• TABLE OF CONTENTS
HIDE
 Title Page
 Dedication
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Abstract
 Introduction
 Device fabrication
 Experimental methods
 Experimental results
 Summary of thermal and generation-recombination...
 Discussion of results
 Conclusions and recommendation...
 Bibliography
 Biographical sketch
 Copyright














Group Title: Noise in junction-gate field-effect transistors at low temperatures
Title: Noise in junction-gate field-effect transistors at low temperaturs
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00082478/00001
 Material Information
Title: Noise in junction-gate field-effect transistors at low temperaturs
Physical Description: x, 145 leaves. : illus. ; 28 cm.
Language: English
Creator: Hiatt, Clifford Frederick, 1932-
Publication Date: 1974
 Subjects
Subject: Junction transistors   ( lcsh )
Transistors -- Noise   ( lcsh )
Low temperatures   ( lcsh )
Noise generators (Electronics)   ( lcsh )
Electrical Engineering thesis Ph. D
Dissertations, Academic -- Electrical Engineering -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis -- University of Florida.
Bibliography: Bibliography: leaves 143-144.
Statement of Responsibility: by Clifford F. Hiatt.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00082478
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 000580738
oclc - 14080812
notis - ADA8843

Table of Contents
    Title Page
        Page i
    Dedication
        Page ii
    Acknowledgement
        Page iii
    Table of Contents
        Page iv
        Page v
    List of Tables
        Page vi
    List of Figures
        Page vii
        Page viii
        Page ix
    Abstract
        Page x
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
    Device fabrication
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
    Experimental methods
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
    Experimental results
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
    Summary of thermal and generation-recombination noise theory
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
    Discussion of results
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
    Conclusions and recommendation for further study
        Page 140
        Page 141
        Page 142
    Bibliography
        Page 143
        Page 144
    Biographical sketch
        Page 145
        Page 146
        Page 147
    Copyright
        Copyright
Full Text















NOISE IN JUNCTION-GATE FIELD-EFFECT
TRANSISTORS AT LOW TEMPERATURES







By

CLIFFORD F. HIATT


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







UNIVERSITY OF FLORIDA


1974





























To my wife and my children

whose sacrifices and encouragement

made this work possible
















ACKNOWLEDGMENTS


I am deeply indebted to Dr. Aldert van der Ziel and

Dr. Karel van Vliet for their most generous and valuable

guidance and assistance during the preparation of this work.

I thank Dr. Eugene R. Chenette for the opportunity to do

this study and for his advice and encouragement. Dr. S. S.

Li and Dr. Robert B. Bennett have been most helpful during

many phases of this work and their help is greatly appreciated.

The generous help and advice from Mr. Richard King

during design and construction of the cryogenic system was

most valuable and contributed significantly to the rapidity

and the accuracy of the measurements obtained.

I also wish to express my appreciation to all the

members of the Faculty and Staff who have guided and assisted

me during my graduate studies. I am particularly indebted

to Mrs. Roswitha Zamorano for her patience and persistence

in the preparation of this manuscript.

My deepest gratitude go to my wife and my children who

have always had faith, encouragement, and understanding in

great abundance when it was most needed.


iii

















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS . . . . . . . . .iii

LIST OF TABLES . . . . . . . . .. vi

LIST OF FIGURES . . . . . . . . .. vii

ABSTRACT . . . . . . . . .. . . x

CHAPTER

I. INTRODUCTION . . . . . . . 1

II. DEVICE FABRICATION . . . . . . 6

III. EXPERIMENTAL METHODS. . . . .. 13
A. Basic Measurement System . . .. 13
B. Low Frequency Measurements . . .. 15
C. Mid Frequency Measurements . . .. 18
D. High Frequency Measurements ... . 20
E. Cryogenic Chamber. . . . . 21
F. Power Supplies and Monitoring Circuits 23

IV. EXPERIMENTAL RESULTS . . . . . 40
A. Parameters of Devices Fabricated . 40
B. Device UFSFB #13 . . . . .. 41
C. Device UFSFB #18 . . . . ... 45
D. Device 2N4416 #3 . . . . .. 48
E. Tabulated Data. . . . . .. 52

V. SUMMARY OF THERMAL AND GENERATION-
RECOMBINATION NOISE THEORY . . .. 104
A. Thermal Noise . . . . . . 104
B. Channel g-r Noise . . . . 110
C. Depletion Layer g-r Noise . . . 120

VI. DISCUSSION OF RESULTS . . . . .. .128
A. Analysis of Data for UFSFB Devices #13
and #18 . . . . . . . 129
B. Analysis of Data for the 2N4416 Device 138











CHAPTER Page

VII. CONCLUSIONS AND RECOMMENDATIONS FOR FURTHER
STUDY . ... . . . . .... . 140
A. Conclusions . . . . . . . 140
B. Recommendations for Further Study . 141

BIBLIOGRAPHY ...... ............ .143

BIOGRAPHICAL SKETCH . . . . . . . ... .145

















LIST OF TABLES


Table No. Page


1 UFSFB #13 Data 53

2 UFSFB #13 Data 54

3 UFSFB #13 Data 55

4 UFSFB #18 Data 57

5 UFSFB #18 Data 58

6 UFSFB #18 Data 59

7 UFSFB #18 Data 60

8 2N4416 #3 Data 62

9 2N4416 #3 Data 63

10 Range of Parameters Slice #3, UFSFB 65
















LIST OF FIGURES


Figure No.

1



2


3

4

5

6

7

8


9


10


11


12

13

14

15

16

17


Page


(a) Microphotograph of Device UFSFB
(b) Cross Section of One Gate Element
Showing Nominal Dimensions . . . 8

Circuit Used to Measure Pinch-Off Voltage -
V . . . . . . . . . 12

Complete Block Diagram Measurement System 24

Low Noise Preamplifier. . . . . . 25

D.U.T. Test Fixture . . . . .. 26

Noise Diode Circuit . . . . . .. 27

Power Supplies for Noise Diode . . ... 28

Preamplifier Equivalent Noise Resistance
Referred to the Input . . . . ... 29

Simplified Test Circuit for Frequencies Below
80KHz . . . . . . . . . 30

Simplified Test Circuit for Frequencies from
50KHz to 1.5MHz . . . . . ... 31

Simplified Test Circuit for Frequencies
Above 2MHz . . . . . . . . 32

Block Diagram of Complete Cryogenic System 33

Cross Section of the Cryogenic Test Chamber 34

Simplified Block Diagram of Master Control Box 35

Device Under Test (D.U.T.) Power Supplies 36

Noise Diode Current Monitor, M1 . . .. 37

D.U.T. Current Monitor Circuit, M2 .... 38


vii










Figure No.

18

19


20

21

22

23

24

25


26

27

28


29

30


Page

Block Diagram of Temperature Control . 39

UFSFB #13 R vs. Frequency, a Function
of I . . . . . . . . 66

UFSFB #13 R vs. Frequency . . .. 67

UFSFB #13 Rn vs. Frequency . . .. 68

UFSFB #13 Rn vs. Frequency . .. . 69

UFSFB #13 R vs. Frequency. . . . 70
n
UFSFB #13 Rn vs. Frequency ...... .71

UFSFB #13 Composite Plot: R vs. T,
a Function of Frequency . .. . . 72

UFSFB #13 ID and gm vs. T ....... 73

UFSFB #13 and #18, T vs. 1000/T .. ... 74

UFSFB #18 R vs. Frequency, a Function
of VDS . . . . . . . . 75

UFSFB #18 Ieq vs. VDS at 3000K and 800K 76

UFSFB #18 R vs. Frequency, a Function
of VDS . . . . . . . .


UFSFB #18 R vs. Frequency . . .

UFSFB #18 R vs. Frequency . . .

UFSFB #18 R vs. Frequency . . .
n
UFSFB #18 R vs. Frequency . . .

UFSFB #18 Rn vs. Frequency . . .

UFSFB #18 R vs. Frequency . . .

UFSFB #18 R vs. Frequency . .

UFSFB #18 Rn vs. Frequency . .

UFSFB #18 Rn vs. Frequency . .

UFSFB #18 R vs. Frequency . . .
n

UFSFB #18 Composite plot: R vs. T,
a Function of Frequency .
a Function of Frequency


S. 78

. 79

S. 80

S. 81

S. 82

. 83

. 84

S. 85

. 86

. 87


. 88


viii










Figure No.

42

43


44


45


46


47

48

49

50

51


52

53

54


55


56


UFSFB #18 ID and gm vs. Temperature .

UFSFB #18 R'(meas.) and R (thermal) vs.
n n
VDS . . . . . ... .
UFSFB #18 R'(meas.) and R (thermal) vs.
T, at 40KHz . . . . . . . .

UFSFB #18 I (thermal) and I (meas.)
vs. Frequencye'o 20MHz . . . .

2N4416 #3 R vs. Frequency at 800K, as
a Function ofnVDS

2N4416 #3 Rn vs. T . .. . . .

2N4416 #3 R vs. Frequency .. . .
n
2N4416 #3 R vs. Frequency . . .

2N4416 #3 R vs. Frequency . . .

2N4416 #3 Composite Plot: R vs. T,
a Function of Frequency . . . . .

2N4416 #3 gm and ID vs. T . ....

2N4416 #3 T vs. 1000/T .. .

2N4416 #3 R'(meas.) and Rn(thermal) vs.
T at 40KHz . . . . . . .

Mobility of Silicon at Cryogenic Tempera-
tures. (From Morin and Maita) . . .

Cross Section of the JFET:
(a) 2-dimensional
(b) 3-dimensional . . . . . .


Page

89


90


91


92


93

94

95

96

97


98

99

100


101


102



103









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



NOISE IN JUNCTION-GATE FIELD-EFFECT
TRANSISTORS AT LOW TEMPERATURES

By

Clifford F. Hiatt

December 1974

Chairman: Aldert van der Ziel
Co-Chairman: Eugene R. Chenette
Major Department: Electrical Engineering

At room temperatures the limiting noise in Junction-

Gate Field-Effect Transistors is the thermal noise in the

channel. However, at cryogenic temperatures the limiting

noise is generation-recombination noise occurring in the

depletion region and in the conducting channel.

Special JFET devices were made in the Microelectronics

Laboratory having very low pinch-off voltages, high trans-

conductance,and low saturated drain current to facilitate

a closer examination of the g-r noise problem.

This dissertation presents the results of a detailed

noise study of these devices and compares the results with

those obtained from measurements on commercially made JFETs.

The results obtained indicate a departure from classical

theory when channel thickness is very small. Further

investigation is recommended and suggestions for further

study are made.
















