approved for public release. case...

Honor, indua M-102

Servoaeohai u oma Lahore t ory Massachusetts Institute of Technology

C sort; 1 c'.ge Maes achus ir 11 -3

To; Director, Special Devices Conter 6245 Pago 1 of 9

Fron? Jf.y If, Forrester

Subject; Storage Tabo Program Project Whirlwind See Drawing Li at OA l a 9 t PBgO

Datoi! Sopfcuuber 15,, 1947

2A.lt nejrovmdum I s in resporre to r. r o q u e t £roji Mr„ Charles Doavwaa for in.foraat.ion on the present stetua o£ Project Whirlwind Storage Tune research and £cr a eonparioon of the tube bein-j developed at K3ICT0 v l t h the one conotructod a t the Baval 2esearch Laboratory,

Ccmaer.tc wil l be connideved in the following order.

fic M.Z.ff, Storage Tube Program

b„ Coap?.rioon of H»Z,T, and HEL Storage lubes

Mai,!,. Storage lube Programs

The K.I .To storage tube -.-rori; i s organized to produce o da-;a storage aye ten for tho Whirlwind Computers. She whirlwind Computers -\xe to be of the pa ra l l e l data cremsMission tyje and tbi iitor&go nrunt road end record a l l d i g i t s of <•> rajah or •lnoltanao'ualy, "he Whirlwind Z computer wil l require operating storage by abovt .November, 1948 end storage tube and c i r c u i t reeearoh are geared to "hie •chednj.ee

A br ief esplauaticn of tho tube wil l bo given referr ing to Figure 1 which shown an oloufwitary electrode arrangement, Defleotion p l a t e i pos i t ion the electron beaiBfl p r io r to i t s being turned oac to khe point oc the d i e l ee t r lo surface a t which wri t ing 07." reading 1B to tsUr.n place, Zhs di£i«.:> 1 and 0 aro wri t ten as pos i t ive ly and negatively charged areas en the d i e l e c t r i c snrfaoe a t the point of beam iapact0 The d io loc t r i c snrfaoe ctxct hare a secondary SBdsslon r a t i o greater than uni ty (a r a t io of two or iaoro iu des i red) . To / r i t e tho d ig i t 1 the signal gr id i s made pos i t ive to col lect oocond&ry electrons frorr ths d ie leo t r io resu l t ing in a pos i t ive surface. Posi t ive charging l i accomplished by col lec t ing tho exoess electrons above tho nwabor in ths high velocity b®j»a0 To writ© -he d ig i t 0 ; the &:l<jnal grid i s sade negative, no electrons are collectccL(. and the snrfaoe charges negatively a t tho ra t e permitted Vf current flow in the primary beume Secondary eni •anion i s not a necessary factor in negative charging,.

APPROVED FOR PUBLIC RELEASE. CASE 06-1104.

http://2A.lt

http://in.foraat.ion

http://chednj.ee

Mentor Sua •'• •• Sept saber X"• 1947

Reeding signals la accosplii&ed by positioning tie signal grid voltage Did'..-£!,v between toe two writing voltagos and churging the dielectric positively or negatively to ward this voltage. The output signal ppeers tlie signal plate by eapacltlve coupling through tho dielectric.

Peraanenoe of tho stored si/n-i la insured -)y th-3 regenerative eetion of a flood of low energy electrons operating on a principle which van., I undwriatand. used in soae leraaa tubes nad Js saplcyed in the RCA Seleotron tube and ia the ?IRL storage tube, rigors 2 shows ho-? the self-<sustaiulng effect is obtained froa the holding gun, A positive area and e as gatlve area are shown on the dielectric, ihe holding gus Is at shout who sane potet-.tial as ths negative areas end the signal grid i;» at about the potential of the positive areas. Consider the positlvs area: secondary .electrons are ealtted and a nuaber Just equal to those In the prlnary b»)iira are oollected by the signal gride othfr secondary electrons return to tho dieleotrlo surface. Zf the surface should tend to becoae more poeltivei edditio.nal seooadarlee are pulled back to reduce the positlvs charge while, :'.f tho surface should tf.nd to become less positive, no.ro secondarlei will be repelled la the signal griiA to return the dlslsctrlc to its original point 0? stability Considering the negative area* the arss ie approximately at holding gun potential and primary electrons); if they strike at all, will hr.ro low velocities and will salt few ssoondary elect-rone* If the surface becomes sore negative, sll prlaary sleotrons will bo reflected and cksic loakaga will reduce the charge bo the balance point while, If tb-s surface beooaes lose negative 1 electrons from the prlaary bee* will be collected to return, the surface to tho stable negative potential.

