US2436833A - High density beam tube - Google Patents

Description

K. SFANGENBERG ETAL 12,436,833

March -2, 1948;

HIGH DENSITY BEAM TUBE Filed June 15, 1942 3 SheetS-Shet 1 l6 01s TANGE o/v AX/s 9p 7 o s 4 P/u'ssuRs mm mv mo KARL JPENGJVBERG ROBERT new By LESTER M. FIELD ATTORNEY 55AM SPREAD R/ March 2, 1948.

K. SPANGENBERG ETAL HIGH DENSITY BE TUBE Filed June 15,1942

3 Sheets-Sheet 3 5 6/001. mmsu/ zza 6000;. 6422014750- I 370m. Mmsunza ,|6sa76 s+ a 2 [#957554 a 2. K'7887654 PRESSURE r 4 mvem'ok KARL SPANGfNBERG wear/4am BY LESTER Milan ATTORNEY.

Patented Mar. 2, 1948 UNITED STATES PATENT. OFFlC HIGH BEAM TUBE Karl Spangenberg, Lester M. Field; and Robert Helm, Palo Alto, Calif., assignors to International Standard Electric Corporation, New York, N. Y.. a corporatlonof Delaware Application- June 15 1942', Serial No. 447,194

.This invention relates to electron beam tubes, more particularly'to tubes such as the velocity modulation type wherein the electron beam traverses a field-free space for a comparatively great distance;

In tubes of these types it is'desirable to produce an electron beam of high density, that is a beam' haVing as little spread as possible, somewhat analogously to the production of a high concentrated beam of light for projection over a distance.

Such tubes are of great use in the electronic art, examples being velocity modulation tube beampower transmitting and receiving tubes, television projecting tubes, high-intensity X-ray tubes,- and:

other similar devices.

In tubes of the types just mentioned, maximum;

efficiency and control depend, inter alia, upon the production of an electron beam of relatively small cross section, carrying as high a current as pos-- sible at as low a potential as possible, i. e. a rela-- tively, low impedance electron beam. In such 14 Claims. (01. 315-16) a 2 tron beam. 'Experience with cathode ray tubes incorporating drift tubes of substantial length;

lation tube, show that there should be produced sufiicient positiv ions to neutralize the charges within the electron beam, provided that the gas pressure be not reduced much below l0 mm;/Hg, a degree of vacuum near which comtubes the production of a beam of this type is beset with several difiicultiesone greatly detri-. mental factor being the tendency of the beam, once produced, to spread out as it progresses along. the length of the tube, due to the mutually repellant actioncaused by the space charge effect of the electrons in the beam.

As one means of preventing this spreading out of the electron beam, the art has depended to a; large extent upon the ionization of molecules of residual gas within the tube, the positive ions formed bythe impactof electrons from the beam.

upon such gas molecules tending to neutralize the negative charges of the beam electrons and thus tending to prevent the mutual'repulsion ofthe latter. The problem has been complicated by the 4 question of the degree of vacuum existing within the tube,v thenumber of positive. ions produced being dependent upon the number of residual gas molecules present. Several factors have contributed to the production of tubes with increasingly high vacua and atthese higher'vacua fewer positive .ions are produced, so that this remedy for beam-spread has become decreasingly potent in its-effect.

.With the advent of tubes embodying afieldfree region, such as the drift tube of a velocity modulation tube itihad been assumed that the generally practiced enclosure of such field-free space by'a conductive tube would allow theposi tive'ions produced from residual gas to;collect within such tube and therefore wouldcause them to'ex'ert'a greater and more continued restraining actionupon the undesirable spread of the elecmercial tubes are expected to operate.

The discrepancy between the remedial effect ofgas ionization postulated by theory and the actual insufficient effect produced in practice has hitherto remained unsatisfactorily explained, although several reasons for its occurrence have been suggested. Lackingany satisfactory explanation for this undesirable phenomenon, no Way has been' found in the prior'art of overcoming the same by 30' purely electro-static means and recourse to electromagnetic concentrating devices has been found necessary.

Our invention is designed to overcome, solel by electrostatic means, the beam-spread difiiculty" 35'- above described and the modus operandi thereof is hereinafter set forth.

