ENGINEERING SERVICES ON TRANSISTORS. INTERIM TECHNICAL REPORT NO. 7. (CONTRACT DC 36-039-SC-64618).
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
CIA-RDP81-01043R002300090003-0
Release Decision:
RIPPUB
Original Classification:
U
Document Page Count:
27
Document Creation Date:
December 27, 2016
Document Release Date:
October 25, 2013
Sequence Number:
3
Case Number:
Publication Date:
April 29, 1958
Content Type:
REPORT
File:
Attachment | Size |
---|---|
![]() | 2.3 MB |
Body:
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
R
STAT
Next 3 Page(s) In Document Denied
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
W..,
2?????????
??????? 4??????oft. ????? ? AN...
^
Engineering Services on
TRANSISTORS
SEVENTH INTERIM TECHNICAL REPORT
(":tering the Period 16 S?plember to 15 December 1956
Signal Corps Contract DA 36-039 se-64618
Signal Corps Project No. 323A
Department of the Army Project No. 3-19-03-031
This contract is a joint services contrA,t for the Departments of the Army, Navy,
and Air Force It calls for sen ices. fAcilities, and materials to be employed
in studies and iii estigations related to transistors and transistor-like devices,
together with their circuit properties and their applications for military uses.
This report was prepared by Bell Telephone Laboratories Incorporated
_ The foi;ow mg engineers and scientists participated in its preparation:
1ft
W I. Brown J E lw ersen C II Know les J T Nelson J Sevick
? ?
-
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
Declassified in Part - Sanitized Copy Approved for Release 50-Yr 2013/10/25 : CIA-RDP81-01043R002300090003-0
,-!
r ,
?-? ?
SUMMARY OF STATUS
4 -
As ret.orted in Sections 2 and 3. Tasks 1. 3. and 6 have been completed No work was dune on Task 7
during th.s report interval Task S is covered In a separatz series of re-ports. Work done on the other
tasks is summarized below.
Poring the period covered by this report 16 September through 15 December 1956. approximately
5 OUa hours were devated by key personnel to work on this project.
TASK 2 - TFLA.NCIETOR RELIABILfeY
fin account of the heavy demand for ag.ng information from users of transistors the direct determina-
tion of reliability is still being pushed, especially for the neer transistors Teats on time rates of aging
of particular transistor: v.ith temperature and with applied voltage are under way and a pre:immary report
may be meltable next qaarter. Although the results of this type of test are dependent on changes and Im-
provements in fabriea,.tri, it is believed that the greater stability now avidable from new surface process-
ing and the need fur edormatiun on new ty [WS make this work of continuing urgency.
Correlation of the observed aging trends with the physical causes is continuing, under difficulties
r.aused by the inherent sensitivity of the surfaces to contamination by atomic amoiaits of water, oxygen and
other tract- ato,uents An experiment to study ,,imultaneuusly surface properties and aging of function
characteristics in the same ambient has been unde -taken. ;iom %%Filch it is hoped to separate the physical
Losses and thereby improve the understanding and effectiveness of the aging mPasurements.
TASK 4 - NEW AND IMPROVED TRANSMISSION-TYPE TRANSISTORS
D-evele^^..eot of diffused-base germanium and silicon trarristors has continued. Temperature studies
of characteristics and performance have been made- on the T.12039 chffused-tease gern.anium p-n-p oscil-
lator transistor under indt.strial Preparedness Studs Contract DA 36-039 s.--72729 Excellent oscillator
efficienrws were obtained, even at 10:i*C. Development activity on the 1.12055 and h1205S low-level common-
best' and common-emitter an-while. s.gt,shis been expanded under the same contract The character-
istics of initial models of these diffused-base gt rmanium devices are also eery encouraging.
Modifications in the structure of the M2'J36 p-n-i-p siiiconpouer transistor have increased the oscil-
lator efficiency of this device so that efficiencies; above 50 per cent and pow:rs greater than 5 watts have
been obtained in experimental ands at 10 me eracii:ator pertormance testing has bee, accompanied by other
characterization work such as d-c low-frequency and high-frequencv parameter measorements These
results have heen compared with a proposed equivalent circuit and with pulse-measurement data. The
initial design objectives for this unit have been reached
TASK 5 - TRANSISTOR TEST METHODS
Chapter 6 describes the. present status of the 5-mc to 250-me Phase Set and d ..-usses improved crosr-
.alk pi rformance signal to noise ratio and improvemen s in the synchronization -7stem performance.
This report contains copyrighted or copyright:11,1e material (Chapters 1 and 2)
not first produced under a government contract and shall be reproduced or used for
gg.vernmental purposes only .:nd not for sales or do,position to the genei al public.
This report contains intelligence (Chapters I and 21 not first produced under a
govPrnment contract and shall not be given to any foreign government without the
written consent of the contractor
? rot re.'" . . ... ? -
`Y.
TABLE OF CONTENTS
SECTION 1 - PURPOSE . . . . . .
SECTION 2 - ABSTRACT. . . . . .
SECTION 3 - PUBLICATIONS AND REPORTS . . .
SECTION 4 - FACTUAL DATA . . . . .
TASK 2 - TRANSISTOR RELIABILITY
Chapter 1 - Neutron Radiation Effects on Germaniuri Transistors
and Silicon Diodes
1.1 Introduction .
1.2 The Devices and Their Neutron Expcx,ures
I 3 Interpretation of Changes in Device ..:haracteristics
Table 1-1: Changes in 1868 P-h-P Units
Table 1-2: Changes in 1853 N-P-N Units
1.4 Summary. Page 12
Page
1
Page
5
Page
6
Page
7
. Page 7
Page
7
Page
8
Page
8
Page
9
Page
9
Chapter 2 - Noise and Reliability of Diffuse Silicon Rectifiers . Page 13
2.1 Introduction .
2.2 Discussion .
2.3 Measuring Equipment.
2.4 Diode Noise Characteristics
2.5 Measurements of Reliability
2.6 Results .
2.7 Conclusions
ill
?
Page 13
Page 13
Ptige 16
Page 16
Page 17
Page 19
Page 19
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
71411V.P" " 'Am
Declassified in Part - Sanitized Copy Ayko..?e_v_7..c_liirie,..:47,I7!,aillsle, .0., : CIA-RDP81-01043R002300090003-0
?-
Tzt-
-?
Iv
?
TASK 4 - NEW AND IMPROVED TRANSMISSION-TYPE TRANSISTORS
Chapter 3 - Temperature Dependence of the M2039 Transistor
? . Page 20
3.1 Introduction . Page 20
3.2 Method Page 20
3.3 High-Frequency Oscillator Pe-formance Page 21
3.4 Low-Frequency Measurements. ? Page 21
Table 3-1: Variation of Alpha Cutoff Frequency
witi. Temperature . . . . ? Page 22
3.5 Low-Frequency h Parameters and 'Co . Page 24
3.6 Summary . Page 27
Chapter 4 - M2055 and M2058 Transistors ? ? ? ? . Page 28
4.1 Introduction . Page 28
4.2 Fabrication Process Page 28
4.3 Electrical Measurements ? Page 29
4.4 Summary . Page 30
Chapter 5 - Status of the M2036 P-N-I-P Silicon Power Transistor . .Page 31
5.1 Introduction .
5.2 Changes in Structure and Rough Results.
5.3 Electrical Characterization
5.4 Summary and Future Plans
TASK 5 - TRANSISTOR TEST METHODS
Chapter 6 - Review cf Task 5 Activities
6.1 1- troduction .
