Skip to main content

Full text of "DTIC AD0025414: GOLD BONDED TRANSISTOR DEVELOPMENT"

See other formats


ml 



1 



g T:{/iNSIST01i PRODUCTS, INC. 

^ Gold Bonded Transistor Development 

■ Final Report, fi ja^ . 

Uontraot AF 19(6oM-Rlii 

A4« n ^ T't^ i. n A. 




BEST AVAILABLE COPY 





TRANS ISTOB PRODUCTS, mCORPORATED 
Snow and Union Streets 
Boston 35, Masso 



OOU) BONDED TRANSISTOR 

DEVELOPMENT 

8 January 195ii 

FINAL REPORT 

R^ A. Langevln 
Pro.lect Manager 



Prepared for Contract AF 19 (SOh) - 81 {4 
sponsored by the Air Force Cambridge Research 
Center 




TRANSISTOR PRODUCTS INC. 



ADMINISTRATION AND H£HSC»I,'IEL 



Personnel : The present contract began under the direction of 

Dr. C, Q. Smith who had joined the company as project manager 
of tlie gold bonded transistor development prograjm- Dr. Smith 
left the employ of the company in mid July and upon his depart- 
ure Hr. R. A. Laiigevin was assigned as project manager and lias 
continued in that capacity to the present time. 

The assignment of senior personnel to the project has been in 
the areas and for the {Periods indicated below; 

Dr. P. R. Rlcliards, Fac ) to face transistors 
(July - Deceinber) 

Pr. 0. Knight, Theoretical studies and humidity studies 
(July - Deceinber ) 

Dr. R, Johnson - Gold bonding studies 
(July - September) 

Mr. E. Gschwind, Mechanical design 
(July - December'' 

Mr. Do Hume* - One sided (type A) transistors 
(Sep'tember - December) 

General Comments; Generally spsalclng, no deficiencies of mat- 
erial, equip'^ nt or personnel liave hindered the progress of the 
contract with tlie exception of tlie need that developed midway 
through tlie contract for very lightly donor-doped gold wire- 
Considerable reluctance was found on the part of com-’ercial 
suppliers to fabricate donor-doped gold cf the type found nece- 
ssary for making transistors with tlie desired characteristics. 
Significant progress, however, has been made by the combination 
of de-doping techniques developed liere and specially doped gold 
wire which has been secured from commercial sources. 




TRANSISTOR PRODUCTS INC 



Page -2- 



Conferencea and CoTTmunlcatlons i During the course of tlie ccwjtract, 
several conferences were held between Mr. Boue, the Air Force ^ojeot 
Engineer monitoring the contract, and Mr. Langevin, Project Manager 
for Transistor Products, Inc. The fir&l of U^se conferences, held In 
late July, was intended for general orientation and at that time our plans 
for work on the contract were presented to Hr. Bowe. A subsequent meeting 
in late September was attended by Messrs. Bowe, Ryan and Mueller from Air 
Force Cambridge Research Center and Messrs. Langevin, Richards and Rumes from 
Transistor Products, Inc. This September meeting served to review progress 
of the contract up to that date and at tliat time an outline was given of 
the work planned up to the termination date of the contract. 

Mr. R. Nelson, Comptroller of Transistor Products, Inc., accompanied by 
Mr. langevin, visited Mr. “adile, Uie Air Force Contract Officer, to discuss 
general aspects of contract langxjage, and audit policy etc. 

On December 2 , 1953, Mr. Bowe and Lt. Rutherford from Air Force Cambridge 
Research Center visited Transistor Products, Inc. and a general conference 
was held covering all phases of the work done on tie contract. Transistor 
Products, Inc, was represented at this conference by all of ttis senior project 
people previously listed. At the same time, Mr. Bowe was provide! with a 
sample of fifty-four gold bonded transistors which hiad been made by the 
techniques developed during tlie extract period. An additional fifty-seven 
gold bonded transistors have been supplied at the terraiucition of the contract. 
In aauition twenty-five sample lace to face transistor assemoiies have been 
supplied to the sponsoring agency for their expermmental use.. Fiaally, a 
sample of twenty to twenty-five symmetrical gold bonded transistors have 
also been supplied to the sponsoring agency. These transistors, altliough 
not specifically called for by the contract, are considered of consiilerable 
theoretical and practical Interest, particularly for discriminator applica- 
tions in the megacycle region. During the course of the contract, a proposal 
was submitted to Mr. Bowe outlining a development program for a ** junction-like” 




TRANSISTOR PRODUCTS INC. 



Page -3- 



translBtor idiich viU give usefal gain at frequencies of the oorder of ten 
negaeTolee. This propcaal, which was mbmltted on 28 Septeinber, grows, 
naturalljr, o\xt of and, we believe, is tiie logical next step to follow, 
work done on the present contract on "point contact" gold bonded transistors. 
The proposal contenpXates a program somewhat more general in scope than the 
present contract to permit a study of materials other than gold in the 
bonding proosas and It is further proposed Uiat the decision as to the 
speoificatlana of tbs transistors to be fabrloated be deferred until the 
oowplstion of a preliainary etudy phase. It is belisTsd that by such 
■sans the aost rapid progress can be made toward useful end derioas and 
that this procedure will provide the fullest poesibis exploits ticn of bonding 
techniques. 




TRANSISTOR PRODUCTS INC. 



R EyiEW OF THk: DEVr.LOWE Wr 

Pro blem* t Tl»e coitrae of the development work on the present contract 
car. most readily bs outlined by cliscuasing tlie work done on the fol- 
lovriLng nine major problems which liave arisen in the course of ihe con- 
tract work : 

1. Theory of alplia in two-slcieu transistore. 

2. Theory of freqxxency cutoff. 

3. Study of gold bonding techini-iuea r 

h. Methods of obtaining thin base layers. 

5- Huridity studies. 

6. ?'ethods for obtaining lightly doncr-doped gold wire. 

7 . Fabrication of experimental tranalstorso 

8. Methods of packaging two-sided units. 

9. Special apparatus developw’ent 

a. Airbraslve unit 

b. Gold bonding set 

c. Noise figure measuring set 

•' Methods of Atta cj-: Significant progress has been made on all of the 
above probloma with the exception of S'* '-‘ethods of packaging 2 sided 
units. The theoretical problems 1) and 2) h»ave been solved ior the 
case where diffusion is the only active transport nei.l>anl8ra . Idealiatnl 
geometry has een utilized as much as possible to rake the probiejns tract- 
able although it Is believed that the idealization still retaijis the relevant 
features of the geometry of the actual devices. Details of these theoretical 
studies will be found in Appendix A. 




TRANSISTOR PRODUCTS INC. 



The gold bondirxg study, 3) 1ms beencarrled out primarily on diodes- 
This has seemed to be ti» best way to minimize the number of active 
variables in the problem, Riperimenta have been made primarily to obtain 
a comparison of "resistance bonding", in which tha gold wire, initially 
separated from tlie germanium is advanced until an arc strikes effecting 
the bond. Details of this work will be found in Appendix B. 

Theory indicates and experience confirms that in order to obtain a 
high^;7^ and high cut-off frequency in the two-sided units it is 
necessary to have base layer thicknesses of the order of »00l" or less. 
Attempts to secure base thicknesses of this order by masking and etching 
and by lapping either fail or produce structures so fragile tliat they 
cannot be handled. This problem has, however, been very adequately solved 
by the use of an S. S. V/hite Airhrasive unit. With this machine it is pos- 
sible to bore wells or troughs in germanium wafers in .such a way that ihe 
thin section is essentially supported by thick material and hence can be 
handled wit-h ease. Base layers of less than ,'X)1" can readily be obtained 
by this process. 

Concurrent with the development work on gold bonding and gold bonded trans- 
istors, a limited study lias been carried on to determine the effects of 
liumidity on gold bonded contacts- At tiie initiation of the contract tiiere 
tiad been some hope that gold bonded contacts by virtue of tlie intimate 



contact attained between the gold and the germanium ndgtit be less sensitive 
to humidity effect.s than pressure contacts. Unfortunately, humididty 
studie,s on gold bonded contacts have not suoported tliis hope- Generally 
speaking, the rectifying gold bonded contacts are similar to pressure 
contacts in the adverse effects of high humidity; however, sane success 
in minimizing these effects has been had with siiecial surface treating agent 




TRANSISTOR PRODUCTS INC. 



Oeta-lls of the work earrlod out in tills humidltj study will be found 
in the Attached Appendix C. 

As work on tbs ccntrect progressed and variously doped gold vlree were 
studied experimentally it beoaiee liicreaeingly evident tliat transistors with 
characteristics as contemplated in the contract could not be obtained unless 
it was possible to obtain donor doped gold wire with extremely low concen- 
trstions of the doping agent. Initial efforta to seoure antimony dc^%d gold 
with doping paroantagca ranging down to .OOli or less from comnercial sources 
were not suocessfal. Consequently, several a^proaclies were utilised in an 
attempt to fabricate such wire oorselvee. The first apfshsecb was based 
on the use of comwrolally available li antlsony-doped wire, olnce the vapor 
pressure of sntlnaay is higher than that of gold, it should be poesibls to 
"de-dope" such, gold wire to any desired extant by beating the wire in a 
vacuus and essentially boiling off sone of the antimony. 

Experlaienta on this de-doplng enabled us to secure smsU 8aq)leo of de-doped 
wire with which transistore of desirable characteristics could be fabricated. 
However, because of inhonogeneity in wire doping and strain configuration, 
this procedure has not proved practical. Coneaquently, doping was attempted 
by usli% essentially pure gold which had been dipped in various antimony 
solutions. Results obtained by this method were promising but were abandoned 
in favor of a method which oonslated of moving the gold wire continuouely 
through a furnace. In the neantine, we were successful in obtaining antimony 
doped geld wii^ in which tbs antiisony dope was supposed to be = 01,^ = with t.hi 
wire, moderate success has been had in tr'anslstor fabrication but the deeir- 
abllity of still lighter doping has led us to continue our de-doping efforts 
to the termination of the contract. Details of the doping studies will be 
found in Appendix D. 