Ourinvention' has for one object the productlon of an electron beam of low impedance and the propagation of such beam over a relatively great distance through afield-free space enclosed by a conducting shield;

Another object of our invention is the provision of an electrical circuit and determination of the constants thereof, which'may be applied to an electron beam tube in order substantiall to reduce or practically wholly to prevent undesired electron beam spread within such tube.

Yet another object of'this invention is to pro vide a method andmeans for trapping and confining within a predetermined region positive gas ions produced in an electron beam tube.

Yet another purpose of our invention is to provide quantitative data from which the design,

construction and operating constants of an eleotron beam tube may be predicatedin such fashion A still furtherobjectis to produce upon-the target or screen of a cathode ray tube, an illuminated spot reduced in size to substantially that of the cross-section of the beam as initially generated, without the necessity of employing electromagnetic beam concentrating mean's'to secure this result.

Another object is to providemeans and a method for limiting the cross section of an electron beam to any desired degree commensurate with the current carried thereby, by the employment of purely electrostatic means.

Still another object is to provide means for maintaining a dense electronxbeam'in extremely good vacua, such as 10-? mm./Hg.

Fig. 1 is a diagrammatic representation showing one form of cathode raytube, embodying our invention and also showing the axial potential distribution thereof when connected according to our invention.

Fig. 2 shows a family of curves illustrating the beam-spread effects observed-with a tube constructed and operated according to the prior art.

Fig. 3 shows the beam-spread effect due to mutual electron repulsion, in the absence of positive ion effect, of. an initially parallel electron beam.

Fig. 4 is a nomograph showing the beamspread effect corresponding .to'a given potential, total current and beam length in the absence of positive ion effect, of an initially parallel electron beam.

Fig. 5 shows graphically beam-spread effects as observed and as calculated according to the theory of operation of our invention.

In deriving the data upon which our conclusions as to the cause of the difiiculties encountered in the prior art with electron beamspread were based and in deriving formulae useful for the design of electron. tubes for overcoming these difficulties according to our invention, the following conclusions were first reached from theoretical considerations and were then tested experimentally as to the correctness thereof.

For a typical electron beam, one having the following characteristicswas-chosenas an example:

Area1 cm. Current-0.05 ampere Potential-6000 volts Produced in a as pressure-of mm./Hg

The firstpoint to beconsidered is at what pressure enough gas molecules will be found present so as to furnish a sufficient quantity of positive ions completely to neutralize the electron charges of such beam. Formulae well known in the electronic art give the number of electrons existing molecules to supply a quantity: of-positive ions.

great enough to neutralize all the electron charges present in the beam.

The next question arising is concerning the actual number of ions which will be formed under the above stated conditions. This value may be found by the formula set forth by S. H. Bennett in an article entitled Magnetically'self focusing streams and published. in'Phys. .REV., 'vol. 45, No. 12, June 15, 1934, pp. 890-897.

."According to this formula the number of ions formedper cm. length of beam per second is equalto 200 pi V0 in which .1) is: the pressure expressed in uum/Hg;

i is the current expressed in E. S. U.; V is the potentialexpressed in E. S. U.; and e i the electroniccharge, also expressed in E. S. U.

Applying this formula to the instant assumed case, it is found that 314x10 ions are formed per cm. length persecond. Although the life period of ions under such conditions and with an allowance for any possible existence of.recombination, is of the order of one second, the foregoing values derived from the electron count and the speed of production of ions, show that each ion need live only =2.17 X 10 seconds I result in therebeing present too few moleculesto provide the necessary ions for complete electron neutralization. Now commercial tubes of the'electron beam types usually operate at pressures ofabout 10- .mm./Hg and yet his well known thatsome tubes actually do display detrimental beam-spread effects as soon as the pressure is reduced below 10 .mm./Hg. "We have made careful experimental determination upon such tubes provided with a field-free space or drift tube especially well shielded and yet :the beam-spread effect reached. noticeable proportions asisoonas'rth gaspressure was reduced tocarpoint' lying: between. 10- andlo" rum/Hg. i .-In-: Fig. 1 therexisshown as oneexample of our invention an electron 'tube having a cathode structure I'll, ;a,.l'fi1$t accelerating electrode ll.