6.2 Measuring- Path Switches .
6.3 Amplifier Modulation Test Noises
6.4 Synchronization .
Page 31
Page 31
Pagi 31
Pagz 38
. Page 39
Page 39
Page 39
Page 40
Page 40
??..??????? 1/1,?????",?.
6.1.1.???
? ..11?????.... "tom, ????(...111.??????? ??? .
? ? .??
SECTION 5 - CONCLUSIONS .
? ? ? . Page 41
SECTION 6 - PROGRAM FOR NEXT INTERVAL . ^ . Page 42
SECTION 7 - IDENTIFICATION OF PERSONNEL . . ?age 43
DISTRIBUTION LIST . . Page 45
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
4V-A
Declassified in Part - Sanitized Copy Approved for Release 0 50-Yr 2013/10/25 : CIA-RDP81-01043R002300090003-0
11-Aif
1,? ? ??
c
?
usr OF ILLUSTRATIONS
1 - "Shakiness" in breakdown .
2 - Proposed mechanism of failure
3 - Noise-measuring circuit set to measure noise currents at constant reverse voltage
4 - f'soise circuit for constant-current measurements
5 - Constant-current generator
6 - Expected diode noire spectrum. showing addition of 1/f noise to shot noise
- Noise spectrum of a very noisy M2045 transistor under -6.0 volts bias
8 - Change in VB versus reverse current
9 - Change in VB versus mean noise at -1.0 pa reverse-current bias
10 - Change in VB during aging versus range of noise
11 - Change in VD versus mc n noise rt constant-current forward bias
12 - Chang? n VB versus range in aois0
13 - Portable heating apparatus . . ?
14 - Oscillator efficiency as a function of temperature
15 - (1-ao) as a function of temperqture
16 - Collector reverse current at Ve = 18 volts as a function of temperature
17 - Improved structure of the M2055 and M2058 diffused-base transistor .
18 - Maximum available gain characteristics with bias and frequency for M2055 transistors
19 - Revised structure of M2036 transistor.
20 - Common-emitter static characteristics
21 - Collector capacitance as a function of voltage .
22 - Common-emitter current gain a& a function of collector current .
23 - Common-emitter current gain as a function of frequency.
24 - Proposed enuivalent circuit (showing common emitter) .
25 - Schematic diagram of oscillator circuit .
26 - Block diagram of the 5-mc to 250-mc phase set.
vi
..4.11 ?????.
V t?
- ? " ^ - ? "" - r . -
SECTION 1 - PURPOSE
The generol purpose of this contract is to make studies 3nd investigations
related to transistors and transistor-like devices, together with their circuit
properties and applications, with a view toward demonstrating and increasing the
Page practicality of their use in operating equipment. rhis contract is a successor to
two preceding contracts of a similar nature: Contract W36-039 se-44497, herein.
after referred to as the First Engineering Services Contract on Transistors; and
Contract DA 36-039 sc -5589, hereinafter referred to as the Second Engineering
Services Conti-Act on Transistors. Both of these contracts were included in
Department of the Army Project 3-19-03-031 and Signal Corps Project 27-323A-1.
The reports on these contracts contain much reference material, particularly on
dual-stability c:Jitrhing cir7.?uit.3 suitable for data transmission and also on other
properties of various transistors and related devices. Unless otherwise stated,
all referencesk previous reports are to those produced under this, the Thi-d
Engineering Services Contract.
These contracts call for services, facilities, and material to be employed on
mutually acceptable tasks. Of the eight tasks assigned, three have been completed.
Task i involved the development of an oscillator of very low power drain, the Final
Report on this task was isseed on 1 June 1952. Task 3 covered a symposium on the
circuit properties of transistors; the Final Report on this task was issued on 1 Feb-
ruary 1952. Task 6 called for theoretical and experimental studies :eading to the
development of photocell blocks as set forth in the specifications for p-n junction
photocells and photocell blocks (NRL Problem R05.54); the Final Report on this
task was issued on 31 January 1955.
The other assigned tasks are outlined below. Although Task 8 is being covered
by separate reports, an outline of this task as agreed to between the contracting
parties is also included here.
14
14
14
15
15
16
17
18
18
18
19
19
? 21
. 23
25
27
. 29
. 30
? 32
? 33
? 33
? 34
34
. 35
. 37
. 39
The terms "zero-order exploratory development" and "first-crier feasibility
development" employed in the following outlines are defined in two reports issued
under the First Engineering Services Contract un Transistors: the Sixth Quarterly
Report, pages 11 to 13, and the Final Report, Section 4, pages 5 and 6.
TASK 2 - TRANSISTOR RELIABILITY
The contractor shall conduct a broad program in transistor reliability designed
to accomplish the following objectives:
(a) Provide quantitative definition's of transistor and transistor circuit
reliability figures of merit.
(b) Evaluate available transistors in the light of the above definitions.
(c) Explore the possibilities of developing more reliable transistors in
terms of (a) above.
1
nes laccif ri in P rtsanitized C A
V
ed for Release ? 50-Yr 2013/10/25 ? CIA-RDP81-01043R002300090003-0
lAnkt ,
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
t_
'
?
2
TASK 4 - NEW AND IMPROVED TRANSMISSION-TYPE TRANSISTORS
The contractor shall make theoretical and experimental studies leading to zero-
order exploratory development and, upon mutual agreement, to first-order feasibility
development of:
(a) New transistors using new or previously untried principles.
(b) New transistors obtained by studied modifications of existing types.
The new transistors shall be primal ily intended and suitable for application to
voltage, current, and power amplifiers, and associated electronic transducers. In
particular, it is desirable to carry through the zero- and first-order development
of a point-juncticn transistor for micropower radio-frequency applications.
In general, a-c amplifying devices for voltage, current, and power amplifica-
tion, which are of particular interest, have transmission properties in the following
ranges:
(a) Frequency range: two cycles per second to 100 megacycles; this range
need not be achieved in one transistor.
(b) Voltage ranges: one microvolt to fifty volts; same comment.
(c) Current ranges: one microampere to one ampere; same comment.
(d) Power ranges: one microwatt to ten waits; same comment.
(e) Noise figures.at 100 cycles per second; two decibels as a goal.
Class A and AB operation is intended. Per cent harmonic distortion is indicated
at three per cent as a goal. In order to arrive at useful devices it will be necessary
to keep in mind the eventual importance of simplicity of operation and construction,
interchangeability of transistor units, drifts of important circuit parameters, such
as gain, selectivity, distortion, etc., with time, temperature, and climatic conditions,
in that order of importance. Avoidance of unusual circuit requirements, such as
inordinately low per cent regulation in power supplieL nr cumbersome associated
equipment is also an ultimate desideratum.
For CW oscillator service, device prnperties in the following ranges are of
particular interest:
(a) Frequency range: twenty cyclta per second to 100 megacycles, not
necessarily in one unit.