TRANSISTOR PRODUCTS INC. 



ExperiiMntal gold tended tx^iuiiBtors of two general types have been 

nadet The flrat of these, the ^'one«sided” or type A ml%, is essentlall 7 
a oonventioDal Vfeatem Electric eartrldge type point contact transistor In 
ehleh the phoepher bronse and beryllium copper whiskers have been replaced 
by variously doped gold vires of .002" to .003** In diameter. A more uncon- 
ventional but more Intereatlng type of device is the 2 aided or "face to face" 
unit in which emitter and collector wire, are bonded to opposite faces of the 
germanloai wafer. These unite are aeaenbled on 3 pin glaaSHvetal headers 
almilar to those being used by tnaijy nanufaeturexa for junction trenelators. 

A limited attwnpt has also been leede to fabricate the one added units on 
the three pin glass metal headers. A detailed dlseussioD of the transistor 
fabrication will be found in Appendix E. 

In additioc to tlie problem of building experimental transistors of the 
face to face and one sided type for the evaluation of germanium resistivity, 
verioQS doping agents, etc.. It tiss been necessary to consider the prte>len 
of packaging the gold bonded transistor into structures that are mechanical ly 
rugged, stable, resistant to humidity, etc. This has been adequately accom- 
plished on the one-sided unite by the use of the Type A cax*tridge. The two- 
eided units howerer have proved to be too fragile for packaging by any 
means so far tried. This fragility is piri'narlly attributable to the fact 
that with the donor doped gold wire so far used for the collector It Is not 
possible to obtain mechanically strong bonde and at tite same tinte secure 
transistor characterlstlca in the desired range. In addition to the point 
contact-like transistors called for by Exhibit A of the contract a limited 
number of syTuetrical transistors of the face to face type have been fabrica- 
ted using 2S gallium doped gold wire for tlis emitter and collector. Samples 
of these xmits have been supplied to the sponsoring agency. 




TRANSISTOR PRODUCTS INC 



Special Apparatus and Equipment 

^*ention has already been made of the S. S. Wh^te Alrbraslve unit. This 
machine, made by the S. S, White Dental ^'fg. Co,, N* Y. C-, hae been 
indispensible in cutting wells in gemaniam wafers to obtain tliin base 
sections. Considerable control of the cutting process has been obtained 
utilijiing a timer to give short (.01 to 1 second) blasts of abrasive. 
Empirically, a nun>ber (5-10) of short blasts of equal duration seems to give 
the best control of the cutting process . 

To facilitate the bonding and testing of the experimental "face to face" trans- 
istors a standard gold bondirig ptilser lias been modified to permit bonding 
and approximate measurement of (X. and in the same set. 

In order to obtain noise figures on the experijnental transistors fabricated 
in the course of the contract, and on the sanples supplied to the sponsoring 
agency, a transistor noise figure measuring set has been constructed. De- 
tails of the gold bonding set and transistor noise measuring set are con- 
tained in the Appendix F. 




TRANSISTOR PRODUCTS INC. 



Status at Termination of Contract 

Theory ; Calculation of the maximum attainable(^ in a face to face trans- 
istor suggest that a limiting value of about OC ~ • S best tfiat 

car; be expected for collector and emitter of equal cross section and accepter 
doping In both emitter and collector. Cut off frequency calculations, assum- 
ing diffusion to be tlie only active transport neciianism suggest that 10 mega- 
cycles can be attained for base layer thicknesses of the oilier of .001". The 
fabrication of experimont^il transistors, as will be described subsequently, 
bus provided general confirmation of these results. 

Gold Bo n ding Studies ; The work done to study the comparative behavior of 
resistance versus "arc" bonding has been very inconclusive. Results obtai.ned 
by the arc tecluiique have not at any time been significantly better t;.an 
those obtained with ccnventional res' stance banding tecluiic,ues . On this 
account further work on this problem was discontinued. 

Thin Base Layer : As already mentioned, thin base layers ranging down to 

.001" can be easily obtained with Uie Airbrasive unit. While no specific 
reproducibility studies have been made, it is our strong impression tliat 
the base thickness can be controlled to ±)out ,0005" in this process. 

Kuwidity Studies ; The basi., object.ive of the humidity studies has been to 
ascertain wl;ether gold bonded contacts exhibit in the presence of moisture 
any characteristics significantly different from those observed with conven- 
tional pressure contacts. A general discussion of results in Af^pendtx C shows 
that geld hondod contavCts are unfortiinat^’iy no better in this regard. 

Methods for Obtaining Lightly Donor Do ped Gold Wire ; Ti/ree methcnls liave been 
utilized for the local preparation of lightly donor doped gold wire; 

1. De -doping antimony doped gold wire by passing current 
tlirough the wire in vacuum 

2. Doping "pure gold wii^ by dipping into various aqueous antimony 
solutions . 

* 3« De -doping antinemy doped gold wire by moving it continuously 

titrough a resistance furnace in air. 




TRANSISTOR PRODUCTS INC. 

The firet of the bo techniques has proved unsatisfactory du<* to •narkei in- 
honogeneity In doping level and vire diameter resulting iron Uie process. 

The dipping technique, while promising, has been abandor-od in favor of tlie 
third, continuous processing technique. By this latter means It tias beer 
possible after considerable experimentation to de-dope antimony doped gold 
wire in relatively hon»ogeneou8 fashion without introducing notlcable varia- 
tions in wire diameter. 

Fapricatlon of Experimental Transistors ; By utilising the techniques 
developed in the course of the deveiopment work it has been possible 
to fabricate gold banded transistors as contemplated in the contract. 
Experimental transistors have been of two types, a one sided unit similar 
structure to a conventions] "Type A" point contact transistor and a 
two sided or "face to fane" unit which offers structural and electrtcal 
adventa^s. 110 samples of the one slde.i units togetlj^T with canplete date 
JiSi'e been supplied to the sponsoring agen: ,/ In addition two sided trams - 
istors, of a spieclal "symmetrical" type he'iv’ been supplied in sample quantities^ 
These units are characterised by the fact i at they Jiave an emitter and col- 
lector with nearly identical electrical c . iracteristics . 

Packaging of Two Sided Units ; Gold bonded ransistors of the type contemplated 
in the contract require the use of suitably donor doped collectors. Very 
light bonding is reqxiired with such wire to :>btain the desired electrical 
characteristics. In the two sided units th fragility of this bond lias 
made it impractical to "package" these unit: satisfactorily. As a con- 
eequence it riHB nut be-jn feasibis to supply t . spies of this second type of gold 



bonded transistor before the termination of a contract. 

Special Apparatus Dev elopment ! No difficult! >5 were encountered in modifying 
a gold bonding pulse set and fabricating a tr insistor noise figure measuring 
set for the purposes of this contract. The ei{uipment has been used in the 
course of the work and performs satisfactori . 





TRANSISTOR PRODUCTS INC. 



SUHKARI CP RiiSUrrS 



During tha course of devslopment work it has been established that trans- 
istors of the type contemplated in the contract can be made using 
galliuB or indium doped gold wire for the emitter and lightly antimony doped 
gold wire for the collectoi*. It lias been determined that the preferred 
antimony doping is less than .01$. Ccxxsiderable work has been done on 
specific techniques for preparing antimony doped wire with the desired doping 
ISTel. Bonding techniques have bean developed for obtaining transistors 
with the desired characteristics. Transistors have been built using the 
methods and materials developed in the course of the contract and a siz- 
able sample together wlUi data supplied to the sponsoring agency. In ad- 
dition it has been poesible to supply samples of a "symmetrioal" transistor 
which, although not contemplated in the present contract, is considered of 
considerable theoretical a nd practical interest. 

REC0M”ENDATI0HS 



On the basis of the work carried out to bring tlie contract to a successful 
conclusion it has become evident that in order to fully exploit the knowledge 
aixl techniques developed to this point it would be desirable to vuidertake 
farther development work in directions outeide tlie scope of the present con- 
tract. In particular, the following specific recommendations appear to 
be warranted t 

1. Evidence has accumuLited that the process of gold bonding may 
well involve primarily a recrystallization pi'ocess from the 
germanium -gold eutectic. Such a process would be critically 
dependent upon the nature of the current pulse utilized in the 
bonding operation. While the bonding equipment used in the 
course of the present contract has made it possible to fabricate 
satisfactory transistors the possibility exists that much more 
elaborate bonding techniques affording full control of the bonding 
pulse Tray permit more adequate control of the recrystalllBation 
phase of the bonding operation and algnlfloantly Improve the yield 
in the fabrication of gold bondad transistors. 




TRANSISTOR PRODUCTS INC 



2o The techniques developed in the course of the piresent 
contract for obtaining lightly antimony doped gold wire 
for collector use Itave been e ff active on the practical 
level. It would be desirable to detf^mine quantitative- 
ly the optimal doping level. In addition the present 
tectmiques for de-doping leave tlie gold wire extemely 
soft. While this interjxjses no real barrier to the fab- 
rication of gold bonded tranei store such as have been made 
for the present contract, should such transistors be re- 
quired in largo quantities it would be helpful to develop 
soi table means for hardening the wire without upsetting the 
doping level. 

3. It will lisve been noted in the preceeding discussion that it 
was not considered feasible to adequately package tlie two 
sided gold bonded tran8ld>ors because of tlie fragility of the 
collector bond. While the one sided samples supplied are 
considered adLquata there are meciianical and electrical 
advantages to the two sided units which would make it desir- 
able to undertake iiie necessary packaging development to 
learn hew to adequately package the two sided units. 




TRANSISTOR PRODUCTS INC 



A ppendix A 

Cut’^ff Frequency Calculation 

Problem: To determine tl»e cat-off frequency for tlie transport of 

lio't’r from emitter to collector assOTlng the transport to tai>, 
place solely by diffusion. 