1' held at .1000. volts: positive potential relative. to

\ the seathode, asecond acceleratingelectrode i2,

held 'at. 5000 volts: additional positive potential and. provided withan .apertured end structure l3,:-adrift tube M'helclat about 15 volts negative relative to the electrode [2, and-a target or 1001- lectingeiectrode 15, held. at the :same potential aselectrode l2.

The axial potential distribution, :as shown. by thegraph, "startsxat point :16 .at zero potential,

rises-asshownv at I! when electrode] I isrea'ched',

rises again sharply as shown at I8, until. at

structure l3'thezmaximum potential. is reached,

showrr'by peak 1:9. There .is .then av reversalof potential gradient .to the level 20,. existing 2 throughout the vzd'rift tube. l4; and again a rise-of gradient to pointizi, corresponding to the .potentialsof target 15.8110]; rec ualv to that applied. to..electrodezl 2. v

Fora. betterzunderstandingof the operation of'the device ofeFig. 1, reference is made. toFig.

2, where are plotted the experimentally determined beam-spread effects found inan electron tube similar to that of Fig. 1, but having the drift tube maintained atthe same poten-- tial as the electrodes situated adjacent to each end thereof, respectively, this mode of connection being that commonly used in the prior art.

From the graphs shown in Fig.2 it will be noted as the gas pressure is decreased while the potential is held constant, the beam-spread effect increases at a rate which varies according to the particular potential employed. The drift tube potential shown in Fig. 1 was varied in the derivation of these graphs. At lower potentials the beam-spread effect is barely noticeable until the pressure is reduced to about 10 mm./Hg and increases comparatively slowly as the pressure is further reduced. At higher potentials the beamspread effect starts at about the same degree of vacuum, but rises comparatively rapidly to a greater value, reaching its limiting value of approximately R/Ro=3 for a 6100 volt, 100 milliamperes beam before the pressure has been reduced by the factor of 10. These effects of beam-spread are partially explicable from the fact that over this range of voltages the probability of ionization varies inversely as the voltage and undoubtedly a still higher maximum beam-spread would be reached if the vacuum were pushed further. When the voltage is increased above 6100 volts the beam tends to reach its limiting spread quite rapidly and this limiting spread is considerably smaller than'that reached around 6100 Volts, this latter effect being that which would be expected from ordinary space-charge beamspread considerations, independently of any positive ion phenomena.

As previously stated, mathematical analysis shows that there should be substantially zero beam-spread at pressures below 10- mm./Hg, due to the positive ions formed. Yet the results plotted in Fig. 2 clearly indicate that for some reason the positive ions fail to operate in the manner assumed by the considerations upon a which the prior art has been based.

In Fig. 3 in which Z is the axial length of the beam there is shown a graph illustrating the theoretical beam-spread effect in the absence of positive ionization. Theory shows that spreading beams are reducible to the shape illustrated by the graph by a suitable change of vertical and horizontal scale factors. Further in any given structure if the potential be changed by a given factor the current, if from a space charge limited source will in turn change in such a way that the shape of the beam is not in any way altered. We have experimentally verified this rather astonishing conclusion. The graph of Fig. 3 can likewise be used for initially convergent beams, provided that the beam electrons are initially so directed that in the absence of electrostaticrepulsion they would converge to a point, since such beams become parallel at some point and thereafter diverge.

In Fig. 4 there is shown a diagram which allows the beam-spread efiect to be computed for a given potential, total current and beam length. The laws illustrated in Figs. 3 and 4 serve to explain some of the beam-spread effects shown by the graph of Fig. 2, but still leave unexplained the occurrence of very appreciable beam-spread at pressures lying between 10- mm./Hg and l-' mm./I-Ig, since considerations previously discussed indicate that over this range of pressures there exists a sufllcient quantity of positive ions confined within-the drift tube, to neutralize com-- pletely the negative ..charges. of the electrons in the. beam. These considerations, then, which have been assumed to becorrect by the prior art, show that comparatively tcomplete neutralization ofthespace-charge efiect in the electron beam should obviate the occurrence of. appreciable beam-spread at pressures in the neighborhoodof 10- mm./Hg l We are thus confronted with serious discrepancies:v between the assumptions upon which the prior art structures were designed. and connected for operation,.and the actual results obtained in practice,- suchresults including the extremely detrimental and undesirable spreading of the electron beam to a .very greatdegreawith .consequent failureto .maintain the desired, high density and .lowimpedance of the electron beam. After the consideration of several possible tentative explanations for this -discrepancy,'all yielding negative results,twe arrived at the conclusion that thepositive-ions were removed axially from the portions of the electron beam adjacent to the entranceend, of the drift tube, and-that as the positive ions were so removed: other ions. from themorecentral-portion of the beamwere fed axially along the. beam so as to reach the:

portions thereof near ,-the ends of the drift tube, wherethese ions in turn were swept out and removed. This process just described amounts then to a continual leakage of positive ions along. the electron beam in an. axial direction until such ions reached points lying without the drift.