(b) Power ranges: ten m crowatts to five watts.
(c) Frequency stability- Variations of device characteristics which affect
frequency should be clinparable to those of electron tubes commonly
used in electron-coupled oscillators.
The transistors to be ultimately developed for manufacture are irterded to
improve the electronic contrivances now available using electron tub.
matter of:
(a) Operating parameters, such as gain, selectivity, noise, drift,
operating life;
(b) Physical size and weight (reduction);
?
V ?
?
?
?????? ? ????-? ????????? ?????-?? ???? ? ?? ???? ?-? )???11. eall.0111.4???????41.41air....e r. t?-???????
(c) Power requirements (reduction);
(d) Cost;
(e) Simplicity of construction and operation;
(f) Maintenance problems;
(g) Effect of extraneous environmental iniluences, such as temperature,
humidity, shock and bounce, and pressure.
TASK 5 - TRANSISTOR TEST METHODS
Task 5 involves the formulatior of requirements for transistor test equipment
and the fabrication of an experimental model.
TASK 7 - TRANSISTOR CIRCUIT COMPONENTS
Task 7 specifies studies and investigations of suitable miniaturized components
for transistor circuit applications aimed toward the following four major objectives:
(a) To determine appropriate electrical ratings and the most suitable
mechanical characteristics for component apparatus such as inductors,
transformers, capacitors, resistors, potentiometers, power supplies,
etc., presently needed in transistor circuitry.
(h) To determine what types of components meeting these requirements
are currently available in reasonably satisfactory form.
(c) To report on the status of work being carried on at Bell Telephone
Laboratories on the development of miniaturized components to
meet the stated requirements.
(d) To formulate recommendations as to areas of component development
work not being adequately co-/ered by Bell Telephone Laboratories
and on which the military services might concentrate fruitful additional
development and rcsearch effort elsewhere.
TASK 8 - IMPROVED CRYSTAL RECTIFIERS
The contractor shall make theorelical and experimental studies leading to
zero-order exploratory development. and upon mutual agreement, to first-order
feashility d?welopment and to second-order development of:
(a) New crystal rectifiers using new or previously untried principles.
Ca) New crystal rectifiers obtained by studied modifications of existing
principles.
The new crystal rectifiers shall be primarily intended and suitable for appli-
cation to broadband microwave mixers.
Since Task 8 is being supported solely by thn Signal Corps Procurement Agency,
and is of interest in a separate group of applications, it is being issued as a separate
report under Ccntract DA 36-039 sc-5589.
3
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
41111K:k' V T. 1 i
Declassified in Part - Sanitized Copy Approved for Release 50-Yr 2013/10/25 : CIA-RDP81-01043R002300090003-0
.NO4-110:-i.1, .72; ? ?:?:".
NZ e;41.1....-4:46.;
4
The zero-order research and exploratory development shall be carried on in
three phases:
(a) The development of a comprehensive design theory of the crystal rectifier
by extension of the p-n theory of rectifiers to the required frequencies and
signal levels shall be carried on in the first phase. The theory should
relate the large- and small-signal impedances to the resulting conversion
characteristics. Moreover, it should relate the large- and small-signal
impedances, the noise, and the power capabilities to the material and
structural properties of the device.
(b) The second phase shall include the exploration and evaluation of materials
and techniques. Examples of mater:ais 'night be silicon, germanium,
silicon-germanium alloys, or the intermetallic compounds. Examples
of techniques might be diffusion, alloying, remelting, or bombardment.
(c) The third phase shall include the fabricat:ou of exploratory crystal
rectifiers and a determination of their characteristics to check the
zero-order theory.
Upon mutual agreement the Task may include first-order feasibility develop-
ment which shall be carried on in the following manner:
(a) The first phase shall involve a refining of the zero-order theory in light
of the findings of (c) above. The theory shall be extended to include a
study of reliability factors, such as temperature, humidity, shock, and
vibration.
(h) The second phase shall constitute a choice of material and technology.
(c) In the third phase sufficient crystal rectifiers shall be fabricated to
determine feasibility.
The performar objectives to be sought by these designs shall be:
(a) Improvements in the order of two to three db for conversion loss and a
two-to-one reduction in noise ratio over present-day crystal rectifiers.
(b) Crystal rectifiers capable of operating over a 12 per cent bandwidth.
(c) Crystal rectifiers having substantial improvement in regard to burn-out
by high level pulses.
(d) Balanced crystal rectifiers and reversed polarity crystal rectifiers shall
be investigated.
Upon mutual agreement, the Task may include second-order development,
based on the information gathered in the zero-order and first-order development,
applied to the design of one or more specific crystal rectifers. These crystal
rectifiers would be specific in the sense that they would be designed to operate at
a fixed-frequency range and bandwidth between specific impedance levels and under
specific power levels. Particular attention would be given to the reproducibility
and reliability of the crystal rectifier. No development of the mixer plumbing
would be antizieated.
The specific development would lead to the gathering and writing of second-
order design information and the supplying of 200 sample models.
-rag! 41.
SECTION 2 - ABSTRACT
TASK 2 - TRANSISTOR RELIABILITY
This report contains two chapters: one on neutron radiation damage to certain
transistors and diodes, and one -in an attempt to correlate aging behavior with noise
measurements. Although neither of these studies was made on the subject contract,
they are reported because they are closely related to the field of Task 2.
A preliminary study of the neutron radiation sensitivity of five types of germanium
transistors and five types of silicon diodes is reported in Chapter 1. All measurements
were performed outside the neutron envirc:iment several weeks after the neutron ex-
posure. The extent to which the various devices were affected varied enormously.
Changes in resistivity, junction capacity, and lifetime (as manifested by change in
alpha for the transistors and change In forward current for the diodes) can be reason-
ably v.ell explained by the expected radiation damage changes in the bulk properties.
The universally observed increase in reverse junction current cannot similarly be
explained and apparently results from degradation in the device surfaces.
In Chapter 2, net!e measurements are reported that were made on silicon diodes
in an attempt to correlate these measurements with aging behavior in a short test. If
sucessful, the correlation would be an important aid in quick prediction of reliability.
Although no correlation was found, it is still believed that physical mechanisms pro-
ducing aging trends must be related to those producing noise; consequently, a cor-
relation should appear some day when the phenomena are more completely analyzed.
TASK 4 - NEW AND IMPROVED TRANSMISSION-TYPE TRANSISTORS
The work on development of transmission transistors is reported in three chap-
ters. Results on germarium diffused-base transistors are to be found in Chapters
3 and 4, which give, respectively, oscillator performance and parameter measurements
on the M2039 at 100'C and preliminary results on the low-level amplifiers M2055 and
M2058:' The preliminary studies of the P-N-I-P Silicon power transistor (M2036) are
in Chapter 3.
