOS) is distributed by MTU along with the DDDC board. The CODOS software provides many utility functions as well as a Service Call Processor and a Visual Memory Terminal driver program (MTU, 1981) which enhance the overall utility of the SPICA system.







24

The CODOS and its associated programs occupy address space in the range of $5000 to $5FFF, and $8000 to $9FFF. The DDDC board also provides an additional 4-kilobytes of RAM which is mapped in the address range $4000 to $4FFF. Although the disk/CODOS system is an integral part of SPICA-IV/LODAS, it is modular. Both the disk drive and the DDDC board may be removed from the system without impairing the data acquisition capability of the LODAS.

The final MTU board is a bit-mapped dynamic 8-kilobyte visual" RAM high resolution display, model number K-1O08 (MTU, 1980b). The display has a horizontal resolution of 320 dots (40 bytes) and is 192 scan lines in length. Thus 61440 bits may be individually controlled on the display within a 192-by-320 dot matrix. The hi-bit of the lowest byte on the board's address space is displayed on the upper-left of a Video Display Unit. The lo-bit of the highest byte occupies the extreme lower-right corner of the display. This board is memory mapped for the address space $6000 to $7FFF.

All three MTU boards were designed with on-board voltage regulators to derive +5 Volts from an unregulated +8 Volt supply. Since the SPICA-IV power supply provides regulated +5 Volts, the +8 Volt regulator on each MTU board was bypassed, and regulated +5 Volts distributed to each board. Other than this and the addition of a BNC connector on the visual memory board, the MTU boards are unmodified.








25

An additional board. referred to as the

Analog-to-Digital Converter/Clock (ADCC) board used in the SPICA-IV/LODAS system, was built by Oliver specifically for high speed occultation photometry. However, the devices provided on this board have been conveniently memory mapped and are available for other application programs running in any SPICA system at RHO. An examination of the

SPICA-IV/LODAS system memory map, as shown on the LODAS assembly listing in Appendix A reveals that the system memory space would normally be fully occupied with no addressable locations available for the placement of this board. The AIM-65 reserves the address space $AOOO to $AFFF for its own on-board devices, but only a small portion is actually address decoded (Rockwell International, 1978). A minor modification on the AIM-65 board was made to free up normally undecoded address space within this range, and has subsequently been made to all SPICA-III and SPICA-IV computer systems in use at the RHO.

The ADCC was laid out on an MTU prototyping board,

K-1020 (MTU, 1980c), and holds a real time clock and three analog-to-digital converters (A-to-D's). All components on the board were wire-wrapped. The clock circuit is based on a National Semiconductor MM-58167 chip which keeps time in month, day-of-month, day-of-week, hour, minute, second, and hundredths of a second. The clock is software readable and writeable, and can generate interrupts at either predetermined intervals or at a specific time/date









26



5 iK(ii1)




5+?
V00
Si psi i l 1

AO- RSO P8 -- 03 5
Al S' 6 P89-01 37 z T o

DO 2 PSG- U
0i 0 2~




05 141







Fi ~ DOe 2-82 Sceai ioamo h PR~~I/O lc
07 PA rcu i t












Fi~~ure 2-. ShmaiAilamo n f h he





Siue24.Shmtc ioa o h PCA-IV/LODAS clockdoia onete icut.











combination. For ease of operation the clock chip was interfaced to the SPICA-IV system through a 6522 on the protoboard. Thus the LODAS control software commands the clock through the protoboard VIA rather than controlling it directly. A schematic diagram showing the implementation of the clock circuit is presented in Figure 2-4.

The ADCC board also holds three 12-bit Analog Devices AD-574 analog-to-digital converters, each of these memory mapped into two contiguous bytes. A voltage conversion is initiated by writing to the A-to-D's. A digitized representation of the presented input voltage is obtained by reading the two memory mapped data bytes. The settling time for these A-to-D's is 35 microseconds. All three A-to-D's may be used in either a bipolar or unipolar mode, as selected by switches placed on the front-left edge of the ADCC board. In the unipolar mode the dynamic input range of the A-to-D's is zero to 10 Volts. In the bipolar mode the range is

-5 to +5 Volts. Since the DC output of the photometer amplifiers used at RHO produce a zero-to-I Volt negative going signal, buffer amplifiers (AD-741L's) are used between the A-to-D inputs and the actual signal input to the protoboard. Figure 2-5 shows one of the three A-to-D converter circuits. The circuits for all three channels are identical.

Ancillary equipment. Special I/0 signals, such as the analog inputs to the A-to-D's, are connected through the system to a 44-pin connector on the back-left edge of the








28

protocard. In addition inputs to each of the A-to-D's may be provided through miniature phone plugs mounted on the front-left side of the ADCC board.

The AIM-65, the three MTU boards and the A-to-D/clock board are mounted on an MTU K-1005 Card File and 5-Slot Motherboard (MTU, 1980d). A Zenith Data Systems 12-inch model ZVM-121-Z monitor is used to display video output. A Radio Shack model CTS-41 cassette recorder is used for secondary data storage.

For operational convenience a remote-control paddle was built. The need for remote control can be critical when observing alone, and obviates the need for continually running up and down telescope access ladders to operate the SPICA-IV keyboard. The paddle holds twelve pushbutton switches, which can select up to twenty-two key closures on the AIM keyboard. It is connected to the SPICA-IV/LODAS through its own cable, and specific functions can be activated by the observer from the telescope.