tube. This means that complete neutralization of the space-charge effects tending to produce,

divergence :of a portion of the electron beam lying within the drift tube, will not take place according to the conclusions above reached,

mathematically; since the quantity of, ions available at pressuresbelow 10- -mm./Hg will be insufficient to take care of this. continual axial leakage or loss of ions from. the. interior of the drift tube.

Referring againto Fig. 1, it will be seen that we have established at the respective ends of the drift tube, a reverse potential gradient. We have foundthat this comparatively low reverse potential gradient serves to trap the positive ions within the drift tube and thus substantially to prevent the removal of positive ions from the interior of the drift tube by leakage of ions in an,

axial directionalong the electron beam andthus substantially to prevent-beam-spread atthe degree of vacua. employed in ordinary practice,- for example 10*? rnm./Hg.-v

While we believe-this. theoretical explanation lust given of the trapping action of our reverse potentialfgradients to be a substantially correct one,. yet-gwe-do not confine. ourselves to such theory, but we have found by actual successful experimentation, that a cathode ray tube arranged and connected in .--accordance with the principles illustratedin-Fig. 1, actually does virtually completely prevent the undesirable beamspread effect encountered in the prior art. By

the employmentofourinvention, it is unnecessary to employ *the', magnetic fields found necese sary by the prior art in order to maintain a high density beam having low impedance. 7

According to another viewpoint, an ,explanation of the removal of positive ions from the interior ofthe field-:freespaceconstituted by the drift tube may be found by considering that there exists adjacent the beam opening atone end of the drift tube an external field which extends.

.sneaeae oversassma11edistanceiihitoxiitheefleldsireei.spa

: and whichrfield continuously "to :sremov'e -of.;'thesbeam. will become-imoreinegative impotential,therebymansihgavalf flovfijdiolthef bfian'hiiflfiiions firomaotherssectinns In other-.xword esmechanismaof mayitbe lconsideredt-in.iitheriguise z'ofrsan ion sink pointsimtheiel-ectron Omthe'se considrapparen' thatiionireniovalxfrom tions it iwi ltabe the electrortibea maytake-iailace overfdistances fargreater tham heidi-stancetol-whie esac'tnal external: field pcnetratesiwithin'theidriftltnbe.

We'fhave de-termined mathematically,- fellowing the'data to beaderived fromtheaboveitheory, what would :be S-the iappreximate 'vbeamspread actionof -suchmechanism upon anelectron beam having given-constants. lne -Fig? eareishow'n graphsiillustratingi the beam spreadtaction measure'd =in some oasesand calcnlatedaa'ccording to the mathematical snethodsabove referred to; in othencas'e's. -it -will 'be observed that themeasuredmnd calculated valuesearedneiose accordance with one anbthergehusitending to substantiate the I correctness-pf ouritheory of an ion sink above stated. i

Our mathematieally derived -iconeluslo'ns allow us to predict the general-shape rand "approxim ate values of 'the-curves of==Fie: -'2. risthe voltage-" is lowered-monthly kis'theeffect much th'e'same as an increase ofpressure wouldihave*had 'inreducing beam-spread;-but-a1sdthepotexitialgradient -at= thedrift tube aperture is-lowered in direct proportion to' 'the voltage=-decrease;thus still further acceri tmating the efie'ct. The -effect will be at first almost a function off-voltage squared; thusagreeing with the measured results shown in Fig. 2. ==At*the'higheriran'gescfvoltage,

the maximum-- spread inia perfect Va'cuum would become less and '"ss as the voltage is ra-ised. These two effects ogether explain why higher voltage-beams-reach. heirulti'mate spreads more quitalelyastheg the respective timate walues of --'spread; grow progressively less,- a as -the voltage rises.

