US2016760A

US2016760A - Electron discharge device

Info

Publication number: US2016760A
Application number: US662203A
Authority: US
Inventors: Harold A Wheeler
Original assignee: Hazeltine Corp
Current assignee: BAE Systems Aerospace Inc
Priority date: 1933-03-23
Filing date: 1933-03-23
Publication date: 1935-10-08
Anticipated expiration: 1952-10-08
Also published as: AT147305B; DE751217C; FR770602A

Description

.8, 1935. HA. WHEELER ELECTRON DISCHARGE DEVICE.

2 Sheets-Sheet 1 Original Filed March 23, 1933 HI I H IHHP INVENTOR J7 J5 HAROLD A. WHEELER ATTORNEYS Get. 8, 1935..

H. A. WHEELER ELECTRON DISCHARGE DEVICE Original Filed March 23, 1933 2 Sheets-Sheet 2 b 8 2 Qmfls INVENTOR HAROLD A. WHEELER ATTORNEYS Y Patented Oct. 8, 3 I

UNITED STATES ELECTRON DISCHARGE DEVICE Harold A. Wheeler, Great Neck, N. Y., alaignor to Haseltine' Corporation,

a corporation of Dela- Application March as, 1933, Serial No. maes- Renewed July 11, 1934 18 Claims.

'IiJsinve'ntion relates to electron discharge. devices, and more particularly to devices oi! this character having. a grid element with a non: uniform pitch, or grid mesh.

5 This is a continuation in part oi. my copending application Serial No. 654,327, filed January 30,

It is an object 0! this invention to provide an electron discharge device 01' which the gain may '10 be readily controlled by the application of a biasing voltage to a grid; and a particular objectis to provide a modulator tube of which the conversion gain can thus be readily controlled.

In a preferred embodiment for carrying out 15 the above objects, there is provided a vacuum tube having six electrodes, namely, an electronemitting cathode and an anode in the form of a.

plate, in the space path between which are cated four other electrodes. These four latter 20 electrodes are in the form of screens or grids and are located at successively increasing distances from the electron-emitting element, the first of these screens or grids being designated as that closest to the said emitting element.

25 A feature which makes a :device according to this invention readily subject to a control biasing voltage, and hence well-suited as a controllable modulator is the construction of the control grid.-

) This control grid is formed of a mesh having a 30 variable spacing or pitch, so that the resulting characteristic imparted to the device is in the form of a gradual curve having no sharp bend, or cutofl', in the operating range. It is well-known that a type of characteristic which is well-suited 3 to permit efiective gain control over a range of controlling biasing voltages is that in which'the curve of mutual conductance .versus biasing voltage is a straight line when plotted on log-linear coordinates. This is known as an exponential 40 curve. Control grids for providing'this relation are known as exponential grids.

Ihave round that the curve of mutual conduct- 1 ance versus grid bias can be predetermined by the manner in which the grid spacing, or pitch, is 45 made to vary throughout its length. I have ascertained by experiment that it the grid mesh is varied in such a way that the graph of serial numbers of successive grid meshes plotted against the pitch on log-log coordinates is a straight line, the

serial number 1 being the grid mesh of greatest spacing and the last serial number being the grid mesh orv finest spacing, the curve 01' mutual conductance versus grid bias will be ofv the desired exponential type. The absolute values oi mutual 55 conductance for a given grid bias and its variation with grid voltage, that is, the slope of the log-linear curve or line, may be adjusted as desired by suitably sloping and/or positioning the log-log curve of serial numbers" versus pitch with respect to its origin of coordinates, or by leaving (Cl. ISO-27.5)

the log-log curve in its initial position and suitably varying the log-log scales with respect thereto.

As a practical matter, the grid can be designed to meet the'above requirements by proportioning 5 the lateral pitch of the spacesor meshes between adjacent grid structural elements to be substantially in the inverse ratio of a predetermined iractional power of the number which is assigned to each space by serially numbering the spaces in lo the order oi decreasing pitch.