Solution: For this problem a one-dlinonsional georneti’y shoula be 

adequate for determining tli« correct orxier of magnitude of Uie rut- 
->ff frequincy, 

rne reli-r.lon curre>r i-ti'. ty, j «i. i ,.oif • r-.nt’-ii' luu, 

p, In ♦lie iim ii'~i I « i Va 

, ^ - k t//i j . 



. J» ' "*■ "Ole r-bllfty. •• tre rlead. state must be 

Insiio.^ i.e t ti/ - ■ • rp''f'i’-'.itnat1oi ran b-' negiecteo ao we mav re - 

wri .e V as 




1^“ 



dx 



( 2 ) 



where tne coefficient of ix is a constant . The boundary conJitions 
are; hole current Jp introduced at the emitter (x - O'; hole current 
3p rem V ed (l.e, p " O'* at tlie col.lector (x " L) The solution of 
(2) is Uierefore 




k.T'^P 



(L-X) 



( 3 ) 



If recombination is considered, thie continuity equation must be used: 




U) 




TRANSISTOR PRODUCTS INC 



Piige -2- 

whero is the lifetime of holes In the germanium, combining equations (I") 
and ()j) gives 



P ,-r- h 



The solution of this equation must have the form: 



- 



p = f^Q ~h 6 S 



Differentiation gives 






while applying tf:e oouridary condition p " 0 at x • L gives: 



- 

B 



r - /O 



and the boundary condition 



dp\ ^ ^ 

dxl)t.~o /CT^p 






^ 7 



C H- 



The solution is therefore 






^ _ 



kT^pc<. 



(10 




TRANSISTOR PRODUCTS INC. 



Page -3- 



If CcL. I thia reduces to 






^ — 
G. 



[a-iL- 




(u) 



which is idf^ntical with the solution obtained by neglecting 
recCTnbinatlon. Since may he ^ /O ^sec for a good single 

crystal of gen»ianium (cf. Shockley, iilectrons and Holes in Semi- 
conductors, p. 68) and since klP^C^p ■ Ui cm /sec for jerrnaniura 
at room temperature (Conwell, Proc. 1.8.1., Nov. 1’5>2) 

f 

O.L- < — - — > V ~ (12) 

so recombination can be neglected for (_ « 0.07 cm. 

In order to determine the cut-off frequency, a solution of tlie 
time dependent equation. 



It 

dt 



ilt 



(13) 



is required. If a periodically varying current, 

OJt - J ri J,ap 0 ^ 

wliore lyj ( 'j indicates "ttie imaginary part of ( ' ", flows at x “ 0, 
tJie boundary conditions are 

c/k I,.,, 




TRANSISTOR PRODUCTS INC 



Page • I4 



and p “ 0 at >v “ L. Assuming a solution of th^ form 









we have 


2 CO .UL >^) ~ k 

dK^ 


( 16 ) 


The function 


(.X '' must have the form 






- c e'*' + 0 e'‘”^ 




80 that 


iu) ~ k~r 1 


(13) 


or 


, ( LOi y'‘- 

■* ' \ ) 




Application 


of the boundary conditions g -ves 






t 

' - kic-oj 

M 


(19) 


and 


^ 0 e ' ^ 


iPO) 




D - 
c 


(21) 




.4 6 


(22) 






(23) 




TIUNSiSTOR PRODUCTS INC 



Lti 



, J 






.-/■>(X-L) 



{2h) 



^ ! 






g-ii. 






(^=5) 



Fo(‘ bL^si , this becomes 



(which is identical with the steady steti solution.) 



(i^6) 






!J?;i 



for bL < 1 



4, j -id J I 

I. '^:cihL)^ J 



V c- u ; 



J { ^ ■ - -•-- 

/ -»■ i- / bdL^ \ ^ 



i 



2 I - 

^ kT J 



(i’?) 



Therefore at i. “ L 



//)' 



..^Ar j , + ?• X fci. . f 

/ + -!./ WL> V •■ ST N- 

^ { ~:t“' i t 
' w: / .^, v 



(30^ 



i 



and are cnt oJ‘ phasa^v/hare the Eip: rozinat: o)i is valid to ./ithin IS for 



/ u> 

K /< 







TRANSISTOR PRODUCTS INC. 






Evidently tae magnitade of the C'orreut Et x •• L i;- tir»en 2C^ 

less ttion thr current at-X *-D-» ®rsd itr niiaco tlzn ctiier 

o 

27 • Thus U>e order of niagaityde of trie cut-off freiiuen'iy ia 

•i/y 



>aU _ 






V . 



z 't C'’' 



/ - 



-/ = 



-1 



*/ .<*^7 * /, r T ^.-7 






bj,'- about 



( 31 ) 




TRANSISTOR PRODUCTS INC. 



Appendix B 



Stiidy of Bonding Techniques 



Introducti on -• At the beginning of the pro.lect it was suggested 
that bonding performed by establistii)ig contact between ■noiten 
gold and molten gexTraniun might resu '.t in bonds superior to those 
achieved by first making contacting between gold wire and german- 
ium and then passing curTi^nt through tlie contact. 



The general scheme suggested is shown belcry: 






X 






-I- C 






d-oLp 



.y 



On moving tiie gold toward the germanium an arc will be established 
through 'A’lich the condenser C will b"* discharged. It was hoped 
that by varying C , R, and the magnit ide and polarit> of 2 some 
control of the bonding process could be obtained. 

Poss ible Advantages of "Arc" Bonding 

a) Possibly greater uniformity of bonds could be aoliieved 
since the bond should be largely independent of the initial con- 
dition of the germanium and gold surCccee, In the method present- 
ly used (resistance bonding) variation in >he resistance of the 
initial contact may contritiite to tiie non-uniforinity of the 
resulting bond. 

b) Variation of C , P,, E, and the polarity of E can be used 
to preferentially heat either the gold or the germanium - ttiis in- 
dependent control is not available with iiesistance bonding where 
the heat is developed at tlie gold -germanium interface. It is pos- 
sible that bonds of greater mechanical strength could be made by 



this method 




TRANSISTOR PRODUCTS INC. 



Page -vi-' 



Preferential Heating of Gold or Oerrianium - The energy dissipated in 
the ^Id and in the germanxvin can be calculated if tiie arc is assumed 
to liave a fixed voltage drop, 



Energy in Germanium 



n-/eVe‘ 



where Rg is the spreading resistance of tiie gemaniun. 
theoretical value 



It lias the 




8 r 



Here is the resistivity in olm cm and r is the e ffective radius 

O 

of the contact in cm (say 10 cm). 

Similarly, (wljen the gold Is used as the anode) 



Energy in gold 







Substituting, we have 



Energy = RiC ~ Ec^c) 



and 



Energy germanium " \ ~ a^c) C 



The fact U»at either tlic gold or the germanium cain be preferentially 
heated can be demonstrated by computing the ratio 
E yf ^ ^ yc 

Thus at a fixed value of C the fraction of the total energy which 
heats the gold can be controlled by varying the initial voltage on the 
condenser E and the series resistance R. It is thought that this degree 
of control is not available with conventional resistance bonding. 




TRANSISTOR PRODUCTS INC 



-J- 



The above approxlnate Xvormulae are not suffloiently accurata for 
(tuantitative design of a bonding nethod| since in addition to the 
assumption of a fixed arc voltage drop, we liave neglected the tem- 
pera tiire dependence of the spreading resistance and the arc c^iaracter- 
istice. 

Magnitude of Rlectrostatlc Forces - Originally it was thought that the 
electrostatic forces between the gold and germanium would be on import- 
ant factor In establishing the mechanical bond. Analysis shows that for 
the range of voltages used tlie arc is always initiated at distances much 
larger ttian t.iosc where the electrostatic forces would become apprecia- 
ble. The electrostatic forces arc sufficiently great to close the gap 
when the separation is K ^ where E is tlis voltage, 

K is a constant related to the electried ca, acity of t tie gold-germaniUj.i 
system (K “ Cd where d is the separation), and S is the spxdng constant 
of the p«ciianical system supporting the germanium and gold point. For 
E ■ 100 volts and estimated values of K and S for the system used, elec- 
trostatic closure of the gap would be expected at distances of the order 
5 X 10”^ cm. However, the arc starts when tlie separation is ~ where 

y the breakdown field sti*ength (about ic/* volts/cm) which gives 
-2 

10 cm for the striking point of the ax. Accordingly, the arc starts 
long before the point is close enough for electrostatic forces to be 
of sufficient magnitude to aid in establiealng contact. 

Mi scellaneous Calculations 

A) - The energy required to bring a segment of gold wire of diam- 
eter .003” and length “ .003" to xhe melting point is approximately 

-L 

2x10^ joules. 



B) - The energy required to raise ttie temperature of an equal 
volume of germanium to the melting point is approximately the same 




TRANSISTOR PRODUCTS INC. 



Page -U- 



C ) • The energy aTa.« lable from a 1 uf condenser c^larged 

-3 

100 Tolta is 5 X 10 joules. Therefore, getting sufficient 
energy is not a problem. 

D) The time leqtiired for a heated region of gold wire 
.003 Inches long to lose lialf of its heat by conduction is 
approximately 6 milliseconds. 

E) - Using D above and the breakdown strength of air, 
the gold point must be moved toward the germanium at a velocity 
of greater than 2 cm/sec in order to make contact before the 
gold Is cooled. 

bdt-perlmental Results 

A. Polarity Effect - It has been verified experimentally 
that more energy is dissipated in the gold when it is used as the 
anode of the arc. For E ■ lOOv, R '• 0, and C * 6 uf. None of 
the gold melts when it is used as the arc cathode. On reversing 
pjolarity, quite a bit of gold is vaporized. Under these condi- 
tions a crater is vaporised in the germanium for either polarity, 

presumably because the heating in the germanium is largely due to 
2 

I R loss in the spreading resistance. 

B. Cptimum Bonding with Slow Velocity of Approach - Bonds 
were attempted between etched germanium of 5 ohm cm. resistivity 

and 1 % gallium doped gold wire with the following results. The gold 
was broxight up to tlie germanium slowly in tltese expertnents . 