Power supplies. The DC voltages for the system, except those required by the thermal printer and disk drive, are provided by a Power One, model HBB-512 power supply. This supply can source 5 Volts DC at 3 Amperes, and 12 Volts DC at

1.2 Amperes. The thermal printer is powered by a Power One, model HB-24-1.2 (+24 Volt, 1.2 Ampere) supply. The disk drive employs a Power One, model CP-205 power supply providing 24 Volts DC at 1.5 Amperes, +5 Volts DC at I knpere, and -5 Volts DC at 0.5 Ampere. The CP-205 supply







29

along with the disk drive, is packaged separately from SPICA-IV/LODAS. For use at RHO the disk drive/power supply unit is mounted on the bottom shelf of a rolling equipment cart (see Figure 2-6).

SPICA-IV system configuration. Figure 2-7 indicates the overall system configuration. All major components including peripheral 1/O devices are shown. Figure 2-8 is a photograph of the ADCC board. The polarity-mode switches for the A-to-D's are mounted on the top of the board, as is a trim capacitor to adjust the MM-58167 clock rate.

The MTU K-1005 card file holding the digital electronics boards and the HB8-512 power supply, which comprise the major components of the SPICA-IV/LODAS system, are packaged in a small aluminum chassis. All signal and power cables are brought into the system through connectors on the back plate so as not to be obtrusive during operation. A cooling fan, which can be disabled on cold nights, is also mounted on the back plate of the chassis. Figure 2-9 shows the placement and fucntion of each signal and power connector found on the back plate of the chassis. Figure 2-10 shows the assembled SPICA-IV/LODAS system on its rolling cart in operation at RHO.

Portable use. A key element in the system design was

the need for relatively low power utilization. The portable aspect of the SPICA system had to be retained in order to use SPICA-IV/LODAS in the field. Observations of lunar grazing or asteroidal occultations often require setting up a



















0




















ka



C6
CL
0



u


OD
(1)







u 0
(n
Go .C
F




N ap
L
m
0
LL







31









32







TAPE
RECORDER THERMAL
PRINTER LODAS ROM
KEYBOARD
20 CHAR. DISPLAY 7AIM 65 / %,. VIDEO
\ 00, DISPLAY



y8K "VISIBLE MEMORY"

DISK
DRI V E






YDD 0 C .*, 1%
SOO,

I

/16K RAM CLOCK

N 00,

ADC2 ADC3
A D:C 1' E7z E7 PMT

AOCC


LmPLIFIER


IN ol PMT

wwVB
RECEIVER AMPLIFIER



Fioure 2-7. The SPICA-IV/LODAS system confiouration.



























-6
L (0
0



U




u
0


L
W




0
u







0


0








OD
1
04 Go
L CR








34





































0



L
0

u


0
u






ed

CL .Y
u IV J3

W




D










N

40
L


U.









































ir





























Figure 2-10. The SPICA-IV/LODAS rolling rack and Casseorain photometric equipment in use at the Rosemary Hill Observatory.






































0 fig
I ml I IRALN Tf














lk












IN







39

portable photoelectric station in a dark, secluded site where the availability of AC electric power is often non-existent. Thus one reason for using dynamic (as opposed to static) RAM is its lower overall power utilization, drawing higher current only during periods of active write cycles.

For observing at a remote site AC power is required to operate not only the SPICA-IV/LODAS system, but a Kepco model ABC-2500M high voltage power supply, an Astronomical Time Mechanisms model 240V DC electrometer amplifier, and a True Time Instruments WW..B receiver as well. To provide AC power a Nova model 1260-24 DC-to-AC inverter, running on two 12 Volt DC automobile batteries, has sufficient capacity to operate the entire photoelectric station for 35 hours. The AC inverter can supply approximately one Ampere at 120 Volts AC. Thus, to conserve power, the DDDC board and the Shugart 801 disk drive are not used. Data are saved to cassette tape after an observed event. Although the inverter can also provide power for the ZDS 12-inch monitor, this additional load reduces the working life of the portable power supply system considerably. Hence, for field use a Gold Star 12-inch black and white television with an RF modulator that had been built-in is used in its place. Though the power required for the television is no less than that of the ZDS monitor, it can be run directly on 12 Volts DC. The source of 12 Volts can be derived from one of the two batteries supplying the DC-to-AC inverter. In practice, however, the transporting vehicle's 12 Volt car battery is







40

used to power the television as well as the telescope drive corrector, slewing motors, electric dew cap, and ancillary equipment.

Figure 2-11 shows the SPICA-IV/LODAS system in field use while observing the asteroidal occultation of 14 Piscium by Nemausa on September 11, 1983 (Dunham et al., 1984). In this case AC power was available at the observing site. Figure 2-12, taken on November 13, 1984, shows the SPICA-IV/LODAS system when it was powered by the portable supply system while observing the asteroidal occultation of BD +08 471 by Ceres from the middle of the Florida Everglades.

Analog Electronics

The WWVB receiver. Nather and Evans (1970) have pointed out that the reduction of photoelectric observations, in principle, can yield the times of geometrical occultation with an uncertainty of only one or two milliseconds. As Table 2-4 has shown, a primary design criterion of SPICA-IV/LODAS was to have a data acquisition rate of least one point per millisecond. The inherent degree of accuracy in overall system timing depends upon both the AIM-65 phase-2 clock and the clock on the ADCC board. Thus, in order to realize absolute timing accuracy of one millisecond a standard time calibration source must be employed.

This requirement precludes the idea of using High

Freguency (HF) transmissions from the National Bureau of Standards' radio station WWV (located in Fort Collins,






















0;




0



0 0




4-- u
~.i0 0, 0'
00tb Z- L Li
0 o a ol. 0
0
c010




m ~3 40 0a Z'0.