In carryingioutiiexperinrenta tests serving to substantiate the correctness o'f our:theory-'-and serving as a basis for the embodiment of'wo'ur invention, apparatusarranged-as"shownin'Fig; 1 was used. "Thevnpward lrihks*1 and 21 of the axialpoten-tial curve=were*obtainediby-maintainingthe-drift tube at --'a; potential*negative "with respect to boththe second accelerating-"electrode and' 'the' collecting electrode; by'an'amountslightly greater than the voltage existing between the edge and the-center-'of= the*beam'wherr the latter consists only of unneutralizedlectrons. Due to the fact that. 'the' drift--tubecollects a current of stray electrons, merely-placing aj-re'sistorin an external circuit 'between "the drift tube and the :aperture --provides the-necessary *potential. Should stray 'eleetron currerit--prove-=insufiicient for this purpose, an auxiliary voltagesupply could be used.

immediately stoppedspreadingWhea es little as ressure isadecreased;'butthat 1'5svolts was applied between drift tubezand adj acents-electrodes in .such direction as to reverse the gradients at thesepoints. Since 15 volts variationgfrom 7300 can haveno .appreciable electronilens 0r focusing efiect per se, this result may be considered as successfully verifying thetheory. Values of the spread variation with change of trapping voltageare given in Table I, below.

"Table'I.--Beam spread ion trap data Asjan auxiliary check, increasing the pressure to 5X10 also served to remove thesprea'd as illustrated in Table II, below.

Table IIBeamspread data Of course, in actual use increasingthe pressure is impractical since tubes are normally used as sealed off devices, so this latter Table II merely servesato demonstrate that the efiect of extremely good vacuum inproducing beam spreads could be removed by the use of anion trap .as-suggested here.

It: was noted in the test of ion trap voltage just described. that for the particular beam used 15 volts removed practically all spreading and higher negative voltages produced very little increased efiect, but no detrimental effects were observed when such higher negative voltages were used. It is to be remarked that this voltage is only slightly greater thanthe difference in voltage-between the edge andthe center of the beam. when it consists of unneutralized electrons only. This voltage difierence'for .any section s on thebeam would be where p is the number'of unneutralized electrons .a, negative gradient down which ions canifiow will-always be present until thedrift-space is loweredby this ten volts plus enough more to reverse the gradient produced by field penetration.

-As shown by the-above 'Table I,-the voltage difference between the-drift tube and its ends (back plateand aperture) to form an ion'trap mustbesomewhatgreaterthan 10 volts for e milliampere- '6000-volt beam. 1 The value necessary-would increase directly as the current 'in the-beam'and vary inverselyas'thesquare root ,9 of the voltage of the beam. Thus for volt, 500 ma. beam it would be $3 33 144 volts For a 5 ampere, 6000 volt beam it would be 5o 1 V For beams of this magnitude the ratio of radii of the beam and surrounding drift tube also bex 1000 volts comes of some importance.

necessary is given by the following relations for a 6000.volt beam: Let the volume available=N cm. =Sem.Acm. .1 Let the length of the beam =Scm. 1.

Then

i amperes (Where the volume used has the same length as the beam, A gives the area necessary.)

Thus a 1 ampere beam at 10- mm./Hg needs only a volume the length of the beam with an area of 3.75 cm. For voltages other than 6000 volts this volume should be multiplied by What we claim is:

1. An electron beam tube including means for generating electrons, means for producing a beam of said electrons, means for accelerating said beam, means for defining a substantially field-free space through which said beam passes,

'means for collecting said beam after, passage through said field-free space, and means for maintaining said field-free space defining means at an electrostatic potential slightly negative with respect to said collecting means and with respect to the portion of said accelerating means adjacent said defining means.

2. The method of substantially preventing beam-spread of an electron stream within a field-free space which includes directing a beam of electrons within said space, establishing and maintaining a reverse potential gradient at least at one end of said field-free space, so as to trap positive ions therewithin.

3. A cathode ray tube system including means for generating and for collecting an electron stream in an atmosphere of not more than 10"- mm./Hg, a field-free drift tube through which said stream passes, and means for maintaining the direct current potential of said drift tube tive with respect to said accelerating electrode so as to prevent leakage of positive ion from said 10 drift tube. and means for collecting the electrons.

