US2459283A - Positive grid oscillator - Google Patents

Description

Jan. 1s, 1949.

J. W. MCNALL POSITIVE GRID OSCILLATOR Filed April 13, 1944.

Orr/mas .16'

2 Sheets-Sheet l I l I I 'INVENTOR Ms w14 Mmm ATTORNEY Jan. 18, 1949. J. w. MCNALL POSITIVEl GRID OSCILLATOR 2 Sheets-Sheet 2 Filed April 15, 1944 INVENTOR I hf. MEA/4u. BY W MMM.

ATTORNEY Patented Jan. 18, 1949 POSITIVE GRID OSCILLATOR John W. McNall, East Orange, N. J., assigner to Westinghouse Electric Corporation, East Pittsburgh, Pa., a corporation of Pennsylvania application April 13, 1944seria1No. 530,889

, 11 Claims.

This invention relates to positive-grid oscilla- .A

that article explains for a simple case that by substantial equality of cathode to grid and grid to anode spacing, proper grid bias and cathode and anode voltages, the transit time of anelectron from the cathode to the gridv and from the grid to the anode can be made to correspond to a half cycle oi the generated frequency.

Explanation of the operation is given first by considering motion of an electron in phase with the oscillating field, that is, an electron which leaves the cathode at the instant that the oscillating potenial on the grid is Zero but increasing from zero to a positive value. The electron thus traverses a more than normal accelerating field as it proceeds toward the grid, arriving there with a higher velocity (more energy) vthan it would have in the absence of the oscillating field. As the electron passes through the grid, the cycle of oscillation of the field has progressed to a change of polarity, so that the electron passing through the grid now traverses a less-than-normal retarding eld, arriving at the plate with an energy including that which it has absorbed from the oscillating field during the whole cycle. This represents energy lost, as the electron merely gives up the energy to the anode in the form of heat, and is therefore referred to as an unfavorable electron with regardto the operation of the device. l

Useful energy in the operation of the prior art positive grid oscillator is obtained by electrons which leave the cathode 180 (a half cycle) later than the one referred to above, and are referred to as favorable electrons.- A favorable electron passes through an accelerating field which is less than normal due to the negative R. F. potential on the grid, and therefore gives up energy to the oscillating field and passes through the grid with less than normal energy. As before, by virtue of the oscillation cycle reversal, the electron enters between lgrid and anode as that field becomes progressively retarding, andthe electron continues to give up energy to the oscillating field. Since the electron has given up energy to the iield throughout its travel, it vhasless than normal energy as it approaches the anode and does not reach the anode, but stops land reverses its course, returning toward the positive grid. As the electron came to rest, however, and was reversing its course, the oscillating eld between plate and grid also changedpolarity so that the electron again encounters a less-thannormal accelerating eld. The electron accordingly continues to be a favorable one, giving up energy to the oscillating eld. It may again pass through the grid and continue l.contributing to the oscillating eld, again reversing and repeating the procedure until it nally strikes the grid and is lost for further useful purpose.

It is axiomatic that more energy must be imparted to the' oscillating field by the favorable electrons than is absorbed from` the field byfthe unfavorable ones for the oscillatingeld tobe self sustaining, so that eflciency -is increased by increasing the favorable electrons ,or by decreasing the unfavorable ones. In practice, a benencial proportion has been obtained by employing emission maintained temperature limited rather than `space charge-limited. This, however, prevents increasing the input power or current, and prior art retarding `grid oscillators have a low efciency and a low power operating range.`

vIn its general aspects, the present invention proposes improved efficiency. and higher input power for positive grid oscillators. y

Another object of the invention is to convert unfavorable electrons to be favorable.

Yet another object of the invention is to free secondary'electrons in phase with favorable priy mary' electrons.

Further object of the invention is to enable a. positive grid oscillator to handle much larger peak powers than heretofore and to yoperate progresses, both by direct recitation thereof and by implication from the context.