U) .0 L
41 .


Go E



N I ~


3 EZc
U- 0












Ig.le , I I










ell,




















0 c"

-W4 10 .0
E 40 UGi 40 A u0 0 E- L



0 ~L








uqtA





U4 .60




O'D
0i L



4
0
-.c 0
0' i0 U) 0L C- mo.C G



N. 00L40







44








45

Colorado) as a suitable time base reference. The uncertainty in the HF propagation path arising from variability in ionospheric conditions between Fort Collins and Bronson can lead to timing uncertainties as large as a several milliseconds.

Fortunately, NBS provides a Very Low Frequency

Coordinated Universal Time broadcast, via radio station WWVB, which transmits at a standard carrier frequency of sixty kiloHertz (Kamas, 1977). At sixty kiloHertz the mode of propagation is strictly by ground waves; hence, the variability in propagation time is removed (Department of the Army, 1953). The propagation path will simply follow a great circle from transmitter to receiver, amounting to a fixed light-time delay of 7.4 milliseconds at the RHO.

WWVB transmits timing information in a tristated time code. The strength of the carrier wave is modulated by reducing output power for 0.2, 0.5, or 0.8 seconds each second. Encoded in this modulation envelope are the time, date, and current UTI correction. Each ten-second period and the start of each new minute are identified by encoded framing references.

Detection and interpretation of this signal are

precisely what is needed to provide the timing accuracy desired. Several avenues of approach were debated. Rather than having a receiver built, a commercially available unit well suited to the task was procured. The unit, a True Time Instruments model 60-T receiver provides not only a detected carrier output, but a TTL compatible code output as well.







46

A small modification made to the TTL output, dividing it down to 0.8 Volts, permits feeding the code signal directly into one of the three A-to-D converters available in the SPICA-IV/LODAS system. The digitized time code is sampled simultaneously along with the photometric channels. The receiver is mounted on a 19-inch equipment rack, shown in Figure 2-13, along with a WWV receiver and the high voltage power supply used by the PMT's.

In actual use the LODAS system clock is set manually by the observer to an audio WWV signal. This procedure results in a clock setting accuracy of a few tenths of a second. it is then noted if the clock was set fast or slow. Digitized WWV'~B second transitions then provide a correction to the nearest mill isecond.

The signal strength at RHO rarely exceeds 125 microvolts per meter (True Time Instruments, 1974). An active antenna, model A-6OFS, also manufactured by True Time instruments is currently used at the observatory. This is marginal under circumstances of unfavorable reception, and an alternate antenna design is being considered for future use at the observatory. However, it has been found that eighty percent of the time a decodable signal is available while observing. During nights of signal fading, time code is digitized before and after the event as conditions permit and time corrections to the computer's internal clock are interpolated in post-observational reduction.

The receiver introduces a measured electronic delay time of 19 milliseconds from the time of reception of the WWVB



























Figure 2-13. The WWV and WJWVB receivers, and PMT high
voltage power supply.








48








49

carrier to code output. This, however, varies slightly as a function of signal strength. By attenuating the input signal from the antenna it was found the delay is lengthened to 21 milliseconds at a level where the time code cannot be reliably detected. This then sets the limit of the absolute timing determination to +/- 1 millisecond, meeting the occultation program's allowable tolerance. All reductions then have a final correction of 27.4 milliseconds subtracted from the determined time of geometrical occultation, with an additional error of +/- 1 millisecond added to the formal error of the solution.

Radio antennas. For use at the observatory the receiver is mounted in the telescope main-power distribution rack, immediately above the WWV~ receiver. These two receivers share an antenna cable, so only one receiver can be used at a time. The antenna connector must be switched from the WWV receiver to the LJJ)B receiver before use. Approximately one minute is required by the WWVB) receiver after being powered up to lock onto the time code and produce a readable decoded output. A twenty-five foot cable terminated at one end with a BNC connector and a three conductor phono plug on the other is kept on the rolling cart with the SPICA-IV/LODAS system. The BNC end is connected to the CODE output of the WWVB receiver, and the phono plug end connected to one of the SPICA-IV/LODAS signal inputs. The input channel selected to receive the time code should be switched to unipolar mode.








50

Because of the frequent lightning strikes, unavoidable on one of the highest hills in Florida, the WWV and WWV~B antennas are disconnected at the base of the antenna tower at the end of a night's observing. A PL-259 connector can be found at the tower base to which the WWVB active antenna connects via a plug, and the WWV long wire antenna connects by means of an alligator clip. The LJ4VB antenna and its 15-foot antenna cable are removed from its 6-foot high mounting stand and stored above the desk on the first floor of the observatory building.

The photometer amplifier. Since millisecond time

resolution is desired, the photometer amplifier used must have a response at least as fast at a range of gains useful to the occultation observing program. An Astronomical Time Mechanisms model 240 fast photometric DC electrometer amplifier was choosen. This amplifier described by Astronomical Time Mechanisms (1980) is based on a circuit by Liver (1976) designed specifically with lunar occultation observations in mind. Amplification is achieved in two stages: first in a current-to-voltage conversion stage; and second in a buffer amplifier. A similar amplifier had been in use at the Rosemary Hill Observatory fourty-six centimeter telescope for several years. Caton (1981) carried out a program of UBV photometry on RS CVn stars using this amplifier. He found eighth magnitude stars could be observed giving a full scale reading with the amplifier switched to the so-called "C" gain. This gain setting employs 10 megOhm








51

feedback resistance in the first amplification stage. Using the seventy-six centimeter telescope would yield a gain of approximately one magnitude over the forty-six centimeter tel escope.