Besults of research to date ire compared with objectives of the program in Figure 3, Figures are basod on hasio research results and testi on our complete tube of saall size which was constructed ceverrl weeke p.goc

1, Feaslbl'!.1 ty of using the tubes In banks for pari lei storage is shown by pro•out results,

?, Signal levels obtainable froa the present tube are aore than adequate So:ae of i-he preeent signal magaitudf! can be exchanged in future den-for faster writing speed, Signal level Is stem far two nethods of observation in Figure 4. 21*0 voltage division aethod giveo higher output signals and will irobahly be used with k! ir. tube.

3, Writing rpeod in the present research tube if lower khan will eventually hi required. It Ls expcited that speed con be iaoreaeed a faetor of •% by oh£n in,i tc a wurfaoe with secont'.cry oxii.asioa equal to 3 or moro In place of the proccnt phosphor surface with necoMary 'jmiaelou of 1.16; e. factor of 2 to 4 by obtaining e lover first cross«ovsr vo3.»age of the oecondary sales!OA curve which osn be reduced froa the present 350 volts to lero than 90 volts; a factor of 2 to 3 by increasing the primary beam current froa tho present 30 alcroeaperes; end n factor of 2 by lucres sing the dtolootrii thicks sit in about the proper writing speed.


http://no.ro

http://hr.ro

£34:3 Meaoraadua M-102 -3- Septoaber 15, 1947

•£, Reading speed o£ tho pro ••at bribe is satisfactory for one polarity of signal and the highe:? secondary emission aentlonad above will pr.rmit the srsie spood for both signal polarities.

50 Tho holding goa la -Mia present tobe appear* to prori.do permanent storago, QaaatatiTe to ita w n no'; complete and further tests tdll oo mado,

6. Hollnbi.llty of fchs firs; tube ii good. Indications are that a tube of excellent reliability will re suite Behavior does; not appoa? eritloal vith respect to any parameters.

7. Daring writing a signal Is produced at the signal pints which can b*> need to check that storage ua.3 properly accomplished.

8. Resolution of the present tuba is liaitsd for geometrical reasons. .The* dleleotrlc surface i.s of a cellular texture ns t;hovn lr Figaro 5,. Thfi proBtmt i?.oaaJ.c Is ccarso audi if retained in principle, vil] be reduced in no^h flizo„

9. Storage capacity of tho present iu.be *» lled'ted by the target diameter of about ;. inch,,

10, For greater reliability in reading, the tube and circuits arc boing designed to give opposite Output polarities :'or the "vo binary digltac

llo Definition ou? the experimental tub? is Halted as dincosuad under Item 8. Storage tubsn 'or coripuior U:JO do not require tho fine definition which 1* dosirod in rich applications cs radar. Sinoe piotorlal definition secas to conflict with essential corrpu!-»er character! £tiose it le probable ;hat nz at leapt vdll bo siado to obtain definition of pictorial quality. Usable ooaputer definition will bo liiaitod by defleotion ciroult problems,

12. Reading and writing era bo accomplished by strttcMng the potential of only OEO electrode, eitbor the signal grid ox* signal plate.

The experiaeatal tube froa which much of the abore data '-'as obtained is shown la figure S* The storage rairfacj ic at '.he snail 2nd of tho tubo ar.i tho w i tin --reading jua and the holding gvn .arc la separate ijlasp necJtc, The final tubo dosign will roscaiblc Figaro) 7 or Figure U0

Dceiga, construction, find totting of oonplete ;tcrago tubes rep:r«cent only a fraction of the effort devoted to th» otorago problem,, liiioh r*e&arch has beoc direct-5tl toward tho gathering of basic iaformation( the study of deflection control circait30 tubn construction tcohniquen, and th" doei/pa o:? cpecla.1 test equlpjneat. All testing has been ione on a pulse b:.ciu corresponding tc the condition? vnder whLch the tubo cjust evontuailj operate-


http://iu.be

.. :\

6340 Memorandum 4--.L02 •4" September IE, 1947

Approximate division of the Project fclrlwlnd etaff follows.