Referring to the accompanying drawings in which like numerals of reference indicate similar of several lvoscillator tubes whereinmy inven-V tion is employed. y i

An oscillator embodying the present invention may be designated by Weib-known symbols -of the art, as in Figures l to 4, as an evacuated en-y velope l containing asource of primaryeelec- 4 whereas the anode is at a potential suiciently negative with respect to the cathode or to the selector grid to repel electrons which enter the anode-selector grid region, D. C. potentialis apy plied from a source, such as battery 20, between cathode and control grid Il, and a likebut opposite y.potentialis, applied between kcontrol grid and .-"s'elfe"ctor ,grid'ffrom a source such as Ibattery 2l.

trons, indicated as a cathode lli therein, vspaced Y from which is a usual control ori-,positive .grid rIl and beyond which is an electrode .designatedanf ode or electron pulling electrode l 8. Oscillators of the prior art have containedfth'ese'*three'electrodes, namely, cathode, grid and anode, withfaV spacing proportioned to transit time of electrons for taking advantage of preponderance of energy .'o'f'favorable electrons overvunfavora'ble ones for `inairiiiain'ing oscillations. The 'elect' as explained more fully in. `the' above-identified article `in .mies 5a. .and 5b.

A somewhat more negative D. C. potential is applied to the anode from a suitable source, such as battery 22 if so desired.

` "Clp'eration of the oscillator shown in Figure 1* vis .indicated incormection with diagrams of Fig- Referring rst to Figure 5% an electron which leaves the cathode it at a favor-L able instant gives up energy to the oscillating eld' for thetwo half Vcycles while in transit from ",Elect'ronics,` dependsv uponthe electrode dimeni sions or spacing and voltages yat which they 'are operated, andfsuifers "from both structural. and operational limitationswhich results in low' eili- According tojthepresent invention, the anode `el'e'ctrcide of the priorA art'is replaced loyan' Lelectrode here ldesignated as a selectorgridjlaand an electrodejherein designated the 'anode or electron pulling electrode fl'8, isprovided beyond 'the said selector grid., Anode-rep'l'acing or selector g.'

grid 19,. maintained :at 'a .negative'potentia1 corresponding to use of' .negative potential on -the kas an 'anode 'for electrons which'leave the' cath- `ferred to as anode or electron Ipulling `electrode ,'I't'i therein; isisp'aced a predetermined distan-ce v'beyond the selector grid I9 'and un'ctions asan `'anode 'orele'ctron .pulling electrode -loi electrons 'ode in` favorable Vpli'ase relation. "Electrode rewhich Aemanate fromthe cathode in unfavorable 'phase relation 4and*converts them to JJI "a'vora'lrile I'electronsl By 'this construction, the initially ,favorable electrons vremain,favorable andjthe un- 'favorable welectrons bec'eine*favorable.` Efilciency isfthereby greatly increased; "not Vonly by-presence of the increased fr iuin,biery offavorable .electrons,jj but" 'also "by ll'hje l 'elimination of 'the A cancelling e'iect of nnfavorable electronsupon 'favorable ones.

llesc'ribed` more Vin'deta'il,Figure lLrepr-e'sents an*'oscillator'wherein-'cathode I0 and positive `or lcontrolgrid i1 Y"have thensual-spac'ed relationship and potentials of cathode and grid as generally employed 'in ipositive- .grid oscillatorsf Selector grid I9 hasusual spacing Aand'approxinria'te potential vger'ieral-lychara'cteristic 'or the usual 2anode of a ipositive grid oscillator. "Theseftwo spacings are commonly equal, and in relation "to :electron #transit time andv phase, laref'spacelsuch :ra 4distance that an? Ielectron :leaving ythe:icatli'c'id'ev 'at Iade'lnitev phase tini'eiareachesf'the upositive Ior control-gridr-whenthe phase is advanced. 180