The effective time constant of the amplifier is limited by the high precision megohm feedback resistors and the capacitance of the input signal cable (added to a capacitance of 5 pf., the value of the feedback capacitor used on the signal input). In order to observe ninth magnitude stars with a 2 kHz half power response, twice that of the target time resolution, the capacitance of the input signal cable cannot exceed 45 pf. RG-58 A/AU co-axial cable has a capacitance of 28.5 pf./foot. Thus, to achieve this time resolution for stars of ninth magnitude a cable of this type no longer than 18 inches must be used. Rather than RG-58, which is commw1only used as a signal cable, the occultation program uses RG-71/U, with a measured capacitance of 13.1 pf ./foot.

For practicality, the amplifier is mounted on the side of the photometer cold box as may be seen in Figure 2-9. This allows a short cable run (only eight inches is needed), and is in an extremely convienient place for an observer operating the photometer. Having the amplifier fixed to the photometer also permits the signal cable from the PMT to be securely fastened down thus preventing any possible cable flexure. Such flexure would result in charge redistribution along the signal cable leading to erroneous fluctuations in the observed signal level.








52

The amplifier coarse gain steps are in increments of 2.5 magnitudes, and fine gain steps in increments of 0.5 magnitudes. After initial use at the telescope the amplifier was modified to provide a 0.25 magnitude gain switch to boost the effective gain at any coarse and fine combination. This was done to provide the observer with a bit more flexibility in chosing the amplification factor used for the purpose of real-time photometric data display.

The output of the amplifier is connected to the

SPICA-IV/LODAS system by means of a fifteen foot signal cable terminated on both ends with phono plugs. This cable is kept on the rolling cart along with the taAAVB signal cable. One end is connected to the amplifier output marked RECORDER, and the other end is connected to a SPICA-IV/LODAS input switched

to unipolar mode.



Limitations of the Occultation Photometric System

Once obtained, the new ampl if ier was used to assess the limitations of the seventy-six centimeter telescope photometric system and to confirm that stars of reasonable faintness could be observed while preserving a system time constant on the order of a millisecond. Stars over a range of six magnitudes were observed on the moonless. night of May 23, 1981 U.T. wi th the Astromechan ics photometer cool ed wi th dry ice, and a Johnson V filter. Measurements were taken at both the nominal operating voltage of the PMIT of 1200 Volts DC, and at the maximum operating voltage of








53

1600 Volts DC. Five minutes of settling time was allowed after switching voltages before readings were taken. The observations, listed in Table 2-5, give the star name, U.T. of observation, the star's V magnitude, the photometer diaphragm used. For both vol tages the amplifier coarse and fine gains, and the signal level due to the star (normalized to a full scale value of 255) are listed. In all cases the dark current and sky background have been subtracted from the star-plus-sky readings. Gain settings were adjusted to give readings as close to 65 percent of full scale as possible.

TABLE 2-5
STELLAR INTENSITY READINGS WITH RHO SPICA-IV/LODAS


1200 Volts 1600 Volts
Star Name U.T. my Dia. Gain Star Gain Star

Gainuna Leo-a 0315 2.6 J 87 193 82 173
57 LUla 0338 5.2 J C7 173 87 153
88 Leo-a 0401 6.1 J C9 165 89 143
88 Leo-b 0413 8.6 J 010 158 CIO 136


The fine gain steps of 0.5 magnitude run from 'I" to "lla. Hence stars with a V-magnitude as faint as approximately 7 can be observed at a coarse gain setting of 'C', at a PMT voltage of 1200 Volts DC. In order to gain one magnitude observations can be made at 25 percent of full scale. With twelve bit digitization this is still roughly one part in one-thousand, or a photometric precision of 0.001 magnitude. Alternatively, a gain can be achieved by increasing the PMT voltage. As can be seen from Table 2-5 increasing the Pt-T voltage to 1600 Volts provides a gain of approximately 2.4 magnitudes. However, the penalty of








54

increased thermal noise (dark current) must be oaid if this option is taken. Fortunately, the RMS amplitude of the dark current for the 1600 Volt observation of 88 Leo-b was only

2 percent of the star signal level.

These observations tend to lead to over-optimistic

results, as occultation observations will not be made in dark skies; indeed the telescope will be pointed in the direction of the moon. To assess a "worse cs"condition, the star SAO 098723 (mV=-8.7) was observed on May 11, 1981 when it was only 0.1 degrees away from the dark limb of a 55 percent illuminated moon. Using the same diaphragm and filter as the observations in Table 2-5, a gain setting of D7 was required to bring the star-plus-sky level up to 67 percent (170 counts out of 255). The sky contribution was 130 counts, so the star signal was only 16 percent of full scale. In order to bring the signal into the range of 'C" gains, an additional gain of 0.5 magnitude is needed. This would place the gain of the amplifier at C11. At 1600 Volts the star-plus-sky signal was just above half scale on a gain setting of C7. Hence under bright sky conditions the photometric system would allow detection of stars down to ninth magnitude with timing accuracy of one millisecond. The limit is more stringent for observations with either of the intermediate bandwidth filters, whose integrated bandpass is roughly one-tenth that of the Johnson V filter. In this case the limiting magnitude is reduced to roughly 6.5.