In the drawings:

Figure 1 illustrates an electron discharge device in accordance with this invention, a portion 01' the glass-enclosing vessel being cut away to show more clearly the positionot the electrodes within{ Fig. 2 is a vertical sectional view oi. the device showing in detail its construction and the positions of the elements; r

Fig. 3 is a broken diagram showing in perspective the construction ot the elements and also indicating the manner-"oi connecting operating voltages which place the device in an operative' condition;

Fig. 4 is a plan view, in section, of the electrode construction; a

Fig. 5 is an enlarged illustration, partially in section, of the cathode electrode and the means for heating the same;

Fig. 6 is a graph showing a relation between a mutual conductance and grid bias which provides a desirable type of characteristic;'

Fig. 7 is a graph illustrating a method of designing a grid to provide a desired relationship similar to that of Fig. 6, in accordance with this invention; and

Fig. 8 is .a graph similar to that of Fig. 7 illustrating the manner of .obtaining a closer approximation to the desired characteristic. 40

Figure 1 illustrates an electron'discharge device similar to that illustrated in my above-men tioned application Serial No. 654,327, but differing therefrom in having an'improved form of control grid. The device comprises an evacuated glass bulb II, the stem end of which is fitted" into a bakelite base Ii in which are embeddedseven projecting terminal lugs, numbered :ll to i II, respectively, adapted to flt into a suitable' socket. An eighth terminal is is situated at the top of the bulb in theiorm of a metallic cap cemented to the glass. The electrodes of, the device are supported by a number of verticalsupporting wires embedded in a glass stem 20 formed on a reentrant portion II at the stem end oi the bulb (see Fig. 2). For the purpose or maintaining the supporting wires rigidly in their upright position, there are provided two

fiat guides

22 and 23 of an insulating material, such as mica, through which the supporting wires project. The

ployed: A cathode I in the form of a sleeve of very small diameter is centrally located within the bulb. This sleeve is preferably constructed of nickel coated with strontium and barium oxides for the purpose of causing it to copiously emit electrons when heated. For the purpose of heating the cathode sleeve, there is located within the sleeve a heater filament 38 of fine wire, preferably of tungsten, coated with ceramic insulation and threaded back and forth withinthe sleeve in the formof a W as shown in Fig. 5. The ends of the heater filament are connected to the large-diameter terminal lugs I2 and I 3 at the base. g

Surrounding the cathode at a small distance therefrom isan inner grid electrode 2 in the shape of an elliptical cylinder formed by a helix of ilne wire and supported on two of the vertical supporting wires 2| and 25. This grid electrode inner screen, isconstructed in theform of an elliptical cylinder by winding a helix of wire which surrounds the first grid 2, the second grid-like electrode being supported on the two supporting

wires

26 and 21. This second helix is of fine mesh, that is, has a small pitch, which is preferably uniiorm throughout. As thus far described,- the device is the same as that of my application Serial No. 654,327; r

The improvement over the tube of my said original application, provided by this invention, resides in. the construction of the third grid electrode I. This third grid surrounds the electrode 3, and is likewise formed by a helix of wire' in 'the shape of an elliptical cylinder supported on

v supporting wiresv

28 and 29; but the pitch or mesh of this electrode is-not uniform throughout its extent.. .As this is the grid upon which it is intended .to'impress a biasing voltage for controlling the conversion gain of the tube, it is desired that this grid be proportioned to provide a gradual 'cutoil, of a type in general characterizing the so-'-called variable-mu tubes, so that wide ranges or control-biasj'will not deleteriously alter the tube performance. I v v It isthe: purpose of the grid construction about to be'described' to providethe most advantageous tube characteristic for this purpose. It is known that such" a highly satisfactory characteristic exists-when the curve of grid "bias versus mutual conductance is a sloping straight line when plotted on log-linear coordinates. desired type of curve is illustrated in Fig. 6 wherein the 'abscisas represent negative bias plotted on a in this case is located beyond the range of bias voltage.