- .and icon-iiifnui'rfig `tharetl'i-rough.arrives -atthe Fselecl-tor-grid whenY the Iphase has :advanced another y 11806.vr linelectron- ,passing through the` selector grid approaches .theelectron pulling ,electrode 18,

,and its .position is such thatthe electron .reaches thes-electronspulling .electrode when the phase has Vfurther 'advanced an s amount of 90 jor quarterrzy'cle transit time". `The' selector jgri'd is .preferably a't or approidmately'atcathode potential,

.thefcathode to the vicinity of the selector grid. Due to the fact that-:these favorable electrons have lost some of their energy in transit to the retardingfield,` they arerepelled bythe selector vgrid 19 which' acts Las a virtual anode for those electrons.' The action is repetitions, so the electrons scllate 'back and. forth 'through the posintive grid L1 giving up Aenergy to theretarding ffield. .'Thetfoscil'lat'ingjpath of the favorable elecf tron 'isindi'cat'edby lille 23.

Next; referring to Figure 5b, an `electron which leaves thecathode at an unfavorable instant absorb's .energy from the koscillating field for the :initial two. half cycles of transit, arrivingat the selector gridy i9 with a velocity determined by the amount lof .energy it hasA absorbed.k This added 'velocity results in theelectron continuing vthrough or past the 'selectorgri'd finto the `retard- `ing field 'between the selector Vgrid i 9 and electron pullingelectrode Iilfwlii'c'h `slows the electron xuntil it' comes 'to restlbefore arriving at'and in` the vicinity of the electron 4pulling electrode. lDue v 'to the D. `C. field, Vthe electron thereupon reverses lindirection and is vaccelerated baci: toward the J l Jselector grid, to again pass' through "that grid.. Y

C'f ccursezsome'electrons will strike the said sevhector gri'dj'and are lost `for further use, but conf sideraition is here ygiven 'to the majoritywhich l.pass through the selector grid on the return path,

and' which are byfar the larger `number'oi elecvtronsbecauise of4 the open nature-and the' focus-v "ingV action VJo'ffthe'electric'field existing in the vneigh'tborl'iood'of theselector grid. Since the elec- .tron""pullingv electrode is situated,` at electron ,travel distance from the selector gridv agreeing vwithsubstxetntially'il" phase cycle shift, the round trip transit o'ftheelectronin 'the'region between selector 'grid :and electronpulling electrode occupies "a `ilightft'inie of `substantially 180 phase s'hift. passesthrough the' Vselector grid '1809 later than Accordingly, the electron Areturns and on its forward flight, and due .to the' 180 delay, is then Iaiavorab'le electron 'in the region between Athe cathode vand selector 'grid and actsyasthe above-described 'favorable electrons, oscillating back -andi'orth through controlgrid l1 giving up energy to the retarding field. r Path of'this initially'unf'avorable "and ultimately favorable electron isin'dicated lby the linek 24 in Fig. 5b. It is nowfapparentthat I'have Vprovided means for `'increasing the ratio of the number of favorablel `Ithe-'i'nvention to `increase' the proportion of faha'l'icycle 'electrontransits to the number of unfavorable lones,"and accordingly obtain a corref "sponding increase inefficiency of oscillation.

Asiindicated heretofore, it is a desideratum oi? vura'ble electrons. In 'addition to Vaccomplishelectrons to favorable ones, a further multiplication may be effected by employmentl of secondary emission. Exemplary of this feature, electron pulling electrode I8 of Fig. 1 may be composed of material having a relatively high secondary electron emission ratio. In order that the primary electrons shall travel to the electron pulling electrode or anode for releasing the secondary electrons, the potential of the anode is preferably at or near the potential of the selector grid I9. Thus the extra potential applied by battery 22 in Fig. 1 will be reduced, reversed or eliminated Where secondary emission is desired, and exemplary of the total absence of potential difference, the showing of Figure 2 indicates the selector grid and anode as at the same potential, but spaced, as before, an electron transit distance apart affording approximately a quarter cycle phase shift, or more accurately, to provide ya round trip transit time from and back to the selector grid of substantially 180.