5. In the electron beam art, the method of pro- .iecting a beam over a distance without substantial spread thereof which includes the steps of establishing a substantially'field-free space, passing through the space said beam, and establishing at each end said field-free space of a relatively slight potential gradient in the reverse direction to the normal potential gradient along the course of said beam, whereby positive gas ions are held within said field-free space so as substantially to neutralize the space-charge efiects tending to cause spread o'f'said beam.

'6. A vacuum tube system including an accelerating electrode, a field-free drift tube insulated from other electrodes and means for keeping said drift tube slightly negative with respect to said accelerating electrode so as to prevent leakage of positive ions from said drift tube.

7. In electron optics, a system for reducing the spread of an electron stream due to space charge effects, including a source of positive ions, a fieldfree chamber within which the ions are produced and through which said stream passes, and means for retaining said ions within said chamber, said last means comprising electrostatic means so connected as to establish at least at oneend of said chamber a slight di'rectcurrent potential gradient along said stream in a direction reverse to the gradient prevailing along said stream before the entrance thereof into said chamber.

8. The method of passing an electron beam through a field-free space without substantial beam-spread, which includes passing said beam through a slight reverse direct current potential gradient, immediately passing said beam through said field-free space and supplying gas molecules to said beam during its passage therethrough, whereby positive ions are formed, are substantially confined within said field-free space, and act substantially to neutralize the space-charge effect tending to cause spread of said beam.

9. A system for projecting a low impedance electron beam including conductive means defining a field-free chamber through which said beam is projected, gaseous means within said chamber producing positive ions under the impactof said beam, and electrostatic means maintaining at one end of said chamber a direct current potential gradient along said stream in a reverse direction to the potential gradient used in establishing said stream, whereby said ions are restrained within said chamber and act to neutralize the spacecharge effect tending to spread said beam and increase the impedance thereof.

10. An electron beam tube for the production and projection of a beam of high density, including an electron source, an accelerating electrode establishing the beam, a drift tube through which said beam passes, means for keeping said drift tube at a slightly negative fixed direct current potential with respect to said accelerating electrode and means for collecting the electrons of said beam after passage thereof through said drift tube, whereby said negatively charged drift tube tends to retain therewithin positive ions formed from the residual gas within the beam tube.

11. A cathode ray tube system including means for generating an electron beam, means for accelerating said beam, means for projecting said beam through a tubular electrode and means for main taining said tubular electrode at a direct current potential negative with respect to the potential of said accelerating means adjacent said tubular electrode by an amount which is less than one 121 half. of one percent of thepotentialwof said lastmentioned: accelerating; -means- 'so. -as toprevent leakage ofv positive ions:- from :said 1 tubnlan electtrode.

12. A cathode ray tubessystem includingrmeans for generating, an. electron beam, .means==for accelerating said beam, means forzprojectingasaid beam through atubular electrode; means-for coilecting electrons from said :heam -after.its passa'ge through said tubular: electrode, and means for maintaining said tubularelectrodeiat-a'direct current potential negativeewithrespect to said col lecting means by an amountwhich islessthan one-half of one percentof the. :potential of said.

collecting means so as to prevent leakage of positive ions from said tubular electrode.

13. A cathode ray tubesystemdncludingmeans for generating an-electron-Joeam; means for acceltrating said beam, means for'projecting, said beam through a tubular electrodeand-means for maintaining said tubular electrodeaha-slightly nega tive direct current potentialwith respect to said acceleratingmeans, the potentialof said. tubular electrode being sufficiently'negative withwespect to the potential to said-accelerating means: to prevent leakage of ionsfrom-within said tubulartelec trode vbut not sufficiently-negative to substantially retard the speed-0f the electrons ,passingtthrough said tubular electrode 12 14. A mztlmde ray/tube system according to claim 13 wherein the voltage diflerence between the voltage of the said tubular electrode and said acceleratingjmeans is .p 6:94X10" where p is the number of unneutralized electrons per centimeter length in the beam at the section where itkpassesthrough ithe tubular electrode.

SPANGENBERG.

LESTER M; FIELD. ROBERT HELM.

REFERENCES 7 CITED Theiollowing references areof record in the flle o'f this patent:

I UNITED STATES PATENTS 7 Date varianjhhnncu..- June 17, 1941

Images