Although it hasbeen known that the grid bias versus mutual conductance characteristic (and hence the cut-oil. characteristic) is dependent 5 more or less upon the form or pitch of the control grid, it has not been known how to accurately predeterminethis. characteristic. Consequently, heretofore the design of a grid was worked out by the cut-and-try method which even then did 19 not produce a very close approximation to the desired result. I have ascertained that the desired straight line log-linear relationship between grid bias and mutual conductance is obtained when the relationship between the serial number of thelfi grid turns and the pitch of the turns is a similar linear line when plotted on log-log coordinates. In arriving at this relationship, the mentioned serial number of the grid turns is expressed in the order of decreasing pitch, regardless of the 10- 20 cation of the turns in the grid. The construction is represented graphically in Figure '7 wherein the logarithmic coordinate 121. represents serial numbers of turns and the logarithmic coordinate p represents pitch per turn 25 in mils of each turn (numbered m) The dotted line 50 represents the desired log-log relationship (this being a typical straight line log-log characteristic curve). Inasmuch as no less than a single turn of grid winding is considered, the actual 30 graph is represented by the irregular line 5 I. The numbers It, 2t, etc., at the left of the graph indicate the number of turns in each group of turns having the same pitch; and the designations 69 mils, 54 mils, 41 mils, etc., represent the pitch of 35 each turn of the corresponding group. Thus the portion of the graph designated I t, 69 mils, indicates that one of the grid turns has a 69 mil pitch (the largest pitch of the grid) and the portion marked I612, 19 mils, indicates that sixteen of the grid turns have a 19 mil pitch. It follows, then, that the pitch used in this particular construction varies from 69 mils of which there is one turn down to 19 mils of which there are 16 turns. It is observed that each group includes one or more turns to which are assigned the serial numbers indicated by the ordinates of the graph. Although the line 5| is drawn as adiscontinuous curve, there are actually experienced no sharp bends in the tube characteristic; for,

as is well known, vacuum tubes are not characterized .by a pronounced bend for each grid turn or group of turns.

As stated above, it has been found that, if a grid is so constructed that a plot of the grid-tum serial numbers versus pitch on log-log scales is a straight line, the desired type of control grid characteristic may be obtained. That is to say, by proportioning the grid meshes in accordance with a straight line as plotted on log-log scales, the grid will provide a curve of mutual. conductance versus biasing voltage which is exponential. In other words, the latter curve will be a straight line when plotted on log-linear scales; the same type as the curve of Fig. 6. Having thus provided this type of characteristic, it is obvious that the-identical characteristic of Fig. 6 may be obtained by suitably adjusting the position and/ or slope of the line 50, Fig. 7, or by leaving the line 50 in its original position and'suitably varying the scales with respect thereto.

It is recommended, although it is not essential, that the lower serial numbered turns be located at the middle of the grid, and the higher serial numbered turns be arranged alternately in sec- 2,016,760 tions toward the two ends of the grid, so that the smallest pitch is at the ends. A preferred arrangementis indicated in Fig. 7 by the designashown by the following table, wherein the groups are'listed in the order of their position from the upper to the lower end of the grid, the terms "upper" and lower" corresponding to the terms above and below" used heretofore.

v serial Pitch of n'umbm No. turns in turns in Axial length of m gr p gr p p (m s) 17-32 16 10 304 -8 4 32 128 2 l 54 64 l l 69 60 3-4 2 41 82 oo s as 200 32- 837 (total turns) (length of grid) The grid 4 of Figs. 2 and 3'has been proportioned in accordance with the above table.

Figure 8 shows how a closer approximation to the ideal characteristic canv be obtained. This figure is a chart having the same coordinates as Figure 7, and similar" designations, the principal difference being that the dotted straight line 52,

taken as ideal, is moreclosely approximated by means of a finer grading of the pitch throughout the extent of the grid, as shown by the discontinuous line 53. The following table indicates a recommended arrangement of a grid constructed in accordance with Fig. 8. 1

S Pitch of n 9 No. turns in turns in Axial length of group a group group (mils) p (mi s) 32 v i 829 total turns) (length of grid) I If desired, the grid, instead of being formed in groups of turns of equal pitch as indicated in the tables associated with Figs. 7 and 8, may be formed by placing the serial number 1 turn at the middle of the, grid and proceeding alternately in serial numbers toward both ends. The location of the serial numbers from one end to'the other, in this arrangement, is indicated as follows: v

31, 29, 27 .5, 3, 1,2, 4, 6 28', 30, 32. Alternatively, the grid may be designedbyusing two setsof serial numbers, each set commencing atthe middle with serial number 1 and extending toward either end. The order of the turns will then be:

In the above-described type of grid construction the following relationships obtain:

mg: I 1

. lines in Figures 7 and 8.