With regard to the assembly wherein the electron pulling electrode or anode is emissive of secondary electrons and considering electrons leaving the cathode at favorable moments, the operation is identical with that given above in connection with Figures 1 and 5a. For electrons leaving the cathode I9 at unfavorable moments, operation is analogous to that given above for unfavorable electrons, differing in that the primary electrons reach the electron pulling electrode or anode and in that secondary electrons make the return passage and in greater number than primary electrons. Operation of the oscillator of either Figure 1 or Figure 2 when utilizing secondary emission from the electron pulling electrode or anode is shown in Figure 5c.

Referring more especially to Figure 5C, an electron which leaves the cathode I6 at an unfavorable moment absorbs energy as describedabove from the oscillating field during the initial two half cycles in transit from cathode I6 to selector grid I9. Considering only electrons which pass through the grids, the primary electron from the cathode is accelerated by the absorbed enerby from the oscillating field and strikes the electron pulling electrode or anode I8. Upon impact with the electron pulling electrode or anode, the primary electron releases a plurality of secondary electrons, which, due to the D. C. field, travel to and through the selector grid, where they find themselves in a favorable phase because the oscillating field has advanced 180 since the time the unfavorable primary electron passed through the selector grid. The secondary electrons are then favorable electrons and act as above described, oscillating 4back and forth through control grid I'I giving up energy to the retarding eld. Path of the unfavorable primary electron is indicated in Figure 5c by solid line 25CL and path of one of the secondary electrons is indicated by dotted line 25h. As therear-e many secondary electrons to each primary electron, it. will now be apparent that a corresponding increase in efficiency of oscillation is obtained.

The foregoing examples of the invention in Figures 1 and 2 as applied respectively to derivation of increase energy from either primary electrons or from primary and secondary electrons have been made upon a consideration of oscillating potential on the control grid. 'I'he invention is equally applicable with the control grid constant or grounded and the cathode and anode, either or both, electrically oscillating. One specific example is indicated in Figure 3 wherein the secondary emission type of electron pulling electrode or anode of Fig. 2 has arbitrarily been selected, but it is to be understood electron pulling electrode structure and bias of Fig. 1 is equally applicable. In Fig. 3, by connecting the electron pulling electrode I8 and cathode I6 to opposite ends of the transformer secondary 26, the oscillating potential thereon rocks back and forth, the control grid I1 being neutral as to oscillating potential in View of center tap 2l' to the said secondary 26. Positive D. C. potential is supplied to said control grid I'l by suitable source, such as battery 28. Spacing from cathode I6 to control grid Il and from control grid Il' to selector grid IS is now mad-e transit time distance, whereas spacing from selector grid I9 to electron pulling electrode I8 remains at 90 transit time spacing as before. Thus, as before, unfavorable electrons are impelled with their absorbed energy through the selector grid substantially out of phase With the favorable ones, and those primary electrons in traversing the approximate 90 spacing to the electron pulling electrode there producing secondary electrons which retrace that distance, make a round trip distance of substantially y180 as before and thereby enter the field from selector grid to control grid in phase with the favorable electrons.

Tracing the unfavorable electron more minutely, it will be observed that it leaves the cathode at a moment when the oscillating potential is increasing from 0 at the cathode to 90 when the electron reaches control grid Il'. The electron is speeded by that repelling potential, absorbs energy to do so, and is therefore unfavorable. As the electron passes into the field between the grids, the oscillating potential on the selector grid is at its maximum positive potential which reduces the retarding field. In traversing this less-than-normal retarding field, the electron absorbs energy from the oscillations, arriving at the selector grid as the oscillating potential on the said selector grid decreases to Zero, but the electron having acquired energy, will continue to the electron pulling electrode or anode and there release secondary electrons. The secondary electrons are attracted by the control grid, the D. C. field of which sufliciently penetrates the selector grid and proceeds on what is here referred 'to as the return path. The forward and return path transit time of electron travel between selector grid and electron pulling electrode consumes substantially 180 phase shift of the oscillating period so that the entry of the secondary electrons into the selector grid-control grid region is at a time when the selector grid begins to be affected by an increase of the oscillating positive potential, which algebraicallydecreases effect of the positive D'. C. potential and creates thereby a lessthan-normal accelerating field. The electron loses speed, thereby giving part of its energy to the oscillating field, passes to the region between the control grid and cathode where it is again in a retarding field, is turned back and oscillates back and forth through the control grid as in the cases previously described. Conversion and use of originally unfavorable electrons is accordingly accomplished in the several embodiments illustrated and referred to or described above.