Storage Program

Circuit* and Systsrao

February, 3.347 October, 104?

s taff Morten

£

20

Per CeJ Sotal

3o$e

so*

^t of Staff Kami) or a

8

32

Per Cent of 2otal

2<#

80£

Division of activity in the etoxago tv.to yorfc ie approximately;

Po:.* Gent, of Storage yfc« Staff

Per Cent of Total Staff

1. Design of Test Ecjuipsont end oircvits

2t Study of 2ubo Conatraetiou Method! *md Techniques

3c Daoio Research into dielcotrf.ca, secondary woiBcioa. etCc

4„ Circuits to co used vith tubes for deflect ion output readings etc,

5, Actual construction, .and lent of Storage Tabe Hodelo

30

15

23

10

80

^rom the abovo table, d-iaign, of test equipment (pulua generators, gato and delay circuits sweep generatore- and pove:.* supplies) has required more effor", than any other phase of tho storage division, Aa a remit ">f thin worfc( tontr. have boon possible on a puloo basin and tube sharacterlatles are available directly in terms of the soaputor amplication,, Results of this work centri™ buto aleo to oth g JB of the computer program.


J

Q0VO

Memorandum M-102 -5- September 16 .'.547

2. Tube conatructJ sn methods ere of the upmost Importance in ftoragf tube work and techniques .*r.ust be mora carefully controlled than in conraer:.cal vacuum tub© conetraction,, liking orene, vacuum systems and an induction heater imv© beca constructed or procured at 3hown in the laboratory vieve of Figures 9 and 100 Slaae lathe vork a* in Figure 11 io required for tuba construction, Personnel and the necessary background of emparlance will be available at K.ICT. to produco tho tubes required for tac Whirlwind I computer. For larger quantities of tuben the required training can b^ given to a tube manufacturer without delaying conputor vork *hlle a now grcugp eol7K3 the same difficulties already ovarccne by the re-ssaroh laboratory, With available equipment ?jiii loersonnel t ta.be3 of any type required for tectin& can be oonetruoted in a f&w days. The rate of advancement iw set by procurement and analysis of datac ret by tr.b<j construction,

3. Research studies hare boon oosduoted into the probleno asocciated with storage tubes, Sh^ uso of Elumlntns oide an a dielectric and the preparation cf the oxide surface hca for oasy month3 occupied the tirao of one staff aomoere Surfaces ar shown in Figure ? -'zre 'Bboesed on alumiaum and tinodizei to prodv.ee a dieleotrlo covaring„ The nosh of the dielactrio tsurfacc in small oomparod to the beam diameter,, Figure 12 uhevs th* esperimer.tal anodising apparatus and Figuree 13C ld„ 16r and 15 ntrpz in the preparation of samples for testing;. Since reliable i;o«t3 of the dielectrics ct?\ only be made ifc high vacuum, it ie neeeooary thafc samples be sealed 5.nto vacuum tuber. for taking measurements.

4„ Defloction of the cathode rty be in ie one of the »e;"or problens in the type of tube being considered. Several kinds ol circuits have been tested and the desired results eppcar within reach, Final selection of a deflection oyoira from aaoc£ BCTeral available will be made on the basis of properly balancing tho requirements cf accuracy and reliability against tho amount of equipment involved.. Test results show that deflection of ban\;t o:? storage tubes ia parallel can be achieved in 2 mi jrooecondo. Design cf a defloction system haa not lieen completed because available staff time van more urgently needed on other circuitsr

5, Some 20$ of the tiaie devoted io ;he etorage program or 4$ of the project staff time haa boyn e:qpended on actual, atorEgo tubes. Tubes have been constructed a3 required for testing and coordinating tho results of other reeearch. Some of these tubes are rhowa ii Figures 6C 17„ and 18, Zhe tube of Figure 6n firut assembled in August of this year ie the only on© to be constructed to incorporates all the principles to bo used ia the final ttxb©„ It in the typo of tube described last year at the Harvard Computer Syxposiura. Operation of this tubo corresponds* in all rospocto to predictions which can bo made on the basis of geometrical and circuit considerations, Zt vaa


http://ta.be

http://prodv.ee

6346 Memorandum M-102 >6- September 16, 1647

constructed for etudy purposes and after teste will ba uesd for demonstration of principles aod for lifo rtuiicc* Similar tubes will be built e.s required and vrtll incorporate changes toward & final design