55

The Lunar Occultation Data Acquisition System Software Software Design Considerations

The design criteria specified in Table 2-4 were binding not only for the choice of hardware to be used the SPICA-IV system, but applied equally, if not even to a greater extent, to the design of the data acquisition and process control software. The execution of a repetitive task at a precisely defined rate, which must interact in real-time with the "outside-world" is best accomplished by the technique of interrupt processing. Thus, the major process control task of the LODAS software, real-time data acquisition at a rate of at least one-thousand 12-bit samples per second, in at least two channels, was relegated to an interrupt service routine. Yet, some functions of LODAS do not have the need for either repetitive or regularly scheduled execution. For example, scanning the keyboard for user requests, or updating the 20-character alphanumeric display with system status information can be done at the microprocessor's leisure. Hence LODAS actually operates two concurrent programs. A foreground program handling all input to, and output from the observer runs continuously in a relatively quiescent mode, calling upon system services only when required by the observer or program logic. This program is repeatedly interrupted by the aforementioned interrupt service routine. referred to as the background task, on a regularly scheduled basis.








56

System timing,. The limiting factor which played a major role in the development of the LODAS operating concepts was the execution speed of the 6502 microprocessor. The system (i.e. microprocessor) clock rate for the AIM-65, as for all 6502 systems, is 1-megaHertz. A sampling rate of 1-kiloHertz would require the evocation of the interrupt service routine every millisecond. This constrains the process control tasks to a maximum of one-thousand system clock cycles. However, the system cannot spend all of its available clock cycles executing the interrupt service requests. Some percentage of

the total system throughput must be allocated to the foreground task. Fortunately, the reaction time of any person is much longer than the cycle time of a 1-MegaHertz computer. This undeniable physiological fact allows a very low bias in time-slicing for the foreground routine. A comfortable allowance of a minimum of 10 percent was deemed more than adequate.

On average, the execution cycle time for a typical machine instruction on a 6502 microprocessor is 3.5 microseconds. This means that approximately three hundred instructions, at most, could be issued in the interrupt service routine before interrupt pile-up would occur. The need for rapid execution speed is clear. For this reason the LODAS software was implemented in 6502 machine language, rather than a more user-oriented, but slower, high level language.







57

It was found that while the task of data acquisition and storage requires less than a half-millisecond, other interrupt service requests (such as servicing a video "strip chart" display) would demand a total number of machine instructions well in excess of the three hundred maximum. Because of this the interrupt service routine was multi-phased, handling data acquisition and storage in each phase, and pieces of other service requests in successive phases. In breaking the background task into four phases the execution time for any one phase required less than 450-microseconds. Hence, the basic system interrupt rate was defined as 500-microseconds. This allows data to be sampled in successive pairs and averaged together in real-time before being stored. Thus a 1-kiloHertz data sample is actually comprised of two 500-microsecond pair-averaged samples. This not only effectively increases the signal-to-noise ratio of the acquired data by 41 percent, but the stored 1-kiloHertz samples have a Nyquist cut-off frequency of 1-kiloHertz (half the actual data sampling rate) as well.

Memory usage. The area available for storing data in a circulating event buffer is 18-kilobytes in length. In order to optimize the use of this limited (but sufficient) resource, two 12-bit acquired and averaged data samples are bit-packed into three 8-bit bytes. This packing, accomplished by the interrupt service routine, saves 25 percent over storing the 12-bit data, unpacked, into two 8-bit bytes. The 18-kilobytes of available RAM are







58

partitioned into three 6-kilobyte regions., each to hold one channel's data. Four seconds of pair-averaged data, acquired at an effective rate of 1-kilohertz, can be stored in each 6-kilobyte region. Hence, the system as built and programmed is capable of holding twice the amount of data originally envisioned, and in an additional data channel as well.

A modified version of the LODAS program, called FASTDAS (Fast Asteroidal Data Acquisition System) partitions RAM into only two data storage areas, thereby gaining 50 percent in the data buffer circulation length. This program has been used in observing asteroidal occultations of stars from remote sites (Dunham et. al, 1984).

The LODAS/EO7 program has been assembled to reside in ROM at an address space of $DOOO. This allows co-residency with AIM-65/FORTH, which is often used on RHO SPICA systems.

Supplementary program documentation. The overall logic flow for the LODAS foreground and background (interrupt service) programs, as well as the program initialization procedure are shown on the operational flow charts presented as Figures 2-13 and 2-14. A fully annotated assembly listing of the LODAS program is contained in Appendix A. This listing reflects LODAS program revision number E07, the seventeenth incarnation of LODAS since its inception. The assembly listing is preceeded by a detailed accounting of the LODAS memory space in terms of I/0 addressing, AIM-65 monitor utilization, program variable space, and an overall system memory map. Following the assembly listing is a symbol-table








59

Label Description of Program Step

(INITIAL PROGRAM ENTRY)

COLDST [Establish Address of Interrupt Service Routine) LODASO [Print/Display Program Name and Version Number] SETVIA [Initialize T1 for 500 Microsecond Interrupts] CLINIT [Setup Access Control for Internal Clock]

SETGAP [Initialize Cold Start Variables]

[Interactive Setup of All Control Parameters]
4
[Initialize Video Display for "Glass Recorder"] WARMST [Initialize Action Code to "Quit"]

SETCTR [Initialize Flags, Counters and Pointers]

SETSTL [Initialize Data Buffer Addresses]

[Clear Possible Interrupt Sources: Clock, T1I
4
SETBT [Setup Bit Pattern Table for Video Display]

[Clear and Draw Background Pattern on Video]

MAIN (Foreground Program: Command Service, Display)

SCAN [Scan Keyboard for Command Entry]

<"Exit to Monitor" Command Received?>--NO-[Clear both VIA Interrupt Flaq Registers] 4
4. 4
(STOP! EXIT TO AIM-65 MONITOR) 4 NEWCDE [Execute User Request: See Table 2-84--DSPCHK
4
DISTME


Figure 2-14. LODAS foreground program logic flow chart.