While emphasis of adaptation of the invention as an oscillator only may have been implied by the extended description above based on the simplicity of explanation achieved by consideration in that connection, it is to be understood the 'in- .above thesaid 4cap 35.

assunse 7 vention is'applicable in more generalusa'ge', for instance as an amplifier. Thus in Figure 4 isillustrated an input oscillating means or circuit 29 within the oscillating period of the-above-de- .scribed oscillator for supplying the oscillation potential of the oscillator. This is then amplified bythe oscillator and passes for useful purpose to an output circuit or means 3D. The particular oscillator shown in Fig. 4.- is one having kan velectron pulling electrode or anode f8 emissive of secondary electrons and spacing of the several electrodes therein is, as in Fig. 4, electron transit time distance of 96 from each to the next.

Observation is made that whereas Athe oscillator described in the Electronics article must of necessity operate with the cathode non-space lcharge limited, the present invention is not so restricted but may function either space charge or emission limited, because the unfavorable electrons as described herein ultimately vbecome favorable electrons.

In addition to increase of eliiciency .and eifectiveness of an oscillator obtained by the improvements above described, it is also a feature of the invention to enable use of higher currents and of higher voltages both for taking advantage of these higher currents and for increasing power. I propose .to pulse the grid voltage, thusY increasing the non-space charge limited currents, and

`to use a considerably higher potential than heretofore found possible in positive grid oscillators, thus allowing the oscillator to handle much larger peak powers for the same amount of electrode heating. Pulsing the grid voltage will not only allow much larger peak voltages to be used for the same average power but will result in a considerably, and desirably, shorter wave length of radiation'for a given tube structure.

Positive grid oscillators may be made in accordance with my invention with various physical arrangement of tube parts, and my conception kcontemplates use of coaxial electrodes as illustrated in Figure 6, or hollow body resonant structures of which Figures '7 and 8 are illustrative. In Figures 6 and 8 all electrodes are insulated from each other for applying any desired potential to any electrode, whereas in Figure '7 the cathode and selector grid are necessarily at a common potential. Figure 8 furthermore shows tun-ing adjustmcnt for cathode, electron pulling electrode and selector grid with respect to the control grid. In Figures 6 and '7 the electrodes are in predetermined fixed position so that tuning is by .adjustment of applied potentials. l

Oscillator tube of Figure 6 depicts coaxial electrodes comprising a rcylindrical cathode 16a containing a heater 3| therein consisting of a coil carried upon an insulating rod 32. A spacer 33, shown of plug type, is provided at and fits into the upper end of the cathode and receives an' end of said rod. Another spacer 34 is provided yat and iits into the lower end of the cathode and receives the other end of said rod. Both said spacers also make supporting Contact with ends of a cylindrical control grid lla, the upper end of which is shown having a, metallic cap .35 which overlies the end of the spacer as well as a part of the peripheral side of the same. The lower spacer protrudes from the end of the cathode and yshoulders outward below the grid cylinder and aords .thereat a supporting engagement for the lower end of a cylindrical selector grid 19a. The upper end of the selector grid is engaged at its inner peripheral margin by a third spacer 36 mounted A cylindrical electron coaxially varound wand of the cathode to positive grid, v.control Ygr-idto selector gridand selector vgrid to electron pulling electrode is made -inaccordance with description l given -Vabove in the discussion Vof Figuresl to 5.