Continued research in oevoral directions will be roquired to dsvolop the desired tube,, l) Work to the present has "seen on storage surfaces of aluminum oxide or one of the floreaeeat phosphors. Kclthe.r of thoee has a sufficiently hi,~n secondary emiseion ratio. Satisfactory resuit8 seam indicated from the resuita obtained on oxides of beryllium end other metalu which ?iav9 beon studied at other laboratories. Some evaporation tests have not yet been completed. I low firct crossover voltagec Va Figura 19, on tho secondary emission curve Is desirable. This characteristic will in general accompany the greater secondary ealsaioii, A high second cross-over voltaget V0, ie desirable in order that a high primary bean volta&o may be used to permit higher beam currente J3aeed on the little information availablet a high 6econd crose~over voltage is azpocted to be a 'Sharacteristic of bsrylliua oxide life studies muet be initiated on tho secondary emission surface as noon as a tontativo Bolection is madee SOBS surfaces change their secondary emission behaviors due to contamination from the tube cathode isd from residual gas la the tube Additional circuit design will be required V.o develop output detection circuits which will receive the tube output but exclude tho transients which are produced from grid switching during reading and writing* Resolution of the tubo over that thua far r.erc-onetrated will be required bui. the problem appears lose difficult than others and can b» approached through, entailer be>ja dirjneterr.c closer grid opacingc and a fine storage surface mOsaio if the mosaic is rotained in profo.rfi.nco to a plane surface-, St is probable that better olectron guns should be procured than now available^ Present work io done with commercial cathode ray tube guns, better guis are in use for other purposes in this country and in iicgland and may prevj satisfactory. An Increase in gun current without loss of definition is desired. Special gu»a procured from England are available in the laboratory but have not yet been tested.

Comparison with the IBL Tube.

Full details on the operation and dorelopment piano for the SEL tube aro not available at the Scrvomachauisas Laboratory but the following comments are in answer to the requent„

The HBL tube was developed initially for tho otorage of pictorial information in radar applications. It shows considerable promise for this purpose although the tubs is in the very early research stages and an ovaluatior. would be difficult for some timet, The characteristicB which make the tube ueeful for radar ari» of doubtful value if not hazardous for computer application,,

Figure 20 repeats tho information of Figure 3 on tht M,ICTC tubes and adds a column on the NHL tube0


http://profo.rfi.nco

102 -7- September 150 1947

Ho information has been released and the tube has probably not been tried for parallel digit transmission. The requirement for parallel digit transmission was not known to the Bill group while the tube was under development for radar and the concept of parallel operation has only recently been introduced eince the tuba was taken over by the 3ureau of Standards on a contract to Raytheon, Difficulties in reading to a parallel hue with a stationary beam are discussed in the next paragraph. Tests to date have been for Bteady state scanning for picture presentation.

Signal levels thuc far obtained from the BBL tube are the order of 1 millivolt. Thin will require signal detection amongst the switching transient noise of computer operation or else changes of a fundamental nature in the tube must be introduced. The voltage division method of signal detection pictured in Jlgure 4 ie probably unsuited to the tube because Bignal and beam current are carried through the sane electrode. Figure 21 show3 the output signal comparison between the NBL tube and the two methods of signal detection in the M.I.T. tube. It le possible that changee in the NBL tube can be made for computer use to increase the present reading beam currents of about one microampere to the 100 microampere level of the M.I.T. tube.

Bo data is available on the pulse writing speed of the BBL tube although some conclunions pointing to the "microsecond range" have been drawn from steady state scanning tests.

The same comment epplies to reading speed.

Storage time of the BBL tube is controllable over a limited range, This appears to be its principle unique feature which is valuable In some radar applications but of no value in computer work. In factB precautions must be taken to insure that thin feature cannot cause erasure and adversely affect the reliability of the tube for computer application.

Signal, storage of the tube is presumed to be permanent under proper conditions but movies and demonstrations to date have shown unexplained skipping and erratic operation of the vrlting beam which must be overcome.

So far as known, the problem of checking signals has not been considered with respect to the BBL tube.