60

LabelI Description of Program Step

(INTERRUPT SERVICE ROUTINE)

INTRCV [Save A. X. and Y-Registers on the Stack)
4.
SNAPCK (Did a Quit After Delay Just Occur?>--NO-4, 4,
(RGET< ---- [Transfer $0000-$0020 to TAPBUF] l,

ADCONV. (Start A-to-D-Conversi onsJ4 ---------YES -------(INCMSC<--[Clear Interrupt Flao for TIO4

CLCIRG [Clear Clock Interrupt, Set Display Update Flaqj*-4,
--NO----------4, 4
CNTDWN [Disable Clock Interrupts at VIA)]4

(RGET< ---([Clear Tape and Clock Request Flaas) 4

RENI*1I (Reseed Ti Timer in Synchronism with Clock](4
[Enable 500 Microsec. Interrupts, Reset Counter]
4
INCM-SC (Increment 500 Microsec Counter]
4,
(Has a Millisecond Elapsed?>--NO-4 4
(Disable Ti Interrupts] 4

4
INTTST (Data Taking in Progress?>--YES-4 4,
(RGET< ---- NO-(Has 0.1 Second Elapsed?> 4, 4, 4
(RGET(-------------ERead Clock] 4


SAMPLE (STPND2<--NO-
4
SAMPLO STEPP, STEP2, STEP3(- 4,
STEPO [Put A-to-D Readings in Average Registers]
4



Figure 2-15. LODAS background program logic flow chart.











STEPI [Read A-to-D's, Average with Previous Value]

[Store Averages in Circulating Data Buffer)



--YES----

STEP3 [Read A-to-D's, Average with Previous Value]

[Store Averages in Circulating Data Buffer]

[Update Data Buffer Pointer]
4
STPEND [Increment Current Storage-Byte Pointer]
4
STPND1 [Increment the Interrupt Service Step Count]
4
STPND2 [Increment the Sample Counter]
4
RGET [Restore Y, X and A-Registers from Stack]
4
(EXIT INTERRUPT SERVICE ROUTINE)
- ---- -----------------------------------------------Statements in the flow chart are interpreted as follows:

[Unconditionally Execute Statement In Brackets]



(PRIMARY ENTRY OR EXIT POINT)

Follow flow lines either unconditionally, or propositionally as is applicable:

[Statement] ------- --YES/NO-4 or 4

Figure 2-15----- -------Continued.----------------------Figure 2-15--Continued.








62

and reference list. The LODAS makes extensive use of internal subroutines and data tables. Tables 2-6 and 2-7 provide a synopsis of these routines and data tables respectivyelIy.

Peripheral Input/Output

User (Observer) 1/0. The LODAS program was written with ease of operation in mind. An observer at the telescope often has enough problems confronting him or her, and an unfriendly computer need not be among them. After powering up the SPICA-IV, and evoking LODAS, the observer is prompted on the 20-character alphanumeric display to enter the parameters salient to the observing session. The date, data acquisition rate and channel assignments are among these input requests. These parameters may be changed at any time

by issuance of a LODAS command. The LODAS commvands are actuated by a single keystroke (or for safety, by depressing the CTRL key simultaneously with the command key). The corwnand-key assignments are, in most cases, mnemonic to the command request. A list of LODAS commands is given in Table 2-8. Some of the LODAS commands are acted upon immediately (as in the case of EXIT) and some require additional information from the observer. Examination of Table 2-8 will reveal the proper responses to any LODAS command request prompt. These responses are identical to those required on system initialization when the parameters are first established.

Once the LODAS system initialization is completed the

observer is kept abreast of system status on the 20-character







63

TABLE 2-6
LODAS/E07 SUBROUTINES


Name Description of Subroutine
- ----- ----------------------------------------------------CLENUP Post tape writing clean-up. Reset interrupt count,
clear VIA interrupt flags, re-enable 0.1 second
interrupts, restore "Action Code*.

CLEAR Clear the 20-character alphanumeric display.

COMENT Input comment from user, up to 40-characters in
length. Display, print, and store in the tape header
buffer, (TBUFF).

DECHEX Convert packed BCD in A-register to hexidecimal.

DISBUF Output contents of display buffer (DISBUF) to
20-character LED display.

DRSET Interactive data acquisition rate setting. Get data
acquisition rate from keyboard.

FRZSET Interactive delay time (freeze) setting. Get delay
from keyboard. Validity of entry is checked.

GETKEY Input a character from the AIM-65 keyboard, if no key
pressed then wait.

GOCLCK Reset clock with values stored in MILSEC.

HIOUT Convert left half-word of A-register to ASCII and
store in DISBUF at offset indicated by X-register.

HXIBCD Convert hexidecimal byte in A-register to 2-byte BCD.
Puts MSD into Y-register, and LSD into A-register.

HXASC2 Convert hexidecimal value in A-register to 2-byte
ASCII and store in DISBUF at offset indicated in
X-register.

HXASC3 Convert two byte hexidecimal value in A, and
Y-registers to three byte ASCII and store in DISBUFF
at offset indicated by X-register.

KEYCK Modified AIM-65 keyboard scan routine. Does not exit
to monitor on ESC.








64

Table 2-6. Continued.

Name Description of Subroutine
- ---- ----------------------------------------------------LHWOFA Halfword shift to A-register to the right. Zero out
the left halfword of A-register.