1. ) At low voltages, (E ■ 22v)with R • 0, bonds can be made 

with the gold as tiie anode. Increased meclumical strength and increased 
reproducibility were observed with increasing capacity from 1 to 10 uf. 
However the strength and reproducibility were not good under any con- 
ditions . 

2. ) With E - U5 volts, R - 0, Au ♦, C • 2 to 10 uf, good 
bonds were achieved with fair C(Mialstency. Microscopic examination 




TRANSISTOR PRODUCTS INC. _ ^ 

Page -5- 

showed both gold and germanium had melted. 

3.) With E • 90 Tolta, R ■ 0, Au +, C “ 2 uf or more, no 

>2 

bond was made although a large (3x10 ) cm dia.) crater was 
melted In the germanium and Uie gold was obviously melted. Evid* 
ently, the gold never came In contact with the gemanlura. Pre- 
sumably, using a greater velocity of approach would do better here. 
Thie introduction of series resistance made It possible to achieve 
bonds altho\;gh Uie reproducibility was not nearly as good as in 2) 
above. 

U.) At higher voltages large quantities of gold were vaporized 
and no bonds achieved with R ** 0. 

C. Sane bonds were made using a higher velocity of approach 
(about Scm/eec) with fair to good mechanical strength and repi*oduc- 
Ibility. However, this method did not appear s\ii table for mass pro- 
duction - particularly at the narrow point spacings required for 
transistors . 

D. Electrical cliaracteristlcs of diodes made by tliese arc 
bonding techniques were checked and found to be essentially sim- 
ilar to those obtained on diodes rtrde by resistance bonding. 




TRANSISTOR PRODUCTS INC 



APPENDIX C 

EFFE-rTS (F HUMIDITY ON BONDED G'^U -CiERMANIU>^ CONTAC TS 

In connection with ttie development of gold bonded traiiaistsjrs 
it has seemed important to undertake a limited program to determine the 
effe: .3 of liiimidity on oonded gold -^jer-maniu*' c ontacts ana tc make a 
curs sry search for effective moans of reducing th' 'e effects. / 

The most Ijnportant effect of humidity on bonded gold-gemanium 
contacts is a marked decrease in reverse res stance which occurs when 
moiGT.ore comes into contact with the bond. The effect, similar to that 
observed for pressure contacts is shown in the accompanying Figure in 
which the value of reverse current at ten volts is plotted against the 
percentage of units having less than tijat value for ten gold bonded 
diodes on successive days of h'jmidity cycling. Although these units 
were protected by enca^jsuiating in a glit filled plieiioliC cartridge, 
the gradual deterioration of the reverse cfiaracteristics is evident , 

A limited amount of work has been unde taken in an effort 
to prevent or at least to retard this deterioration. One has 

involved an attempt to reduce the sensitivity of the bond itself to 
moisture, wh ]*■ a second .'p,.! ^ ; t has souglit means of preventing the 
penetration of moisture to the sensitive region. 

In the first part of the study, elementary, unencapsulated 
diodes !iave been utilized. It has been found tfiat the response of a 
bare diode to water vapor depends critically on the treatment wliich the 
ger-’snium wafer receives prior to bonding. In particular, diodes which 
are not washed after the wafer is soldered to the ohmic lead z^spond im- 
mediately to the presence of water vapor whereas diodes which have been 
washed with a detergent such as '’Alconox" do not. Surprisingly, a group 



TRANSISTOR PRODUCTS INC. 



of diodes washed with detergent exhibited no significant deterioration of 
reverse characteristics during a three day exposure to a relative humidity 
of approxinately one hundred per cent, at 70 °C, These diodes, however, 
regained their sensitivity to water vapor after a drop of deionized water 
had been placed at the bond and allowed to evaporate. Perhaps the most 
surprising feature is that this water drop treatment improved both the 
forward and reverse diode characteristics In many cases. 

These observations suggest the possibility of finding a s’irface 
treatment which will indeed reduce the sensitivity of ttie active part of 
the diode to moisture, A number of possibilities, for example, tiie h/drolj’sis 
of water repellent chlorosilanes onto the surface of Ljk germ?ni’ui!, could be 
tried, 

tittle has teen done on the second piiate of olie ;;ludy, concerned with 
keeping moisture away from the c-ensitive region. L' ttlo or nc Impx'ovoment 
has been achieved by using a variety of rasix.s or rubber coatings although 
the use of a silica filler in an epoxy resin shows 00:111 promise. The most 
successful moisture barrier studied is achieved by enil using a clastic beaded 
diode in a glass tube filled with dessicant end sealing the ends of the tubes 
with an epoxy resin. Presumably, this technique of a double seal against 
moisture could also be applied to tron 8 l 8 tcr.s. 




TRAN3fSTOR RRODUCT3 /A/C. 




REVERSE CURRENT DETERIORATION (JU\TH 
HUMIDITY FOR GOLD BONDED D10DE5 






TRANSISTOR PRODUCTS INC. 



APaHDH D 
DOPIMQ STUDIES 



Initial experiments were carried out on face lo face gold bonded 
traneietors using .003" gold wire doped as follows i 

EMITTER DOPING COLlliCTOR DOPING 

None None 

15 Ga 1^ Ga 

S% In iSb Sb 

15ln IX Sb 

1% Ga It Sb 

These init'' studies were carried out on units with a nominal .OGC” point 

spacing, utilising germanium ranging in resistivity from 3 to 30 o.hnrj cm. 
Measurements of ti*e small signal parameters of the resulting transistors 
established the following! 

1. Units with It Gallium doped emitter and collector exhibit 
junction like properties with the always less tl*an 1. 

Values of ,5 to .7 are typical. CoUe ctor impedances of 

10 to 20 thousand ohms are readily obtainable with cutoff 
frequencies as high as 7 megacycles and power gains uf 15 db. 
or more. 

2. No difference in emitter eii«if«cteristirs could be discerned 
between tlit* 1% gallium and It indium doping, kl lie the 5>S 
indium doping also affords satisfactory emitter characterise ‘ cs 
SOTie difficulty is experienced in bonding successfully when 
using this wire. 

3. Antimwiy doping (presumably arsenic would do as well) of the 
collector is essential if cuirent gains greater than one are to 
be obtained. VRille IX antimony doping makes it possible to obtain 





TRANSISTOR PRODUCTS INC. 



current game as high as S to 6, tliese liigh current 
gains are slwa/s accnmpanleci by collector InptxlanceB 
of the order of o few h>indi*ed ohms. Thle ie interpreted 
as indicating ttjat tiu* antimony doping level is much too 
high, 

Followlr^ tiie discovery U»at tlie donor collector doping required for 

satisfactory operation is less than offorts were -^lude to secure leas 

heavily doped vflre fron comercial sources . Because there was initially 

no readily available cowmercial source of antimony dc^d gold wire in the 

doping levels desired a fsrograni was undertaken to devise ireans of providing 

wire of the desired cliaracteristics locally. The first approach was based 

on the fact tliat the vapor pressure of antimoby is much higher tlian that 

of gold and that therefore it should be possible to effectively "boil out" 

soma of the antimony by heating 12 antimony doped gold in a vacuum. 

Initial studies of this process confirmed tiwt the antiiiiony could indeed 

be "boiled out" of antimony doped wire by ^£ating In vacuum to temperatures 

o 

of approximately 600 Centigrade for times of the order cf twelve to 
twenty-four hours. Although the dofjing levels attained by these methods 
gave improved results in transistors, oti»er difficulties arose in connection 
with this process. Specifically, the heating (a) destroyed the temper of 
the wire and thus made It difficult to liandle in Uje transistor assembly 
cperntlcn and, more imporcaritiy, (o) relieved local strains (due presumr^bly 
to tt'*e wire-drawing operation) thereby leading to considerable non -uniformity 
in both doping level and the wire diameter,, 

Because of Uieee difficulties another approach was explored, the introduction 
of antimony dope onto "pure" (99.99$) gold wire by clipping the wire into an 
aqueous solution containing antimony. At first, solutions of SbCl^ were usee 
but ths results were found to be very variable. This was traced to the fact 
that SbCljj reacts with water to fona SbOCl which in turn is essentially In- 




TRANSISTOR PRODUCTS INC. 

i t 

soluble but which forms a colloidal suspension in watar. Thus, the 
antlnooy strength of the resulting •’solution” Is very dependent on tine 
and methods of handling. To obriste this difficulty solutions were 
prepared using antiBiony*potaBsiuoi>tartrate. This forms a true solution in 
water and it was found tiiat 99.99^ gold wire dipped in such a solution 
(10 mg of K-Sb-tartrate to 3p cc of distilled water) yielded, when bonded, 
results very similar to K Sb-doped gold wire. 

The reproduoibility was at first poor but became quite good upon aiding 
to the above solutian about 0.1 graa oX "alooncsc* detergent which apparent- 
ly bad the effect of cutting throu^ residual die -lubricants etc. cm the 
surface of the wire. 

At this stage in the Inveetlgatioo, we were fortunate in obtaining some 
0 , 01 % antimony doped wire from a eonnerclal supplier and the dip-doping 
was discontinued in favor of this predoped wire which was felt to be more 
reliable and more easily handled in assembly. A number of one sided 
transistors were fabricated using this wire. With this lightly doped wire 
it proved possible to obtain transistors meeting the contract specifications. 
Difficulty was still, however, experienced with tlie strength of tlie bond with 
the result timt de-doplng studies were resumed using the .01^ antimony 
doped wire as a stax*tlDg materially. 