Fo data is available on resolution of the tube in computer terms since so far as known no tests of this type of storage have been made. Results of pictorial testing would indicate satisfactory resolution of the screen with the limit probably set as in other deflection tubes for computer use by the deflection circuits,


1' *

6345 Kesao.Trnc.iun M-102 -8- September 15, 1S47

9„ Storage capacity la probably adequate srd the preceding C0E3\e;it3 apply.

10,. Since the signal 1V> aupo-'ispoijed on the part of the beaa current vfhlch is collected by the eigaal scroen5 only variation in jia.'pai-tude of this current is available for eigcal detection. Problems will e:d.»t in obtflidng a olga&l variation for parallel tute operation and the situation will be aggravated by variations in primary beam current due to transients rnd cathode changes. Operation of the Project Whirlwind storage tubo la joesible vikh wide v;.u-laiionB in beam current,

11, Pictorial presentation is a feature of the E8L tube vhioh probably will not be pesalble with the M.Z.T. tube.

12r. Operation in the F.c£in««r thna far described by ItRL requires switching the potential o.f tho entire gan and anoc.e oirouite. High capacitances toast be driven 'oy short voltage pulaoa anc. different control methods may be required,

Prior to the award of a contract to Saytheon, I diocuaauc. the 5EL tub«» with Mr, S, N, Alexander of the I/uroau of Stani&rdn and agree vfith him that the tubo is sufficiently different from the K.X.T. work to .Justify Its pursuit. I feel however that the tube has been g.roeely "oversold" with roaject to itt- present etatun. The practical problems of neing the tube hare Dcen dieclsscd as "cicro engineering details" by its HPX tiponsora yhiio the tube is still subject to mo?;; of tho research problems outlined for the H.I.ff. tube in the laet paragraph of th?> first section of this nesorandTCJ and fcao before it in addition the difficulties discussed in thiti eocticn„ Thos? can doubtless bo overcome but aome research and considerable develop-nent still appeer necessary.

iho only new principle ehovn in tho 821 tube 1 B tho variable docay tiise fox-radar work which ic not required in the OOKputer application,. Peirnanence of storage through the use of low energy clcctronn 1 B cosnion to vork abroad, the RCA Solectron, and Project Whirlwind. It is probable that the EEL tv.be will hav« io: i difficulty with cecoadary electron redistribution than the H.I.T, tuba in vfc~.ch it b?.o been neceecary for eoae work to go to a matrix etorage surface with electron barriers.

rubes built at HEL lievo had three electron gone for reading, writing and holding, 'ince computer use -sould require the reading end writing guns to trade nearly *-; ;ly for the sene deflection voltages, vhioh vould be almost impossible for good defi.. 'on, the Swytheon version of the tubo will combine tho reading and writing operations into one gun ae is done in the K.X.f, tubo.

Examination of the table gi«reu previously for the distribution of effort on tho storage program will ahov that 30.40 20% hao been devoted directly to tubo building and testing. Thie part of the work corrocponds roughly to the ERL tube program nince no tet.t equipment for pulse stuiier- has been used, tube construction was 8ubcontractod to one research group and recently transferred to Raytheon under tho Bureau of Standards to that to dnte the work has not resulted in personnel trained in construction methods, and no --research and development hs.s been directed toward use of the tube in computer circuits,,


http://Kesao.Trnc.iun

http://tv.be

6345 Memorandum M-102 - 9 - September 15, 1947

I have dwelt on the difficulties to be encountered by the ERL tube in some detail, not because of doubt that it has a useful future, but to call attention to what I feel is an unjustified impression gained by many that the tube is now ready for use. The history of storage devices has been one of overoptimism repeated as each new idea or design has been presented. The NRL tube appears to be among the brighter prospects for radar use and should be developed along its original lines for simulator use. In simulators where data is to come from a computer to be presented on a radar type screen, the separate write and read guns and adjustable decay can save much numerical computing time.

I am pleased to see the NRL tube continued under development and feel that it will receive oareful guidance and evaluation from Mr. Alexander of the Bureau of Standards.