The exponent c is the actual inverse slope of the dotted lines 50 or 52; and must be substantially less than unity, otherwise the above equations fail. This exponent should usually be held within the ranges of .3 to .4. The values 01' c, n and b are given in Figs. 7 and 8.

An inspection of Equation (1), above, shows that the grid is properly constructed when the pitch of the spaces between adjacent grid turns is proportioned substantially in the inverse ratio of the cth power of m, that is, in the inverse ratio of a predetermined fractional power of the number which is assigned to each grid mesh, or space, by serially numbering the spaces in the order of decreasing pitch.

In the case of the grid structures represented by Figs. '7 and 8, Equation (1) becomes The exponent applied to m is therefore substantially M which is the cube root.

Surrounding the' third grid, and supported on supporting wires 30 and ii, is a fourth grid-like electrode 5, herecalled an outer screen, also constructed of a helix in the form of an elliptical cyl- I inder. This fourth grid-like electrode preferably has a relatively fine mesh which is uniform throughout.

Finally, a plate 8, in the form of a cylinder, is provided which surrounds the entire gridelectrode structure, the plate being supported on supporting

wires

32 and 33. The plate is pref erably constructed of carbonized nickel and its axial length is not as great-as that of the other electrodes.

constructional data for the electrodes other than the third grid electrode, or outer grid 3, which have been found very satisfactory, are as follows:

' Inner grid 2-major axis, 0.121 inch; minor axis .096 inch; .wound with 25 turns of, .004 inch diameter wire with a .033 inch uniform pitch.

Inner screen 3-major axis, 0.236 inch; minor diameter wire, with a .025 inch uniform pitch.

It is usually desirable to carbonize the inner wall of the glass bulb or vessel, as indicated in Figure 1 by the dotted portion of the glass; this, however, is a refinement which may be omitted.

The connection of operating voltages which place the device in a very advantageous operative condition, particularly for use as a modulator, is illustrated in Figure 3. There is connected across terminal lugs l2 and It! a battery 34 (or other source of filament heating current) which serves to heat the W-type filament 38 within the cathode sleeve, the said cathode itself being nominally grounded by the ground connection at terminal l4.

Batteries

35, 36, and 31, or other sources of direct voltage are connected between ground and the terminals l6, l1, and I8, respectively, for applying positive voltages to the inner screen 3, the outer screen and the plate 6, respectively. A negative bias battery 39 is connected between ground and the cap I 9 for placing anegative bias upon the non-uniform mesh, outer grid 4.

With the above-described system of operating voltages, the action of the device is as follows:

the heated cathode I emits electrons which are attracted to the inner screen 3 by virtue of the positive voltage applied to the latter. The electrons upon reaching the screen 3 are traveling at a high speed so that most of them pass on through the screen and approach the outer grid 4, the negative potential of which retards the electrons and causes most of them to be attracted back to the positive screen. This retarding action causes a cloud of slowly-moving electrons to accumulate between the

electrodes

3 and 4. Hence, the position of this cloud may be termed a virtual cathode because of the fact that the electrons can be easily drawn away from the cloud in the same manner that they were originally drawn away from the vicinity of the actual cathode I. Since the screen 5 and plate 6 are at a positive voltage, this is exactly what happens, a stream of electrons being drawn from the virtual cathode, through the meshes of the grid-

like electrodes

4 and 5 to the plate. Electrodes'l, 5, and 6 function similarly to the respective control grid, screen and plate of screen-grid tetrode tubes. The outer screen 5 will usually have applied to it a source of positive voltage somewhat less than that of the plate in the preferred manner of screen-grid tubes.

The electron-discharge device in accordance with this invention has been found to be very useful where it is desired to exert a control upon I! will be repeated at the output between terminalsl8 and I4.