^his time the problem was approached via the "boiling out" technique but 
with the differences tljat tlie "boiling out" was carried out in air instead 
of in a vacuum and wais acconplisbed by drawing t^ic wire continuousii'^ through 
a 900°C resistance furnace. This continuous operation appears to provide 
good uniformity of doping level and little alteration of the wire diameter 
contrary to previous experience in the vacuum de-doping operation. The 

* 

successful operation of this de-doplng system was only obtained in the very 
last stages of the contract and wiiilc experimental transistors confirm its 
value in obtaining tlie desired transistor characteristics along with a 




TRANSISTOR PRODUCTS INC. 



mechaniCciLly strong bond it orrived too Iste to be utilized in Uie 

samples supplied to t he sponeorlng agency » 

A slcetcb Cf tJie dedoplng furnace is shown below: 




T7p>loiFI operstlcn of the furnace involved the following parameters: 

Furnace temperature at center - approximately 900®Cc 

Pull Rate - 1.8" /hour 

Oold Wire - .003" diameter, oOlJt Sb Gold 

The extremely slow pull rate could be obviated by modifying tie above 

structure to provide for the gold wire to traverse the furnace a nuri>er 



of times. 




TRANSISTOR PRODUCTS INC. 



APPENDIX E 

FABRICATION CF FXPERIWPJITAL TRANSISTORS 

Experimental transistors of two types have been fabricated. These fiave 
been one sided units in a cartridge similar to -he Westex’n Electric 
Type A and a two sided unit constructed on an in line transistor base 
as tentatively proposed by the RI'^TEA. It has been found that either 
IS gallium or It indium doped gold wire .0T2'» to .003" in diameter 
makes a very satisfactory emitter. Typical diode cliaracteristics of 
the emitters are given in the attached data sheets. For the colL' 'tor L» 
and .01.t antimony doped gold vxire has been used. Other experimental urits 
have been fabricated using "de-doped" gold cbtained by methods described in 
Appendix D. While the optimal de -doping has not been quantititively detenrd 
experiments using .01’', antimony doped gold wire ,'X)3" in diameter as 
starting material liave established a {proximate ly i. 3" /hour as a satis- 
factory pull rate in the continuous dedoping process. Such dedoped wire 
gives promising results in the fabrication of transislors with 
^ 'T V- > and cutoff Cl - 

The rtjsistlvity of the germanium liiis been variej in the range from 5 
to 20 ohm centivieters without materially affecting the cl-iaracteristice 
of exreirimental units, Tliere are, however, some indications in the 
accompanying data that a material resistivity in the vicinity of 10 
ohm centimeters may give somewhat better results. 

One hundred and tern sample transistors of thie one sided type have been 
fabricated, tested, and supplied to the sponsoring agency. The test 
data on these units will be found in the following pages together with 
an information sheet which shows the emitter and collector materials end 
tiie germanium resistivity used in each sample transistor. 




TRANSISTOR PRODUCTS INC. 



It should be emphasized t)iat the sanple units supplied to tije sponsor- 
ing agency have been selected with a view to illuatrating the range of 
transistor characteristics which can be obtained by Uie gold bonding 
teclmique. It is, tlierefore, only units later in the series {incid- 
entally the late.'it chronoxogically) which meet the contract specifica- 
tions. 

In addition to the neasurements tabulated in the succeeding pages, noise 
figure measurements iiave been made on units 1'? tlirough 110. Tie fol- 
lowing histogram gives the result of these measurements. The noise 
figure here is measured at 1000 cyclea/second for a nominal one cycle 
bandwidth. 



do 



io4- 




/ol 



—I > 1 T 

7o So i>e 

NO/SE F/QURE /A' 



The above distribution and values of noise figures are substantially 
similar to those obtained from measurements on conventional point con- 
tact transistors. Thie high noise fig\ire of these transistors is not 




TRANSISTOR PRODUCTS INC. 



iinderstood but may be associated wltli the fact that Uie collectors were 
only very lightly bonded. 

Noise figure measurements were also made on a number of units on which 
emitter and collector current were varied. There is observed a noticeable 
trend towai'd Increasing noise figure with increasing emitter cux*rent. This 
trend is of the general form, noise fig’ure ■ F + • F ■ 55 db and 

2 db Aa are representative of the valves obseirved over the range ^ 

^ 1^*2? U MA. Relatively little dependence on the value of was 
observed. 

Dxiring the course of the development of a few " symmetrical" transistors 
were constz*uctedo These units have emitter and collector wires of 1^ 
gallium doped gold . 003 " in diameter and utilise a germanium wafer with 
approximately 5 olsn centimeter resistivity. The units are so bonded that 
emitter and collector characteristics are made as nearly identical as 
possible. As a result either gold wire can be used as tl« emitter or 
collector and the units are true symmetrical transistors. Typical char- 
acteristics of the units are Yx-,. ^ /O ad Coro^^ >y>/C. 

While these units are not specif ically c ailed for by the contract a 
sample of ten unite has been suj plied to the sponsoring agency as a 
matter of general interest. Typical static characteristics of these 
symmetrical units are shorn in the following figure; 




TRANSISTOR PRODUCTS INC. 



Traneiator No-> Resistivity Otm Cm. 


Emitter Wire 


Collector Wire 


1,79 


15 


.003” 


1^ Oa 


,003” l^b 


77,78 


5 


o002" 


1^ Ind. 


,003" l$Sb 


80,81 


9 


.003” 


Itoa 


.003" l^Sb 


82 


15-20 


. "02” 


135 Ind. 


.003” WSb 


111 


10 


.002" 


li Ind. 


.0Q3" litSb 


83, 8U 


7 


.002" 


H I 


.003" 1$Sb 


85,86,87,88,89. 


6 


.•902" 


1^ I 


.003" llSb 


90 


10 


.002” 


1.5 Ind. 


.003" liSb 


3,7,9,12.15,29, 

30,91,92,93 




.OCE " 


lat Ind. 


o003" .OliSb 


11, 12, 13, lu, 15 
16, 1.', 13,19, ?0 
99,1'10.101,102, 
103, lOU, 105, 106 
21,22,23,2L,9l4, 
107, 108,109] 110 
95,96,97,98 


7 


c002” 


1:^ Ind. 


.003" . 013 ; Sb 


25,26,27,28 


10 


. X2” 


1^ Ind. 


.•X)3” .ou Sb 


29,30,31,32, 33, 3ii 


9 


.002" 


Ind. 


..003” 1% Sb 


35, 36,37, 38,39,U0 
hl,U2,h3,liU,U5,U6 
Ii7,li8,l9,50,5l,52 
53 


V 


.002” 


1$ Ind. 


. 003” 1< Sb 


51, 5U, 56,57,58,59, 
60,71,62,63,611,65 
66,67,58,69,70, 71, 
72, (3, (U, V5,Y6 


13 


.002” 


Ind. 


. 003" 1% Sb 




TRANSISTOR PRODUCTS INC. 






TRANSISTOR PRODUCTS INC 





-ij 


XA 


-3 


VO 


CO 


Os 


‘O 




1’^ 


H 


a? 


'M 


uT\ 


o* 


f“4 


f 


• 








* 


(X 


e 


ft 




i 


t 




• 


t 


(\3 


04 


04 




04 


m 




rn 






CM 


o 




*•4 


o 








r\» 


a> 


-d 




00 




o 












o 


CM 




CU 


• 

r^- 


<30 


CD 


HI 


CN 


CM 

CM 






CV) 

rM 


26 


T'- 

r-t 


?! 


f 

00 


On 


Cv 


O- 




«> 


r' 












O 


o 




a'' 


CO 


vM 


•w 




rH 


fu 


CM 


o> 


t^4 




-4 


i-l 


CM 


CM 




.-1 






a 








O 


O 


o 


a 


o 


r*> 


■ o 


o 


Ok 




o 




O 


O 


O 


Jil 


vO 


p- 




7^ 






o 




Co 


aj 


o 




-:3 


OD 


03 


04 


H 


04 




r< 


H 


r4 






-a 






O 

C 






























C 

vt? 


O 

vO 


o 

cc 


8 


C.' 

oc 


00 


c*; 

-•I 


S:* 


o 


o 

<30 


*\ 




c::; 


.-:r 


’• ^ 






r 4 


CM 








H 


0- 


r^. 




cV* 


' ‘j 


/•n 


-4 


i 




'N* 


Ov 


rV*. 


<D 




50 


04 


vO 




CJ 


C J 






r 


■ * 




(V 


(VI 


(\i 




f“H 


r-i 


1"^ 


f’.' 


( 4 


o 


t \ 








tx; 


Cv> 


vfl 


(•^ 


Cm 


o 


CM 


CO 


C> 


M) 




a; 


w> 


V •> 


,, 


0 


C'J 


C'J 


T 


>o 


mC 


r-4 


t^ 












f * 


y ’ 




••* 


1) 














* 




-J 




_J 












. 


,ir^ 


i ' 


• 


xf 


l*'4 


i,.’l 


. J 


!• 


ir. 




•*!/ 










r-» 


r ( 




•:\i 


ft 


t\ 


rvi 




in 


* '* 


. :? 


, 1 




•'v.’ 












O 


O 


c» 






» > 


o 


f 


( 




■"y 


Z'f 


(; 


o 


O 






(S 


o 




f 4 




•a 


Cvl 




' ^ 




C' 




•n 


ro 


ly 


C-4 




<'•> 


ri 






1 < 


'■’ 






■ 


C-’ 


C' 


ry 




O 


ij* 




O' 






o 


(!» 


I' 






•s^ 




i* 


'O 


rr. 






■•> 


, -t 


•-'T 


• '» 


cv 


r*> 








< 


• -I 




c^ 


r « 


tH 


fv 




CM 


ri 


M 


CM 


t*4 


•* J 


r 


•4 









x,y 


Oj 


























• •< 


^ ' 


r\ 


* 






■:> 


H 


( J 




r N 






'1 




1 


fC- 




< 4 








•a 


H 




■ 0 




r -t 


•‘ 4 




. N 



^ i 


‘•i 


OJ 

Q 




r4 

O 


.H 

Q 






CM 

o 


c^‘ 

i" 


• •♦ 
-t) 




■s 




'^'i 




D 


O 


V 


r> 


(f' 


•f) 


o 


'*1 




r“> 




• 




o 


O 






o 




» 


• 


< 








o 


C5 







r-f 

i 


o 




O- 


p- CD 

* / 


o 

• 


# 


ȣ) 

f 






t\l 




cr 


|-( 


- 


o 


''A 


CM 


H 


3 

0 




o 


O 


c.O 


c 


o 


H 


l*M( 







ro 


o 

(1 


NO 


O 


• 




vO 


VO 

■•<. 


o 


vO 


03 


Ai 


O 


H 


H 


CM 


CM 


r*> 


H 


O 


-■3 




O 


•’*'1 

















♦J 

b 

O 

£ 


O 














*1 

U 

o 


l>3 

Cj 






CM 


rA 


u 

a 

o 


CM 


U> 


a. 