CHAPTER I

INTRODUCTION



The purpose of this study is to investigate the

mechanisms causing noise in Junction-Gate Field-Effect

Transistors (JFETs). Such a device consists of a

conducting channel in an epitaxial layer, bounded by the

substrate and the junctions formed with a diffused-in

gate layer. The channel thickness can be modulated by

biasing the gate with respect to the channel. Current is

drawn through the channel from source to drain.

Fluctuations in the drain current ID stem from various

causes. First of all, there is thermal noise in the

channel, as for normal conductive materials. The noise of

the individual volume elements of the channel can be de-

scribed with thermal noise generations as given by Nyquist's

formula of 1928. However, the overall noise observed in ID

is more complex because of the spontaneous fluctuations in

channel thickness caused by gate-channel bias fluctuations

induced by the thermal noise. The overall thermal noise

was first computed by van der Ziel in 1962 [1]. His results

were experimentally verified by a number of investigators.

Others have given an alternate form of thermal noise theory

applicable both to JFETs and MOSFETs, compare e.g., Klaassen









and Prins [2]. For a review of these theories, see the mono-

graph by van der Ziel [3].

It was soon recognized, -both theoretically and experi-

mentally, that other noise effects exist. In particular,

since the device consists of semiconductor material, genera-

tion and recombination of carriers should contribute to

current fluctuations (g-r noise). For homogeneous semi-

conducting crystals this noise is well known. It has been

observed by many investigators in germanium and silicon;

for a summary of early experimental work, see van der Ziel's

book on Fluctuation Phenomena in Semiconductors [4]. This

noise has subsequently been observed in a great variety of

materials, such as PbS, CdS, CdHgTe, ZnO, GaAs, etc. The

theory for this effect has well been established, cf. van

Vliet [5] and van Vliet and Fassett [6].

In the early sixties, soon after the thermal noise

theory for JFETs, the occurrence of excess noise due to

carrier transitions, i.e., generation and recombination or

trapping in the channel of JFETs was predicted by van der

Ziel [7]. The effect was made responsible for measurements

of excess noise, showing the typical g-r noise relaxation

spectrum, Si(f) T/(I+2 T2), reported at about the same

time by Halladay and Bruncke [8].

Subsequent analysis indicated, however, that carrier

density fluctuations caused by generation and recombination

in the channel are not easily observable. This is due to

the fact that most donors and acceptors in silicon have









very small ionization energies. Hence, statistical fluctua-

tions of the free carriers are only occurring at very low

temperatures, where "freeze out" of the free carriers (i.e.,

recombination with the ionized impurity sites) starts to

occur.

About a year later, Sah proposed a theory leading to

noise spectra which have similar features as the above

mentioned g-r noise, but stemming from carrier fluctuations

of a different origin [9]. He suggested the occurrence of

trapping fluctuations in the space-charge layer of the gate-

channel junction. Spontaneous fluctuations in the trap

occupancy give rise to a fluctuating space charge, which

thereby modulates the width of the depletion layer and conse-

quently also modulates the thickness of the conducting

channel. Though this noise in essence is still of the g-r

type, the location of the noise source is now in the space

charge layer, and the electronic processes responsible for

the noise are entirely different. Also, the conversion of

these noise causes to the external current fluctuations is

very different. The measurements of Halladay and Bruncke

referred to above, as well as the measurements of others

were now attributed to this "depletion-layer g-r noise,"

cf. van der Ziel [3]. This noise is quite prevalent in

JFETs, and observable over a wide temperature range in

well selected units.

The originally proposed "channel g-r noise" phenomenon

has subsequently been observed by Shoji at the University of










Minnesota [10]. A major problem for observing this noise

is the temperature rise caused by the power dissipation of

the drain current. To eliminate this heating effect, Shoji

made.low duty cycle pulse.measurements at 770K. Even so,

this noise is not easily discernible and certainly not

amenable to detailed study.

The object of the present study of noise in JFET's was

to study in detail the various types of g-r noise processes;

in particular, an attempt has been made to observe the channel

g-r noise described above. To that purpose, special devices

were built in the Microelectronics Laboratory of the University

of Florida, which would permit operation at 770K with much

lower dissipation than commercially available devices. This

was mainly achieved by making devices with a very low pinch-

off voltage. The noise results for these devices are reported

in this thesis and show that large effects of g-r noise at

770K are observable, which presumably stem indeed from channel

fluctuations. The depletion layer noise is also present in

our devices as well as in subsequently measured commercial

devices.

The division of the work described in this dissertation

is as follows. In Chapter II we discuss the device fabrica-

tion techniques and problems encountered in achieving a low

power dissipation, while in Chapter III the experimental

measurement procedures are described.

The experimental results are presented in Chapter IV.

We report in particular on two devices fabricated here,










selected from a large number of devices having different

parameters. Also, for comparison, results are reported

on a Texas Instruments commercial JFET.

A survey of the theory pertaining to thermal noise

as well as to the various g-r noise effects is given in

Chapter V, with special emphasis on the temperature

dependence of the noise and the time constants involved.

Finally, in Chapter VI the experimental results are

discussed and compared with theoretical predictions, the

conclusions are summarized in Chapter VII.

It is hoped that this dissertation will show

abundantly the importance of g-r noise in JFETs, both

for the fundamental study of the electronic processes in

these devices, as well as for the limitations and advantages

of these devices in engineering applications. The experi-

mental data indicate that at all temperaturesthe g-r noise

well exceeds the thermal noise in the frequency range mainly

studied (1Hz-100KHz), so that this noise indeed determines

the device sensitivity limit.















CHAPTER II

DEVICE FABRICATION



The major impetus for fabricating special devices

for this noise study was a desire to study a device having

very low pinch-off voltage, Vpo ; low saturated drain cur-

rent, IDSS; and a reasonably good transconductance, gm

The above parameters would result in a device having low

power dissipation thereby reducing device heating which

often masks the g-r noise being sought.

It is known from JFET theory that:

ZNDd
g c D (2.1)
-m L

where Z = sum total of gate widths,

ND = channel doping,

d = channel thickness,

L = channel length,

and
IDSS c gm (2.2)

V cN d2 (2.3)
po D

It is immediately evident that there are several options

available which will reduce V and IDSS. The one method

most quickly ruled out involves increasing L. To do this

would increase the channel resistance resulting in.increased










thermal noise. Thus L must be kept small. Therefore,

referring to Eqs. (2.1), (2.2) and (2.3), two solutions

become apparent:

(a) Reduce d: The result is a linear reduction of
gm and IDSS' and a squared reduction
of V
po

(b) Reduce ND: The result in this case is a linear
reduction of all three parameters.

Since we require a good gm together with a low Vpo,

it was decided to reduce d in the devices to be fabricated.

Increasing Z was not considered due to the detrimental effect

of increased capacitance of the metal overlay, and the lack

of any effect on V
po
The geometry chosen for the devices closely resembles

the Texas Instruments SFB8558. This device has a multiple

gate geometry with interleaved source-drain contact windows.

The production SFB8558 has a gm of approximately 20-30 mmho.

Figure la is a microphotograph of the UFSFB device. Figure

lb is a simplified sketch of the cross section. The Z/L

ratio is approximately 240:1 which yields a very high gm'

assuming a reasonable value of ND is used.

Commercially produced wafers were obtained which had

approximately the same parameters as the wafers used on the

SFB8558. These parameters are

1. N/P Epitaxy, (111) lattice orientation

2. n epi: 0.45-0.49 Q-cm.
1.8-2.1 microns thick

p substrate: 0.0048-0.0059 P-cm.




























































Figure 1 (a) Microphotograph of Device UFSFB.
(b) Cross Section of One Gate Element
Showing Nominal Dimensions.








Ruby-lith mask masters were cut and photographically

reduced using a 3-step reduction technique resulting in a

500:1 reduction of the original master. This 3-step

reduction was found necessary in order to take advantage

of the higher quality of the lens center area, a require-

ment when fine line resolution is desired. The first and

second reductions were done using a Henry Mann type 1003

camera. The final or third reduction was done using a

Henry Mann type 1080 step-repeat camera which produced the

contact mask for each of the five photo-resist steps in

the device fabrication schedule.

A computer program was written to calculate diffusion

depth versus time as a function of furnace temperature.

These profiles proved to be fairly accurate. However, a

similar treatment of the up-diffusion from the substrate

did not yield usable data. Since the substrate was quite

heavily doped compared to the epitaxial layer, the up-

diffusion progressed at a rapid rate, except at very low

temperatures. A wafer was then cut up and subjected to

different furnace temperatures, with oxide thickness, sur-

face doping concentrations, and furnace time as parameters.

The resulting data yielded a satisfactory profile of both

down-and up-diffusion as a function of several parameters.

Groove and stain measurements were made on all sample runs

to measure both up-and down-diffusion.

Many departures from the University of Florida Micro-

electronics Laboratory standard procedures were required in









order to produce usable devices. Thinner oxides were used

to obtain better line resolution, and subsequently, thinner

photo-resist was used to further improve this resolution.

Ultimately, consistently clean, sharp 2.5 micron windows

were obtained using photo-reduced masks and photo-resist

coatings.

The goal was set for a device with a Vo of approxi-

mately 0.3 volts. After initial diffusion, the Vpo was

measured using a test pattern on the lower corner of the

device area. Subsequent furnace drive-in steps were then

used to reduce the Vo to the desired value. A test chip

which was carried along with the wafer was then grooved and

stained in order to measure the channel thickness; however,

the optical equipment considered accurate for 0.3 micron

layers was inadequate to measure the resulting thin channel.

A rough estimate would place the channel thickness in the

range of 0.1 micron.

The finished wafer was then probed using micromanipu-

lator probes with leads to a curve tracer and the parameters,

Vpo IDSS' and gm, were measured at different points on the
wafer. The results indicated a very large variation of these

parameters; also, many devices on the wafer were shorted out.

Parameter Range Predominant

IDSS 28pA. to 6mA. 500pA.

V 0.2V to 0.8V. 0.3V.
po
gm 180pmho to 24mmho 6mmho










It is believed that the very wide variations of these para-

meters could be caused by several factors:

1. Uneven silicon substrate surface prior to
epitaxy.

2. Uneven thickness of the epitaxial layer.

3. Nonuniformity in the diffusion process,
both up-diffusion from the substrate and
down-diffusion from the gate area.

It therefore appears obvious that the channel thickness

d varies considerably over the entire wafer. All commercial

devices exhibit some variation in parameters but since the

channel is considerably thicker, the variations in thickness

have less effect on the device parameters. It is possible

that a more workable compromise would be a combination of

thinning down the channel thickness (d) less than before,

and decreasing the channel doping (ND) slightly. This should

permit working with thicker channels as well as obtaining a

lower V .
po
Figure 2 illustrates the method used to accurately mea-

sure V The meter used to measure VGS was a high input
po
impedance Digitec digital voltmeter and Vpo is read directly.