Drawing List Instructions - Put Figure Numbers on Prints

Figure Figure Figure Bigure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure

1 2 3 4 5 S 7 8 9 10 11 12 13 14 15 IS 17 18 19 20 21

Drawing Drawing Drawing

SA-30892 A-30885 A-30886

Drawing AAA-30887 Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing Drawing

A-31324 A-31325 B-30888 B-30889 A-31326 A-31327 A-33 *28 A-31329 A-31330 A-31331 A-31332 A-31333 A-31S34 A-S1335 A-30893 A-30890 B-30891

JWFtvh


A-^0£S2. USEO IN ^ 4 5 MEMO NV 102

DIELECTRIC SURTACE

ELECTRON OUN

DBTLECTION PLATES

FIG. I SIGNAL GRID

SIOIAL PLATE

SIOIAL OUT

ELECTRC STATIC STORAGE TUBE

Ban


A - 2 0 £ > 6 5 USED IN <2B'S M E M O M-JOS

/

READING AND WRITING CUN

SIGNAL GRID

.+ t2

O ^ <S.S S^M

fZXr.

092

4#^ HOLDING GUN A T - POTENTIAL

» ^ »

e

rr DIELECTRIC

+ f

FIGURE 2

-V VOLT 3


•

•

i •

1 0

! 1 hi

in

v5

a tu in

9 i

<

<n 5 3 3

u 01

Ul

Ul

Ul

D O H-ui

3 2 * <*

c/>u.

2t Si? ul 01 a u. o i o 5 <

2 8

x._

4

5 8

M id *

0 2c-

SI - %

X

o

•

Ul >

<0 ul M*

>

3 g 5>

a. -

•

88

55

o3 > 7

O

h 41*

lid *

i O i i 0 It*

* o •J

s 5 N

O

i

5 3 tri

•

i 8 £ N

0 H

« U Ui «0

s 2 N

F

Hi 0-<0

1 <

i 2

s ft.

2 Uf 2 <

ui

a

ai

m

O

5 in

O O i9

2 Ui

a X ui

7*

i § df

r' • *

0 m 5 v

0 2

& m 7>

i • 2 5

2

0 H

in

o

to y

5 <J

2 * Pzui

<J2f

^

• ^

ul

01

m

5 5

g

p

i si

!

<tf

0 tti 7*

ul 7»

^ ^i < Mm

vS

1

2

•

o

o 2

to a-4 5 21 a

2 o

2

s 8

•

0 2

0 2

J

i 5 >* ?3 2 K

10

• c*

on

UL


.- .-

A- ooseTJ U S E D IN <S3<5 M E M O M- l O *

C^tCAPAClTV OF I TO 4 yuyufd &ET\NE£M S 1 6 K A L PLATE AND AREA. OKI DtELECTPAC WMEVRE A S\NGLE D\G\T \S 5TOCGD.

I

Sl&NAL PLATE

- • OUTPUT S\C>MA\. P R O P O f c m O N A L . TO TOTAL CVAAP.GE TVfcAKASPER.REO T U e U

STRAV AMD D\STR\fcUTEO OUTPUT CAPAC\TANGE Of SOWNE \SO jujatd.

BEAM FROM *EAO\NG AMD Vs)R\T\NG GUM

V O L T A G E D\\AS^OM M E T W O D

OUTPUT S\&MAL, P t t O P O R H O M M . TO D\SP\-ACEKAENT CURR.EKAT T U 2 U C*

"T\ME COMSTANT R C 0 <=>VAOR.T COMPARED TO fcE*0\NG POL.SR

CUBBENT DETECTION MtTUOO

OUTPUT VOLTAGE

FOR. CONSTANT CURRENT

THRU c:

VOLTAGE 0 W i S \ O K l METUOD

CURRENT OETECTVOM KAETHOO

PUL.SE LENGTH

SIGKIAL LEVEL DETECTION F\G>. 4


O

O

z

c I t

« c r r


csl O

O

o UJ CO

m CVJ

to ro i

<

CO «

CD


• B - B o a S 6 U S E D IN & S 4 S M K M O Nft- l o t

- j ti

DEFLECTION TERMINALS

STOR^ ASSEMI

ELECTRdi

ANODE?

R E A D AND WRITE- GUN RING TVPE HOLDING CATHODE:

F IGURE 7

POSSIBLE PINAL STORAGE T U B E DESIGN


U S E D IN G 9 4 3 M E M O M - 1 0 2

READ AND WRITE GUN

STORAGE ASSEMBLY

ELECTRODES

MOLDING GUN

F IGURE e

POSSIBLE INTERMEDIATE STORAGE T U B E D E S I G N


I

o 10

10 <M ro rO

i

<


O

2

(0

O lO

a z

o UJ t/> 3

o o T.