OBHXAS Convert a one-digit hexidecimal number to ASCII. PACK2 Get two digit BDC number from keyboard, display
number, store as packed BCD in A-register. PNDM Print and display an in-line message. ROCLIN Set up clock to accept a read request. RDCLK Read current time from clock and store in MILSEC.
A. X, and Y registers unaffected.

ROLACT Clear the 20-character alphanumeric display and
restore old mAction Codea.
TICSET Set up tics, data channel markers and screen lines on

video display.

TIMEGO Start clock running from keyboard command. TIMSET Interactive clock setting routine. Get time and
date from keyboard. Start clock if commanded.
Validity of entry is checked. TOGPRT Toggle AIM-65 printer on/off. TAPINT Write occultation observation to cassette tape.
interrupt request servicing disabled. TVCLEA Clear the video display. TVDISP Output next hi and lo resolution data points on video
display.

TVSET Interactive set-up of video display parameters. Get
channel assignments and display rate from keyboard. TVSETX Clear video display and redraw background. WRCLIN Set up clock to accept a write request. WRTAPE Transfer a contiguous block of data to tape.










TABLE 2-7
LODAS/E07 DATA TABLES


Name Description of Data Table

CLCTBL A table of packed BCD numbers from decimal 00 to 99,
inclusive. Used by clock and intensity display
routines to 20-character LED display.

TVTICS A bit mapped data table containing the video display
background pattern.

TVTBLH The hi byte of the address of each video display
line as mapped on the visual memory.

TVTBLL The lo byte of the address of each video display
line as mapped on the visual memory.








66

TABLE 2-8
LODAS/E07 KEYBOARD COMMANDS



Key=5 LODAS Cold Start:

This command, if executed from the AIM-65 monitor, will transfer control to the Lunar Occultation Data Acquisition System and begin execution. The system will respond by flashing LODAS R65/E07 on the alphanumeric display and logging this on the printer. Program default parameters will be established, and the observer asked for variable set up parameters. This command is valid only from the AIM-65 monitor and will be ignored once LODAS is in control and running.


Key-6 LODAS Warm Start:

To re-enter LODAS from the AIM-65 monitor while preserving previously set parameters use this command. It is assumed that LODAS was previously cold started, and exited (i.e. to CODOS or the AIM-65 monitor). Warm start will not require resetting the internal clock/calander, selecting data acquisition rates, display channels or the 'video strip chart" display rate.


Key-Cc Enter a COMMENT:

This command will allow the observer to enter a comment of up to 40 characters in length (two lines) on the printed observing log. The most recent comment is also retained in the data buffer header to be saved on disk or tape on command. Entering a RETURN in the comment field will terminate comment entry.


Key=Dc Exit To The DISK Operating System (CODOS):

This command will terminate LODAS and boot the Channel Oriented Disk Operating System. If a CODOS system diskette is not in the disk drive, and the drive door closed the system will freeze up. CODOS must be entered in order to save observing data to a data diskette, or load another observing program from the system diskette. Note: To go from CODOS to the AIM-65 monitor strike the ESC key. To go from the AIM-65 monitor to CODOS, after CODOS had been previously booted strike the F3 key.







67

Table 2-8. Continued.



Key-Fc Save FILE On Cassette tape

Data may be saved on cassette tape instead of a CODOS disk by issuing this command. Be sure the cassette tape recorder is set to RECORD, and that a non-write protected cassette is in the tape recorder. LODAS will ask for a data FILE name. Any name up to five characters in length may be entered. Data transfer to tape will begin immediately after the entry of the file name. As each block is writen the current block count will be presented on the alphanumeric display. When all data has been transferred the message TAPE WRITE COMPLETED will be displayed.


Key-Gc Restart System U.T. Clock To Preset Time: (GO):

If the system clock had previously been set using the U.T. Clock Set command, but not started, this command will start the clock running at the preset time.


Key=4Ic INITIATE Data Taking (Integrate/Sample):

Upon receipt of this command LODAS will immediately begin sampling the three Analog to Digital Converters at the preset sample rate. Acquired data will be stored in the all three channels of the 4096 sample circulating data buffers. The Initiate Data Taking command will reset the data buffer pointers to the beginning of the each of the three data storage buffers will overwrite any previously buffered data.


Key-Pc Toggl e PRINTER (On/Off):

Each time this command is entered the 20 column printer will toggle from on to off, or off to on. If toggled off command interaction logging will still be displayed on the 20 column alphanumeric display, but will not be printed. The words ON or OFF will be flashed on the 20 character alphaneumeric display to indicate the new status of the printer.







68

Table 2-8. Continued.



Key--Q QUIT Data Taking:

Upon receipt of this command LODAS will cease data taking after the previously specified delay time. When data taking stops all system status information related to data taking will be saved in the data header buffer for possible subsequent storage to disk or tape.


Key--Rc Select Data Taking RATE:

This command is used to select the data acquisition rate. LODAS will ask for the desired rate, which is to be entered in milliseconds per point. Three digits must be entered. Thus if data is to be taken at five points per millisecond the entry should be 005. Data may be taken at any rate from one to 256 milliseconds per point. An entry of 000 will result in the 256 millisecond per point rate. Data will actually be taken at twice the specified rate and pair averaged before being stored.


Key--Tc Set Delay TIME:

The Delay TIME parameter set by this command affects the system response time to a QUIT command. LODAS will prompt for the desired time delay to be waited before the system will QUIT data taking when commanded to do so. Delay times are entered in units of 100 times the data acquisition rate in milliseconds. Two digits (00-99) must be entered. Thus an entry of 12, with a data acquisition rate of one point per millisecond, will cause LODAS to wait 1.2 seconds before halting data acquisition on a QUIT command.