These devices were fabricated in the University of

Florida Microelectronics Laboratory.













DEPLETION REGIONS


limit


Figure 2 Circuit Used to Measure Pinch-Off Voltage V .


V
S






V
po















CHAPTER III

EXPERIMENTAL METHODS



The goal in this study was to investigate the generation-

recombination (g-r) noise mechanism in JFETs as a function

of frequency and temperature. In order to accomplish these

studies, a special measurement system was designed and

built. This chapter will discuss various portions of the

system and explain the measurement techniques employed.



A. Basic Measurement System

Referring to Figure 3, the basic measurement system

consists of all those components to the left of the dotted

line. This section of the system remained the same for all

measurements regardless of temperature or frequency. The

Device Under Test (D.U.T.) is placed on the end of a probe

and inserted into the cryogenic chamber (to be discussed

later). At the top of the cryogenic chamber rests a box

which contains the low-noise preamplifier (Figure 4), the

test fixture for the D.U.T. (Figure 5), the R.F. tuning

assembly (Figure 5), the noise diode and associated filters

(Figure 6), and the 40db fixed attenuator (Figure 5). These

sections are all contained in a copper box with each section

shielded from the others by copper panels. All power to the










unit shares a common return bus which is shielded to ground.

The voltages necessary for operation of the preamplifier,

D.U.T., and noise diode are batteries located in a shielded

steel cabinet, except for the noise diode high-voltage which

is obtained from a separate high-voltage power supply. Fil-

ter capacitors are used extensively in order to reduce low

frequency pickup from the AC operated equipment. Small

value (0.1pF) capacitors are used extensively to eliminate

RF feedback paths.

Heater current for the D.U.T. heater is obtained through

a constant current control (Figure 7) from two 12-volt auto-

mobile batteries. D.U.T. temperature is monitored by

attaching a thermocouple junction directly to the device

header, near the base. The thermocouple leads are compensated

using an Omega thermocouple reference junction, the output

of which is measured using a Leeds and Northrup Potentio-

meter Bridge.

The D.U.T. current for each element may be monitored on

the Master Panel along with the noise diode plate current.

Notice in Figure 5 that the D.U.T. may be operated in any

mode: common source, common gate, or common drain. In

addition, the test circuit will accommodate n-channel or p-

channel FETs and PNP or NPN bipolar transistors. This is

accomplished by using plug-in bias and load elements which

may be changed as desired.

The output from the D.U.T. is fed to the shielded low-

noise preamplifier. This preamplifier has a gain of approx-









imately 300 and has an equivalent noise resistance (R )

referred to the input as shown in Figure 8. The input

impedance is 200 Kohms and the output impedance is 30 ohms.


B. Low Frequency Measurements

When measuring the noise in the range from 3.15 Hz to

80 KHz, the output of the preamplifier is passed through a

600 ohm attenuator to a selectable gain amplifier which

further amplifies the noise signal to a level suitable for

processing by the G-R 1925-1926 Real Time Spectrum Analyzer.

The G-R Real Time Spectrum Analyzer contains a set of

45 third-octave filters, ranging in center frequencies from

3.15 Hz to 80 KHz. The output of each filter is sampled for

32 seconds and the db level of the RMS voltage of each fil-

ter is computed and displayed on the G-R 1926 Display Scope

and simultaneously printed out by an MDS800 tape printer.

A G-R 1381 Noise Generator followed by a 600 ohm attenuator
2
provide the noise calibration signal, ecal

The measurements are performed as follows: The tempera-

ture and proper terminal voltages of the D.U.T. are first

established and allowed to stabilize. [Temperature control

will be discussed in a following section.] The attenuator

(#1), Figure 9, following the noise generator, is turned to

maximum attenuation and the resultant noise observed on an

oscilloscope and on the G-R Display Scope using an 0.25

second integration time. The gain of the selectable-gain

amplifier together with a 600 ohm attenuator allow adjustment










of the noise display to a level in the lower portion of the

G-R Multifilter Dynamic Range. Once the levels are set, the

G-R Analyzer is put in the 32-second integration and print-

out mode of operation. Three successive integration are

performed and printed on paper tape. The integration is then

stopped and the G-R Noise Generator Attenuator (#1) is adjusted

to give a level approximately 20 db above the original noise

signal. A second set of three 32-second integration are

recorded. The two different sets of three integration are

averaged and the noise signal (dbl) is subtracted from the

noise plus calibrated noise signal (db2) for each channel,

giving

db2 dbI = Adb (3.1)

Now we define:
volts
e = noise voltage of D.U.T. in v
n fHz

volts
e a = calibration noise source in vots
cal /Hz

We can now equate:

V = en (3.2)

V2 = eca + e (3.3)

and since

db = 20 log (3.4)


we can write
Adb
e + e (2)
en+ = 10 (3.5)
e
n










and thus

2 2 Adb
e + e
n cal 10
n= 1 10 (3.6)
2
e
n

which can be written as

2
2 cal
e cal (3.7)
n db
10 10 1

Since the output of the noise generator has been calibrated

and plotted as

2 e
S volts ng (3.8)
V Hz Af

where Af is the bandwidth of a particular filter channel.
2
We cannot, however, use this value for ecal since we have

inserted two attenuators between the noise generator output

and the D.U.T. input. We add the value of attenuator (#1)

to the value of attenuator (#2) and call this ATT(db). Our

expression for e now becomes
cal

SSV Af
2 V
e S (3.9)
cal ATT(db) -(
L10 10 J

Therefore for each series of measurements we must record the

value of attenuator (#1), and add it to our 40 db fixed

attenuator (#2). Since we are interested in equivalent noise

resistance we may substitute in Eq. (3.7) Eq. (3.9) and

e = 4kT R Af (3.10)
n n









The complete equation is thus


Rn(eq) = ATT(db) Adb (3.11)
10 i0 1

The noise inherent in the following stage is small and

since the D.U.T. is a gain device, the preamplifier noise

becomes negligible. Referred to the input of the D.U.T.,

the preamplifier noise is
2
-- e preampp)
e preampp) = n (3.12)
n (gm RL)
[at input of DUT]

this value being only a few ohms for the worst case where

the gm of the D.U.T. is low. The resultant values of R are

plotted versus frequency.


C. Mid-Frequency Measurements

For the spectra between 50 KHz and 1.5 MHz a different

and simpler technique is employed. Figure 10 is a simplified

diagram of the test circuit employed.

The G-R Analyzer is replaced by an HP-310A Wave Analyzer.

An attenuator is placed between the preamplifier output and

the analyzer input to facilitate careful adjustment of the

noise signal amplitude thus obtaining a convenient reading on

the analyzer meter.

The HP-310A consists of a tuned BFO which tracks with

a tuned input circuit. The input signal is passed through

an input attenuator, a tuned circuit with a selectable band-

width, and finally to a square-law detector giving a reading










calibrated in RMS volts or db. The HP-310 BFO output may

be used as a signal source as shown in Figure 10; however,

a much simpler method is normally used to supply the

calibrating signal. The method employed is as follows:

Since we wish to observe the noise in the channel, the

gate contact is short-circuited and the device allowed to

pass current, the magnitude of this current being a function

of VDS until we reach Vo where IDSS remains essentially
2
constant. The calibrating signal Ial is injected at point
cal
(B) from the noise diode. If system gain is low, RF tuning

may be employed by switching in the tuning network which

contains tuned circuits for selected measurement frequencies.

However, in this case, the gain of the system is essentially

flat up to 3 MHz and tuning is not employed.

When the D.U.T. temperature and operating voltage and

current have stabilized, the attenuators are set to obtain

a (-3db) reading on the analyzer RMS meter. The noise diode

is then turned on and its heater current adjusted until

the Analyzer RMS meter reads (0db). The plate current of

the diode (monitored on the master panel) is now equal to

I of the D.U.T. Noise resistance Rn may then be calculated
eq n
if desired or I may be plotted directly.

The measurements described above are taken as a function

of frequency and VDS starting at a value much less than Vp

and progressing through Vpo up to a value short of breakdown

using selected spot frequencies within the range 50 KHz to

1.5 MHz.









D. High Frequency Measurements

The technique employed for frequencies above 2 MHz is

essentially the same as that described above for the mid-

frequency range. The main difference being that the HP-310A

Wave Analyzer is replaced by a Collins 51S-1 receiver as

seen in Figure 11. The noise signal from the preamplifier

output is again passed through an attenuator and then

injected into the antenna terminals of a modified Collins

51S-1 receiver. The modification consists of disabling the

receiver AGC circuits. Receiver RF gain can be adjusted

manually on the front panel; however, the IF stage now has

a fixed gain. The receiver IF output is then fed to an

HP-3400A True RMS meter.

In operation, the D.U.T. temperature and voltages are

set and allowed to stabilize. The RF tuning section is

switched in and set to the spot frequency of interest and

the receiver is tuned for peak amplitude of the signal at

that frequency. Attenuators and IF gain controls are then

adjusted to give a convenient (-3db) reading on the True

RMS meter. The noise diode current is again increased until

the meter reads (Odb). The plate current as monitored on

the main panel is now equal to Ie of the D.U.T. noise as

before. These readings are also taken as a function of T,

freq., and VDS.

It is essential that all impedances be properly matched

and careful attention be given to a good grounding system.









E. Cryogenic Chamber

In experiments measuring low levels of noise at

cryogenic temperatures, it is.very important that accurate

and stable temperature control be available. For this

reason the following description of the cryo-system used

for these measurements is presented.

A block diagram of the complete cryo-system is shown

in Figure 12. The cryo-chamber itself is shown in detail

in Figure 13. Referring to Figure 12, N2 gas is used to

pressurize the liquid N2 Dewar to 5 psi thus enabling

transfer of liquid N2 to the chamber when needed. The

level of liquid N2 is sensed in the inner chamber (Figure

13) by a level sensor element which signals the opening

and closing of the liquid N2 supply solenoid. As the liquid

N2 turns to gas (boils off), it is vented through a liquid

trap which has provisions for attachment of a vacuum pump

should it be desired to pump down the liquid N2 in the inner

chamber in order to obtain temperatures a few degrees below

liquid N2 (770K).

The outer container of the cryo-chamber is pumped down

by a vacuum pump to reduce thermal conduction between the

room ambient and the inner chamber temperatures. Not shown

on Figure 13 are several layers of insulation around the out-

side of the outer container, inside of the outer container,

and surrounding the inner container and sample chamber below.

A heater consisting of four 25-watt resistors is placed around

the D.U.T. chamber to facilitate raising the D.U.T. tempera-









ture as desired. An additional feature is the He gas pre-

cooler used as a fine temperature control.

In operation, the inner container is filled to the

set level with liquid N2; as the tests progress to higher

temperatures, the heater is used to raise the temperature

to the desired point, slightly more heater current being

used than would be required in a steady state condition.