FIG. 10


CM O

O

o Ul (A 3

CO

fO

IO I


01

o

o

o toJ CO

en

10 i

<

-

. . . . .

-

i V.

• >

'

— Ml

mr

'9 "k I I

£> - U

. .

'

HBwS

>

• > -

i H

v~ * \

, i i

1 1 • - I

-•'

.

@

•

SlS

-

^m

M

*

•

. '.'•'

I

1!

!

"3 >

» v l i

#

. . .ajMMWWHM

1

•

H • /

» ^ — 1 jfij ' 1 i

,;jg

- : l

' • % • • •

1 1 ,1 1

tr 1

i * ggKff^. m

• • "

0 I r

r

I i

i

1

11 •

» 3 9

k

r

^^™

\ • * T i ^

O


01 o

o ro

a as

. I

; )

ro

Li_


o

4 >

i J


-d\56Z USED IN M 130 a M 102

FIG.15

FB 154 9 PY 77


USED IN M 130 & M 102

* <$*

FIG. 16


'

M O

O 10

a 3

ro ro ro i

<

6


o

o

o di

3

O

1

GO

u.

• t


A - ^ 0 & S 5 USED IN (23*"? MSMO /

M- /bit

• i •

INCIDENT (ELECTRON VOVJTS)

FIG. 19 SECONDARY EMISSION RATIO AS A FUNCTION OF THE ENERGY OF \UC\DEUT PRIMARY ELECTRONS.


\

I

(V 0

5

o 2 lu 2 io v tn \9

Q UJ

D

o (P dO O

•

•

•

- *k

i/>

5

2 j

uJ

at 5 Z uJ

Of 25 ° a

UJ

r< .1 *-

4!

4

2 O 2 O

S a 5 0 0

,\ .

uit; 4 i/1

3 a* F t0

ui °*

S i * iv Z

j k

a1

. tfl < Ul

10 UI ui «/>

? IS a d

0 I*

5 ui

*+»

m % O

ui

ui 4 1 0

D ul

01 ^

O Z

ui

1/1 ul 7»

I O

ui

a

«•"•

ui[i

•7 mm*

8

oo

0 0 0 0

O O

Z ?

U>£

is 0

H§

^3 2 0 J ,

ti

4.

4 O

0 Z

0 ui

O of 0

5

0 ip

0

<A

0 at 0

5

O

s ui a

Z

at 3 (O

4

0 z

u ul

0 at 0

5

0 v-

6 ul

</>

0 61 O 5

O F

s ui a-

1 3 < UJ at <

4 0? 1

z Ul z 5 At

a.

z Ui Z <

ui a

5 F Ul (9 4

O

5 uS

J

5 0 0 D

O 0 0 v9

Z ul J

ui 0 ul

7* V-

J ui

vSi

h-0 Z

3 10 4 10 O

ul 01

0 Z

ui 7*

** Z v9

Z

5 0

H

s'

i5

2 O

P 0

ul ac

< )

# t

0

0 Z

4

0

ul

0

^ui

22 So 0.0

0 z

• • * H ul

<

J <

y

5 5

0

> y

0 <r D.

O

ul

2 O V-

<r»

0 Z

UJ 7*

ul 7»

^1 < Sim

In 1

2 O

ul J (0

^

0 a-

O Z

2

s a

z 0

z IV

t i s 0

01 ^

0 z

0 z

J

V %

5 ui

z 1-

31 D ui

5 ^

e

o CM

LL


3063\ USED IN G34S JEMO M -loa

MOTE BREAK IN SCALE

OUTPUT SI6NAL (VOLTAGE) READING ONE (CAPACITY DIVIDER METHOD)

OUTPUT S IGNAL (VOLTAGE) FOR READING DIGIT ONE

( C U R R E N T METHOD)

HEADING BEAM CURRENT

NOTE BREAK IN SCALE

CURVES IN THIS REGION TO 5 T I M E S SCALE OF THOSE ABOVE

TIME

OUTPUT Sl&NAu 6/OLTAGE) FOR READING Z E R O ( .CURRENT M E T H O D )

DOTTED C U R V E S N R L T U B E SOLID C U R V E S M I T TUBE

O U T P U T S I G N A L READING 21ERO

CCAPACITV DIV IDER METHOD)

NOTE BREAK IN SCALE

FIGURE Z\


approved for public release. case...

Documents