Said electronpulling electrode may be reflectiveV or capable of secondary .emission likewise in accordance with .the foregoing dissertation. Y Y

Support of the electrodes .andsealing'as an evacuated envelopemay be accomplishedinapproved manner. As shown, the `anode or electron pulling electrode constitutes part vof .the envelope wall, and at its ends has'rings 31- solderedthere; around, said rings hav-ing'flanges of metallic' collars 38 welded or otherwise .secured-thereto,,the collars projecting as cylindrical continuationsof said electron pulling electrode.. Said collarsare of appropriate material, such as that manufactured in accordance with patent to Howard Scott, No. 2,062,335, and sold in the trade-as Kovanr The upper `one of the collars has a. glassrcap39 sealed' thereto, said cap also being ,seated at its outer end portion around a lead-in metallictrod support 4.0 coaxial within the collar and .supporting, at its lower end, the metallic cap ,35 heretofore described as part of thelcontrol grid.` Thelower collar 38 has a glasscap M sealedfthereto, this cap being shown with a re-entrantfsternjd therein which seals yaround lead-in wire. supports 413', 44 and 45, by which ,heater current is vconducted to the heater and connection afforded for the cathode and selector grid. Y

Adaptation of the invention to'hollow'body .resonant oscillatorsis indicated by inclusion herein of structures of that character inthe showings of Figs. 7 and 8. Separately considered, Fig. .ishows a hollow bodyV resonator, the hollowbody 50 whereof is all metal, provides a cylindrical wallr '5i and end walls 52, v53, which respectively provide re-entrant portions 54, `A55 coaxial with the cylindrical wall 5I and directed toward each other but with space between.l At the end of-and transverse to one re-entrant-portion 5411s a cathode lb which isv caused to emit from heatappliedby heater 32". Within the `other re-entrantpolttion v55 neXt the innermost end thereof is an electron pulling electrode or anode I Bb'in the form of a'flat disc transverse to the said re-entrant portion and parallel to thecathode. At the endv of thereentrant portion containing theelectrony pulling electrode Ais a .selector grid I9b parallel to .the anode, and between .said selectorgr-id andthe cathode is Va control grid Il,b also parallel to thev electron pulling electrode, cathode and selectorl als in Figure 3 are Aaccordingly .duplicatedV but with exponent C applied thereto for more readyreference. The showing is here included for indicating the adjustability of electrodes on Vopposite sides of the control grid .for tuning purposes, the structure being kmore fully described and claimed in Va ycopending application.

I claim:4 g 1. A positive grid oscillator 'having electrodes in parallelism to each other, said electrodescomprising a cathode, an electron pulling" electrode c,459,2ss

and two grids therebetween of which one is a control grid adapted to be positive with respect to the cathode and the other is a selector grid, the spacing of said electrodes being related to the electron transit time and oscillating cycle such that round-trip electron transit time from selector grid to electron pulling electrode and back to said selector grid is substantially the same time required for 180 phase shift of the said selector grid.

2. A positive grid oscillator having electrodes in parallelism to each other, said electrodes comprising a cathode, an anode-like electrode and two grids of which one is a control grid adapted to be positive with respect to the cathode and the other a selector grid, the spacing of the selector grid and anode-like electrode being substantially a quarter wave-length distance apart referring to the wave-length of the oscillating cycle of the oscillator.

3. A positive grid oscillator having electrodes in parallelism to each other, said electrodes comprising successively a cathode, a grid adapted to be positive with respect to the cathode, a selector grid and an electron pulling electrode, the spacing of the selector grid and electron pulling electrode being substantially a quarter wave length distance apart referring to the wave length of the oscillating cycle of the oscillator, and the other electrodes being spaced substantially equal distances apart from each other.