As the D.U.T. temperature reaches the desired setting, the

flow valve controlling the He gas to the precooler is ad-

justed to maintain the set temperature. Once stabilized,

very fine control of temperature is possible by using the

flow meter valve. The voltage to the drain of the D.U.T.

is applied through a 2 Kohm load resistor, and in satura-

tion, any change in device temperature causes a rapid

change in VDS. This VDS is constantly monitored and good

stabilization of temperature is possible by adjusting the

gas flow. The main advantage of this system is the rapidity

by which a new temperature may be reached and the stabilized

condition obtained.

As seen in Figure 13, the D.U.T. rests on a saphire

chip which gives very good electrical insulation while

maintaining good thermal conductivity. The D.U.T. lead

wires and the thermocouple wires are brought down the center

tube to the D.U.T. chamber in a grooved, segmented plastic

holder assembly which reduces thermal loss up the tube to

a minimum. A short piece of #33 wire is inserted in each

D.U.T. lead wire as a further step toward reducing thermal

conduction from the chamber.










This system was used for all low temperature measure-

ments and spot checks on measurement repeatability were

excellent.



F. Power Supplies and Monitoring Circuits

The power supplies, control, and monitoring circuits

for all parts of the test system are shown in Figs. 14

through 18. The flexibility of the complete measurement

system has not been fully utilized in the measurements

performed for this work. It is obvious that many different

arrangements of test and types of devices could be studied

using this system.






2 1500
e ________________--_--------------
ecatl ATT. i
r"^ _--I- ---
9 Q 600o .2 -
cal 50

ATT ATT.
- - -R F .
s ..
TUNIN h
.U.T PREA i

BIAS









BCHASER I i B ERS
ATTEND
.T









CONTROL NOIS DODE
S MASTER CONTROL, :






CRYOGENIC
S3 H___________________ .U. POWER



Figure 3 Complete BlockDiagram Measurement System.






+18V


+12V


OK2


681


I


0.1
VF


25pF

SMEAS.
SYSTEM


= SFB8558
= 2N4416
= SFB8558
t%
U,


Figure 4 Low Noise Preamplifier.


100o
Vw.w
w.w.


.6K w.w.


D.U.T.


Q1,Q2

Q2,Q4
Q5






VDS


6000 --
2 6 ,
6___ O TUNED LOAD
e2 ____ -- - a -- - - --
cal
49.95Q 1
50(
I 0.5 A = e2 from Noise Diode I
I cal

L B cal from Noise Diode


SILDSHIELD
j -IELD- I

SHIELD =


Figure 5 D.U.T. Test Fixture.













200V -dc



Large
RFC



0.1 100

R2 pF

0.11F To Points
------- ---- A or B on
D.U.T. Board


5722


10o-- 0. 0- A .1 1..-10
pF*T I RFC RFC PF -1




To Filaments
Current Control


Figure 6 Noise Diode Circuit.



















DIODE


lPl.i, -


PLATE
CIRCUI'

de) COMMON
7 ------- gs


2KQ
+...o----.. --...... ...
2mh.
+ 100
200V dc T F
~T


1001 _0.1

T


I,- - -


02


RF = Current Limit


0 =


0.7
[RF + Rcoarse + Rfine
F coarse fine


Figure 7 Power Supplies for Noise Diode.


I -__.


4700


12V dc


DIODES


R
coarse


Fine


- -- -


--


- --- - --- -





a g----B~


r












R n()


Freq (Hz)


102


Figure 8 Preamplifier Equivalent Noise Resistance Referred to the Input.


102










10






















GENERATOR


Figure 9


Simplified Test Circuit for Frequencies Below 80KHz.


2
ecal
R n


(4kT RnAf)/2


I -_









.2
VDD ical
FROM
NOISE
DIODE

RFC
HP 310
ANALYZE
lpF IpF
-- -PREAM --ATT. >0
D.U.T. i --IN


>200K


TUNED
2
CIRCUIT ca i
cal
S. e -- ATT.


502







Figure 10 Simplified Test Circuit for Frequencies from 50KHz to 1.5MHz.











2
cal
NOISE DIODE


Figure 11 Simplified Test Circuit for Frequencies Above 2MHz.


VDD









He GAS PRECOOLER


SOLENOID
VALVE
N2

GAS (5psi)

RUBBER CORK SEAL


20 LITER
II DEWAR


k INSULATING CHAMBER
CHAMBER VACUUM PUMP
VACUUM


Figure 12 Block Diagram of Complete Cryogenic System.











LEADS TO D.U.T. BIAS BOARD


TO HEATER
POWER


TO
VACUUM
PUMP


SEAL PLUG


LIQUID
N2


HEATER RESISTORS
(4-25 watt)


COPPER BOX
CONTAINING
D.U.T.
CIRCUITS


VACUUM
FLANGE


FLANGE


- OUTER
ISULATIl
VACUUM
VESSEL


RUBBER
VENT


HEAT SINK


THERMO-COUPLE
JUNCTION







- 12V dc

S18V dc


PREAMP
BATTERIES


VDD

VSS

VGG


TO
PREAMPLIFIER


TO
D.U.T.
TEST
CIRCUIT


FILAMENT


NOISE DIODE
CIRCUIT BOARD


Figure 14 Simplified Block Diagram of Master Control Box.


BATTERY
12V dc


200V dc




D.U.T.
POWER

--f---.- C-
1.5VJf
each .,-


VGG

PANEL JACK


6V each


Figure 15 Device Under Test (D.U.T.) Power Supplies.





NOISE DIODE

H.V. SUPPLY


VGG


METER FUNCTION


Figure 16 Noise Diode Current Monitor, MI.


VND OUT
VND








DD ---






SS^
Vss ~ -












CURRENT
POLARITY,.
____-----y(

+

M2 0-10
2 1


METER FUNCTION
SWITCH


I


SWITCH


Rshunt (1)


Figure 17 D.U.T. Current Monitor Circuit, M2.


VDD OUT


VSS OUT


METER




























THERMOCOUPLE


T-'SET'
REFERENCE
CONTROL


DIFFERENTIAL


C AMP, OVER
,ED ON TEMPERATURE
VERSS -- UNDER


LIGHT EMITTING
DIODES


AMPLIFIER


AMPLIFIER


TEMPERATURE
'SET-READ' DIGITAL
AMPLIFIER VOLT-METER


DRIVER
AND
PASS
TRANSISTOR


POWER
SUPPLY
+5V, 15V


Figure 18 Block Diagram of Temperature Control.


HEATER
















CHAPTER IV

EXPERIMENTAL RESULTS



A. Parameters of Devices Fabricated

The devices used for these measurements were taken

from slice #3. Table 10 lists several of the devices that

were bonded to headers along with the measured room tempera-

ture parameters. Note that the parameters IDSS and gm vary

over a large range while changes in Vp are not as drastic.

Devices #13 and #18 were selected as test devices since

they were representative of two different ranges in device

parameters.

The room temperature spectra of several other devices

were measured on the G-R Spectrum Analyzer and all exhibited

the same g-r type noise curve which is seen in devices #13

and #18 at 3000K. One device, #33, was measured down to

1500K where freeze-out occurred. This problem was present

in several other devices when subjected to a liquid N2 test

using a curve tracer to check for transistor operation.

Devices #13 and #18 did not show freeze-out at 770K and were

retained as primary test devices. Measurements of the noise

spectra were made for each 100K change in T; however, only

for device #18 all curves are presented. For devices #13 and

2N4416 #3, curves are presented for each 200K change in T.










The noise curves presented in the figures were plotted

from R calculations based on T = 3000K. This resulted in
n
simplifying the calculation of the very large amount of data

being processed. Where absolute values (corrected for D.U.T.

temperature) were needed, the term R' was used to designate
n
temperature corrected data. The correction is applied as

follows:

R'(meas.) = 3 R (meas.) (4.1)
n T n



B. Device UFSFB #13


1. Measurements of R vs. Freq as a Function of I at
Room Temperature.

Figure 19 shows the behavior of the g-r noise curve as

a function of ID at 294K. It is apparent that there is

very little, if any, dependence on current of the corner

frequency. However, the overall noise curve amplitude reduces

with increasing ID.


2. Low Temperature Behavior of Device #13

Considering the range of temperatures from 800K to 1200K

as seen on Figure 20, the noise resistance of #13 is very high

( 107 ohms) and does not exhibit any clear g-r noise bumps.

There is a suggestion of a g-r corner frequency on the 1200K

curve at about 5Hz. This slight bump appears again on the

1300K curve, but does not show up again. Therefore, these

two bumps are questionable. The next indication of a g-r bump

occurs on Figure 21, the 1600K curve. This g-r bump increases










in frequency and becomes more dominant in Figure 22 (the

2000K and the 2200K curves). Note that as T increases, the

flat band noise, that is, the noise in the frequency range

from 1KHz up to 80KHz, decreases steadily until the second

g-r bump begins to affect the mid frequencies at about 2600K

(Figure 23).


3. High Temperature Behavior of Device #13

In the range of temperatures from 2600K to 3600K the

major g-r noise effect becomes dominant. The corner

frequency of this g-r effect increases continuously up to

the 3600K curve. As the corner frequency increases, the

amplitude of the plateau (frequencies below the g-r corner

frequency) reduces and at 3200K (Figure 24), there appears

to be the beginning of a 1/fn curve at low frequencies.

This 1/fn curve is very dominant at 3600; however, the g-r

bump has not completely disappeared.

A compositeplot of R vs. T for several frequencies

is presented in Figure 25. The major g-r peaks and the

effects due to temperature can easily be seen.


4. Behavior of ID and g

It can be seen from Figure 26 that gm and ID decrease

quite rapidly as a function of decreasing temperature.

There also appear on the ID curve, three or four plateaus.

In particular, there is a long,low temperature,constant ID

region (1200K-1700K) which may be responsible for the

specific low temperature g-r processes seen on Figures 15










through 19. The gm curve does not show the expected rise,

followed by a drop as temperature is decreased, usually seen

in JFET's.


5. Time Constant T

The time constant as used in connection with the measured

data is obtained from

T (4.2)
T2rfT

where fT is the frequency on the g-r noise curve which is 3db

down from the g-r noise plateau. The values for the frequency

fT' noise resistance Rn(corner), and noise resistance

R (flat) are obtained by plotting the function


F = -1 (4.3)
1 + x

on graph paper and making a plastic template with the f corner

point marked with a vertical line. This curve thus matches

closely the g-r noise modification to any noise curve since

I (g-r) c K T (4.4)
eq I+W2T

Accurate reading of the curves is now possible leading

to a determination of T and hence to the activation energy E

of the process involved. Note that all curves presented in

the figures are marked by a vertical line giving the corner

frequency from which T is determined. This line is numbered.