The high-temperature studies on the M2039 in Chapter 3 show that oscillator
efficiencies at 200 me decrease from 44 per cent to 3?. per rent as temperature is
increased from to l00?C. Parameter measurements also reported in Chapter 3
show that collector trvcrse current increases to about 150 microamperes as tem-
perature increaacs to 10trC. The 10(1-mc ce:rmon-emitter current gain decreases
about 1 to 2 db in the same temperature rise. Ohmic base resistance is substantially
?Although this work is being done under Industrial Preparedr.ess Study Contract DA 36-039 sc-62529,
It is reported here because of its bearing on future work of the Development Contract
5
flr-IQifi
'n Pirt
A
d for Release ? 50-Yr 2013/10/25 CIA-RDP81-01043R002300090003-0
Declassified in Part - Sanitized Copy Approved for Release
Mit\i/Arildija
'-i,-.4,47?4,44.41.111934.41MA.'!
50-Yr 2013/10/25 : CIA-RDP81-01043R002300090003-0
6
????? ??????
?
unaffected by the temperature increase. Low-frequency h-parameters vary In ap-
proximately the manner expected from theory.
A number of low-level germanium diffused-base amplifiers have been made, as
reported in Chapter 4. The mechanical design of the M2055 common-base low-level
r-f amplifier unit and the M2058 common-emitter broadband unit has been modified
to simplify fabrication and improve reliability. These changes have permitted the
fabrication of about sixty units with very encouraging electrical characteristics. Of
these units, 60 per cent had common-emitter short-circuit current gains (8) at 100
mc in excess of 12 db, and 33 per cent hate gain greater than 14 db. Low-frequency
1 + h21 is unsatisfactorily high at present, with a center around 0.05. Work on all
three of these diffused-base germanium types will continue under the Industrial
Preparedness Study Contract.
Laboratory models of the M2036 p-n-i-p power trans?stor are described in
Chapter 5. The structure has been modified by reduction of the collector area and
of the base-to-emitter spacing, and the base region has also been made thinner.
These changes have produced wafers which deliver more than 5 watts r-f output
power at 10 mc. Characterization studies have shown alpha cutoff frequencies in the
60- to 90-mc range as extrapolated from common-emitter measurements. A simple
equivalent circuit for this device has been proposed and approximately verified by
radio-frequency bridge measurements. Comparison of pulse and small-signal meas-
urements indicate discrepancies possibly associated with collector series resistance
or with nonlinearities in the device. The high-power performance capability of the
M2036 design has been demonstrated. Continuation of this development is planned.
TASK 5 - TRANSISTOR TEST METHODS
Chaptei 6 descrltes the present status of the 5-me to.2s0-mc Transistor Phase
Set. The following items are discussed:
(I)
(2)
(3)
Improved crosstalk performance in the measuring path switches
Signal-to-noise ratio data on the amplifier-modulator
Improvements in the synchronization system performance
SECTION 3 - PUBLICATIONS AND REPORTS
No publicatiols or reports were issued during this period.
ge.e..? ? :...??? g ? ?????? ? ?..
SECTION 4 - FACTUAL DATA
TASK 2 - TRANSISTOR RELIABILITY
Chapter 1
NEUTRON RADIATION EFFECTS ON
GERMANIUM TRANSISTORS AND SELICON DIODES
By W. L. Brown
1.1 INTRODUCTION
!Ti the interests of exploring the sensitivity of semiconductor devices to a neutron
environment, a group oi germanium transistors and silicon diodes were irradiated in
the Brookhaven National Laboratorics nuclear reactor. In addition to the author, a
large number of people contributed assistance and stimulation to the work. Their
contributions are gratefully acknowledged.
Pile radiation produces effects on semiconductor devices In several different
ways:
(1) Through transmutations of the semiconductor atoms to tfr-^nical donor and
acceptor atoms
(2) Through displacements in the atoms of the lattice (tannbardment damage),
the vacancies and interstitials thus produced acting like donors, acceptors,
and recombination centers
Through ionization in. the lattice, which gives hole-electron pairs in much
the same way that light does
4) Through surface effects, either by direct interaction of the radiation with
the semiconductor surface or through interac'ion with the encapsulating
container or its filling medium
The rate of generation of hole-electron pairs is epproximately g = 2 x
101611/cm3sec where R is the radiation field in roentgen per hour. With a life time
r and diffusion constant D this results in a junction current ij = 3 x 10-9R
amp/cm3 which for a lifetime in germanium on the order of ten microseconds is a
current ij = 1 x 10-1?R amp/cm2. None of the measurements to be described here
were made in the pile, so that ionization occurs only as a result of the radioactivity
of the units following the neutron radiation. By the time the units were measured,
R was much less than 1 r/hr, so that the current developed in this way was negli-
gible. In the pile, where the radiation field may be 106 r/hr, the added junction cur-
rent might be quite important.
(3)
This chapter contains copyrighted or copyrightable material not first produced
under a government contract and shAli be reprodt ced or used for governmental pur-
poses only and not for sales or disposal( n to the general public. This chapter con-
tains intelligence not first produced under a govel nment contract and shall not be
given to any foreign government without thp written consent ..;! the enntrArtc,r.
7
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
f.t1,4
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/10/25:
^ ^
8
?
???????? - ? "N?
With the usual spectrum of pile neutrons, and with temperatures sufficiently low
(less than about 200`C), so that a substantial part of the bombardment damage has
not annealed, the eff.:ct of transmutations is generally small compared with the ef-
fect of bombardment damage.
? The present experiments are thus concerned with body effects of radiation damage,
about which considerable is known and more can be reasonably estimated, and with
surface effects, about which essentially nothing is known. We will attempt to assess
to what extent one or the other of those is dominant.
1.2 THE DEVICES AND THEIR NEUTRON EXPOSURES
Five types of germanium transistors were irradiated with integrated thermal
neutron fluxes of 1013. 1014, and 1015 neutrons/cm2. Two 1893 point-contact units,
two 1868 p-n-p alloy units, two 1353 n-p-n alloy units, two 1859 n-p-n grown-junction
units, and one 1777 p-n-p alio.; unit were subjected to each of the three neutron doses.
Five types of silicon diodes were given exposures of 1015 and 1016 n/cm2. Two
2048 and two 2032 double-diode voltage regulator units, two 2025 and two 2029 power-
diode units, and two solar batteries wcre each given the two neutrons doses.
The flux level of the exposeres was about 2.5 x 1012 thermal neutrons/cm2sec.
Radiation damage is produced by the last neutron flux, which is roughly 30 per cent
as large. In the remainder of this report, the flux figures given represent the inte-
grated thermal flux as provided by Brookhaven National Laboratories. The temper-
ature during radiation was less than 60?C, and probably less than 50?C, in all cases.
Nuclear reactions in the semiconductors and their leads and encapsulating
envelopes made the units highly radioactive, particularly in the devices given larger
integrated fluxes and involving massive pieces of copper for cooling purposes. No
electrical measurements were macie on the units until several weeks after the neutron
exposures, to allow the radioactivity to decay to a safe level. Even short of studying
the tlevices during actual radiation, there is a large time interval which has not been
explored in the present experiments and in which there might be ionization-induced
junction currents and perhaps important effects of readjustment in surface chemistry.
1.3 INTERPRETATION OF CHANGES IN DEVICE CHARACTERISTICS
In this section the various types of devices are discussed individually. A large
number of parameters were measured for some types and only a few for others.
Particular attention has been paid to changes in reverse current, alpha, and junction
cap c it; for the transistors, and in reverse current, forward current, and capacity
for the diodes.