4. A positive grid oscillator having electrodes in parallelism to each other, said electrodes comprising successively a cathode, a grid adapted to be positive with respect to the cathode, a selector grid and an electron pulling electrode, al1 of said electrodes being spaced from each other a substantially equal distance comprising a predetermined fractional part of a wave length of the wave length of the oscillating cycle of operation of the oscillator.

5. A positive grid oscillator having electrodes in parallelism to each other, said electrodes comprising successively a cathode, a grid adapted to be positive with respect to the cathode, a selector grid and an electron pulling electrode, all of said electrodes being -spaced from each other substantially a quarter Wave length distance apart referring to the wave length of the oscillating cycle of the oscillator.

6. A positive grid oscillator having electrodes in parallelism to each other, said electrodes comprising successively a cathode, a grid adapted to be positive with respect to the cathode, a selector grid and an electron pulling electrode, the spacing of the selector grid and electron pulling electrode being substantially a quarter wave length distance apart referring to the wave length of the oscillating cycle of the oscillator, and the other electrodes being spaced substantially half wave length distances apart from each other.

7. A positive grid oscillator comprising a cathode, a, selector grid spaced in parallelism to said cathode, another grid between said cathode and selector grid constructed and arranged to establish an oscillating eld for electrons from said cathode and oscillating the same therein, and another electrode beyond said selector grid having spacing therefrom less than spacing of said cathode from the selector grid and adapted to return electrons to the field between said cathode and selector grid in phase with oscillating electrons in said eld.

8. A positive grid oscillator comprising a cathode, a selector grid spaced in parallelism to said cathode, another grid evenly spaced between said cathode and selector grid constructed and arranged to establish an oscillating eld for electrons from said cathode and oscillating the same therein, and another electrode beyond said selector grid having spacing therefrom less than spacing of said cathode from the selector grid and adapted to return electrons to the iield between said cathode and selector grid in phase with oscillating electrons in said eld.

9. A positive grid oscillator comprising a cathode, a control grid adapted to be positive with respect to the cathode, and spaced from the cathode in the direction of normal flow of electrons from the cathode, a selector grid beyond the control grid having substantially an equal spacing therefrom as said spacing of the control grid from the cathode, said cathode and grids being constructed and arranged to establish an oscillating field for the electrons with attendant cycle of phase reversal, said selector grid passing electrons therethrough out of phase with said cycle, and means in the vpath of travel of and for returning the outof-phase electrons back through the selector grid in phase with the oscillating cycle.

10. A positive grid oscillator comprising a cathode, a control grid adapted to be positive with respect to the cathode and spaced from the cathode in the direction ofvnormal llow of electrons from the cathode, a selector grid beyond the control grid having substantially an equal spacing therefrom as said spacing of the control grid from the cathode, said cathode and grids being constructed and arranged to establish an oscillating field for the electrons with attendant cycle of phase reversal, said selector grid passing electrons therethrough out of phase with said cycle, and a secondary emission electrode beyond the selector grid for receiving the out-of-phase electrons passing through the selector grid, said secondary emission electrode being productive of secondary electrons and adapted to promulgate said secondary electrons through the selector grid in phase with the said oscillating cycle.

11. A positive grid oscillator having an electron source for promulgation of electrons and having grids spaced therefrom and from each other successively with equal spacing, said grids adapted to have operating voltages and currents applied thereto, a, source of pulsed voltage connected with one of said grids for increasing the non-space charge limited currents, and means in the path of and for returning electrons passing through the outer spaced grid into the field of the pulsed voltage.

JOHN W. McNALL.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 1,844,319 Hatt Feb. 9, 1932 2,115,866 Lux May 3, 1938 2,128,232 Dallenbach Aug. 30, 1938 2,170,219 Seiler Aug. 22, 1939 2,287,845 Varian et al June 30, 1942 OTHER REFERENCES Electrical Engineers Handbook by Pender McIlwain, 3rd ed., published 1936, by John Wiley and Sons, Inc., New York city. (Copy in Div. 54, U. S. Patent Omce.)

Images