This method of marking the T in each curve helps increase

the certainty and accuracy of a suspected g-r bump.









It should be noted that for the curves in Figures

20 through 24, this corner frequency (or T) changes with

temperature. As mentioned in section B.1., T is affected

very weakly by changes in ID. A more accurate reading of

the curves on Figure 19 might eliminate this very small

change.

If we plot the various T's (log scale) vs. 1000/T

(linear scale) as shown in Figure 27, we can find the

activation energy of the centers of interest from the slope

of the resulting line. This is accomplished in the

following manner:

a) Plot T on the vertical log scale, and 1000/T on the

horizontal linear scale.

b) Select some convenient number of decades between T1
2
and T2, say 102

c) Mark off on the horizontal line the distance 1000/T1

and 1000/T2 which correspond with T1 and T2.

d) The activation energy is now calculated as follows:

T is defined as

qEA/kT
T = e (4.5)

thus,
qEA/kT1 qEA/kT2
Ti = Te = Te (4.6)
o 2 o

If we divide T1 by :2, we obtain


= exp[qEA/k] ] (4.7)
T A T1 2









Take the logarithm base 10 and convert right side from

Zn to log10;


l 0 = 0.43 -- A E0 (4.8)
1 10k L 1 T2 k

then substituting,

log = 4.99E r1000 1000
Tlog-499E(4.9)

then;

logl0 C / 1 ]
E = 1 -1 (4,10)
A 4.9 1000 1o000
4.99 U T2
I J- T2

The activation energies of all measured T vs. T were

calculated in this manner.

For device #13, we find

slope #1, EA = 0.17eV

slope #2, EA = 0.34eV

slope #3, EA = 0.6eV



C. Device UFSFB #18

1. Measurement of R with V as a Parameter
----- 'DS-
In Figure 28 we see a family of curves taken at 3000K

with VDS as the changing parameter. Note that we can

immediately make three observations:

a) The flat top of the g-r curve rises and reaches a

maximum at 0.6V.

b) The corner frequency remains fairly constant.










c) The higher frequency noise at 40KHz first drops rapidly

and reaches a low at a VDS of 1.04V.

In Figure 29 can be seen a log-log plot of Ieq vs. VDS at

3000K, and 800K for f = 40KHz. The resultant points on

the 3000K plot have a slope of 5/2 which is expected

according to the theory. The 800K plot probably does not

go down to low enough VDS values (experimentally lower ID

measurements were not possible).


2. Low Temperature Behavior of Device #18

Figure 30 consists of data curves taken at 800K.

Again we plot R vs. Freq as a function of VDS. Note that

the high frequency flat portion of the curve changes very

little with VDS; however, the 1/fn tail at low frequencies

changes slope with changes in VDS. At VDS = 0.8V (800K

curve) we can see the beginning of a g-r effect at about

7.5Hz. This observed turnover is the beginning of the g-r

process marked #1 on Figure 31. The turnover frequency of

each g-r process may be observed by reading Figures 31

through 37. In Figure 33, g-r effect #1 is moving toward

the high frequency end of the curve and g-r effect #2 is

just beginning to appear at about 3.5Hz. The g-r effect #2

turnover frequency progresses upward until it leaves the

curve on Figure 37.


3. High Temperature Behavior of Device #18

Referring again to Figure 37, a new g-r effect #3 can

be seen at the low frequency end of the spectrum. This g-r











#3 turnover frequency moves upward as T increases to 3700K

where the measurements were stopped (Figure 40). Note on

Figure 40 (the 370 curve) that the g-r effect is still

present at about 3KHz, and the 1/fn slope is present below

40Hz.

A composite plot of R vs. T for several frequencies

is presented in Figure 41. The major g-r noise peaks and

the effects of temperature can be easily seen.


4. Behavior of ID and

Figure 42 is a plot showing the effects of temperature

on ID and g As was the case for device #13, gm of device

#18 does not rise, then fall as temperature is lowered as

is usually observed in commercial devices.


5. Time Constant T

The same comments given in section B.5. for device #13

apply to device #18. The high temperature slope (#3) is

identical with that of device #13 (Figure 27). A departure

is noted between devices for slope #2 and slope #1. The

slight departure of slope #2 could be due to errors in reading

the curves. There is also a small but insignificant difference

between #13 and #18 for slope #1.

The resulting activation energies for device #18 are

slope #1, EA = 0.19eV

slope #2, EA = 0.36eV

slope #3, EA = 0.6eV










6. Comparison of Measured Noise to Calculated Thermal Noise

A comparison between R'(meas.) and calculated R (thermal)
n n
is seen in Figure 43. These data were taken at a device power

dissipation ranging from 0.24p watts (0.2V VDS) to a maximum

power dissipation of 122p watts. It is easily seen that

device heating is not a factor in the magnitude of R'(meas.).
n
The difference between these curves is the excess (g-r) noise.

The curves plotted in Figure 44 verify that R (meas.) is

n
always higher than Rn(thermal) at all temperatures. This

difference is about an order of magnitude at 800K, which is

verified by Figure 45 which shows I (meas.) and Ieq(thermal)

vs. frequency up to 20MHz. There was no detectable deviation

in the I noise up to 20MHz where stray capacitances became

too high to permit circuit tuning. The time constant involved

in this spectrum is apparently too short to show up even at

20MHz.



D. Device 2N4416 #3

Device 2N4416 #3 is a commercially available JFET manu-

factured by Texas Instruments. The geometry is different and

the device is smaller than that of the UFSFB, resulting in a

lower average gm. However, the silicon slice used to process

both devices is very nearly the same. Major differences in

process procedures and channel thickness would be expected

to account for some unusual differences in performance.

The results of the noise measurements of the 2N4416 are there-

fore presented for comparison, since the 2N4416 is a widely

accepted low noise JFET.










1. Measurements of R with VDS as a Parameter
n --DS
The noise performance at 800K as a function of VDS is

presented in Figure 46. In these curves the g-r noise

corner frequency moves upward in frequency as VDS is in-

creased. This change of fco r is not due to temperature
corner
but is probably a result of device heating. A further

explanation of this problem will be discussed in Chapter VI.

Figure 46 also shows that R (flat) and R (40KHz) decrease

with increasing VDS. The curve for VDS = 4V clearly shows

the appearance of a 1/fn slope at the lower frequencies.

This 1/fn slope does not occur at VDS below 3 volts. The

pinch-off voltage (Vp) of this device is very close to 3

volts, thus the 1/fn slope only shows up above Vp. Since

device heating is probably occurring in the device (even

at VDS = 1 to 2 volts) it is doubtful that any immediate

conclusion can be drawn regarding the correlation of 1/fn

noise and Vp (e.g., these could be heat transfer fluctua-

tions).


2. Low Temperature Behavior of 2N4416 #3

The turnover frequency, corner' at 800K and VDS = 3V

will be considered as the first indication for a g-r

process labeled as process #2 (see Figure 47). The same

g-r process is noticeable in the 1000K curve, also given

in Figure 47. However, a large change occurs in Rn(flat)

from 800K to 1000K. At 1000K there is a strong 1/fn slope

at the lower frequencies, and the g-r corner has moved
corner Id-










upward in frequency to 40KHz. The 1100K curve (not shown)

shows no evidence anymore of g-r process #2.

A new g-r effect appears on the 1200K curve (Figure 47).

The f of this effect (labeled #3) continues very
corner
clearly up to 1900K, see Figure 48. The curves from 2000K

through 2400K (Figure 49) show a questionable suggestion

of the g-r bump #3. These last few points could be neglected

with very little, if any, sacrifice of accuracy in the T vs.

1000/T plot.


3. High Temperature Behavior of 2N4416 #3

Beginning with Figure 49 and moving up in temperature

to 3000K (Figure 50) a very definite 1/fn spectrum begins to

be dominant at lower frequencies. At 3000K this spectrum is

exactly 1/f, going over into white noise at 5KHz. No

significant data were expected above 3000K and the measure-

ments on 2N4416 #3 ended there.

A composite plot of R vs. T for several frequencies

is presented in Figure 51. The major g-r noise peaks are

readily visible as a function of temperature.


4. Behavior of ID and g

Figure 52 is a plot showing the effects of temperature

on ID and gm for the 2N4416. It is seen that the ID and gm

curves are fairly closely matched until T goes below 1300K,

at which point the gm curve slopes downward while the ID

curve goes up and levels off. Note that the maximum gm is

almost twice the gm at room temperature, and this maximum










occurs at about 1300K. This behavior is entirely different

from that of the thin channel devices discussed before.


5. Time Constant T

In Figure 53 the time constants were plotted (T vs.

1000/T) in the same manner as those of devices UFSFB #13

and #18. Note that slope #1 is vertical. This is due to

the change of corner and consequently of T when VDS (and

consequently the power dissipation) is increased. This is

possibly a device heating effect mentioned before. Slope #2

was plotted for T obtained from the 800K, 900K, and 1000K

curves. Some inaccuracy is expected in the slope of this

curve due to device heating.

Slope #3 is the slope found if the questionable points

above 1900K are included. Slope #3A is the slope resulting

if the questionable points are deleted. There is a 0.03eV

difference between the two plots.

The resulting activation energies for device 2N4416 #3

are

slope #1, (not applicable)

slope #2, EA = 0.15eV (subject to error)

slope #3, EA = 0.18eV

slope #3A, EA = 0.15 eV


6. Comparison of Measured Noise to Calculated Thermal Noise

A comparison between R'(meas.) and R (thermal) at 40KHz
n n
can be seen in Figure 54. Note the large value for Rn(meas.)

at low tLemperatures. This excess noise drops rapidly to a









low at 1700K where a bump in R'(meas.) is seen. This corre-
n
sponds to the second g-r effect noted from the individual

curves. It appears then that the questionable g-r process #3

on the curves from 2000 to 2400 does indeed exist.



E. Tabulated Data

The operating parameters, measured data, and calculated

data for each device are listed in Tables 1 through 10. The

data from these tables were used to generate many of the

curves showing comparative measurements. The values for the

mobility were taken from Morin and Maita's paper [12].

However, since the devices (#13 and #18) which were studied

in this work have very thin channels, it is possible that

the mobility data given in Figure 55 may be in error by up

to an order of magnitude. To our knowledge, no mobility

measurements have been made to date on very thin channels.