1.3.1 1893 Germanium Point-Contact Transistors
At exposures of 1013 and 1014 n/cm2 the characteristics of these units
were practically unchanged. At 1015 n/cm2 the collector current at zero emitter
current rose by a large factor, and the devices lost transistor action. These units
are made of four- to six-ohm cni n-type germanium (donor concentration -1 4 x
1014 /cc). Radiation damage by neutrons,accordir,g to Cleland,Crawford,and Pigg?in-
troduces net acceptors in n-type material at a rate 2 acceptors/cm3/neutron/cm2
Cleland, Crawford, and Pia:, Physical Review, Vol. 98, 1955, p. 1742.
? ? ??????????? .--????????
(two ac.ceptors/neutron cm). Consequently one would expect the n-type material to
conve^t to p-type between 1014 and 1015 n/ce12. The observed changes bear out
this expectation. Surface effects in these units seem to play no important role.
1.3.2 1868 P-N-P Germanium Alloy Transistors
At 1013 and 1014 n/cm2 these devices suffered a progressive increase in
reverse collector current, a progressive decrease in alpha, and, at 1014 n/cm2, a
measurable decrease in collector capacity. At 1i.;" n/cm2 the units failed, es-
sentially with shorts between emitter and collector. The decrease in collector ca-
pacity at 1014 n/cm2 indicates the introduction of acceptor centers at a rate approx-
imately two per neutron cm. The failure at 1015 n/cm2, as in the preceding ease,
can be attributed to conversion of the base material from n- to p-type. If the changes
in alpha (which before radiation was about 0.95) are interpreted as merely due to a
decrease iti electron lifetime in the base region, assumed to be 1-mil thick, at 1014
n/cm2 T 1 microsecond. The liietime due to radictior damage at 1013 n/cm2
(the change in net lifetime is so small the result lacks act. ,racy) is about ten micro-
seconds, agreeing with the notion tint bombardment introduces recombination centers
at a constant rate. There areerio measurements availa)le of ouip: lifetime change pro-
duced by neutron radiation, but one can make a judicious, if somewhat unreliable esti-
mate by analogy with experiments of lifetime degradation by electron radation damage
carried out by Loferski and Rappaport* in n-type gerinaniu n. Using the rate of change
of conductivity by electrons and by neutrons as a means of comparing their ellec-
tiveness in introducing ;:amage, and assuming the same ratio applies t.) degradation
of lifetime, one calculates a lifetime of about 0.3 microsecond for 1014 n 'cm2. The
agreement is at least as good as could be expected, and makes it plausable that the
alpha degradation does result from body-lifetime change.
Using the val le for the lifetime obtained from alpha (the change in base
resistivity is too small to he considered) one can calculate the change in 'co to be
expected due to change in hole current flowing across the collector junction. At
1014 n/cra2 the ncrease should be about 3 microamperes, about an order of magni-
tude less than observed. This it seems that while capacity and alpha changes are
consistant with expected change in body properties of the base material, the change
in collector current is dominated by some surface effect, details unknown. Clearly
one would like to re-etch he radiated units to check this point. Such measurements
have not yet been made.
Table 1-1
Changes in 1858 P-N-P Units
Parameter Measured
n/cm2
alcc Alpha aCe
1013 5tta 0.91 no greater than 10 volts at the maximum
;--
z
7100
.I collector characteristics of the transis-
tor are shown in Fig. 20. Little difficulty is encountered in meeting the maximum
1
;--
.-
...
E 9? voltage requirement. Negligible collector
2
current is drawn with zero base current,
gc I out to a breakdown in excess of 100 olts.
0 5X The high current performance does not
quite meet the requirement. There is a fall-
off of current gain at high current with
low collector voltage, as well as a series
collector resistance of approxiinately 40
ohms. Difficulty of control over the intrinsic-layer thickness is at present partially
responsible for the series resistance.
20t, 300 400
C0LLE"Cs0 J00ENT Vq....???.apEGE
F g. 2n - Corn:non-emitter
static characteristics
5.3.3 Collector Capacitance
Collector capacitance as a function of collector voltage for the revised
structure is shown in Fig. 21. The collector capacitance has been reduced by a fac-
tor of two from that encountered in the previous structure (see Chapter 11 of the
Sixth Interim Report). More careful etching of the collector junction v ill permit a
reduction of from three to four. As in the previous structure, there is no leveling
off of the capacitance as the space charge is swept into the collector region, as would
be expected with a step junction between the collector and intrinsic layers. Rather,
the presence of the diffuEe4 collector results in lower slope in the variation of ca-
pacitance with voltage tnan would be present witi.uut tiic collector (n-i junction only).
200
a
? I. tr,
-
-
70
60
3 50
4!) -
30 -
e
a
20
ci
3 4 5 t .? a 9 ': Z0 NO
COL L. E T00 I3ASE AI.
4.3 5:.? to: 7c as
Fig 21 - Collector capacitance as a function of voltage
000
flc.rI
rt
A
V
d for Release ? 50-Yr 2013/10/25 CIA-RDP81-01043R002300090003-0
140/WA.
Declassified in Part - Sanitized Cop Approved for Release
50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
'47.1-
34
It should be pointed out that this Is the total capacitance between the col-
lector and base. It will be shown later that only about a third of this capacitance
limits the amplifier performance of the transistor when used where neutralization
is poiAlble.
5.3.4 Small-Signal Current Gain
Figure 22 shows the small-signal common-emitter current gain as a func-
tion of collector current. Collector junction voltage was maintained at 30 .olts over
the current range. The current gain appears
Z 30[
.76
to remain high, from 20 milliamperes out
measurements by pulse techniques tend
to 500 milliamperes, where power dissipation
terminated measurements. More recent
25
to indicate small-signal measurements
made in this f,.1.ion are apt to be misleading
because there is a strong dependence of
current gain on temperature. Since the
z current gain increases with temperature, a
3 ,5
0 ?00 200 300 .400 500 rall-off of gain with current may be masked
by the temperature effect due to greater
power dissipation at high current. It will
therefore be necessary to investigate vari-
ation of current gain with bias current in
such a manner that, the average power dissipated in the transistor will remain con:rant.
The variation of small-signal common-emitter current gain with frequency
is given in Fig. 23. If the grounded base fa is calculated by the approximate relation
fa = f emitter/(1 - ao), an 1,, of 60 megacycles per second is obtained for transistors
UNIT 24 k -$7
COLLECT OP CuP ENT
Fig. 22 - Common emitter current gain :Ls
a function of collector current
50
Sc
----
VS
30 0
25 0
a
D
20 5
2
,
I . I
I
..0. T 235-'3
-
242-17
?
I
h--
-- ._
.
.
COvVC.,
EV "69
30B DOWN
G.4,...
__ -
___....--- -.
.
?
I
I
.
...
t
I
Z:Zr.
..... ???? / ? ? ord... .n..? ? ? ?
?
242-4 and 242-17, and an fa of 90 megacycles per second for 238-13. (The base-
layer width is 1.611 for 242-4 and 242-17 and 1.4ti for 238-13.) The actual common-
base fa is probably in excess of these values.