The desired modification of the input signals can then be effected by applying the modifying voltage to the inner grid 2, that is, between terminals l4 and II. This modifying voltage may often be a simple alternating voltage, which then causes the entire device to operate as a modulator. The eifect of the voltage on the grid 2 is to modify the electron stream which flows to the virtual cathode and hence indirectly to modify 5 complished without afiecting the action of anymodifying voltage applied to inner grid 2. This variation of the negative bias of the outer grid 4 is a very convenient expedient inasmuch as it permits a very simple control of the responsiveness of the device.

The non-uniform spacing of the meshes of control grid 4, in accordance with the abovedescribed design, permits a wide range of bias voltage to be applied with the least amount of signal distortion, and practically eliminates interference with the normal action of

electrodes

2 and 3.

The outer screen 5 may be omitted, but its presence greatly improves the results, particularly in obtaining the greatest responsiveness. 3 This screen, in general, performs the function of a similar screen in the well-known .screen grid tetrode, and is therefore usually given a positive voltage somewhat less than that of the plate.

Although the form of grid construction in accordance with thisinvention has been applied to the preferred embodiment, which is a hexode oscillator-modulator form tube, it should be understood that this form of grid construction is 40 not limited to hexodes or oscillator-modulator tubes, but has application in any tube to which it is desired to impart a gradual cutofi characteristic.

Where the control grid operates upon a virtual cathode, as in the preferred oscillator-modulator embodiment, the exponent 0 should be held fairly small, of the order of .3 to .4. Somewhat larger values of 0, however, may be employed when the control grid operates upon a real cathode, and when a very broad cutoff is desired.

The graphical method of predetermining the mechanical dimensions of a grid, as herein employed in the design of exponential grids, is also applicable to the design of grids which are intended to provide predetermined relations other than exponential.

I claim:

l. A vacuum tube grid structure composed of a helix of an electrical conducting material of which the adjacent grid turns are spaced in relation to each other so that the ratio of the space between any turn to the space between the most widely spaced turns is inversely proportional to a fractional power of the serial number which is assigned to the spaces by serially numbering these spaces in the order of decreasing space width.

2. The structure of a grid electrode for a vacu- 7 um tube, comprising a plurality of laterally spaced elements, the lateral pitch of the spaces between adjacent elements being proportioned substantially in the inverse ratio of a fractional power of the number which is assigned to each inter-element space by serially numbering the spaces in the order of decreasing pitch.

3. A non-uniform-pitch helical grid electrode for a vacuum tube, having the pitch of its turns proportioned substantially in the inverse ratio of a fractional power of the number which is assigned to each turn by serially numbering the turns in the order of decreasing pitch. I

4. A non-uniform-pitch helical grid electrode for a vacuum tube, having the pitch of its turns proportioned substantially in the inverse ratio of the cube root of the number which is assigned to each turn by serially numbering the turns in the order of decreasing pitch.

5. A-non-uniform-pitch helical grid electrode for a vacuum tube, having the pitch of its turns proportioned substantially in the inverse ratio of a predetermined fractional power of the number which is assigned to each-turn by serially numbering the turns, starting with unity in the middle of the grid and proceeding alternately toward both ends.

6. A non-uniform-pitch helical grid electrode I for a vacuum tube, having the pitch of its turns proportioned substantially in the inverse ratio of a predetermined fractional power of the number which is assigned to each turn by serially numbering the turns outward from the middle of the grid, using odd numbers from unity increasing toward one end, and using even numbers from two increasing; toward the other end.

7. A non-uniform-pitch helical grid electrode for a vacuum tube,- having the pitch of its turns proportioned substantially in the inverse ratio of a predetermined fractional power of the number which is assigned to each turn by serially numbering the turns from the middle toward each end of the grid.

8. A non-uniform-pitch helical grid electrode for a vacuum tube, having the pitch of its turns proportioned substantially in the inverse ratio of a predetermined fractional power of the number which is assigned to each turn by serially numbering the turns in the order of decreasing pitch, the turns of smallest pitch being located nearest the ends of said grid electrode.

9. A non-uniform-pitch helical grid electrode for a vacuum tube, having the pitch of its turns proportioned substantially in the inverse ratio of a predetermined fractional power of the number which is assigned to each turn by serially numbering the turns in the order of decreasing pitch,

the turns of greatest pitch being located near the middle of said grid electrode.

10. An electron discharge device comprising an evacuated bulb having a cathode, an anode, and at least three grid-like electrodes located in the space path between said cathode and said anode, each of said electrodes being located at a different distance from the cathode and the third of said electrodes in point of distance from said cathode having a non-uniform pitch of meshes, said pitch being proportioned substantially in the inverse ratio of a fractional power of the number which is assigned to each mesh by serially numbering the meshes in the order of decreasing pitch.

11. An electron discharge device comprising an evacuated bulb having located therewithln a centrally located cathode, at least three grid-like electrodes of a cylindrical form surrounding said cathode, said three electrodes being of successively increasing diameters, each successive electrode surrounding those of smaller diameter, and

a plate surrounding the said electrodes, the said electrode of intermediate diameter having meshes of non-uniform pitch, said pitch being proportioned substantially in the inverse ratio of a fractional power of the number which is assigned to each mesh by serially numbering the meshes in the order of decreasing pitch.

12. An electron discharge device comprising a closed bulb containing a centrally located cathode,a cylindrical plate laterally surrounding said cathode, and four cylindrical electrodes of a mesh construction also laterally surrounding said cath- Ode but laterally surrounded by said plate, each succeeding electrode surrounding the one preceding, and each of said meshformed electrodes being constructed in the form of a helix, the helix forming the third of said mesh-formed electrodes, in point of distance from said cathode,

having a pitch'of winding which is proportioned substantially in the inverse ratio of a fractional power of the number which is assigned to each mesh by serially numbering the meshes in the order of decreasing pitch.

13. An electron discharge device comprising a closed vessel containing a cathode, a cylindrical plate laterally surrounding said cathode and four cylindrical grid-like electrodes laterally surrounding said cathode, but laterally surrounded by said plate, each succeeding electrode surrounding the one precedfng, and each of said grid-like electrodes being constructed of a. helix of wire, the helix forming the first, second and fourth of said grid-like electrodes, in point of distance from said .cathode, having a, uniform spacing between adjacent turns of wire, and the spaces in the order of decreasing pitch.

14. An electron discharge device comprising a closed vessel having located therein a centrally located cathode, four cylindrical grid-constructed electrodes surrounding said cathode at successively increasing distances therefrom and an anode surrounding said electrodes, the third of said grid-constructed electrodes, in point of distance from said cathode, having spaces between successive grid elements which are proportioned substantially in the inverse ratio of a fractional power of the number which is assigned to each space by serially numbering the spaces in the order of decreasing pitch.

15. An electron discharge device comprising in a closed bulb a plurality of co-axial concentric cylindrical electrodes of graduated diameters, the fourth from the inside being a helical grid of non-uniform mesh along its axial dimension, the spaces of said mesh being proportioned substantially in the inverse ratio of a fractional power of the number which is assigned to each mesh by ser ally numbering the meshes in the order of decreasing pitch, the third and fifth electrodes from the inside being helical screens of uniform fine mesh, and the outermost of said electrodes being a cylindrical plate.

16. The method of constructing a vacuum tube grid which comprises winding a helix of wire and spacing the adjacent grid turns so that the spaces vary in the inverse ratio of a fractional power of the number which is assigned to each space by serially numbering the spaces in the order of decreasing pitch.

HAROLD A. WHEELER.

6. CERTIFICATE OF CORRECTION.

Patent No. 2,016,760. I I October 8, 1935.

HAROLD A. WHEELER.v

It is hereby certified that error appears in the printed specification of the above numbered patent requiring correction as follows: Page I, first column, line .48, strike out the word "the"; page 2, second column, lines 27-28, strike out the words "the desired log-log relationship this being"; and line 29., after "curved" strike out the parenthesis; and page 4, first column, line 37, after "screen" insert 3; and that the said Letters Patent should be read with these corrections therein that the same may conform to the record of the case in the Patent Office.

Signed and sealed this 19th day of November; A. D. 1935.

Leslie m (Seal) Acting Commissioner of Patents.