03 




-3 


CM 




A< 




«0 


3 


;\i 



C 


f-i 


CM 






J^ 


o 




CD 


O 


o 


4 


CVI 


r»0 




•,r. 






<X' 




CM 


CM 


CM 


<VJ 


4M 


CM 


O. 


CM 


CM 






r<\ 






r'% 


r»\ 


r\ 


r**> 




91 nn 051 L9€ 



ftODUCTS INC 



>o o\ 

• 9 

fa CM 

- O 

C> ^ 

Ov 

vn 

♦ 

rsi CM 

o c* 



o 

* 

rsi 

«• 



O 

CNJ 

ca 



O 

ac 



o 

ir. 

.\4 



C^J C4 



00 o 

i> :.A 



r> 

ri 



CNJ 






<x 



aj 

c\i 

•" \ 

H 



O 

v*’) 



iH 

r\( 



wT. 

O 



-:j o 

m) 

# •< 



O O 

r^> 

rvj -I 



o 


f»> 




CO 


ra 


c^ 


NO 


NO 


(A 


CO 


o 


o 


• 


» 


• 


o 




4 


• 


•M 


• 


9 


J^ 






rH 


rK 






%SI 


lA 


rn 




JH 


vO 


CM 












03 






• 


• 


O 


>o 




-cr 






o 


o 


r-i 




i-» 


r -4 


#H 


r-« 


iH 


NO 


<% 


rH 




-a 




oo 


>- 


fH 


'C 


CO 


o 


O 


s 3 


.■A 


A 




H 


fH 


i*A 


.::t 


ro 


oA 


..r* 




o 


o 




o 


O 


o 


C! 






n ■ 


O 


\A 


on 


rt 




'-A 


fa 


/N 














vO 


»-« 


|H 


ifH 


.H 


*a 


■■J 




-t 


o 


o 


o 


o 


o 


<»«« 


* 


o 


,"-S 


r~ 


o 




NO 


4 


'N’l 


C\j 






>o 


f ^ 


,r. 


•''g 




CM 


CM 


rA 






.Jt 




t'w 


Ml* 


a 


fA 




nO 




J\ 


V 


o 




< ^ 


r . 


-a 


CM 




CNJ 




A* 








♦“ t 




.a 




OD 




O 


fg 








o 






NO 


• 


o> 




•• 


•" 


c::- 


u 


Cl* 


■A 


r-l 


■ A 


r-< 




A ^ 


»A 


ra 














fV> 














4 


fA 




c • 
u^ 




• ^ 
fj 


r' 




A. 


•a: 

Cs. 










o 


? 


o 


o 


O 


o 


c 


o 


sT> 


O 


O 


4 


•-A 


nn, 


>o 


CM 


c^^ 


nO 


b 


X 


'O 


«H 




r-* 




A 1 




r-^ 


o 


o 


o 


c* 






/— 


o 


o 




r.T 


A 


or 


o 


C”) 


U\ 


»N 


■_• 


Ov 


NO 








f-C 


r< 


fa 


•H 


fj 


fv 


Oi 


rj 


rg 


r\ 



? > 



o 

>0 



C'- 

r~l 



O 






.H 



o c 

r*\ 






V» , N C 
- ? Oi *: m 



3 

o 



r-i 

c: 



o 






^ 

'* d d 



> 

o 



rA 

o 



CVi 



<t 






C5 



C> 

d 



Pf 

6 

t 



. rt 






-I 



lA 



sO rH 

c 



a. 

* 

OJ c 



c: nO 

d 



r- vrv 

iH 



CO 


NO 


o 


CM 


O 




" 


• 


V 






fA 


IM 


r^ 


•A 



CO 


CO 


vO 


OD 




c 


o 

CM 


V-* 


• 

ri 




«:3 


-3 



iH 



u\ 

d 



o 

.H 

o 



U^ 



a> 




V 

'• l/\ 



fH 






b> 

8 



t 



o 

J3 



n 



<M 



O 



-=r w »<> 



C\ 



o 



CM 






-a 






s 



CK 

-a 



o 

VA 



(A ^ 



lA 



UV 



Vf% vo r~ 

VP> lA lA 




TRANSISTOR PRODUCTS INC 



«» 


o 




-4 


M 


M3 


o 


•H 


00 


m:> 


O 


r> 


o 




UN 




IM 


oo 


(M 


• 


• 




4 


• 


•» 


• 


• 


• 


$ 


« 


• 




• 


• 


• 


9 


• 


• 






CM 


CM 


UN 


<»N 




O 




•d 


-a 


ۥ*> 


4^ 


CM 


CM 


cn 


CM 


rr 


fp> 


a 








cn 

H 






00 














SO 










v\ 

fH 




a 




o 

rM 


• 

Os 




o 

1-4 


o 

H 


a 


a 


H 


ft 

c^ 


73 


C 

r-» 




r>-. 

H 


^4 


05 


U> 

prs 


oo 

•M 


a 




O 

t*\ 


a 




lO 


a 


OD 

-N 


eo 

--N 


O 

CM 


o 

CJ 


CTi 

4»N 


rM 

PM 




n 

-d 


o 


O 


O 


O 


O 


o 


o 


o 


Q 


o 


O 


o 


o 


8 

CM 


8 


'3 


o 




o 


a 


0C> 

H 


rg 


a 




««N 

H 




oo 


O 

rM 


00 

-M 


r4 

<M 


M) 

fN 


M3 

-3 


c. 

CM 


CD 

r-4 


I 


o 

CM 


s 

<v» 


O 


O 

CD 


270 


o 


o 

CM 


r-i 

O 

rN 


o 

r»s 

CM 


O 

a 


B 

fn 


O 

—4 

-3 


O 


8 

Ox 


8 

-5J 


C) 

-4 


CO 

CJ 

CK 


o 

A 


c 

a: 

f*'* 


o 

CM 

r> 






fSJ 


CM 


CM 


r-' 


CO 

41 




T 




■O 

f 


«< 


-d 


rH 


•• 4 
• 


-1 


CM 

{ 


ft^* 


a-j 


>. 


CJ 


CVI 


<M 






CM 






fN 




CM 




CM 


Cm 


■' \ 


CM 


r^. 


(4 


o 












c 

r-^ 




c 

H 


rvi 


o 






CM 


-T 




CM 








r-4 

v5 


O 

rH 


C7 

f*1 


Os 




a 

UN 


X 


r 

.>» 


a 


a 


• 

r- 


t 

X 


a 


Os 


\s\- 


O 

rM 










CM 


R 








O 


Li « 


X 


c 


<v 


a 


•^) 


/> 


cC 




X 




is) 


fM 


CJ 


(M 






r\i 


CM 


f*N 


• d 


r* 






—3 


Pvl 


o 


O 


o 


o 


o 


c 


O 


O 


o 


O 


8 


O 


o 


o 


O 


rj 


O 


C 


o 






CM 


a 


'O 


•M 


4^ 


o 


CO 


?< 


Q 


CM 


r-4 


^f> 


«o 


00 


Pi 




<■< 


f\l 


fi 


' ^ 


r-< 


d 


H 


I'M 


CM 


sd 


CM 


<-> 


.-1 




r*^ 


(M 


t: 


o 


O 




o 


o 


O 


o 


o 


o 


O 


o 


r- 


o 


O 


o 


O 


C 


c 


04 


vj5 


U*v 


o 


a 


n; 




c: 


>o 


CM 


oo 


o 


CO 


M3 


C*J 




Wj 


o 




C*l 


fo 


Cv 






■X 


4^ 


rsi 


CM 




vO 


r»N 


t-N 








f*N 







vr\ 








oo 


o 


o 




ac 


XA 


O 


OO 


c 


-A 


C' 


cv 


•o 


■ 


r-t 


Cu 




M 


•J 


rM 


UN 


-d 


r— 


rH 


Ci 


rg 


rH 


CM 


04 


tv 


rH 


CM 



too 


TOO 


001 




• 


O 


c 


too 


8 

• 


012 


8 


r*f 

B 


rH 

C‘ 


g 


1 


r| 

8 


rH 

r* 

o 




8 

o 










iM 


c 


• 


• 


c 


• 


p 


• 


d 


• 


r 






o 




o 


r- 


.d 


NO 


rH 


r- 




CM 


f- 


CM 


o 


CM 


NO 


r '• 


d 


-d 


CO 


CM 


o 


CM 


rH 


rM 


rM 


rM 


d 


rM 


d 


d 


a 


rH 


«-i 


rM 


r-f 


rH 


rH 


rH 


rM 


r-i 


NO 


NO 


ec 


o 


O 


co 


O 




dT 


« 


<C 


CM 


o 


o 


« 


o 


<M 


C 


CV 






• 


■ 


4i 


•» 


'» 




« 


• 


• 






• 


•> 




<h 


o 




CVJ 


CM 


cv 


CM 


NO 


rC 


CM 




■UN 


di 


CM 


PN 


CM 


CM 


CM 


CM 


CM 


-3 




IJ*. 


r“* 


\r. 


































CM 


% 

cJ 




UN 




UN 




c 

rM 


XA 


UN 


UN 


UN 




o 


O 


TA 


Xft'A 






o 


o 


h- 

O 


o 

rH 


o 




8 


^- 

r> 


O 


o 


o 

r\ 


s 


B' 


o 


8 


C 

rH 


o 




6 


d 


d 


d 




d 


d 


d 


d 


d 


d 


d 


d 


o 


d 


o 



CM 




-d 


CM 


CO 


rM 


<n 




PN 


91 


UN 


A 


CM 


CM 


oo 


Ul 


-3 


CN 


_d 




Ov 

XA 


5 


61 




63 




'fi 

nO 


n6 


67 




S 


o 


71 




PN 

P- 


t2 


lA 


so 




TRANSISTOR PRODUCTS INC 





CJ 

gl 


u^ 

• 


o 

• 


M 

• 


• 


OI 

• 


vO 

• 


<1 


■3N 

• 


m 

« 


CO 

• 


CC> 

r 


o 

• 


• 


LA 

• 


OI 

* 


C.V 




I.- \ 




x/\ 






CM 


OI 




OI 


OI 


•H 


1 *1 


ro 


OI 


rH 


OI 


fA 


CO 


CO 


"O 


-rt 


rv 


CD 


CM 


OI 


CO 


CO 


CO 


CM 




_3 


CO 


flO 


LA 


vO 


• 


* 


%• 


* 


« 


« 


• 


* 


• 


» 


• 


• 


• 


* 


• 


• 






tM 




.X 


fX 


Vf\ 




UN 


fH 






LA 


U> 




fO 


so 


Os 


3 


LA 


CO 


oo 




H 




\A 

tsi 




CO 


vO 










_3 










•« 

0\ 


a 


iH 

CVI 


t 

A- 


• 

'O 


• 

US 


n 

r-l 


21 


• 

Ov 




• 

Ov 


r~l 




r* 

(H 




r— 

or 


o 

-* 


o 

s 


8 

•H 


O 

Os 


o 

2* 


8 

r-i 


<s 

o 

LA 


8 


O 


8 


s 


8 

*-■1 


O 

CP 

rr\ 


o 

vO 

r— 4 


Ct 


O 

OO 

rH 


8 

i-i 


c 


Cj 


o 




O 


o 






8 

-3 


O 


o 


o 


o 


Q 


o 


o 


o 


o 


-3 


U\ 






iii 


-3 

OI 


0 

01 


• 

N 




-3 

OI 


5X 


CO 

LA 


O 

.3 


vO 

rS 


-3 


'O 

CM 


fA 

tH 


-3 




VA 




VJ 


CO 


OC’ 


vO 


OI 


>o 


o 


vO 


SO 


-3 


sC 




Os 


LA 












• 


• 


« 


' 


n 


e 


e 


• 


t 


!• 




•» 


« 


Cvj 


-3 








AJ 


r ( 


r^ 


rH 


CM 


oi 


#H 


r< 


f^S 


CM 




r*H 


CVI 


(M 




CM 


VO 




<o 


X) 


o 


O 


CM 


oo 


o 


O 


•o 


X. 


3 


o 










• 




* 




«• 


• 


• 


P 




a 


• 


•» 


« 


•» 




fVJ 




r\j 


rH 




u\ 


CM 


.H 


-3 


ro 


-=J 


LA 


'O 


Oi 




3s 




fA 


t'i 




fC 


O 






C: 




O 


*o 


M5 


.3 




<c 






CSI 


vD 


U\ 


“ 




vC» 




uA 


r 




• 


' 


t 


« 


o 




2M 


xs 




» 


Tw 






rH 


iN 




UN 


oc 


:> 


30 


r-t 


vO 




r-l 


vA 


CO 


o 


o 

u\ 


C 


O 

t.v 

r< 




'I 


■“ 1 

C’' 


O 

IT: 

f 1 


O 

2i 


o 

21 


OCT 


8 


8 

.-1 


Cj 

5X 


3T 


o 

•M 


C' 

OO 

rH 


8 

ri 


30 


o 


o 


r 


c 


O 


Q 


.■ 


»■ *' 


8 

-3 


Q 


c 


O 


o 


O 




O 


O 




vN 

.••1 


— ' 


•-1 


O 

CM 


o 

C4 


0^ 


r' 


H 

tsi 


3 

rsi 


CM 

• -> 


a: 


;3 


-3 


rA 


21 



V> 


C-'' 




A •» 




•J 




>o 

a 


o 

a 




cr 


o 

• 


rH 


U> 

9 


CM 


lA. 


3 

a 


CM 

• 


fi 


rO 


CM 


rH 


rH 








rr> 

r-» 


H 


'O 


3D 


nj 




ro 


CM 




CM 



g 


<A 

O 

C' 


x! 




r'>. 

O 

o 


8* 

c 


6 

» 


r~* 


iH 


a 

• 


rH 

< 


rH 






f 


vO 

8 


8 

a 


O 










t 


o 


o 


e 


d 


o 


C 


6 


a 






% 


o 


« 


3 


3 


O' 


CT 


vO 


oo 


<c 






o 




UN 




m:' 




C- 


sO 


H 


• 


• 


• 


a 


• 


a 


• 




• 


a 


a 


t 










a 


t 


GO 


rH 


l-^ 


Os 


vO 


CM 




CD 


OI 


rH 


AS 




rH 


lA 




.-H 


LA 


ro 


-O 


00 




Cu 


•D 


3 


CO 


cc 


OI 


sC 


O 


CM 


CM 


O 


CM 


-3 


CO 




• 


• 


• 


• 


• 


• 


a 


■a 


a 


a 


a 


•• 


> 


<* 


* 


a 


a 


CM 


3 


Cj 


LTl 


f'N 


CM 


rH 


rH 




rA 


CM 


rH 


rH 


CM 


CM 


rH 


rH 


rj 


Q 


O 


VA 


o 


UN 


U\ 


o 


U“l 


\A 


O 


UN 


c 






U** 


o 


UN 


O 




rH 

O 






O 


o 




& 


8 


» 


8 




o 


8 


8 




8 




• 


» 


• 


• 


a 


a 


a 


•• 


a 


a 


9 


• 


a 


• 


a 


a 


• 




o 


O 


O 


o 


o 


O 


o 


o 


o 


c 


o 


o 


o 


o 


o 


o 


o 


d 



Ov 


fA 


so 


os 


OO 


3 


LA 


15 


LA 


fA 


LA 


fA 


fH 


16 




fA 


c 

rH 


10 


77 


78 


79 


O 

CT 


£l 




fA 

OO 








87 


« 

X 


89 


o 

Ov 


|H 

O 


OI 

Os 


93 






TRANSISTOR PRODUCTS INC 



o 

• 


CM 

ft 


VS 

• 


r1 

ft 


CO 

# 


o 

• 


00 

« 


o 

ft 


o 

• 




CM 

• 


o 

ft 


-a 

ft 


cc 


o 

c* 


rr\ 

t 




-a 


CM 


r»s 


CM 


CM 


H 


CM 


-a 


CM 


-a 


CM 


CM 


ri 


«c 


CM 




oo 


O 


CM 


sO 


SO 


O 


o 


CM 


-a 


CM 


CM 


SO 




CM 


s© 


• 


• 


• 


• 


• 


• 


0 


ft 


• 


• 


ft 


* 


ft 


y\ 


« 


-a 


>- 




-a 






Cl 




.a 


o 


<n 


CO 


UA 


r-i 


c; 


-a 












.a 


o 


• 


















• 


CD 

CM 


o 

H 


» 


o 

H 


• 

Os 


• 






LA 

*-» 






00 


t 

1 1 


« 

r*< 


o 

-» 




O 


O 


C3 


O 


O 


5? 

ri 


o 


p 


O 


o 


o 


O 


8 

f 1 


Cj 


•*> 




SO 

iH 


vO 

r-< 


iH 

fH 






H 


o 

CM 


iH 




cc 

fH 


CD 


LM 




3 


O 

CO 


o 


O 

-a 


S 


3 


O 


3 


O 


O 


o 

o 


n 


O 

CM 


o 

>o 


C • 
Cs: 


o 

f 


iH 


CM 


iH 




H 


CM 


CM 


rH 


rM 


r*' 


cn 


CM 




4 


o 


-a 


-a 


-o 




OS 


sO 


00 


-a 


CM 


OD 


O 


rH 


fs> 


C 


o 


• 


• 


• 


ft 


• 


ft 


# 


• 


« 


« 


• 


• 


ft 


ft 




• 


IN 


rr\ 


CM 


CM 


CM 


r*» 


fH 


H 


rr\ 


cvj 




CM 


CM 


rH 


f j 


cu 


<e 


CD 


-a 


-O 


CM 


SO 




CM 


CO 


CM 


O 


o 




sO 


CM 




» 


ft 


« 


• 


ft 


• 


• 


ft 


ft 


ft 


ft 


• 


ft 


ft 


ft 


* 


<v 


U\ 


CM 


CM 


«»S 


CM 


CJ 


\s\ 


CM 




CA 






OD 


tM 




o 




















O 


o 




r 


o 






o 


CM 


-a 


-a 


fM 


CD 


CD 


NO 


ft 




o 


ft 


ft 


CO 


ft 


CD 


fl 


a 


• 


• 


ft 


m 


ft 


» 


*« 


c*\ 


« 






• 


o 




s© 


r— 


■ ”■ 


-T\ 




Ov 


> 




r-C 


r4 


On 




N."** 


i$ 


P 


3 


S 


o 


g 


10 


120 


O 

CD 


120 


O 


c-* 

OD 


o 

OD 


o 




o 


f-4 


a 


r-i 


i~4 




r4 


fH 


fH 


rH 


rH 




CM 


74 


o 


o 


O 


O 


O 






o 




c 




O 


fj 


o 


O 


o 


-4 


or 


o 


o 


LTN 


o 


O 


CM 




CV) 


O 


CM 


t ^ 


-r3 


rc 


('u 


r< 






•H 




<r 


fM 




c-f 


CM 




-M 


CV 


ri 

































o 
































o 


o 




so 


r— 


CC 


-a 


\IN 


o 


a: 


ft 


•-C 


M ' 


fV 


-j 


ft 


♦ 


ft 


« 


• 


• 


r 


ft 






ft 


nO 


ft 


a 


ft 


- 


'M 


CM 


CM 


CM 




rH* 


.-H 




fH 


_a 


rH 


rH 


CM 


o 


CM 


lA 



n 




3 




016 


TOO 


2 


Too 


g 

Ift 




p 






F> 

* 






ft 




ft 




ft 


• 




c 




o 


A 


C" 