5.3.5 Proposed Equivalent Circuit and Measurements at 10 Megacycles
Since the transistor will operate at or near 10 megacycles, we desire an
accurate characterization at this frequency. Figure 24 shows the proposed equivalent
circuit, which is seen to be the usual high-frequency "T" pit-1 an element, Co, called
the outer capacity, to distinguish it from C? the inner capacity. Co + Ci = C, the
total collector capacity, which is the quantity mea.sured by the usual means. We have
dotted in a series-collector body resistance (r'o) in the equivalent circuit, since the
static characteristics show that one exists and is cf the order of :ens of ohms. How-
ever, this meas,,r2ment is made at a few volts and, if r'o is due to unswept w type
material, as seems reasonable, it should decrease with voltage and, in a properly
constructed transistor, be negligible at "3.. operating point. In any case it can be
treated as part of the termitnition, so no account of it has been taken In the following.
Without measuring trarismission properties, such as transfer impedances,
which is difficult, we can still get three intiepenuent parameters by measuring the in-
put and output open-circuit impedance. (zit and z22) and the input and output short-
circuit admittances (Y11 and Y22). The relationsip z11y11 = Z22Y22 allows a check.
These parameters are, neglecting negligible terms:
Z u -
Z22 ?
Co t-t
rto + r, + [rib (1 - a) + re]
Ct
?c-T + Jr, b o
1 - a ?Co
+--feb (1 - a) + re]
jcoCt C1
Ct .
+ Jr% totto
Ct
1 - a
Yn r'b (1 - a) + r, + icoCo
Y22 ? rib (1 - a) + r,
(r?b + r,)
+ jcoCo
where
a =
a0
ce
1 +j?.
toc,
Co
go
t j
.-a
VIO
3
15
0
3
01 0.2 0.3 0.4 05 06 08 2 3
...t.C.AcvCLES PER SECOND
4 5 6 5 9 .0
Fig 23 - Common-emitter current gain as a funct.on of frequency
?
Fig 24 - Proposed equivalent circuit (showing common emitter)
35
T ri in Part Sanitized Copy Approved for Release ? 50-Yr 2013/10/25 CIA-RDP81-01043R002300090003-0
601Airi
Declassified in Part - Sanitized Copy Approved for Release 50-Yr 2013/10/25: CIA-RDP81-01043R002300090003-0
- pt.
36
Measurements weee performed on unit 242-17. (This unit was selected at
random from the first few to meet the 5-watt, 10-megacycle objective, not for any
reason having to do with if to theory.) kTiqI, = 0.13 ohms at room temperature
(i, = 200 milliamps.) bo:, making allowance for the operating temperature and some
small contact and series resistance, we assume re = 0.16 ohms. Ct is taken from
Fig. 21 at the bias point of 30 volts where it equals 28 micromicrofarads. cda
Sc11 x 106sec-1 and I - a?= 0.0426 are taken from Fig. 23. The ratio Ct/Ci is
estimated equal to 4, from the ratio of the total collector area to the emitter area.
(The capacity under the space between emitter and base contacts should be connected
to the base contact, since almost all the base resistance occurs under the emitter
and very close to it.) Figure 19 gives this ratio as 3.3 but the value 4.0 was meas-
ured on the actual transistor. The only parameter fitted to the measurements is r'e
however, this procedure is fairly direct, since the conductance portion of yii is
closely identieal to lieb. All these numbers, together with r'b = 35.5 ohms, yield:
zii =
Yli =
znYii =
Z22
Y22
z22Y22 =
10.7 + j3.8 ohms
0.028 + j0.0020 ohms-1
0.292 + j0.127
86 - j35 ohms
0.0024 + j0.0025 ohms-1
0.294 + j0.131
Measurements of these four parameters were made on a Boonton R-X meter,
using appropriate transmission-line transformers. The measured values are:
Z
Yii
ZI1Y11
z22
Y22
= 10 + j1.5 ohms
= 0.028 + j0.002 ohms-1
= 0.28 + j0.062
= 65 - j25 ohms
= 0.0034 + j0.0028 ohms4
0.29 + j0.10
Z22Y22 =
The difference between the products of the measured values (zilyil and
z22Y22) shows that there are inconsistencies in the measured values comparable to
the differences between the measured and calculated values. The fit is remarkably
good, especially in view of the a priori nature of four of the five components of the
equivalent circuit. Error in mpactiromPnt5 resits from thr difficulty in providing
bias to the transistor and at the same time providing an a-c open circuit (zii and
722). For the citilittance measurements (yil and y22) it is necessary to provide an
a-c short circuit which is small compared to the input and output impedances. For
example, in the y22 measurement the base must be shorted by an impedance small
compared to 10 j1.5 ohms. This could not be done without great difficulty, since
the transistor header and sicet lead inductance has a reactance of about 1 ohm. In
order to obtain closer agreement between the equivalent circuit and the measure-
ments it may become necessary to include tht lead inductance in the equivalent
?? ?
circuit, if owever, until the inconsistencies in the measured values themselves can
be reduced, t!,ere Is little point to complicating the equivalent circuit.
In spite of the fact that eeondary effects such as lead Inductance have
been neglected, there is such good agreement betweeo the calculated and measured
values that we intend to adopt the equivalent circuit of rig. 24 as representative of
the M2036 at radio frequency.
The unilateral power gain calculated from the measured parameters for
unit 242-17 gives 95 or 19.8 dh -it 10 megacyclec pr compared to about
6 or 7.8 db given for transistero reported on in the pfevious quarter.
5.3.6 Oscillator Performance
An oscillator has been built to see if the M2036 transistor would meet the
objective of delivering 5 watts at 10 megacycles per second. The oscillator shown in
Fig. 25 has been treated az a common-emitter circuit with a pi feedback network,
(C3L3C ), between collector and base. The collector is grounded to a heat sink.
Power output was determined by measuring t1-1, voltage across the load resistor R3
by means of an oscilloscope. The values of components in the feedback network were
estimated so that feedback would be of proper phase and magnitude, using the meas-
ured real parts of the input and output impedances and an assumed 90 phase shift
from base to collector. The calculated values did not give the best oscillator per-
formance. This might be expected, since the calculations were based on crude as-
sumptions and on small-signal linear measurements, while the oscillator operates
over the entire nonlinear range. However, the values were within a factor of two
of the optimum and gave oscillation within 25 per cent of 10 megacycles per second.
By adjusting the components in operation for maximum output, over 5 watts at 10
megacycles per second was obtained.
Oscillation has been detected at 102 megacycles when the feedback network
was reduced to a minimum, i.e., wiring capacitance and lead inductance.
5.3.7 Pulse Measurements of Parameters
Because of the temperature dependence of current gain, it has been found
difficult to determine meaningful a - I,, and static curves of the transistor at several
??? C.-.0AES
AS
D!rt,f
? 12 ?
1_
8+
? ,}L'
- - -
S,E0 DO*' AND
p..AsE /ANC.
.fED,34c?
'P NE' WORK
C.'1.