N
aHw
U> :
H





H.-.
rso

Su
^ ~


o N 00c)oJ 0 Q


o 0000


00in in

000


inL O I

000


;:L ;: 17 ;n.: ;::
in in in 0 Mn
L: ;I-_ ;. I. ; .

r-MNrm) Mm m)m m --:r





LnCmm4 oren H Co44 Nr


5 E
oin
*





Ln Ln

000


NoC in C n mm
oEM aE
He n nMm
* * **r~


0lo 1o
000


i0 Ln 0
Q0Ot.0 kD
ooo


3- a
Mr- olmz OMO
~OnO(OON
HHrHNCN


M0 Mcii


* U *in*
N tlo C) IZ

ItT It, Lr)


Ln Q6 co
mem)L


0000000
C; o ; ooo 0 o


gEs






in n in

000


e E

in n onM co


o in inH r- n
o mn ""l r-l r- H 00 0

44 44 144 r r


r- *T -1
CYO M M
C; o C


CO en
00 0o Y
000000
0 0


00000000000000000000000000000
cMO O0 r4 HN m I lr M o Nr- co w M 0 H lN mIqTn in o r- M 0 rH N en r
H H H MM H H H HH N N N N N N N N N N m M m


Sr r-I 111 11111111111111111 t I I I I I -
ooo00000000000000000000000000

XXXXXXXXXXXXXXXXXXXXXXXXXXXXX
OrW10 MO-OnM"mON Or-m^Mf C r, -ON M OH-lr"mHr-iQ
iMNGHNHHH H MNNHNNNNNNNCnMMW





0000000000

X XXXXXX X osem~mw~mwmwmNNNHH



C 0 C) 0mC) 0C) 0'C HHHH 0HHH H
HrH H COin NrHrHri







X X XX X X X X X X X X X !X Ln N r-i lo NiM r, 00r to- 0o M







00000000000000
HHr-HHH HHOI r-I r-r-r-OOOOOO
XX XX XXX XXxXXooXoX X s cnoi
N . . . .


N
X
-ZM
: o



*



In N

(0 ff
e o
rd -^
a1^









Table 2 UFSFB #13 Data


R (meas.)
n
40KHz

200

230

260

300

410

lxl03

3xl03


R'
n
40KHz

204

235

265

306

418

1. 02x103

3.06x103


R
n
Thermal
(calc.)

113

113

113

113

113

113

113


eq
40KHz

4.0x10-4

4.6x10-4

5.2xl0-4

5.9x10-4

8.1x10-4

2.0x10-3

6.0x10-3


300
T


TOK


294

294

294

294

294

294

294


1.02

1.02

1.02

1.02

1.02

1.02

1.02


VDS
V


0.65

0.65

0.65

0.65

0.65

0.65

0.65


6201.

500u

4001u

3001

2001

1001O

5011


gm
mhos

6.2m

6.2m

6.2m

6.2m

6.2m

6.2m

6.2m


1


v










Table 3 UFSFB #13 Data


80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
310
320
330
340
350
360


f
corner
Hz


12.5
11.1
10
9.1
8.33
7.69



6.25
5.88
5.56
5.26
5.0
4.76
4.55
4.35
4.17
4.0
3.85
3.7
3.57
3.45
3.33
3.23
3.13
3.03
2.94
2.86
2.78


R
n
flat


#1
3.1xl0-2
-3
8.8x103


#2
1.2x10-2
-3
2.2x103
1.7x10-4
2.9xl0-4
1.lxl0-4
6.6xl0-5
2.3x10-5
1.9x10-5


#3
1.3x10-2
5..1x10-3
3.1xl0-3
1.4x10-3
5.9xl0-4
3.1x10-4
1.1x10-4
5.3x1.0-5
4.8x10-5
-5
1.7x10-5
7.1x10-6


f
corner
x R' flat
n


2.34x108
3.96xl08



2.84x108
2.37xl0
1.47x108
6.6x107
3.45xl07
2.06xl07
2.17x107
1.12x107



1.44x106
1.2x106
9.9xl05
4.14 x107
4.32x106
3.5x106
4.5x106
4.5x106
4.6xl06
5.6x106
8.2x106


TK


1000
T


5.2
18



13.5
74
920
550
1.5K
2.4K
7K
8.6K



12
31.5
52
115
270
515
1.5K
3K
3.3K
9.6K
22.5K


1.8x107
9.7xl06



l.lxl07
1.8x107
9.5x104
7.5x104
1.55x104
6x103
2.3x103
ix103



lxl05
3.4xl04
1.8x104

4
3.5x104
1.6x104
7xl0
3.2x103
1.65x103
1.6x103
680
440


TOK T










Table 3 extended


R'
n
flat


R'
n
corner


KER (g-r)
max


I
eq
(g-r max)


1.0xl013
1.2x1012


4x103
1.1xl03


5.3xl0-10
5.9xl0-10


2.3x1011
1.5x1010
4.8xl08
5.3x108
1.5x108
7.0x107
2.1x107
1.7x107


5.8x109
2.5x109
1.5xl09
6.7x108
2.8x108
1.6x108
5.7x107
2.8x107
2.5x107
8.9x106
3.8xl06


1. 1x103
197
14.9
26.7
11.4
8.2
3.83
4.23


3.6x103
2.1x103
1.5x103
840
425
258
98
53
52.2
19.8
8.97


8.2x10-10
8.7xl010
9.5xl0-10
-9
lxl09
1.lxl09
-9
1.2x109
1.2x10-9
1.3xl0-9
1.3x10


-9
1.6x10
1.7xl09
-9
1.8x109
1.9xl09
-9
1.9xl09
2x109
--9
2.2x109
-9
2.2x109
-9
2.3x109
-9
2.4x109
2.5xl09


4.25xl03
4x103
3.65x103
3.4x103
3.15x103
2.95x103
2.8x103
2.6xl03


2.2xl03
2.1x103
1.95xl03
1.87x103
1.8x103
1.7x103
1.6x103
1.55xl03
1.5x103
1.44x103
1.37x103


Tdr
sec


cm/v-s
cm2/v-s


4.5x107
2.2x107


2.3xl07
1.1xl07


6.6xl03
5.9x103


2.1x107
3.2x107
1.6x105
1.2x105
2.3x104
8.6x103
3.1x103
1.3x103


1.2x105
3.8x104
1.9x104
3.6x105
1.6x104
6.8x103
3x103
1.5x103
1.4x103
584
365


1.0x107
1.4x106
7.9x104
5.9x104
1.2x104
4.3x103
1.6x103
650


5.8xl04
1.9x104
8.6xl03
1.8x104
8.0x103
3.4x103
1.5x103
755
704
292
303










Table 4 UFSFB #18 Data



TK VDS IDS f (Hz) eq

80 0.8 179pA 40K 1.9x10-4

80 0.8 179pA 60K 1.9x10-4

80 0.8 179pA 100K 1.9x10-4

80 0.8 179pA 200K 1.9x10-4

80 0.8 179pA 400K 1.9x10-4

80 0.8 179pA 600K 1.9x10-4
-4

80 0.8 179pA IM 1.9x104

80 0.8 179pA 1.5M 1.9x10-4

80 0.8 179VA 2M 1.9xl0-4

80 0.8 179pA 3M 1.9x10-4

80 0.8 179pA 4M 1.9x10-4

80 0.8 179pA 5M 1.9x10-4

80 0.8 179pA 6M 1.9x10-4

80 0.8 179pA 7M 1.9x10-4

80 0.8 179pA 1.9xl0-4

80 0.8 179pA 20MHz 1.9x10-4


Amplifier Noise (Input shorted) = 10db down









Table 5 UFSFB #18 Data


300
ToK T


VDS
V


V T s


R (meas.)
40KHz
40KHz


300
300
300
300
300
300
300
300
300


3.2m
4.3m
4.95m
5.35m
5.6m
5.77m
5.9m
5.97m
6.0m


87.5p1
120p
135-p
140p
143pi
1471
150p
153up


R'
n
40KHz


R
n
Thermal
(calc.)


I
eq
40KHz


0.2
0.3
0.4
0.5
0.6
0.71
0.82
0.92
1.04


0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8


3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.75


?
0.7m
l.lm
2.4m
4.0m
6.3m
8.4m
10.0m
11.2m


?

1.2m
1.6m
1.8m
2.0m
2.1m
2.1m
2.1m


2.1K
550
180
200
160
130
120
120
105


500
850
800
900
800
740
860
820


2.1K
550
180
200
160
130
120
120
105


1.9K
3.2K
3K
3.4K
3K
2.8K
3.2K
3.1K


1.14xl04
4.63x103
972
350
141
83.3
70.0
62.5


?

972
547
432
350
333
333
333


3.72x10-6
3.0xl0-6
1.59x10-5
3.53xl0-5
-5
7.11x10-5
1.17x10-4
1.65x10-4
1.82x10-4


?
-5
6.4x105
1.1x10-4
1.5xl0-4
1.7xl0-4
1.7x10-4
1.95x10-4
1.9x10-4
1.9x10


gm
mhos










Table 6 UFSFB #18 Data


R
30 I V g R (meas.) R' nI
300 D DS m n n Thermal eq
TK T A v mhos 40KHz 40KHz (calc.) 40KHz

-4
80 3.75 153p 0.8 2.1m 820 3.1K 333 1.9x104
-4
90 3.33 2651 0.8 3.4m 400 1.33K 206 2.4x104
-4
100 3.0 480p 0.8 5.2m 190 570 135 2.7x104
-4
110 2.73 7781 0.8 7.5m 110 300 93 3.2x104
-4
120 2.5 im 0.8 9.2m 130 325 76 5.7xl0-
-4
130 2.31 1.32m 0.8 10.6m 98 226 66 5.7x104
-4
140 2.14 1.6m 0.8 11.6m 88 188 60 6.1x104
-4
150 2.0 2.0m 0.8 13m 70 140 54 6.1x10l
-4
160 1.88 2.25m 0.8 13m 73 137 54 6.4x10l
-4
170 1.76 2.62m 0.8 14.1m 80 141 50 8.2x10l
-4
180 1.67 2.95m 0.8 12.6m 70 117 56 5.8xl04
-4
190 1.58 3.15m 0.8 13.4m 75 119 52 7.0x10l
-4
200 1.5 3.5m 0.8 13.4m 85 128 52 7.9x104
-3
210 1.43 3.8m 0.8 16.8m 90 129 42 1.3x10-
-3
220 1.36 4.05m 0.8 15m 96 131 47 1.1x10-
-3
230 1.3 4.35m 0.8 20m 100 130 35 2.1x10-
-4
240 1.25 4.6m 0.8 12.6m 105 131 56 8.6x10-
-4
250 1.2 4.8m 0.8 11.2m 110 132 63 7.1x10-
-4
260 1.15 5.0m 0.8 10.6m 100 115 66 5.8x10 -
-4
270 1.11 5.2m 0.8 10.6m 110 122 66 6.4x10l
-4
280 1.07 5.4m 0.82 9.5m 95 102 74 4.4x10-
-4
290 1.03 5.65m 0.78 9m 100 103 78 4.2x10-
-4
300 1.0 5.77m 0.8 8.6m 100 100 81.4 3.8x10
-4
310 0.97 5.9m 0.8 7.5m 110 107 93.3 3.2x10
-4
320 0.94 6.0m 0.8 7.1m 130 122 98.6 3.4x104
-4
330 0.91 6.1m 0.8 6.7m 160 146 104.5 3.7x104
-4
340 0.88 6.2m 0.8 6.3m 250 220 111 5.1x104
350 0.86 6.25m 0.8 5.5m 230 198 127 3.6x104
360 0.83 6.32m 0.8 5.0m 250 208 140 3.2x10-
370 0.81 6.38m 0.81 4.4m 275 223 159 2.7x10-4
370 0.81 6.38m 0.81 4.4m 275 223 159 2.7x10