69

Table 2-8. Continued.




Key=-Uc Set The UNIVERSAL Time Clock/Calendar:

This command will allow~ the observe to manually reset the internal U.T. Clock. LODAS will first prompt for the the year, month, and day to which the clock should be seeded. Entry must be in the form of a six digit number. For example 850118 will seed the calendar to January 18. 1985. LODAS will then ask for the day of the week, as a single digit number. Day I is Sunday and day 7 is Saturday. LODAS will then request the hour and minute to be entered as a four digit number. Thus 1820 will seed the clock to 18 hours and 20 minutes. After these entries are made LODAS will prompt by displaying START=ANY EX

Key=-Vc Set Up VIDEO Display Parameters:

This commland allows the observer to select which of the three input data channels is be displayed on the Hi Resolution Graphics Display, and which is to be displayed on the Lo Resolution Graphics Display. LODAS first asks for the input channel number to be assigned to the A (Hi Resolution) display, and then for the channel number to be assigned to the B (Lo Resolution) display. Valid input channel numbers are 1, 2, and 3. LODAS then requests the video display rate. The display rate is coupled to the the data acquisition rate. Entry is in units of the data acquisition speed divided by two per displayed point. Thus if the data acquisition rate is 5 milliseconds per point (1/200 second), an entry of 002 will result in a point being displayed every 1/50 second. Entries must be three digit numbers in the range 001 to 255. An entry of 000 is interpreted as 256.








70

Table 2-8. Continued.



Key=Xc EXIT To The AIM-65 Monitor:

This command will terminate LODAS and return control to the AIM-65 monitor. LODAS may then be restarted with either a COLD or WARM start. This command should be used if CODOS is to be re-entered rather than booted. After exiting to the AIM-65 monitor use the F3 key to re-enter a previously booted CODOS system.


NOTE: A letter designation of "c" postfixing the command key (i.e. KEY=Xc) indicates that the CTRL key must be depressed simultaneously along with the specified key.








71

alphanumeric display. Figure 2-16 shows the format of this display.





HHMSS M 000 000 128
Inest on A-- chne #3
: :Intensity on A-to-D channel # Intensity on A-to-D channel #2 Current (or Last) Commwand Code
Universal Time



Figure 2-16. Format of the LODAS 20-character
alphanumeric system status display.



The A-to-D channel intensities are scaled from zero to 255. With no input on an A-to-D channel the display will read 000 if that channel is set to unipolar mode, or 128 if that channel is set to bipolar mode.

All observer commnands, command responses, and the current status display are logged on the 20-character thermal printer whenever a command is issued. The printer can be disabled by an observer command.

The video *strip chart recorder"M. Without a doubt,

strip chart recorders are the bane of photoelectric observers worldwide. Renowned for clogging up, dripping ink, jammvTing paper or spewing it forth in voluminous quantities, these devices, while undoubtedly useful, often seem more troublesome than they are worth. Since the LODAS saves all observational data in a digital format on disk or tape there is no need for a printed chart record of the observations.







72

Yet, a chart recorder is an extraordinarily' handy too] (when working) on the observing floor, even if used simply to visually monitor the ongoing photoelectric observations. An ideal chart recorder for use with SPICA-IV/LODAS would provide a graphic display of the event for the observer's immediate inspection, without the necessity of plotting it out on reams of paper. A further ideal would be such a device with no mechanical parts to fail, as they invariably do, while observing.

These ideals were transformed to reality with the

introduction of what is now referred to as the video "strip chart recorder'm. Rather than having the photoelectric and/or timing signals drawn on a paper chart with an inking mechanism, data are displayed on a video monitor. The screen is divided into two halves. The left half, called the hi-res, or A-channel display, has a resolution of

0.25 percent (1 part in 256). The right half, called the lo-res, or B-channel display, has a resolution of roughly

1.5 percent (1 part in 64). Data input to any of the three A-to-D's can be selected to be displayed on either the A or B channels. While observing, typically the A-channel is used to display photometric data, and the B-channel is used to display the WWAVB time code. The display rate is software selectable, and can be as fast as one-half the data taking rate, or as slow as 1/256 of the data taking rate. With a millisecond data acquisition rate it is possible to display 500 points per second.







73

The video 'strip chart recorder" is best thought of as either a vertical two-channel programmable storage oscilloscope. or a chart recorder with a fixed paper and moving pen. When the pen hits the bottom of the page, the next point will be plotted on the top of the page as the point previously there is erased.

The LODAS activates the video "strip chart recorder"

when an Integrate command is issued, and freezes the display when a Quit command times out. A sample video 'strip chart recorder' display is seen in Figure 2-17. The number of short dashes at the top left and top right of the display indicate which A-to-D channels have been assigned to the hi-res and lo-res portions of the screen respectively. The short horizontal dashes (on the same line for both hi and lo-res) indicate where the cursor ("chart pen") was when data acquisition stopped. Tic marks on top and bottom indicate the 25, 50, and 75 percent signal levels for the hi-res display, and the 50 percent level for the lo-res display.

Data archival. Post-event observing data may be stored on either 8-inch floppy disks or on cassette tapes. To save data to disk the observer should command LODAS to enter the disk operating system (CODOS), save the observing data by issuing the command: SAVE filename 0700 4FFF, and then re-enter LODAS (if more observations are to be made) with a warm-start. To save data to cassette tape the LODAS Fc command is used. Approximately 10 minutes are required to save the observing data to cassette tape.




























Fioure 2-17. The SPICA-IV/LODAS and the video "strip chart recorder" display.