» 


ft 


o 




o 


CM 


sO 


•o 


o 


t*- 




• 


fH 


CM 




s© 


-a 




ft 


» 


<« 


ft 


ft 


• 


• 


ft 


• 


M 


ft 


ft 




• 


ft 


ft 




a 


or 




-a 


CO 




«*» 






US 


CJ 


’mTs 


o 


rH 


H2 




o 


o 


-a 


aj 


-a 


.a 


A*' 


CC 


o 


O 




o 


o 


O 




ft 


ft 


ft 


• 


• 


ft 


• 


ft 


ft 




ft 


ft 






¥ 


ft 


rH 


W 


CM 


(Ni 


CM 


rH 


fH 


«-« 


rn 


<\i 


-a 


/H 


fv 


r-l 


*o 


fH 


c 


A 


U^ 




u- 


*N 


n 


o 


\r\ 


o 


Vf\ 




•hT> 


U' 


X 




i- 


5^ 












tr« 




» 


&l 






O 


s 




• 


• 


• 


» 


ft 


ft 


ft 


4 


• 




0 


ft 


ft. 


ft 


ft 




C3 


o 


c 


o 


o 


o 


o 


o 


o 


c 


r 




O 




o 


o 







20 


16 


18 


20 


15 


10 


20 


CC 


16 


rr 


US 

.-c 


, rH 


CO 


L\ 


v\ 

Os 


SO 

Os 


97 


oo 

Os 


66 


oot 


101 


8 


S 

iH 


€ 

H 


105 


106 


107 


108 


109 


110 




wmsfOR fftoDuas mL 



APPENPn V 

oFaC DvL APPA RATUS D F VKI O B!HiT 



Gold Bon o.lng Pi.il8 -3 Se t; In order tc f-ioilitate the bondtnjj and ‘■-estini? of 
cxije.. imer.ta3 ijolrt bonded traneis'ioj 5 fa:*ricat€d during the coui eo of the 
contract it wan necessary 'Co modify a etandard ciiodo pulse sec to provide 
in ii. toetirg facilit ios, Tre d.i^^-.iie i scheraatic diagr«.n of the modif 5 3 d 
equipment will be found In Uie folAov^ini?: Figure. 



T ransistor Noitie '"Igure ^ensuring Set: 
the neafjuren'ent oG tronflistor noine 5’i!;' 
C0Tn>7ifcrG3.til^ -ivoii i«bj ii. wan aecasi-^fcvy 
Cor neaanrir.g I'oi ’e fi.guri.’n on tiu: trari 
the cont^^act. Tin equlrrmnt was design 
for ease in servicing I'axxrnin f ler.i oili 
(10 cycl-e bandvidih centered on 1000 c;- 
power -jnE as ur ing output ne-er., The jeoo; 

its asBcciatef contr ^ o:frctiibry e.s we 
trolF for iTitrocuning s^jpialr from an e; 
Gonl^ins power and bi.ri 3\ippli?*s To:;* tl.; 
tor ''oufllng 1>T tween the oi.her tuc 
of tU'is*.- i'igiU’tL; as dracvibecl belc.r l.'i i 
tiva RKTI*A specificstr.on.'j . the g( 

1000 cbm? . ?oi' low noise figures, the : 
x-horeby taiclno .ad>-antHfj? of ttie fs.r-i th^ 
of the noise --b.5nd{fidth of the amplifier 



Because no ftcillties cxi-^ted for 
.ires tjid such oqulp.nGnt waa not 

tc (Jeoigi! ar.a h’-i.ld such <ui insU’-XM-.-iit 
-istorr fabricaWd dui j.np' the co a’sc «.r 
;'d in the for T, oi three s xr it<:' cha.s{ i j 
.■ 7 . One rhassir contains ;■ ne.i r Ivn ■! , 
’L’S^ high -pain nMiiOclfier ’.;iih £ true 
vd contains u noio? gene rr tor di 3de v;it'. 
11 us an Ettenuator and mnterirg cci:- 
:beioul ccnrce; •.•lUilr the thijc -’liass's 
t;-;.in:dstox’ ind apprc;ri.M-o cir-'-uits 
r;.s„ The ine.thai adoydf l ri- r'oarv.rpi'o.v.'- 
!. tai.'d .’J-j, Ai r.-. i. rend c.’ .rn the tei.t., 
r:nerai.nr ■'mpf^riariCo Is piade equal to 
!ioiae diode I 3 used as a r ignal nmirne 
ft th'j vieaKnrsmei'h is then Ind: pindeal 
circuits c 



The noise dlcKie doee not deliver sufficient .signal to directly measure 
noise figures greater than about 25 db.. Therefore provision is r,f;de for 
introducing a knovm sine-wave signal from an extenial sifTial-gonorctor, 




ttANStSTOR ntODUOS INC 



In -thl;3 caee, th-3 noise -bandvidth of the f : llo^7in£ saniplifier b be. knojjn 
and tbs final oit pvt neter must be a tT.'un-pcwer-riiea8uring device. The noise 



boa!ii»idth n.ay ' ' ' ■ ba calibrated by n-easurinf- a low- noioo device 

'such' as a ^.rown-.ivnction tranais toi') using in auccassiorj thr diod.e end 
an ey.tei’na] Sagml source. The output meter is a thermocouple device 
aid. th\U’ ti-ce pairsr prci'ided onl^: thel the amplifier is linear^ 

S.^vcral opecUsI proTlsioiis desG-rve rention. Tho trrnsistor pcv;.:.r-Gud-bins 
supi'j.y has teen ienipned to limit the availHt;?.e pc>:er tr about 1 / 1 ;. watt 
In 1.11 cases in .-'rciar to decrease ths pi’otaoi'iity of burning, cut tY'Si-. unit 
aider lue cutout stage oi the {«:plij.'lor is intentlonf.il;' designod 

to o\arjcac scon aCt-sr full-.scalc* deflentlc.'i of the ovitput r.iuicr in ori.ir 
to protect l.ic t .ei'ircco'aplo o.t the latter, i-.i.rtbcr pre terror is Jt'fcrded 
by nhe ■> oliaje -I'-.r. ltli’'3 aotio.i cf a. njon-bvil: conr.ecvod nerfu c iha primary 
of the output ir irsfoJ-T^r. Ihe ;liiput of tl.e amplifier i.nt;lr.ci£;c; r.ui utteii" 
uatui- to p.rerent overload of the first f;r.id by tJ-tO; (iside band; noise; signal 
frori hjLf.b-pain uoitB undergoi.ng tc/st. The bancividth of the imput sig.sl 
is Jamit.ed by a transformer ia the transietor-chaosis.; the -iir.'jinsforn&r 
coapling: also gives a little gain and allova transistors to be tested at 



high co'J lector c.nreats. Finally, proristen is irsde in the uoiso-diode 
coaiieia (by means of a switch ard. pair of terminalx; a., the back) for run- 
f ing t.’ir. diode .filament from batterice.. This maj occasionally be necessary 



bee. '.'.ISO the dio-'ie fi7'^.rrent tem pe ‘ature has s tendency Ic- fluctuate at ths 
power-line frequency thus giving about 20 rr.v ripple in the diode plate- 
c'irrent., k tert point for obseivi.ng the ar-iplifiei- output wavs f-.>m is 



avstlabl.^; at the back of the emplifier chaaeie. 

To uae the equipment the transistor is inserted and the desirsd bias 
conditions set up. The noise-diode cjrrent is reduced to zero and the 
external signal (if used) is turned off. The amplifier input attenuator 
is set on "low” and tlie amplifier gain advanced until the meter reads a 




TRANSISTOR PRODUCTS INC 



Coirvenient -vaXvB (less tlicn 110 ma). If the gain is iaGiifficienii, ths 
input attenxiator may then be sot on ”high”o The noise diode filament ia 
th»3n turned up until the output metei* refilling is increased by ® .factor of 
l*Ulli (output powei' doubled). If the diode can delivei’ sufficient sigvial 
to do this, tlie noise figure (at 1000 cycles) ia given simply by 

F " 10 log. cOl decibels 
where 1 is the diode u..a-’ 'ent in r,iilliainpares 

If the dio<de noise- signal is insufficient to double the output pct'er, 

an external signal, tuned to tho ce;.iter of the pass-band of the aiKpliiier 

iruf.t be used. The diode current is tva.’ned doim aiid sufficient e-uternal 

signal introduced to double the output pov.er ca previously deecrlte i. 

Tjnder these ctHiditions, read the vc Ltage aci'oss tho -'e-steraai VM'‘ ■ix.’mlnais 

anc: nooe the position oT the eTterrul-:.il: 3 ]:ial-c tteruatoi av’i tehee, 'C:is nr oe 

is then; 

F (ueclbela) - 10 log^o ^ 

loH 

where 7 is tljs (rsis' exteriit:! slgusl iu volts ay meuuiu v.d at the; 
'’-oxternal terminals, A is the ati/euuat.i on read frcT;: the ciiltchouj and ;i 
ia the n oise bsiiciwidth (about 10 cycles) vxhich may be conveniently calibrated 
by measurirg a Icw-ncise device by i>oth irtcthods). Detailu of tne noise 
msssuring eet are given in the last flvu cf the following flgm-oc,. 




MOV 



TRAN5/5TO/R PRODUCTS /A/C 




C/RCUIT D/AGRAM FOR BONDING d TF3TI NO SET (BONDS, MEASURES oC, APPROX 








A05// 9 



TRAN. 








TRANSISTOR PRODUCTS INC. 



IT T 






Q O 



GENERATOR PANEL 




OUTPUT METER 



TRANSISTOR PRODUCTS 



INC. 




AMPLIFIER PANEL