Table 7 UFSFB #18 Data


1000
TK T


f
corner
Hz


R
n
flat


f
corner
x R' flat
n


80
90
100
110
120
130
140
150
160
170

150
160
170
180
190
200
210
220
230
240
250
260

260
270
280
290
300
310
320
330
340
350
360
370


12.5
11.11
10
9.1
8.33
7.69
7.14
6.67
6.25
5.88

6.67
6.25
5.88
5.56
5.26
5.0
4.76
4.55
4.35
4.17
4.0
3.85

3.85
3.7
3.57
3.45
3.33
3.23
3.13
3.03
2.94
2.86
2.78
2.7


#1 -
-2
2.12x10
-2
1.66x10
-2
1.18x10-
-3
3.98x10l
-4
6.5x10-
-4
1.82x10l
-5
6.77x10
-5
3.25x10
-6
7.4x10
-6
3.18x10
#2
-2
4.55x10
-2
2.18x10l
-3
7.96x10
-3
1.99x10-
-4
5.31x10
-4
2.89x10
-4
2.15x10l
-5
4.55x10l
-5
1.99x10
-6
6.63x10
-6
3.39x10
-6
1.45x10
#3
-2
2.38x10
-3
9.09x10
-3
2.89x10
-3
1.1x10l
-42
4.3x104
-4
2.53x10
-4
1.27x10
-5
6.77x105
-5
3.79x10
-5
2.09x10
-5
1.14x10
5.58x-6
5.58xi0


7.5
9.6
13.5
40
245
875
2.35K
4.9K
21.5K
50K

3.5
7.3
20
80
300
550
740
3.5K
8K
24K
47K
110K

6.7
17.5
55
145
370
630
1.25K
2.35K
4.2K
7.6K
14K
28.5K


4
2.5x104
1.7x10l
2.2x10O
9x103
2.3x10
800
370
210
115
100

7x105
5
3.6x10l
1.1x10
3.2x10G
10K
5.4K
2.3K
970
415
235
160
110

1.1x106
5
5x10l
1.5x10l
6.1x10
4
2.75x104
1.6x10
8.5K
4.9K
3.5K
1.6K
1K
580


7x105
5.4x105
8.9x10l
9.8x10
1.4x10l
1.6x10l
1.86x106
2.1x106
4.6x10O
8.8x10

4.9x106
4.9x10O
3.9x106
4.3x106
4.7x106
4.5x106
2.4x106
4.6x106
4.3x10l
7.1x10
9x106
1.4x10

8.5xl06
6
9.7x10i
8.9x10l
9.1x106
1x106
9.8x10
1x107
7
1.1x10l
1.3x10
1.1x107
1.2x107
1.3x107


-









Table 7 extended


K-R (g-r)
max
max


I
eq
(a-r max)


4
9.38x104
5.67x10
4
6.6x10
2.45x10l
3
5.75x10l
1.85x10-
793
420
216
176

1.4x106
5
6.75x10
1.94x10
5.34x10l
1.58x10
8.1K
3.29K
1.3K
540
294
192
127

1.27x106
5.55x10
1.61x10
6.28x10l
2.75x10
1.55xl0
8K
4.5K
3.1K
1.38K
830
470


4.69x104
4
2.83x10l
3.3x10
1.23x10l
2.88x10l
9.23x10
396
210
108
88

7x105
3.38x10i
9.68x10i
2.67x10l
7.9x10
4.1K
1.64K
660
270
147
96
63

6.33xl05
5
2.78x10l
2.68x10l
3.14x10l
1.38x10
7.76K
4K
2.2K
1.54K
688
415
235


11
4.5xl011
1.7x10^
10
7.4x10.
1.6x100
1.8x10
2x10l
8
1.3x10l
5.1x10
1.1x10o
2x106
10
6.9x100
10
3.2x100
lxl10
3x10
8
7x10
3.6x10
1.7x10
7
4.4x10
lx10
8.3x10l
4.9x10
2.6x10
10
3.8xl100
1.4xl10
5.1x10
2.1xl0
8
8.3x10
5.9x10O
3.1x10l
1.8x10l
8
1. xl0
7
7.2xl0
4.5x10O
2.6x10


4
2.8x104
3.1x10
3.5x10l
1.7x10
3.1x10
5x10
2
4.1x10
2.2x10
50
12

3x105
5
1.5x10
5.9x10
1.5x10
4.1x10O
2.3x10l
1.7x10
3.7x10
1.6x10
55
26
13

1.9x105
4
7.1x10
2.2x10
8.3x10l
3.2x10
1.8x10o
870
460
250
130
70
32


-10
2.34x10 1
-10
2.8x10
-10
3.23x10 1
-10
3.7x10 --
-10-
4.26x10
4.77x1010
-10
5.31x10
-10
5.98x10 1
6.62x10-
7.03x10
-10
5.98x10 -
-10
6.62x10 _
-10-
7.03x10 1
-10-
7.71x10
8.27x10 -
8.93x100
9.53x100
xl-9
1x109
-9
1.08x10
-9
1.12x10-
-9
1.25x10l
-9
1.28x10
-9
1.28x109
-9
1.34x109
-9
1.44x10
-9
1.5x10
-9
1.56x10
-9
1.65x10
-9
1.76x10-
-9
1.8x10
-9
1.88x10
1.95x10
-9
2.1x10
2.16x-9
2.16x10


4
1.2x104
4
1x10-
8.7x10l
7.6x10O
6.6x10l
5.9x10
5.3x10l
3
4.7x10-
4.3x10
4x10
4.7xl03
3
4.3x103
4x10.
4xl03



3
3.65x10

3.2x10l
2. 95x10



2.3x10l
2.2x103
2.2xl03
3




2.1x10
1.95x103
1.87x103
1.8x10i3
1.7x10 3
1.6x103
1.55x10i3
1.5x10i3
1.44x103
1.37x10l
1.3x10


R'
n
flat


R'
n
corner


Tdr
sec


1-I
2
cm /v-s


--









Table 8 2N4416 #3 Data


300
T


ToKT


R (meas.)
n Hz
40KHz


R
n
Thermal
(calc.)


80
80
80
80
80
80
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300


3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.33
3.0
2.73
2.5
2.31
2.14
2.0
1.88
1.76
1.67
1.58
1.5
1.43
1.36
1.3
1.25
1.2
1.15
1.11
1.07
1.03
1.0


10.5m
12.7m
13.9m
14.6m
15.2m
15.6m
15.75m
15.3m
15.4m
15.35m
15.2m
14.9m
14.66m
14.34m
13.95m
13.56m
13.2m
12.7m
12.3m
11.9m
11.46m
11.0m
10.6m
10.12m
9.8m
9.6m
9.0m
8.65m
8.33m


VDS
V


gm
mhos


R'
n
40KHz


I
eq
40KHz


1
1.5
2
2.5
3
3.5
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3


3
2.1xl03
1.1x10
660
540
510
740
820
660
740
270
190
150
130
140
140
140
150
190
230
260
260
240
230
220
220
210
210
230
260


3
7.8xl03
4.13x103
2.48x103
2.03x103
1.91x103
2.78x103
3.1x103
2.2x103
2.22x10
736
475
346
279
280
263
247
250
300
345
371
355
313
288
264
254
233
225
238
260


-3
8.7x10-3
4.6x10-
2.8x10l3
2.3x10-3
-3
2.1x10-
3.1x10-
3.5x10l3
2.8x10-3
3.2x103
1.2x10-
8.9x104
7.13x104
4
6.06x10-
6.32x10-
6.13x10-4
5.72x10l4
5.73x10-
6.75x104
7.60x104
7.95x10-4
7.36x10-
6.24x10-4
5.50x10l4
-4
4.80x10l4
4.37xl0-4
3.52x10-4
3.16x104
3.09x10l4
3.22x10


77.78


76.9
76.1
74.8
73.7
72.9
73.7
74.8
76.1
78.6
81.4
84.3
87.5
90.9
94.6
98.6
103
107.7
113
122.8
129.6
137.3
142.9


9m


9.1m
9.2m
9.35m
9.5m
9.6m
9.5m
9.35m
9.2m
8.9m
8.6m
8.3m
8.0m
7.7m
7.4m
7.1m
6.8m
6.5m
6.2m
5.7m
5.4m
5.1m
4.9m


TOK T










Table 9 2N4416 #3 Data


f
corner
Hz


80

80

80

80

80

80

80

90

100


35

155

420

1150

3.2K

9.5K

14.5K

6.6K

40K


12.5

12.5

12.5

12.5

12.5

12.5

12.5

11.11

10


4.55x10-3

1.03x10-3

3.79x10-4
-4
1.38xl0-4

4.97x10-5

1.68xl0-5

1.1x10-5
l.lxlO05

2.41x10-5

3.98xl0-6



2.15x10-2

5.31x10-3

1.1x10-3
l.lxlO3

6.01x10-4

3.54x10-4

2.05x10-4
-4
1.27x10-4

7.4x10-5

6.37x10-5
-5
2.89x105

2.04x105
-6
6.37xl0-6

3.18x10-6

2.38x10-6


TOK


1000
T


R
n
flat


corner
R' flat
n


7.4

30

135

265

450

775

1.25K

2.15K

2.5K

5.5K

7.8K

25K

50K

67K


120

130

140

150

160

170

180

190

200

210

220

230

240

250


8.33

7.69

7.14

6.67

6.25

5.88

5.56

5.26

5.0

4.76

4.55

4.35

4.17

4.0


- - - -- ------- ---- ------ ------------------ ------- - -- --


1x106

4xl05

2.7x105

1.2x105

3.5xl05

9.2x103

2.9x103

1.23x104

1.5x103



4.3xl05

1.6x105

3.5xl04

2.3x104

1.2x104

7.2xl03

3.5xl03

2x103

1.4x103

900

650

400

300

260


1.3x108

2.3x108

4.2x108

5.2x108

4.2x109

3.3xl08

1.6x108

2.7x108

1.8x108



8.1x106

1.lx107

ixl07

1.2x107

1x107

1x107

7.3x106

6.9x106

5.3x106
6
7.2x10

7x106

1.3x107

1.9x107

2.1x107









Table 9 extended


K-R (g-r)
n
max


R'
n
flat

3.8x106

1.5x106

1. 0x106

4.5x105

1.3x106

3.5x104

1. xl04

4.1xl04

4.5x103



1. xl06

3.7x105

7.5x104

4.6x104

2.3x104

1.3x104

5.8x103

3.2x103

2.1xl03

1.3x103

8.9x102

5.2x102

3.8x102

3. 1xl02


R'
n
corner

1.9x106

7.5x105

5.1x105

2.3xl05

6.6xl05

1.7x104

5.4x103

2.1x104

2.3x103



5.4x105

1.9x105

3.8xl04

2.3x104

1.1xl04

6.4xl03

2.9x103

1.6x103

1.lx103

6.4xl02

4.4xl02

2.6xl02

1.9x102

1.6x102


I
eq
(g-r max)