R3. LOAD
RES
(a) c?pc?,T
R3.; C
I C3 di
(b supLt,EO oscr..L?Tna ca?ctot
PL 3DCN4$ 0.1L )
Fig. 25 - Schematic diagram of oscillator circuit
37
cQItIrI I
d for Rel
50-Yr 2013/10/25 CIA-RDP81-01043R002300090003-0
7 41.A0 'ff
Declassified in Part - Sanitized Copy Approved for Release 9 50-Yr 2013/10/25 : CIA-RDP81-01043R002300090003-0
**$
38
"
bias points. Changing the bias point changes the power dissipated within the transis-
tor, which results in a change of temperature and current gal,. This effect was
mentioned in Section 3.3.2 in connection with the current gain versus collector-current
plot. To provide a check on the measurements already taken, and to extend them to
high currents, pulse techniques will be used. Preliminary measurements have been
made which indicate that this method of attack will be effective.
5.4 SUMMARY AND FUTURE PLANS
Changes in design of the M2036 have been carried out and considerable improve-
ment has been noted in the electrical characteristics. Reduction of the area of the
emitter, the space between the emitter alit) base contacts, and the thickness of the
base region have resulted in lowered collector capacitance, Increased frequency cut-
off of alpha, and increased low-frequency current gain. These changes have also in-
creased the unilateral power gain by better than.a factor of 10.
An equivalent circuit has been adopted for the transistor. It is the usual high-
frequency "T" with an added outer capacitance between base and collector. The
validity of the equivalent circuit was checked by a close agreement between four pole
parameters calculated by means of the cquirilent circuit with the measured values
of the same parameters.
Oscillator performance now indicates the transistor is capable of meeting the
design objective of delivering 5 watts at 10 megacycles per second. Measurement
of static characteristics by pulse techniques has been found necessary and has been
initiated.
During the next interval, pulse techniques will be applied in order to mea..sure
transistor parameters at constant temperature. An attempt to reduce the labor in
characterization will be carried out. Fabrication processes will be looked into with
an eye toward decreasing the collector-body series resistance.
- - - ? ?
...nab ? ?????ll,
TASK 5 - TRANSISTOR TEST METHODS
Chapter e
REVIEW OF TASK 5 ACTIVITIES
By H. G. Follingstad and 0. Kummer
? ?????
6.1 INTRODUCTION
Development of the 5-mc to 250-mc Phase Set continued in this period. The set
will completely and accurately define transistors in this freqeency range, with mag-
nitude and phase of four insertion parameters. The latter are fully discussed in
Chapter 8 of the Second Interim Report. Design objectives and a functional descrip-
tion of the 5-me to 250-mc ",ase Set were presented in Chapter 'II of the Fourth
and Chapter 12 of the Sixth Interim Report. This chapter outlines the present status
of development of the set in terms of the block functions of Fig. 26.
6.2 MEASURING-PATH SWITCHES
Measurements on the transfer switch described in Chapter 14 of the Sixth Interim
Report indicate crosstalk between standard and unknown path to be 37 db at 100 me.
The switch has been redesigned and tests on a prototype model indicate crosstalk be-
tween standard and unknown to be lower than 70 db.
r-----, ,---, ?
\ ' f.. 4,..=_,- a. -fh,,
?
40, N...... SiD111.1., >` . Z01 ". -'..i
1 4- i --- , -L
?:_l
7.4.7C1
.TEPi
0 S C .LL ATOR
RE FE,41.CE
- ? -
05C *TOR
F441 ASE
NO
PotA SE
II. UAL
T
SA-ZG
,ND
MOW
,ATOQ =fs- fu
2
t Sr Nr: a
A
Fig. 26 - Block diagram of the 5-mc to 250-mc phase set
39
1-1
accif ri in Psanitized C Approved for Release ? 50-Yr 2013/10/25 CIA-RDP81-01043R002300090003-0
Declassified in Part - Sanitized Copy Approved for Release
WA:VA
50-Yr 2013/10/25 : CIA-RDP81-01043R002300090003-0
f*,
Am,*
40
6.3 AMPLIFIER MODULATOR NOISE TESTS
Over-all measurements on the noise level In the set Indicate a signal-to-noise
ratio of 20 db when the signal Is at the lowest opefalIng level of -80 dbm. Tests on
the amplitude indicator show that an error of about 0.1 db is produced by noise when
the signal is 20 db above noise. This should correspond to an error no greater than
0.6 degree in the phase Indication.
6.4 SYNCHRONIZATION
In the process of testing the 5-mc to 50-mc servo and electronic loops, the
system performance was improved by the following meanc
(1) Reduction in mechanical friction at the slave-oscillator control shaft.
(2) Increased slave-oscillator oatput in the presence of transistor network
loading.
(3) A monitoring circuit to accurately measure
deviation of i-f frequency from
synchronism and to tune the discriminator,
(4) Greater signal-to-noise ratio in the loop.
Important advantages in ...liability of operation are expected by the use of an
improved scanning technique which features indepen,lence of scanning speed from
servo loop gain, elimination of false synchronizing frequencies, and elimination of
discriminator offset toning formerly required for s.canning, purposes.
"p?eirp.
*My ? ? ? I...
? ?????????.???? ? *swot ?...e.+ ? ? ?????'"
SECTION 5 - CONCLUSIONS
TASK 2 - TRANSISTOR RELIABILITY
While a continuation of the direct determination oi dging behavior continues to be
essential it is also of the greatest importance to get a better understanding of the
physics of the processes. From the failure of the attempt to correlate noise with
aging, it is concluded, not that there is no correlation, but rather that the noise cur-
rents and aging rates depend on a complex of factors which need to be better under-
stood before ar.y such correlation will be evident. Work in progress toward such
objectives is mentioned in Section 6.
TASK 4 - NEW AND IMPROVED RANSMISSION-TYPE TRANSISTORS
Temperature studies of the M2039 diffused-base p-n-p germanium oscillator de-
sign have shown that 200-mc oscillator output power is decreased about a fourth by
increase of the operating temperature from 26?C to 100*C. Direct-current, low-
frequency, and high-frequency parameters were shown to vary in approximately the
way expected from theory. Measurements of initial models of the M2.055 and M2058
low-levet germanium diffused-base amplifier transistors indicate that these designs
will be satisfactory. This diffused-base germanium work under Industrial Prepared-
ness Contract DA 36-039 sc-72729 will continue.
The high-power performance capability of the M2036 diffused-base 5-watt 10-mc
p-n-i-p silicon transistor has been demonstrated. Initial characterization studies
indicate a satisfactory equivalent circuit can be developed. The improvements in
performance to give better than 5 watts r-f output at 10 me are attributed almost
entirely to the modifications in design instituted during the contract interval.
TASK 5 - TRANSISTOR TEST METHODS
The present status of the 5-mc to 250-mc Transistor Phase Set has been outlined;
development and prove-in of the brassboard model will continue.
41
no laccif ri in Partsanitized C A
V
ed for Release ? 50-Yr 2013/10/25 ? CIA-RDP81-01043R002300090003-0
_
4irk
Declassified in Part - Sanitized Copy Approved for Release
?
??? ?????
SECTION 6 - PROGRAM FOR NEXT INTERVAL
TASK 2 - TRANSISTOR RELIABILITY
-
tf
50-Yr 2013/10/25 CIA-RDP81-01043R002300090003-0
,,,m_tirie1114?
The direct determination of aging rates continues, since it is needed to predict
reliability in systems, particularly for the newer transistors. Among the tests is one
to separate temperature dependence and voltage dependence of aging rates. Statis-
tical procedures for the improvement of efficiency of data are being developed.
Attempts to correlate junction characteristics with surface properties, such as
surface potential and recombination velocity, are being made.
TASK 4 - NEW AND IMPROVED TRANSMISSION-TYPE TRANSISTORS
Industrial Preparedness Study Contract work on diffused-Lase germanium tran-
sistors will be concentrated on the M2039 oscillator transistor mechanical structure
and on the M2055 and M2058 low-level amplifier designs. Characterization of the
latter devices will be emphasized, with particular concentration on the high-frequency
performance. Noise and linearity studies will also be reported.
M2036 diffused-base 5-watt 10-mc silicon transistor work under Task 4 will
include use of pulse techniques in order to measure transistor parameters at constant
temperature. Efforts will be made to simplify characterization techniques. Modi-
fications of fabrication preeeeeee in order to decrease collector body series resi,A-
ance will be undertaken.
TASK 5 - TRANSISTOR TEST METHODS
Testing of the Phase and Magnitude Indicator as a unit will be completed. The
synchronization function will be proved in using an improved flexible scanning tech-
nique. Design of range switching and standard-unknown path switching panels will
be completed.
Construction of the prototype will be started by Stavid Corporation, according to
information obtained from the brassboard model.
t .
42
eilny Annroved for Release
z?-??????????'..2*
? 4.1.?
SECTION 7 - IDENTIFICATION OF PERSONNEL
Preceding reports under this contract have identified the engineers and scien-
tis;.s whose work has, contributed materially to the progress of the studies and
investigations conducted during the periods covered by the reports. During the cur-
rent period, additional personnel were itsz,igned to v ork on portions of this contract.
Brief biographies of these individuals are provided below.
WALTER L. BROWN
Walter L. Brown received his B.S. from Duke University in 1945, his M.A. in
1947, and his Ph D. in Physics in 1951 from Harvard University. He joined the Lab-
oratories in 1950. His principal interest at the Laboratories has been in the study
of surface states on germanium and silicon surfaces and in the effects of radiation,
primarily high-energy electrons, in introducing imperfections in the structure of
these two semiconductors. His current work is directed toward understanding the
significant mechanisms in the simplest type of radiation damage, namely that
introduced by bombarding particles whose energies are close to the threshold for
the damage process.
CARL HARRY KNOWLES
Carl Harry Knowles was born in Birmingham, Alabama, in 1928. He was in the
United States Marine Corps from 1946 to 1948. He received a B.S. in Physics from
Alabama Polytechnic Institute (Auburn) in 1951, and an M.S. in Physics from
Vanderbilt Un'versity in 1953.
Mr. Knowles has been employed at Bell Telephone Laboratories since leaving
Vanderbilt, and he completed the BTL Communications Development Training Pro-
gram in 1956. During his employment at the Laboratories, Mr. Knowles has teen
employed in semiconductor device development.
JERRY SEVICK
Jerry Sevick received his B.S. in Education from Wayne University in 1940.
From 1942 to 1945 ha was in the U. S. Army Air Force. During his service he
completed pilot training in Texas and radar training at Harvard and M.I.T. At the
time of discharge, he was a project director for radar devices at Wright Field. In
1946 he went back to Wayne University and obtained an M.S. in Physics. After three
50-Yr 2013/10/ . IA-R -
43
43R002300090003-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/10/25 : CIA-RDP81-01043R002300090003-0
I 11)1-
44
???? ???? ?? ? ? ? ??-?...
years of teaching at Wayne University, he returned to Harvard University for grad-
uate training. Mr. Sevick received his Ph.D. in Applied Physics in 1952 for a study
in the application of the variational technique to problems in electromagnetic scat-
tering by coupled objects.
After three more years of teaching at Wayne University and consultant work in
Detroit industry. Mr. Seviek joined the Bell Telephone Laboratories. At the present
time he is a member of the Device Development Department, assigned to the Tran-
sistor Development Group.
During the period covered by this report. 16 June through 15 September 1956,
approximately 5.000 man hours were devoted by key personnel to work on this project.
??????
DISTRIBUTION LIST
150 copies to: *,,ran3portation Officer, Signal Corps Engineering Laboratories
Evans Signal Laboratory. Building 42
Belmar New Jersey
Marked For SCEL Accountable Officer (Inspect at destination)
File No. 144-PH-55-91(1210)
150 copse' to: Departmcnt a the Navy
Bureau of Ships, Code 81613
Aashington 25. D.C.
Attention' Mr A M Andrus
20 copies to- Commanding General. Wright Air Developmait Center
Aright-Patterson A FB
Dayton. Ohio
Attention: Mr Ft. D. Alberts, WCRET-4
5 copies to: A J Busch. Bell Telephone Laboratories
5 copies to R R. Deutsch. Western Electric Company
2 copies to each of the following:
G W Gilman Bell Telephone Laboratories
L C Jarvis. V.estern Electric Company
A C Keller. Bell Telephone Laboratories
1 copy to each of the following:
A. T Day.d. Western Electric Company
F P 1.4ons Aestern Electric Company
H R Strobel A estera Electric Company
E. Temple. V.etern Lleetr;c Company
J E. Tueeddale. Aestern E:.ctric Con pany
11
A Lerkis. Bell Telephone Laboratories
D Lewis. Bell Telephone laboratories
A Lovell. Bell Telephone Laboratories
B. MeDavat, Bell Telephone laboratories
Molnar bell Telephone Laboratories
W. Morrison Bell Telephone Laboratories
HA
Ariel, Bell Telephone Lab-:atories
A. Morton Bell Telephone Laboratories
AE
Arderson Bell Tetephone Laboratories
T Mottram, Bell Telephone Laboratories
AF
Bennett. Bell Te:ephone Laboratories
1.
G D. Paterson Bell Telephone Laboratories
EA
Bescherer. Bell Telephone Laboratories
J. Pietenpul, Bell Telephone Laboratories
AK
Buhren Bell Telephone Laboratories
V.
T r-s Bell Telephone Laboratories
Id E B
H
Cave, Bell Telephone laboratories
Cook. Bell Tel-phone Labozatories
V.
V
E Reichle. Bell Telephone laboratories
L. Ronct Bell Telephone Laboratories
PS
Dainell. Bell Telephone Laboratories
F
J Singer Bell Telephone Laboratories
RL
Mosul& Bill Telephone Laboratories
Id
Spat.... Bell Telephone Laboratories
HT
Fri's, Bell Telephone Laboratories
A
Tradup Bell Telephone Laboratories
K. L
Gou'rl Bell Telephone Laboratories
L A allace. Bell Telephone Laboratories
W H C Higgins. Bell Telephone Laboratories
/4 %%est Bell Telephone laboratories
AG Jensen. Bell Telephone Laboratories
A H White, Bell Telephone Laboratories
51 copies to. R. At Ryder (Project Engineer), Bell Telephone Laboratories
45
r",
i - Sanitized Com/ Approv
for Rel
50 Yr 2013/10/25 CIA-RDP81-01043R002300090003-0