1.lx106

4.5x105

2.5x105

1.2x105

5.2x104

2.1x104

1.6x104

2.1x104

3.1x103



1.2x107

2.7xl06

5x105

2.3x105

1.2x105

6.4x104

3.5xl04

1.9x104

1.4x104

5.7xl03

3.9x103

1.lx103

480

320


Tdr
sec

8.7x10-11

5.8xl0-11

4.3x1011

3.5x1011

2.9x1011

2.5xl011

2.2x1011

3.5x1011

4x10-11

-11
5.3x10

5.9x1011

6.5x1011

7.4xl011

8.2x1011

8.7x10-11

9.5x1011

lxlO-10
-10
l.lxlO-10

1.2x10-10

1.2x1010

1.3x1010

1.4x10-10

1.5x10-10


3.7xl012





1.5x1012

2.1x1011



7.7xl014

1.7xl014

3.2x1013

1.5xl013

8.2xl012

4.7xl012

2.7x1012

1.6x1012

1.3xl012

5.6x101

4.1x1011

1.3x1011

6x1010

4.4x0100


IA
cm2 /v-s

1.2x104

1.2x104

1.2x104

1.2xl04

1.2x104

1.2x104

1.2x104

lxl04

8.7xl03



6.6xl03

5.9x103

5.3x103

4.7x103

4.25x103

4x103

3.65xl03

3.4x103

3.15x103

2.95x103

2.8x103

2.6x103

2.45xl03

2.3xl03










Table 10 Range of Parameters Slice #3, UFSFB



IV 9 (mho)
Device # DSS Vpo(v) g(mho)

1 62pA 0.26 800p

2 28pA 0.4 180O

3 400pA 0.4 4m

4 1751A 0.5 Im

5 25pA 0.3 320p

6 1.3mA 0.5 8.1m

7 1.8mA 0.5 10.2m

8 580pA 0.4 4.6m

9 1.2mA 0.42 10m

10 150pA 0.5 1.2m

11 400pA 0.3 6m

12 350,A 0.6 1.6m

13 650pA 0.4 6.2m

14 230pA 0.35 2.5m

15 550pA 0.5 3m

16 281A 0.3 180p

17 260pA 0.3 2.9m

18 6mA 0.72 24m

19 4.7mA 0.8 20m

20 420pA 0 4 4m

21 900pA 0.4 8m

22 l.lmA 0.45 8.2m

23 220pA 0.3 2.5m

24 52pA 0.4 400p















501 A


6201 A '











T = 3000K


L- -L

- ---


106i


R
n
(ohms)


I I I I I Il


L 10 10 Freq (Hz) 10 10 10
Figure 19 UFSFB #13 R vs Frequency, a Function of Ip.


1 I 1 1 i


I I % ,I


...1 ....,,.~














































, *1*t


Figure 20


.1


10 Freq (Hz) 10-
UFSFB #13 R vs Frequency.
n


108

R
n
(ohms)




107


06









105
10


I t I t ? v I It It
5 i A I a f f I - ---1 1 t I I I -t -- I I 1 1 F I I I 1 .1 1


-. . .


I


I


10


10


104

































1400K


1800K


102 Freq (Hz) 103


I I I 11 I t


I I I I I III!


UFSFB #13 R vs Frequency.
n


R
n
(ohms)




106


1041
1


I t t tI 1


105


Figure 21


I f I II


104



















































I f t I t I t I __


12 3Fq ()
10 Freq (Hz) 10


Figure 22 UFSFB #13 Rn vs Frequency.


106

io


105


R
n
(ohms)


102
1


I t I I I


i


ft I f. t IIt I,









106








5
105


R
n
(ohms)




104


3000K


.1


If l igi1


0 l 2 .F . 10. ,
2 3 4
10 10 Freq (Hz) 10 10 1


UFSFB #13 R vs Frequency.
n


2600


102
1


i


..f


Figure 23






























3200K


3400K #3


3600K


,i I I I 1I I 1I II i _~t I _U


Figure 24


10 Freq (Hz) 10-
UFSFB #13 Rn vs Frequency.


R
n
(ohms)


io2
10


10


A A,


t


I I 1 I(((lf )























Rn
(ohms) 40H/


103 00Hz





r- \


40KHz /



80 100 120 140 160 180 200 220 240 260 280 300 320 340 360
TK


Figure 25 UFSFB #13 Composite Plot: Rn vs. T, a Function of Frequency.












ImA L-lOmmho


gm


4-
-4.
6~~


A


100pA- Immho


/
/,


/
10A 100pmho /


'--I
I




80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 TK

Figure 26 UFSFB #13, ID and gm vs, T.








#2-18


10-2
10


1-13




/;


#3-13,18


i I


.1 1 1 1 1 1 ,1 I 1 1 I I. I I I I I 1 1 -


0 1 2 3 4 5 6 7 8 9 10 11 1
1000
T

Figure 27 UFSFB #13 and #18, T vs. 1000/T.


2 13


#1-13: EA = 0.17V

#2-13: EA = 0.34V

#3-13: EA = 0.6V


#1-18: EA = 0.19V

#2-18: EA = 0.36V

#3-18: EA = 0.6V


I


F
I,/


















c- -


T = 3000K


.___0 3V


..................I


.........................I


0.4V


0.8V to IV -- .._
S I 1 1 I


iI II I l


10 102 Freq (Hz) 10 104 1



Figure 28 UFSFB #18 Rn vs. Frequency, a Function of VDS.


105


*R
n
(ohms)


102


-I














Freq = 40KHz
T = 3000K
I
I
I
/

I
/

/
I


I
I



/

, 1 1 I ..... .


Freq = 40KHz
T = 80K


. . . I


0.1 1
VDS


Figure 29 UFSFB #18 Ieq vs. VDS at 3000K and 800K.
-eq DS


eq



10-5


0.1
0.1


, *a .% A I


^_ __ ~_ ___ ~~_
























T = 800K


VDS: 0.2V

0.4V

0.6V

0.8V


106


R
n
(ohms)


1 10 10' Freq (Hz) 10" 10
Figure 30 UFSFB #18 R vs. Frequency, a Function of V.
n DS


""'


103








106








105

R
n
(ohms)




104 #
#1
#1






103 80K
3


900K

1100K



10 2 -' .. 3 4 5
1 10 102Freq (Hz) 10 10 10
Figure 31- UFSFB #18 R vs. Frequency.
nl








105



1300K
1100K
1200K,


R
n
(ohms)


10 102 Freq (Hz) 103 104 10
Figure 32,, UFSFB #1.8 Rn vs. Frequency.


I t I 1 1 9 1


T I f f f t ? I


t t ~.(f


* 1 t I f p 7I1







































1600K


* I It ( Il1


2I I I 1 a fI I


2I I t A1 I


10 102 103 104 1
Freq (Hz)


Figure 33 UFSFB #18 R vs. Frequency.
n


R
n
hms)


. . ...I


S . ....,I


























1700K


1800K -


". 19o 0 ,
19-00oI
-I-
-.5 -
5%


I a It 1 fa


I I I a a 1 ( 1


a a I rllama! a a as,,1


A.. a .& . -. I It 2 a a


,10(
Freq (Hz)


Figure 34 UFSFB #18 Rn vs. Frequency.


104


R
n
)hms)


102


S . ... I


.


10




































2100K\ \

K


I I I # I t l


I I I I I f i!


t I I t t ll


10 102 103 104. 10
Freq (Hz)


Figure 35 UFSFB #18 R vs. Frequency.
n


105






A.4
104


R
n
)hms)


102


It 9 llt


I I. t 0 2 ..1





















\500K


24001


#2


-..... t


U U U Imas


* a 11a I


II I I 1 II. t t lii I z A r i i i rr i ll
2 3 4
10 10 10 10 1(
Freq (Hz)


Figure 36. UFSFB #18 Rn vs. Frequency.


S 9 a . a AI


I J.*





































800K


2600K


2700K


110K


.I lf


a s I I. a. i


I I 1 1 1 1


10'5


102
Freq (Hz)


Figure 37 UFSFB #18 R vs. Frequency.
n


R
n
)hms)


a I lil


I . IPy I q I














3000K


3100K


3 9 9. 9 991


9 9 9, 9.1~


SI I t t I)1


t t t I t t I1


102 Freq (Hz) 103
38 UFSFB #18 Rn vs. Frequency.


R
n
(ohms)


102


'10
Figure


t ? I t ( f 1


? I ! T ? I I--_-1 -- ~ ~ -- i------ -- ----I-- --- -


















3200K


3400K


K


t t t f 1


I_ i I 1"


f I 1 I I t t I


1


10 102 Freq (Hz) 10j

Figure 39 UFSFB #18 Rn vs. Frequency.


05


10"


R
n
(ohmns)


102


J *


i


1


















































I I I I t I II


f f I i


f I 19 I I ?I


I I 99 I)1


1 I (I a II I


10 102 Freq (Hz) 103 10 14

Figure 40 UFSFB #18 Rn vs. Frequency.


10


R
n
(ohms)


102























R
n
(ohms)


80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380
TK


Figure 41 UFSFB #18 Compositeplot: R vs. T, a Function of Frequency.














10mA
2. "I"









1mA -l0mmh m










Immho
100l A 1I . . 1 I I I I t I I I I I I
80 100 120 140 160 180 200 220 240 260 280 300 320 340 360
T(oK)

Figure 42 UFSFB #18 ID and gm vs. Temperature.





















T = 800K, Freq = 40KHz



R' (meas.)
---- ---- '*- ---


(thermal) calc.


I I I _I I


0.2 0.3


0.4


0.5 0.6
VDS


0.7


0.8


Figure 43 UFSFB #18 R'(meas.) and Rn(thermal) vs. VDS.
-' n n U


(ohms)


10i2
0.


1




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs