US2426681A - High gain amplifier - Google Patents

Description

Sept ..2, 1947. R. ADLER ET'AL 2,426,681

' HIGH GAIN AMPLIFIER Filed May 27, 1944 6 Sheets-Sheet 1 FIG.. 1

INVENTORS 1' ROBERT ADL'E-R- JOHNG. PRENTISS A QAM nSMM' TH IR ATTORNEY P 1947- V R. IADLER ETAL 2,426,681.

HIGH GAIN AMPLIFIER Filed May 27, 1944 6 Sheets-Sheet 2 FIG. 4

INVENTOR SZ RCBERT ADLER JOHNGPRENTISS THEIR ATTORNEY Sept. 2, 1947. I R. ADLER ET AL 2,426,581

HIGH GAIN AMPLIFIER R Filed May 27,1944 6 Sheets-Sheet :5

F l G. 9

80 BI SECOND ANODE CURRENT FIRST ANODE CURRENT CURRENT SECOND'ANODE VOLTAGE PREFERRED OPERATING RANGE INVENTORS 2 R0 8 ERT ADLER J o HN G- PRE NT! 55 B AN THEIR A T TOR NEY Sept. 2, 1947. RQApLER Er AL 2,426,681

HIGH GAIN AMPLIFIER Filed May 27, 1944' s Sheets-Sheet 4 SECOND ANODE VOLTAGE FIG. II

OPERATING RANGE LNEIHHOO EIGONV GNOOEIS INVENTORSI ROBERT ADLER JOHNGPRENTIS s THEIR ATTORNEY Sept. 2, 1947.

R. ADLER El AL 2,426,681

HIGH GAIN AMPLIFIER Filed May 27, 1944 6 Sheets-Sheet 5 ROBERT A DLER JOHNGPRENTISS THEIR A TTORNEY R. ADLER ET AL Sept. 2,1947.

' 1min GAIN AMPLIFIER 6 Shee'ts-Sheet 6 Filed May 27, 1944 FIG. 8

INVENTOR 82 ROBERT ADLER JQHNG.PRENTISS Patented Sept. 2, 1947 UNITED STATES PATENT orric HIGH GAIN AMPLIFIER .Robert Adler, Chicago, and John Prentiss,

Berwyn, 111., assignors to Zenith Radio Corporation, a corporation of Illinois Application May 27, 1944, Serial No. 537,704

27 Claims. 1

This invention relates to amplifiers, and more particularly to electron discharge amplifier devices and circuits associated therewith.

-It is a fundamental object of this invention to provide an improved form of electron discharge amplifier device and circuits especially arranged to produce large amplification.

It has often beenconsidered necessary to provide, in an electron discharge amplifier device arranged to produce high gain, a cathode capable of emitting large-electron currents. It is, accordingly, a further objectof this invention to provide such-an improved form of electron discharge amplifier device and circuit in which a cathode capable of emitting-only a small space current is provided. When such an-electron discharge amplifier device is arranged in accordance with this invention to produce highgain with 'the cathode emittingonly asmall space current, simultaneous savings are effected both in the power required to heat the cathode and in the powerdissipated in the other elements of the discharge device.

It is a corollary object of this invention to provide such an improved form of electron discharge device and circuit in which only a small number of electrodes is utilized'and in which the device and circuit associated therewith is stable in operation and rugged and capable of easy manufacture and assembly.

It is also an object of this invention to,provi'de an improved form of electron discharge device and'circuit in which, in'a single discharge device, gains in the order of at least 100 are readily obtainable. The features ofthisinvention which we believe to be novel are set forth with particularity in the appended claims. Theinvention itself, ,both as to its organization and manner of operation, together with further objects and advantages thereof may best be understood by reference to the following description taken in connection with the accompanying drawings in which:

Figure 1 is a sectional view in elevation of a preferred form of electron discharge device'constructed in accordance with the'invention;

Figure 2 isasectional view-of thedevice illustrated in Figure l, taken-along the line =2-' 2 in Figure '1 Figure '3 is a sectional view of an alternative form :of :the invention, taken :as along the line 2-2 of 'Figure 1;

Figure 4 .is a sectional 'vie'w 'of .a somewhat different form of electron discharge device iconstructed according -to the invention;

Figure 5 is-a sectional-view-of thedevice-il'lus- '2 trated in Figure '4, taken along the line 155 of Figure 4; Figure -6 is a sectional View vof a modified form of :electron discharge device similar to that-illustrated in sectional elevation in Figure 4;

Figure '7 is still another alternative form of an electron discharge device constructed in accordance with the cinventionespecially for operation at low voltages;

Figure i8 is a sectional view .of the device illustrated in Figure v7, taken along the line -8:8 of Figure F7;

Figure '9 is a graph illustrating certain characteristics which may appear in apparatus incorporating .the invention;

Figure 1 0 :is a sectional view of still another embodiment of the invention;

Figure 11 is a sectional view of the device shown in Figure '10, taken .along :the line H.:l:l of Figure '1 0;

Figure 1 2 is a:graph illustrating certain'other characteristics which may appear in apparatus incorporating 'the invention; and

Figures 13 through 18 illustrate circuits especially useful with the various =formsof the electron discharge device of the invention.

in Figure 1 an evacuated glass Vessel lllcontains a filamentary cathode H, a first control electrode 1 2, a first anode 16, a second control electrode l'tanola second-anode +5.

The filamentary cathode H is fastened at its lower-end to asupport lt mounted on a lead-in wire H which passes through the wall of the glass vessel I0,-and is drawn tense through accurately dimensioned holes in the :mica spacers I 8 and I9 by a spring 2-0, to one end of which is fastened'thesupportingtab 2-I on whichthe-o'ther end of the filamentary cathode H is fastened. The other end of the spring 20 is fastened to the support post 2 2 =which'-passes downwardly through the 'mica spacers EI9 and. l8 respectively, and. to the lower-end of which is joined the leadin '-conductor 23 which "passes through the-glass walls of the vessel Ill.

A circuit is formed for h'eati'ng the filamentary cathode H from the lead-in conductor i-l through "the-support Hi, the cathode H, the supporting tab 2 the spring 20, the supporting post 22, and the 1ead-in wire 12-3.

The lfllsl, :control electrode structure is formed by Iwincling :a cylindrical conductor of suitable diameterlaround two supporting posts rz land 225 helically withthe-turns of the conductorrspace'd apart a. suitable distance. The supportingposts Mand -'25 are fastened to each turnrofthe cylindrical conductor, as by welding or by providing slots in the posts 24 and 25 within which the cylindrical conductor is wound, the material at each side of the slot being swaged over the cylindrical conductor. The assembled control electrode I2 is placed between the mica separators I8 and I9 with the posts 24 and 25 extending through suitable holes in those separators and the filament II is thereafter passed through the holes in the separators I8 and I9 and tensed therein, the holes for the supporting posts 24 and 25 and the holes for the filament I I being uitably spaced so that accurately determined distances are maintained between the filament II and the control electrode I2.

The first anode I3 andthe second control electrode I4 are similarly constructed by the winding of a cylindrical conductor around two supporting posts, the diameter of the conductor and the spacing between adjacent turns on the supporting posts being in each case suitable to provide a desired result, as will be explained more fully hereinafter. In each case the supporting posts of the first anode I3 and control electrode I2 and electrode I4 are received within suitably positioned holes in the mica separators I8 and I9 so that accurate spacing is maintained between the control electrode I2, the first anode I3 and the second control electrode I4. It is notable here that, with this construction and arrangement of a discharge device and with circuits to be described hereinafter, it is not necessary to maintain any particular alignment between the turns of the cylindrical conductor of the first control electrode I2, the first anode I3, and the second control electrode I4; In other words, the vertical alignment of the meshes of those three electrodes may be left entirely to chance,

Outside of the second control electrode I4 the anode I5 is arranged cylindrically between the mica spacers I8 and I9 with the axis of the cylinder coincident with the filament I I The anode I5 is supported by and electricall connected to, its supporting post 30, which is placed symmetrically with the supporting post 22, and the upper end of which supports a getter assembly 3I. The lower end of the supporting post 30 is connected with a lead-in conductor 32 which extends through 'the glass wall of the envelope I and provides for external electrical connection to the trical' connection respectively with those elec-' trodes.

A fixed lead-in conductor 35 passes through the glass wall of the envelope I0 between the lead-in conductor I1 and the lead-in conductor 23, and, is connected with the supporting post 25 to provide an external electrical connection for the first control electrode I2.

In Fig, 2, the same numerals are applied to like parts of the discharge device. The central hole through the mica spacerIB can be seen as a triangle with the filament I I at one apex. The fiat sided first control electrode I2 at its intermediate portions lies relatively close to the filament I I and the fiat sides of the first anode I3 lie a substantially larger distance from the filament II, the

structure of those two electrodes being such that the first control electrode I 2 has a relatively large static amplification factor which may be 10 or more.

The second control electrode 14 also has fiat sides substantially parallel with the fiat sides of the first control electrode I2 and the first'anode I3, and the fiat sides of the second control electrode I4 lie substantially closer to the first anode I3 than to the parts of the second anode I which receive the electron stream from the cathode II.

That is, the shadowing effect of the supporting pinge upon only a portion of the anode I5 faced by trodes, and the close mesh of the second control the fiat sides of the inner electrodes.

Purely by way of example, a typical set of dimensions for such an electron discharge device is given to illustrate a specific manner in which the invention may be utilized. The filament I I is made of tungsten wire a little more than inch in length and of such diameter that 50 milliamperes of current flowing through it produce a voltage drop across it of about of a volt. .The helical wire with which the electrodes I2, I3 and I4 are formed is .002 of an inch in diameter. The first control electrode I2 is wound on the supporting posts 24 and 25 with '72 turns per inch, and is so wound that its fiat sides are spaced apart by .018 of an inch. That is, the smallest distance between the filament I I and the control electrode I2 is a little less than .009 of an inch.

The first anode I3 is wound with 50 turns per inch and its flat sides are spaced apart .112 of an inch. The second control electrode I4 is wound with '72 turns per inch and its fiat sides are'spaced r apart .200 of an inch. The diameter of the cylindrical anode I5 is .375 of an inch, and the overall lengthof all of the elements between the mica spacers I8 and I9 is about 1% of an inch. 7

Such dimensions provide that the distance between the second control electrode 1 4 and the first anode I3 is much less than the distance between the second control electrode I4 and effective parts of th anode I5. This spacing between the elecelectrode It, which is Wound with '72 turns per inch, provide that the secondcontrol electrode I4 has a relatively high static amplification factor.

with respect to the anodeI 5, and thatlfactor may be made in'the order of 10 or more.

Such an electron discharge devicemade in accordance with the invention, the usual precaui several hundred times, while remaining stable in 7 operation. At the same time, the discharge device is simple in construction and rugged and is easy g to manufacture.

In general, the construction of the second con-r trol electrode. I4 is critical, the'construction of the remainder of the discharge device being quite similar to that of a normal 'pentode. The second control electrode I 4 may in a preferred form of the invention be placed about'twice as'far fromthe anode I5 as from the first anode I3 and may have openings therethrough ofQsuch area that,

when the second control electrode is substantially at the potential'of the cathode II; space current passing through that second control electrode Id is affected by the electrode 14 in such a way that,

when the potentials on the anodes I5 and Hare substantially equal, 'the 'ratio between current flowing in the first anode I3and current flowin in the second anode I5 is substantially larger than the minimum ratio between those currents which is attained when the potential of the second anode i is increased indefinitely in a positive direction. This relation may be explained somewhat more simply if a discharge device as illustrated in Figs. 1 and 2 be constructed and sub stantially equal positive potentials be applied to the anodes i3 and is, the first control electrode l2 being made slightly negative with respect to the cathode ii to prevent the flow or" unduly large space current from the cathode. With the second control electrode M connected to the oathode H, the current of the first anode it is observed and the current in the second anode i5 is observed and the ratio between those two currents is calculated. Then the potential of the second anode i5 is increased positively and consequently the current flowing in the second anode I5 increases asymptotically toward a maximum. At the same time current flowing in the anode I3 decreases asymptotically toward a minimum, and the ratio between the current in the anode l3 and the current in the anode i5 accordingly decreases toward a minimum. The fact that the ratio decreases substantially upon increase in the positive potential of anode i5 is evidence that the discharge device is properly constructed.

This condition does not occur in a normal pentode in which a third electrode is of relatively open mesh, the mesh being so wide that, with the third electrode connected to the cathode, the third electrode produces substantially no effect upon discharge current flowing from the second electrode to the anode. In other words, in a normal pentode, increase of anode potential, as is very well known, does not substantially increase anode current. Expressed in another way, it is frequently said that the internal anode resistance of a normal pentode is extremely high, approaching infinity.

The construction of the second control electrode according to the invention may be described in still another way. The spacing between the first and second anodes and the fineness of mesh are such that, when the second control electrode is operated at cathode potential, or more accurately at a potential near that potential at which electron current begins to flow into the second control electrode, and when there are substantially equal positive potentials on the anodes, current flowing in the second anode i5 is substantially less than the current which would flow in the second anode 5 with substantiall equal positive anode potentials but with a more positive potential on the second control electrode M. This may be clearer if the steps in testing the construction are explained in detail. Normal positive potentials are applied to the anodes i3 and I5, and a slight negative potential is applied to the first control electrode l2, just sufficient to cause normal space current to flow from the cathode l l. The potential of the control electrode id is adjusted in the region of cathode potential so that electron current just begins to flow in the second control electrode 14. The current flowing into the second anode i5 is then observed and the potential of the second control electrode M is changed in a positive direction, and the current flowing in the anode !5 increases, which indicates that the control electrode is is correctly constructed. The current in the second anode should increase at least and it is actually preferred that the second control electrode be so constructed that the current in the second anode substantially doubles when the second control electrode pot-ential is increased under the described conditions. That is, the current in the second anode to begin with should be at least one-fifth less than the current which flows in the second anode after the second control electrode potential is made substantially more positive, and it is preferred that the second anode current be in the order of half the current which flows in the second anode after the second control electrode potential is made more positive. It is another aspect of this peculiar construction of the second control electrode that it is capable of effecting substantial changes in electron current distribution between the two anodes when its potential is changed in small increments, its average potential being near cathode potential. That is, the static amplification factor of the second control electrode is sufficiently high that, when it is near cathode potential, it has a substantial influence on the distribution of space current between the first and second anodes.

There is still another test which maybe used for certain types of discharge device constructed according to this invention. With the preferred form of second control electrode, when the potentials on the first control electrode [2 and first anode it are adjusted so that substantial space current flows to the first anode l3, and with the second control electrode I 1 near cathode potential, or more accurately near that potential at which current begins to flow in the second control electrode i l, space current does not flow to the second anode as its potential rises from zero voltage with respect to the cathode to a substantially positive voltage when the second control electrode M is correctly constructed. The second anode i 5 may even rise to a positive voltage with respect to the cathode which is ten or twenty per cent. or even more of the positive voltage on the first anode i3 without any measurable current flowing in the second anode l5.

As a general rule, the structures at the first and second control electrodes will be somewhat similar insofar as their mesh and the size of the conductor from which they are wound is concerned.

It should be noted at this point that, while it is preferred to have the second control electrode i i closer to the first anode l3 than to the second anode l5, the second control electrode I 1 can be constructed to produce the requisite results without being sospaced. In this connection, it should be pointed out that the term closer must be differently interpreted when applied to different forms of control electrode. Where the control electrode I4 is fiat sided as illustrated in Fig. 2, the term closer should be given the natural meaning that the actual distance between the first anode. I3 and the control electrode I4 is less than the actual distance from the control electrode M to the anode l5, always remembering, of course, that these distances should be measured along a line perpendicular to the filamentary cathode l l and perpendicular to the flat sides of the electrodes l3 and I 4.

Where the electrode I 4 is made generally cylindrical, the term closer must be given the mean ing that the ratio of the diameter of the first anode l3 to the diameter of the second control electrode I4 is greater than the ratio of the diameter of the second control electrode M to the anode l5,

In Fig. 3, there is illustrated a cross section of a discharge device which appears in sectional elevation like the device illustrated in Fig.1, thesec tion shown in Fig. 3 being taken as along the line '7 240i Fig. 1. In the device or Fig. 3, parts similar to those illustrated in Figs. 1 and 2 are given like reference numerals. A somewhat different 7 second control electrode M3 is arranged between the first anode l3 and the second anode l5. This second control electrode '40 is roughly cylindrical. The second control electrode 40 is made of such diameter that it is closer to the second anode i5 than to the first anode l3. While this is not generally a preferred construction of the invention, it ispossible to construct a device in this manner and achieve results falling within the scope of the invention. Purely by way of example, certain dimensions of the device illustrated in Fig, 3 are given to illustrate how the second control electrode' 40 can be made so that it has a sufficiently large static amplification factor to achieve the desired result. The filament H, the over-all length between the mica spacers l8 and E9 of .the elements, and the size and spacing of the supporting posts may all be like those shown and describedin connection with Fig. '1. Also, the spacing between the fiat sides of the first control electrode l2 may be .018 of an inch and the spacing between the flat sides of the first anode i3 may be .112 of an inch. However, the spacing between the sides of thesecond control electrode 40 along a, line perpendicular to the filament l l and perpendicular to the flat sides of the first anode I3 is made .250 of an inch, and along that same line the diameter of the anode i5 is made .320 of an'inch.-

Such an arrangement as illustrated in Fig.. 3 has in the second control electrode 40 a, smaller static'amplification factor than if the electrode 40 were closer to the first anode I 3 than to the second anode I5, but the second control electrode 40, having '72 turns per inch of cylindrical contials than'the device illustratedin Fig. 1. Many elements are similar to the elements illustrated in the device of Fig. 1 and are given like reference characters. The first control electrode 50, the first anode 5i, second control electrode 52, and

I the second anode 53 are all constructed with different dimensions from the corresponding electrodes of Fig. 1 forthe purpose of providing lower potential operation. The first control electrode is made of somewhat more open meshgso that a relatively lowpositive potential on the first anode 5! is effective to draw a substantial electron current fiow from thecathode l I through the first control electrode'50. Similarly, the second control electrode 52 is made with somewhat more open mesh so that the anode :53 can draw a sub-. stantial electron current flow through the control electrode 52 from aspace charge formed between the control electrode 52 and the first anode but with 50' turns perinchpand the second control electrode '52 may be woundwith the same size conductorbut with 64 turns per inch.

g 7 Byxway of example, toillustratebne manner In Fig. 5. like pat. are'giveni'the sainfe refer- I ence numerals and the shapes of the"el'ectr'odes 50, 5!, 52 and-53may be seen somewhatmore clearly; The control electrodes 50 and 52 and V the first anode 5| are wound with fiat sides, like the corresponding electrodes shown in Fig. 2, and the anode 53 is cylindrical. The spacing between the opposite fiat sides of the first control electrode 50 is .018 of an inch, and the spacing between the opposite flat sides of the first anode 5! and the second control electrode 52 are respectively .086 of an inch and .136 of an inch. The diameter of the anode 53 is .250 of an inch.

With the second control electrode 52 substan-, tially closer to the first anode 5! than to the second anode 53, and with the first anode'5l relatively closer to the first control electrode 50, the two control electrodes being of'more open mesh, the anodes 5i and 53 may be operated at lower potential than is the case with the device illustrated in Fig. 1. For example, anode supply potential for the device of Figs.4 and 5 may be in the order of 20 to 30 volts. open mesh in the control electrodes, it is desirable that the static amplification factor of each control electrode with respect to its associated anode be at least in the order of 10 in order to achieve thenew results of the invention.

It is important to notethat anode potentials" cannot be made indefinitely low. Contact potentials, which are in the order of one volt, and which are due to surface conditions of'the various electrodes,the'surface conditions being. frequently altered by contamination with cathode coating material, may affect very undesirably the anode currents. Such contact potentials vary markedly from one discharge device to another and furthermore the vary during use ofthe discharge device. These contact potentials affect the actual potential of every'electrode in the discharge device, and particularly afiect the effective bias potential of the second control electrode; If the static amplification factor of the second control electrode is high, the supply potential for both anodes must be high compared to the contact" potential of the, second control electrode multiplied by its static amplification factor.

that the contact potential of the second control electrodes affect the current in either anode to any large extent. 7

Bearing in mind, the effect of contact poten-- tial of the second control electrode, where those.

contact potentials are in the order of; one Volt; (as they usually are), it is evident that, if it be desired to operate the anodes from supply potentials in the order of twenty volts, the staticam plification factor of the second controlelectrode should not exceed afactor in the order of 10.

Anode supply potentials lower than about 15 T '7 volts cannot safely be used, unless contact potential is reduced-correspondingly lower than 1 volt.

Even with such' If the supply potential for the anodes is not high compared to the contact "potential of the second control electrode times its static amplification factor, the contact potential may affect the anode current adversely, and might even cut off completely the current flowing" in the second anode; In any case, in the particular I circuits with which this discharge device is par ticularly adapted to be used, it is undesirable Like reference numerals are applied to similar parts. Different mica spacers are provided, of which the lower one 60' is illustrated, in order to support the filamentary cathode I I and the first control electrode i2, first anode i3, and second anode l i'off center with respect to the glass envelope l0. An anode (ii, of relatively small area compared to one of the fiat sides of the second control electrode :4, is supported on asupporting post 62 between the mica separators and is placed substantially symmetrically with respect to a plane passing through the filament H perpendicularly to the fiat side of the second control electrode M.

By way of illustration, the dimensions for a device made in accordance with Fig. 6 may be as follows. The spacings between the fiat sides of the control electrode l2 are the same, being .018 of an inch, and similarly the spacings between the fiat ides of the first anode l3 and of the second control electrode M are the same, being respectively .112 of an inch and .200 of an inch. The distance from the filamentary cathode H to the small anode 6! is .187 of an inch. As is the case with the device illustrated in Fig. 1, the mesh of the electrodes l2, l3 and M is respectively '72 turnsper inch, 50 turns per. inch, and 72 turns per inch.

By making the area of the anode Gl' small, and in fact substantially smaller than the effective area of the second control electrode l t, the internal anode resistance of the anode 6! is made high and the effect of potential changes on the second control electrode M upon the flow of current to the first anode I3 is substantially reduced. This structure of an electron discharge device is especially useful in a special circuit employing regeneration. That is, in a circuit in which potential changes appearing on the first anode [3 (when signal potentials are impressed on the first control electrode l2) are impressed upon the second control electrode M, the discharge device acts not only like a cascaded pair of triode amplifiers, but also acts regeneratively. The regenerative effect, of potentials impressed on second control electrode l t by the first anode I 3, upon the space current flowing to the first anode I3 may be so great as to cause self-induced oscillation. By making the area of the anode 6| small, the effect of the second control electrode M upon current flowing to the first anode i3 is reduced, so that the regenerative effect can be made high enough to give useful results but not so high that self-induced oscillations are produced.

Expressed in other words, the control electrode I 4 is normally used, as explained previously, to cause a change in the current distribution between the first anode i3 and the second anode 5!. When potential changes appearing on the first anode l3 are impressed on the second control electrode M, the change in current distribution between the anodes l3 and 6! caused by theresulting effect of the second control electrode M is in aiding relation on the first anode l3 and is therefore regenerative. This is true because an increase in the total cathode current from the filament. H, caused by a change in potential in the first control electrode 52, causes a reduction in the positive potential of the first anode l3 and, when such reduction in potential is impressed on the second control electrode 14, causes a still further reduction in the potential of. the first anode [3- by reason of the still further increased current flowing into the anode l3 because. of

10:- the change in current distribution between the anodes I.3-. and. 6] induced by the change in potential, on control electrode I 4. This change in current. distribution between the anodes l3. and BI issuiiicient to cause the potential of anode 6| to rise instead. of. fall-as it would if the potential ofsecond, control electrode 14 had not. changed. Because the anode 6| is small, the first anode I3- takes substantially all of the electron current. from the cathode ll over most of its supply potential, andwithout; any sacrifice ingain by increase in mesh size of control electrode asin the. case of the device illustrated. in Figures 4 and:5. In-thisfigure many elements are similar to those illustrated. in Figure 1 and are given like reference character. A space charge electrode 1-0 is provided adjacent to the filamentary cathode II, and isfollowed consecutively by a. first control electrode H, a first anode 12, asecond control electrode-13, and a second anode M. These electrodes are all. held inv proper spaced relation by suitable mica spacers l5 and 76. The space charge electrode 10. is provided with an external connection through. a. lead-in conductor 79 extending through the glass wall of the envelope Ill.

The space charge electrode 70 is constructed so that it may be operated at a small positive potential, with the result that a substantial cathode current is caused toflow through the first control electrode 'H- to the first anode 12, even though the first anode 12 is operated with a very small anode supply potential (for example, as low as fifteen volts) and even though the static amplification factor of the first control electrode H- with respect to the first anode 12 issubstantial. In such a device, as in the ones previously described, the static amplification factor of the second'control electrode 13 may be made high with respect to the anode 14, even though the anode supply potential for theanode 14 is very low, for example, as low as fifteen volts.

By. space charge electrode in the present specification and claims it is understood that reference is made to an electrode of potential positive with respect to the cathode and immediately adjacent thereto without the presence of other electrodes therebetween.

As a specific example, to illustrate exactly how one device of' the type illustrated in Fig. 7 may be constructed, the space charge electrode'lllmay be Wound with 64 turns per inch of cylindrical wire of two thousandths inch diameter, the first control electrode H with 80- turns per inch, the first anode 12" with 50 turns per inch, and the second control electrodeldwith 72 turns per inch,

all being wound with cylindrical wire of two thousandths of an inch diameter.

In Fig. 8, being a sectional view along the line 8-8. of Fig, 7,. his evident that the first control electrode H is. closer to the cathode I l and to the space. charge electrode 10 than it is to the first anode l2, and similarly the second control electrode T3 is closer to the. first anode 12 than it water 7 is especially useful in connection with apparatus such. as a hearing aid in which it is highly desirable, not only that low potential be necessary for all of the electrodes, in orderthat very small wearable batteries may be utilized to supply the potentials, but also that the current andpower drain required by the discharge device be extremely small, so that the small Wearable batteries may have reasonably long life.

In Fig. 9 certain operating characteristics of all,the discharge devices so far described are illustrated, anode currents being plotted as ordinates, against second anode voltages, as abscissae. The characteristics illustrated are present in these discharge devices, provided cathode, temperature is high enough thatelectron emission from the cathode is not temperature limited.

When the second anode voltage is very low, and

with a fixed first anode potential high enough to causea substantial first anode current to fi ow with a zero potential on the second anode, with a small negative bias potential with respect to the cathode n the first control electrode, and withthe second control electrode maintained at cathode potential, or, more accurately at that potential at which the second control electrode just begins to take electron current. all current flows to the first anode. Curve 80 illustrates the variation, in first anode current under those conditions as the second anode voltage is increased from zero volts with respect tothe cathode to a high positive voltage, and curve 8| illustrates the variation in the current flowing in the second anode as the second anode voltage is so changed.

Curve 89 shows that the first anode current is of substantial amount when the second anode voltage is zero with respect to the cathode and remains substantially constant as the second anodevoltage is increased to a substantial positive voltage with respect to the cathode. At some substantial positive second anode voltage, for eX-' ample, a voltage in the order of ten volts, the first anode current begins to decrease as the'second anode voltage increases, and continues decreas ing, finally approaching a limiting minimum cur-. rent which does not decrease as the second anode voltage approaches infinity.

Curve 8! illustrates that the second anode current is zero when the second anode voltage is zero with respect to the cathode, and remains zero as the second anode voltage is increased in a positive direction to a substantial positive voltage, for example, in the order of ten volts. As the second anode voltage increases further in a positive direction, the second anode begins to take current'from the first anode (the total of the first I anode currentand the second anode current increasing gradually), and the second anode current increases through the range where the first anode current decreases, finally approaching asymptotically a maximum current as the second anode voltage increases toward infinity.

There is a particular second anode voltage at which the curves 80 and BI cross, and at which the first anode current and the second anode current are substantially equal. that second anode voltage at which the anode currents are equal and that second anode voltage at which the second anode current first appears that it is preferred to operate the discharge device made according to this invention. With the second control electrode constructed in accordance with the invention, the anode currents and the anode voltages within this preferred operating range are of normal and reasonable magnitude, and the discharge device may be used with various circuits to produce extremely high gain amplification. With the second control electrode so constructed according to the invention, the.

second anode voltage is impractically high, and too high for practical operation of the discharge device, where the second anode current is near maximum and the firstanode current near minimum. It is in that region that the normal pentode operates, and it is able to operate in that region with a smaller anode voltage by reason of a different construction of its third grid, commonly called the suppressor. The suppressor electrode in a normal pentode is made with an open mesh so that it does not substantially afiect the current distribution between the screen electrode and the anode while the suppressor .is near cathode potential.

In- Figure 10 an evacuated glass vessel contains 5 electrodes constructed accordingto the invention, of which the cathode 9| is indirectly heated. With the indirectly heated unipotential cathode 91 more space current is caused to flow to the anodes 92 and 94, and those anodes are arranged for operation at higher positive potentials with respect to cathode 9| than is the. case with the devices illustrated in Figures 1 through 8, that is, the discharge device illustrated in Figure 10 is constructed so that the .anodesupply' potential for the anodes 92 and' 94 may be in the order of 200 volts, and should be in the order of at least 45 volts. The first control electrode 93, between the first anode 92 and the cathode 9l, and the second control electrode 95 between the first anode 92 and the second. anode 94, are both made to have a relatively high static amplification factor. The second control electrode 95 may, for example, be so constructed that its static amplification factor is in the order of 25 or more, thereby necessitating that the supply potential for the second anode 94 be large relative to the static amplification factor of the control electrode 95 multiplied by the contact potential, which is in the order of 1 volt. Except for the construction of the five elec trodes of the discharge device in Figure 10 so that the anodes operate at higher positive poten-'- tials, and so that the control electrode 95 has a higher static amplification factor, the general principles of the construction and operation of this discharge device are much the same as those illustrated in Figures 1 through 8. I

Purely by way of example, specific dimensions for the electrodes are given. The'diameter of the cylindrical unipotential indirectly heated cathode 9! is 0.047 inch. The first control elec- It is between inch. The second anode 94 is formed of sheet metal having flat wing portions on opposite sides of the heated cathode 9 I.

In Figure 11 a sectional View of the device shown in Figure 10 taken along the line HH illustrates more clearly the general shape of the various electrodes of the device in a plane perpendicular to the axis of the cathode 9!. The wings I02 and 13 of the anode Q 1 lie parallel to each other and are spaced apart 0.600 inch. The wings IE2 and we lie in planes which are parallel with the. axis of, the indirectly heated cathode 9|. The second control electrode 95 has flat sides which are generally parallel with the wings I02 and Hit of the anode 94, and the flat sides of the second control electrode 535 are spaced apart. by 0.270 inch. Similarly, the first anode 92 has flat opposite sides generally parallel with the wings Hi2. and use of the anode t4, and the flat opposite sides are spaced apart by 0.172 inch. The distances by which these fiat sides are spaced apart are all measured along a line perpendicular to the axis of the cathode 9i and to the planes IE2 and W3 of the anode M. Along that same line the opposite curved sides of the first control electrode 93 are spaced apart by 0.075 inch. It should be noted that the opposite sides of the first control electrode 93 are not flattened, but are rounded, so that the control electrode 93 is substantially equidistant from all emitting parts of the cathode 91. The shadowing effect of the control electrode supporting posts 95 and Si is effective to suppress electron emission from portions of the surface of the cathode 9i near a plane passing through the supporting posts 95 and 91.

The electrodes 9| through 95- are all supported in symmetrical relation with eachother between mica spacers EM and 55, which in turn are held in proper relation to one another and to the glass vessel 99 by metal caps I06 and it? which are held apart in fixed relation by posts m8 and IE9. A suitable getter assembly l in is supported in any desired fashion in the upper portion of the envelope 9% Individual external connections for the five electrodes 9| through 95 are provided, each extending through the glass wall of the envelope 9* A lead-in conductor l I provides an external connection for the unipotential cathode M. A lead-in conductor H2 provides an external connection for the first control electrode 93. Leadin connection-s H3, H4 and H5 respectively, provide external connections for the first anode 92, second control electrode 95, and second anode 94. Lead-in conductors llii provide external connections through which suitable continuous or alternating current may be. suppliedto a resistance heater within the unipotential cathode 9|.

In Figure 12 certain characteristics of a discharge. device constructed in accordance with this invention are presented in a somewhat different. light than in. Figure 9.. In Figure 12 the currents flowing in the second, anode are plotted as ordinates andv the potentials of the second anode are plotted as abscissae. Curve I20 is similar in significance to curve 81' of Figure 9 being taken with a fixed positive potential on the first; anode, with a small negative bias potential on the first control electrode, and with no bias, potential on the second control electrode. That is, curve I20; is taken with the second control, electrode connected to the cathode, or more 14 accurately, adjusted to that potential at which current just begins to fiow in it.

Curve l2! may be observed under similar conditions, but with the potential of the second control electrode made a little more negative than is the case with curve l2i'3. The curves illustrate that the second anode current is throughout the operating range, with a substantial negative bias potential on the second control electrode substantially less than with no bias potential on the second control electrode.

By referring to points on the curves I28 and :2! on a perpendicular line, such as the line I22, it may be determined how much current change can be expected in the second anode (delta I) in response to a predetermined small voltage change of the second control electrode, the first and second anode voltages remaining constant. Of course, in actual use, application of a small negative bias potential to the second control electrode tends to decrease the second anode current with a corresponding increase in the second anode voltage, because the second anode must be connected to a source of supply potential through a load impedance. Consequently, when a load impedance is connected with the second anode, the actual change, or reduction, in said second anode current upon an increase in negative bias potential of the second control electrode is smaller than might otherwise be expected. EX- pressed another way, the static transconductance from the second control electrode to the second anode is always greater than the transconductance between those two electrodes measured with a load impedance connected with. the second anode.

As pointed out previously, the peculiar construction of the second control electrode makes it possible to have these control characteristics with the second control electrode in the region of cathode potential. The normal pentode with a suppressor electrode operates at'second anode voltages above the operating range indicated in Figure l2, and a suppressor electrode has a negligible transconductance over a substantial range of voltage near a cathode voltage. As a matter of fact, the second anode current does not drop to zero until the second anode voltage has reached Zero voltage with respect to the cathode, unlike the characteristic of the preferred form of discharge device according to the invention in which the construction of the second control electrode keeps the second anode current at zeroeven while the second anode voltage is at substantial positive level,

In Figure 13 the discharge device I30, con-- structed in accordance with this invention, is connected with a special circuit and provides very high gain amplification in that circuit. For convenience in understanding the construction and operation of the discharge device I38, it will be assumed that it is identical with the discharge device described in Figures 1 and 2, and the same reference characters will be used for parts of the discharge device. It should be understood, of course, that any other discharge device constructed in accordance with the invention, and particularly any of the devices illustrated in Figures 1 through 8 or in Figures 10 and 11' might be utilized as the discharge device .l3il, if suitable electrode supply potentials are provided.

One terminal of the filament H is grounded and a suitable source l3! of filament heating current is connected between the two terminals of the filament H. When a filamentary sourc H as illustrated in Figure 1 is utilized, the source I3I should provide continuous current for heating the filament I I, as from a primary or secondary battery or a source of rectified alternating current. If, however, the filament I I be replaced by an indirectly heated cathode, the heating current from the source I3I may be either alternating or continuous current. A source I32 of signal potential is connected between the first control grid I2 and the grounded terminal of the cathode II. The negative terminal of a source I33 of anode current is connected to one terminal of the filament II, and the positive terminal of the source I33 is connected through a resistance I 34 to the first anode I3. A resistance I35 is connected between the second control electrode I4 and ground and a condenser I36 is connected between the second control electrode I4 and the first anode I3. The second anode I is connected through a resistance I31 to the positive terminal of the source I33.

Small potential variations impressed from source I32 between the first control electrode I2 and the cathode I I appear as amplified potential variations upon the first anode I3, and such amplified potential variations across the resistance I34 are impressed through the coupling condenser I36 to resistance I35. Potential variations across the resistance I35 appear on the second control electrode I4, wherefor current distribution betweent he anodes I3 and I5 is adjusted in accordance with such potential variations. As explained previously, the eifect of this adjustment of current distribution between the anodes i3 and I5 by amplified potential variation from the anode I3 impressed upon the second control electrode I4, is regenerative in nature and therefore increases the magnitude of the amplified potential variation on the anode I3.

The greatly amplified potential variations thus appearing on the second control electrode I4 appear in still greater amplified intensity on the second anode I5 and across the resistance I31. The potential variations across the resistance I31 are coupled by a coupling condenser I38 to any desired load I39. Hereafter, when referenceis made to resistance I31, it should be understood to mean the entire load impedance associated with anode .I 5, including load I39, stray capacities, etc.

To understand clearly the operation of the circuit illustrated in Figure 13, it should be borne in mind that the potential variations appearing on the anodes I3 and I5 are opposite in phase. If it be assumed at a particular instance of time that the positive potential of the first anode I3 is decreasing it causes'a corresponding change in the potential of the second control electrod I4 in the negative direction. This potential change generation between the second control electrode V I4 and the first anode I3 should be made as great as possible without causing self-induced oscillation.

For proper operation of this device, the reactance of the condenser I36 should be sufficiently small with respect to the magnitude'of the resistances I34 and E35 that substantially all of the potential variation appearing on the first anode I3 is transferred to the second control electrode I4. Asa general rule, the resistances I34, I35 and I31 should be large compared with the internal resistance of the anodes I3 and I5. Now, if the relative magnitude of those resistances I34, I35, and I31 is made such that the maximum possible gain of the discharge device I30 is achieved b using the maximum amount of regeneration between the control electrode I4 and the anode I3 without producing oscillation within a desired frequency range, the regeneration will be found to change with frequency in such a manner that oscillation is possible at some frequency outside that desired range,

Consequently, with the form of circuit shown in Figure 13, the resistance I34 must be made substantially smaller than the resistance necessary to achieve maximum gain in the desired range of frequencies, to avoid oscillation at a frequency outside that range, with the result that a smaller gain must be accepted in the desired range. This condition of freedom from oscillation can be achieved in either of two ways.

Resistance I31 may be selected to provide a desired frequency response, anode current, output power, etc., and resistance I34 made small enough to prevent oscillation at any frequency, Alternatively, resistance I34 may be made to provide a desired current in first anode I3, or it may be chosen for other reasons, and resistance I 31 made large enough to prevent oscillations. Over a wide range of variation, there is required a minimum ratio between resistance I31 and resistance I34. The whole phenomena can be viewed in another way, in that oscillation occurs because the negative transconductance between the second control electrode I4 and the first anode I3 increases as frequency increases for the reason that the load impedance connected with the anode I5 is not purely resistive buthas capacity effectively in shunt with it. This capacity may be in the load I39 or may be simply'stray circuit capacity, including inter-electrode capacity, and

when very high gains are involved, such capacities can cause oscillation. The dynamic negative transconductance between the second control electrode I 4 and the first anode I3 is suflicient to cause oscillation if it is equal to or greater than on the second control electrode I4 at once does 7 two things. in th' current flowing in the second anode I5, with a consequent increase in the potential of the anode I 5 in the positive direction, and also causes an increase in the current flowing in the first It simultaneously causes a decrease the sum of the reciprocals of the load impedance connected with the first anode I3 and the internal resistance of the first anode I3. This relation of course assumes'that the load impedance connected with the second anode I5 and all sup- .ply voltages are constant. Inspection of the relation immediately indicates that a decrease either in the load impedance connected with the anode I3 or in the internal resistance of the anode I3 makes it less likely that oscillation can occur.

In an circuits of the general type illustrated in Figure 13, the resistances I34 and I31 should preferably be of such magnitude that the current flowing in the anode I5 is less than the current flowing in the anode I3. This relation is illustrated in Figure 9 as'making the device operate in the preferred operating range. It is at least necessary that the resistances I34 and -I31be so adjustedthat the ratio of'curre'nt inthe'anode I to'current in anode i3 is substantially smaller than'the maximum ratio reached when the device is connected as illustrated and-the supply potential for the anode 1'5 is increased indefinitely. This latter relation exists when the device is operated in the indicated operating range of Figure 1-2.

These conditions may be readily satisfiedin a preferred arrangement wherein the current in anode I3 is substantially larger than the current in anode I5, and where the potentials on theseanodes are of the same'order of magnitude, by supplying current-to the anodes from-a common source and making the load resistance I31 substantially larger than the load resistance I34.

Another consideration in choice of magnitudes for the resistances I34 and I3! is that, *even' though the resistance I31 should be larger than the resistance 1341, the ratio of magnitude of resistance I31 to resistance 134 should be not more than the static voltage amplification factor of the second control electrode M with respect to anode =I-5.

An exact-analysis of the 'conditions requisitein this circuit for avoidance of oscillations follows. The analysis-leads to a-choice of load impedances 34, 35 and 1:3? in the return of the 'first and second anodes. By the term -load'impedan'ces'is meant not only the illustrated resistances, but also circuit capacities (such as stray capacity) and circuit inductances,=any of which maybe added deliberately. To obtain a formulaby which'may be'selected load impedances'within which oscillation cannot-occur, it must bebornein mind that the admittance'prevailing betweenthe'first anode I3 and ground must remain positive and finite overthe entire frequen-cy'range. This admittance consists of the sum of the internal admittance l/Rl of the first anode 13 (with all other electrodes'keptat constant potential) theadmittance in. the external circuit of the first anode (which admittance is reciprocal to theimp'e'dance Z1) and the effective trans-conductanceGF from the second control'electro'deld tothe'first anode =I3. Z1 is the external impedance between anode -I=3 and ground-and includes resistances I34 and I35, stray circuit capacities and other effective circuit capacities and any eifective circuit inductances. The trans-conductance G is negative in sign, because a more positive potential on the second control electrode M causes more current to fiowtothe second anode I5 so that'less current flows to the first anode -I 3. Potential variations of the second'anode I 5 alfectthe current to the first anode-I3 in a similar manner, the eflect-of the second anode I 5 being obtained by-dividing the potential Variations at the second-anode I5 bythe static amplification factor from secondcontrol'electrode I4 to'second anode I5. The potential variationsat-the second anode I5 can be derived from those atthe second control electrode I4 by using conventional :triode computations, and substitution leads to an expression for trans-conductance Gr from the second control electrode 14 tothe-first anode I8.

R 1) am In whichGM is the static trans-conductance from control electrode M to the first anode 13, other electrodes being maintained 'at-con'stantpotential (especially-anode I5) and-inwhich Puz'a'nd Zz-are respectively the internal resistance-and external load impedance associated withanode l5, analo- 1'8 gous to'RrandZr. The condition of a finite positive'sum'of admittances can now be written in the following form:

which can be simplified:

3) aa yzacawta and (4) 2) GMR2 It -'is evident by inspection of Equations? and 4 that-oscillations may be suppressed by (a) an increase in the magnitude'of load impedanceZz (as by increase in resistance or inductance or decrease incapacity), (b) adecrease-in GM by appropriate construction of the device (Fig. '6) c) a decrease in internal resistance R2,-othe'rfac-tors remaining constant, (d) a decrease 'in internal resistance R1 by appropriate construction of the device or (e) a-decreasein '21 as by=decrease in its resistance or-inducta'nce oran increase-in capacity.

In utilizing the above'relation in the designcf commercial circuitsjit should be remembered that the relation should lie satisfied within the range of tolerances of commerciall available circuit elements. To achievema'ximum gain, it is desirable furthermore that the inequalit have a minimumwitliin or near the desired ra'ng'e (if-operating frequencies. w

While itispossible to avoid self-induced oscillation in the circuit arrangement cf-Figure -1-3 as described, "(by making the ratio of resistance I31 to resistance 134 large enough) it is preferred to make such oscillations impossible in a-somewhat different manner.

Referring to inequalities '(3) and (4') it'is preferred to suppress oscillation by adjusting not only the resistive components of the load impedances Z1 and Z2 but also the reactive components (usually capacity) so that theinequali-ty (-4) is satisfied at all frequencies.

In Figure "1 4, 'such an'arrangemen t is' illustrated, and since many elern'ents are similar to those i1- lustrated in Figure l3, like reference nu'meralsare applied. A condenser MI) is con'nectedinshunt with the resistance I3 3 and is ofsuch mag'nitude that the timefcons'tant existing between the first anode 13 and ground is similar in magnitude to the time constant existing between the second anode I5 and ground. Expressed in single terms, the condenser I'M is increased until no oscillation-is possible.

The condenser M3 is connected eirectively for alternating "current between the first anode I3 and ground to achieve this purpose, rather than between the second anode Iii and ground, because, as stated' previously, when acom'mon source I33 of anode current is used, the resistance must be substantially larger than the resistance I34 an dconsequently the capacity-associated with the second anode 1-5 must be smaller than that associated with the first anode I 3. Since *stray circuit capacities and inter-electrode capacities associated with the first anode I3 are normally smaller than 'those associated with the second anode I'5,'it is usually necessary to a'ddadd-itional capacity in the form of the'conde'ns'er l lll-inprder thatthe desired relation should be achieved.

Bytheadditicn of the condenser'Mll, so that 19' the time constant or circuits associated with the first anode I3 is similar to the time constant of circuits associated with the second anode I5, the ratio of the absolute magnitude of the impedances associated respectively with the two anodes may be made to satisfy the inequality (4).

The impedance associated with the anode I3 is, by the condenser I40, made to decrease just fast enough in relation to the impedance associated with the anode I that this relation is satisfied.

In Fig. 15 there is illustrated a circuit arrangement which is capable of satisfying that expression more accurately. Again, since many circuit elements are similar to those illustrated in Fig. 13, they are given like reference numerals. A serial combination of a condenser I50 and the resistance I5I are connected in shunt with the resistance I34. The capacity of the condenser I50 with the resistance I5I can be given such magnitude as more nearly to satisfy the relation that the product of the sum of the impedance Z1 associated with anode I3 and the internal resistance R1 of anode I3 and the sum of the impedance Z2 associated with anode I5 and the internal resistance R2 of anode I5 divided by the product of R1 and Z1 is always greater than the product of the internal resistance R2 of anode I5 and the static trans-conductance GM from the second control electrode I4 to the first anode I3. That is, if there is a purely capacitive impedance connected between the first anode I3 and ground, as the condenser I40 in Fig. 14, the impedance associated with anode I3 approaches zero as frequency increases toward infinity. To satisfy the desired relation, however, the impedance between the anode I3 and ground need not approach zero, but must in any case decrease to a positive limiting magnitude as frequency increases toward infinity. That limiting impedance is substantially the resistances I34, I35 and I5I in shunt with each other. The condenser I 50 is of such capacity that, in the operating range of frequencies, the relation is satisfied within the tolerance of available circuit elements whereby the gain of the system is as highas possible. I

In Fig. 16 there is illustrated the circuit arrangement which is especially useful with certain types of loads. Where an amplifier circuit arranged according to this invention feeds a load which may under certain circumstances change impedance abruptly by substantial amounts, it is highly desirable to isolate that load from the amplifier. Since, again, many of the elements illustrated in Fig. 16 are similar to those illustrated in Fig. 13, they are given like reference characters.

The load device in Fig. 16 which is coupled to the second anode I5 through the coupling condenser I38 includes an electron discharge amplifier device I60 of the tetrode type which is connected in normal fashion and with its anode I6I coupled with a load I62. I63 of the device I60 is connected through a resistance I64 to the coupling condenser I38, and the point between the condenser I38 and the resistance I64 is connected through a grid leak resistance I65 to ground and to one grounded terminal of the cathode I66 of the discharge device I60.

Even though the alternating signal volt-age from the source I32 may be in the order of a few millivolts, the amplification in the discharge device I30 is enormous and the. signal potential on the second anode I5 may be in the order of several volts. Such alternating potentials ap- The control electrode.

plied'between the control electrode I63 and oathode I66 of the discharge device I60, where there is no negative bias potential provided, may cause the control electrode I63 'to draw substantial amount of current from the electron stream in the discharge device I60. During normal operation of the device I60, the control electrode I63 does not draw such' current, and the input of the device I60 is accordingly Very high, and the device- I30 can be very satisfactorily adjusted as previously explained for highly stable operation with very high gain. However, if the impedance between the control electrode I63 and the cathode I66 decreases abruptly, as during the short period when current flows in the control electrode I63,

effective to prevent such short bursts of oscillation during each cycle of the amplified signal wave.

The resistance I64 is made sufiiciently large that, even when the control electrode I63 takes current when it is made positive with respect to the cathode I66, the impedance from the anode I5 through the coupling condenser I38 to ground is not reduced sufficiently to cause the device I30 to produce such bursts of self-sustained oscillation. It is evident that, in general, the larger the resistance I64, the closer can the adjustment of the device I30 approach that point at which selfinduced oscillation would persist.

It should benoted that the condenser I61, connected in shunt withresistance I35 in Fig. 16, performs substantially the same function as the condenser [40, connected in shunt with the resistance I34 in Fig. 14.

In Fig. 17 the discharge device I10 is illustrated whichis identical with the discharge device illustrated in Figs. 7 and 8, and for the sake of convenience, the electrodes are given like reference numerals. Since certain of the circuit elements associated with the discharge device I10 are similar to those illustrated in Fig. 13, they are also given like referencenumerals to those in Fig. 13. The space charge electrode 10 is connected to the cathode II through a suitable source I1I of potential, so poled as to maintain the space charge electrode 10 at a suitable positive potential with respect to the cathode II. The first control electrode 1 I, the first anode 12, the second control electrode 13, and the second anode 14. are connected in a manner analogous with the corresponding first control electrode I2, first anode I3, second control electrode I4 and second anode I5 of Fig. 13. a

As pointed out previously in connection with the description of the discharge device itself, as illustrated in Figs. 7 and 8, the potential of the source I33 may be substantially smaller than is the case in Fig. 1 3 to achieve equal gain, or, alternatively, where the same source I33 is used, much higher gain can be achieved in the arrangement illustrated in Fig. 17, because the space charge electrode 10 provides substantial cathode current even when the anodes 12 and 14 are operated at what would otherwise be abnormally low potential.

The eneral adjustments of the circuit are similar to those previously described, and, if desired, the condenser I12 may be connected in shunt with the resistance I35,.so that the arrangement may be adjusted as explained in connection with Fig. 14. That is, as explained previously it is a matter of choice Whether the condenser I12 or its equivalent i placed in parallel with resistance I34 or resistance I35. That is, condenser I61 connected in shunt with resistance I35 in Fig. 16 is of such magnitude that it produces the same desired result as is the case when a condenser I 40 in Fig, 14 for producing the same efiective time constant is connected in parallel to resistance I3 1. In other words, the time constant of the circuit connected between ground and anode I3 maybe adjusted to have a predetermined time constant either by placing an impedance in parallel with resistance I3 1 or-by placing an impedance in parallel with resistance I35. Alternatively, the condenser I'IZ may be-omitted and thecircuit adjusted as'explained in Fig. 13.

In Fig. 18 thereis illustrated acircuit arrangement utilizing the discharge device illustrated in detail in Figs. 10 and 11. It should be understood that this type of discharge device can of course be utilized in the circuit arrangement of Figs. 13 through 17. For the sake of convenience, the-discharge device I80 in Fig. 18 has the same reference numerals applied to its electrodes as are applied to the electrodes illustrated in Fig. 10. A source I8I of alternating potential, having frequencies covering a substantial band, is connected between the control electrode 93 and the grounded cathode 9i. The negative terminal of a source HIE-f anode current is grounded,and the positive terminal of the source I82 is connected to the first anode 92 through a tun'ed'circuit, including in parallel an inductance I83 and a condenser I84.

second control electrode 95, which in turn is connected through a, resistance 186 to ground. The second anode 94 is connected to the positive terminal of the source I82 through a parallel tuned circuit including the primary inductance I3! of a transformer having a secondary I88, a condenser I89 and a resistance I90. The secondary I88 of the transformer is connected to a suitable load ISI, which may, if desired, include a resistance component of such'magnitude as to replace resistance I98, which canthen be omitted.

The operation of the arrangement is similar to the operation of the previously described circuit arrangements. Signals to be amplified may, for example, be a carrier wave, modulated, for example, by audio or television signals or the like.

In order to avoid self induced oscillation, as explained previously, the inequalities previously described are preferably-satisfied for efiicient and high gain operation. In this case, R1, R2 and GM have the same meaning as in .the case of previous circuits. Z 1, for the illustrated circuit,.r.epresents the parallel impedance of condenser I84, inductance I83 and resistance M6, the reactance of coupling condenser I85 being negligibly small. Z2 represents similarly the resistive, and reactive components of impedances between anode 94 and ground (not including internal anoderesistance). Z2 includes resistance I90, condenser IBiLand the reactance and resistance appearing across the primary I187 of the transformer (due at least in part to load IQ'I), To satisfy the inequality, in general, the L/C ratio of the two tuned circuits and the resistances associated with-each may be A coupling condenser I85 is connected between the first anode 92 and the adjusted While maintaining the desired band obvious to those skilled in theart-that'changes and modifications may be made without departing from this invention in its broader aspects, and we, therefore, aim in the appended claims to cover all such changes and modifications as lfall within the true spirit andscope .of this invention.

We claim:

1. The combination with an electron discharge device having a first control electrode, -a first anode, a second control electrode and a second anode lying serially along a single electron discharge stream, said second control electrode and second anode being adjacent, of means for impressing alternating signal voltages on said first control electrode, means for impressing amplified signal voltage from said first anode upon said second control electrode, said last means comprising a resistance which forms at least a portion of the load circuit for said first anode,and means for producing amplified alternating voltages on said second anode, said last means comprising a complex impedance which forms at least a portion of the load circuit for said second :anode and which at some frequency operates to increase the amplification of said device sufiiciently to cause undesired self-oscillation, the time constants of said load circuit for said firstanode and of said load circuit for said second anode being substantially equal so that said device cannot produce undesired self-oscillation.

2. In combination with an electron discharge device having a first control electrode, a first anode, a second control-electrode and a second anode lying serially along the same electron discharge path, said second control electrode and said second anode lying adjacent each other,

means for impressing alternating signal voltages on said first control electrode, means for impressing amplified signal voltages from said first anode upon said second control electrode, said last means comprising a complex impedance formed by a resistance and a condenser connected in shunt with each other and conductively connected with said first anode, and means for producing amplified signal voltages on said second anode, said last means comprising a complex impedance which at some frequency reacts with said device in the absence of said condenser to cause self-sustained oscillations, said resistance and condenser having such magnitude-that their combined impedance changes with respect to changes of said complex impedance in such a manner as to prevent the production of self-sustained oscillation by said device.

3. In combination with an electron discharge device having a first control electrode, a first anode, a second control electrode and a second anode lying serially along'the same electron discharge path, said second control electrode and said second. anode lying adjacent each other, means for impressing alternating signal voltages on said first'control electrode, means for impressing amplified signal voltages "from said first-anode upon said second control electrode, said last means comprising a complex impedance forroed by a resistance and a condenser connected s-i'n shunt with each other, and means for producing amplified signal voltages on said second anode, said last means comprising a complex impedance which at some frequency reacts with said device in the absence of said condenser to :cause selfi- 23 sustained oscillations, said resistance andcondenser having such magnitude that their combined impedance changes with respect to changes of said complex impedance in such a manner as to prevent the production of self-sustained oscillations by said device.

4. In combination with an electron discharge device having a first control electrode, a first anode, a second control electrode anda second anode lying serially along the same electron discharge path, said second control electrode and said second anode'lying adjacent each other, means for impressing alternating signal voltages on said first control electrode, means for impressing amplified signal voltages from said first anode upon said second control electrode, said last means comprising a complex impedance formed by a resistance and a condenser connected in shunt with each other, and means for producing amplified signal voltages on said second anode, said last means comprising a complex impedance which at some frequency reacts with said device in the absence of said condenser to cause selfsustained oscillations, the impedance of said resistance and condenser being substantially larger than said complex impedance and said resistance and condenser having such magnitude that their combined impedance changes with respect to changes of said complex impedance in such a manner as to prevent the production of self-sustained oscillations by said device.

5. In combination with an electron discharge devic having a control electrode, a first anode, a second control electrode and a second anode all lying serially along a single electron discharge path, said second control electrode and second anode lying adjacent each other, means for impressing alternating signal voltages on said first control electrode, means for impressing amplified signal voltages from said first anode upon said second control electrode, said means comprising a complex impedance formed by a resistance and a condenser connected in shunt and conductively connected with said second control electrode, and

means for producing amplified signal voltages on said second anode, said last means comprising a second complex impedance which in the absence of said condenser causes self-sustained oscillations in said device at some frequency, said resistance and condenser having such magnitudes as to prevent such production of self-sustained oscillations at any frequency within the desired range of frequencies.

6. In combination with an electron discharge device having a first control electrode, a first anode, a second control electrode and a second anode lying serially along a single electron discharge path, said second control electrode and second anode lying adjacent each other, means for impressing alternating signal voltages on said first control electrode, means for impressing amplified signal voltages from said first anode upon said second control electrode, said last means comprising a complex load impedance for said first anode having a resistance connected in shunt with a. series arrangement of a condenser and a second resistance, andmeans for producing amplified signal potentials on said second anode, said last means comprising a complex impedance which, in the absence of said serial arrangement of condenser and resistance, tends to cause said device to produce self-sustained oscillations, said condenser and said two resistances having such magnitude as to suppres the production of such 75 impedance changes of said element do not sub-- oscillations at all frequencies within a-fd esired band of frequencies.

'7. In combination with an electron discharge device arranged for operation over a predeter mined range of frequencies, said device having sustained oscillations.

All

plified signal voltage from said first anode upon a first control electrode, a first anode, a secondcontrol electrode, means for impressing amplifiedsignal voltage from said first anode upon said second control. electrode whereby said device tends to oscillate at some frequency within or outside said range of frequencies, and a second anode lying serially along a single electron dis: charge path, a load impedance connected with each of said anodes, and means for making the time constant of said load impedances similar to prevent said device from producing selfof working frequencies, and means for making the time constant of said load impedances similar to prevent the production of such self-sustained oscillation.

9. In combination with an electron discharge device having a first control electrode, a first anode, a second control electrode,,and a second anode all lying serially along a single electron discharge path, a load impedance connected with each of said anodes, means for impressing amsaid second control electrode whereby said device tends to produce self-sustained oscillations at some frequency within a predetermined range of working frequencies, means for making similar the time constants of said load impedances to prevent the production of such self-sustained oscillation, said load impedance connected with said second anode including an element whose impedance changes abruptly in large amount during instantaneous signal voltage peaks, and a fixed impedance connected serially with said T element in said load impedance, said fixed impedance being sufliciently large that abrupt changes in impedance of said element do not substantially affect the time constant of said load impedance connected with said second anode.

10. In combination with an electron discharge device having a first control electrode, a first anode, a second control electrode, and'a second anode all lying serially along a single electron discharge path,'a load impedance connected with each of said anodes, means for impressing amplified signal voltages from said first anode upon said second control electrode whereby said device tends to produce self-sustained oscillations at some frequency within a predetermined range of working frequencies, means formaking the time constants of said load impedances similar to suppress the production of such self-sustained oscillation, said. load impedance connected with said second anode includingv an element whose impedance changes abruptly in large amount during instantaneous signal voltage peaks, and a resistance serially connected with said element and, being of sufficient magnitude that abrupt stantia-lly affect the timeconstant of said. load impedance connected. withsaid second anode.

11. In combination in. an. electron discharge device having a space charge electrode, a first control electrode, a first anode,.a second control electrode. and a second anode lying serially along a single electron discharge path, means for impressing amplified voltage from said first anode upon said second control electrode whereby said device tends to oscillate, and means for making the time constants of circuitsconnected withsaid anodes similar to suppres any tendency of said device to oscillate.

12. In combination, an electron discharge device having a first control electrode, a first anode, a second control electrode and a second anode lying serially along a single electron discharge path, means for impressing alternating voltages having high frequencies lying within a predetermined range of high frequencies upon said first control electrode, means for impressing amplified high frequency voltages in said predetermined range of frequencies from said first anode upon said second control electrode, said last means comprising a damped resonant circuit, and means for producing amplified high frequency voltages on said second anode said last mentioned means comprising a complex impedance of such magnitude as to suppress any tendency of said device toward self-sustained-oscil lations within said predetermined range of frequencies.

13. In combination, an electron discharge device having a first control electrode, a first anode, a second control electrode and a second anode lying serially along a single electron discharge path, means for impressing alternating voltages having high frequencies lyingwithin a predetermined range of high frequencies upon said first control electrode, means for impressing amplified high frequency voltages in said predetermined range of frequencies from. said first anode upon said second control electrode, said last means comprising a damped resonant circuit, and means for producing amplified high frequency voltages on said second anode, said last means comprising a second damped resonant circuit,.the damping of said resonant circuits and the reactances thereof being of. such magnitude as to suppress any tendency of said device toward self-sustained oscillations within said predetermined range of frequencies.

14. In combination, an electron discharge device having a first control electrode, a first anode, a second control electrode and a second anode lying serially along a single electron discharge path, means for impressing alternating voltages having high frequencies lyingwithin a' predetermined range of high frequencies upon said first control electrode, means for impressing amplified high frequency voltages in said predetermined range of frequencies from said first anode upon said second control electrode, said last means comprising a damped resonant circuit, and means for producing amplified high frequency voltages on said second anode, said last means comprising a second damped resonant circuit, the impedance of the resonant circuit connected with said first anode being so related tothe impedance of said other resonant circuit at any frequency within said predetermined range of frequencies that any tendency of said device to produce self-sustained oscillations at such frequency is suppressed.

15. In combination, a discharge device having a first anode,- a control electrode and a second anode all' lying serially along a single electron discharge path, means for impressing amplified signal voltages from said first anode upon said second control electrode whereby said device tends tooscillate, and a load impedance connected with each of said anodes, said' load impedances having similar time constants to prevent" the production ofself-sustained oscillations.

16. In combination, adischarge device having afirst anode, a control electrode and a second anode lying serially along a single electron discharge path, means-for impressing voltage from said first anode upon said' control electrode whereby said] device tends to oscillate, saidmeans comprisinga load'impedance connected with said first anode, and a load impedance connectedwith said-second anode, said load impedances having similar time constants to prevent the production of self-sustained oscillations.

17. In combination, a discharge device having a first anode, a control electrode and a second anode all lying seriallyalong a single electron discharge path, a loadimpedance connected with each of said anodes, means for impressing amplified signal'voltages from said first anode upon said second control" electrode whereby said device tends to oscillate, and a load element forming a part of'the loadimpedance connected with said second anode, said load element being subject to substantial changes'of impedance under certain operating conditions and having a fixed impedance of 'substantialma'gnitude in series therewith, the magnitude ofsaid fixed impedance being sufiicient to prevent substantial changes in impedance of said load'element from afiecting the time constant of the load impedance connected with said' second anode, said load impedances having similar time constants to prevent the production of self-sustained oscillations.

13; In combination, a discharge device having a first anode, a control electrode and a second anode all lying serially along a single electron discharge path, aloadimpedance connected'with each of said anodes, means for impressing amplified signal voltage from said first anodeupon said control'electrode whereby said device tends to oscillate, and a load element forming a part of the load impedance connectedwithsaid second anode, said load" element being subject to substantial changes of impedance under certain operating conditions and having a fixed'resistance of substantial" magnitude in series therewith, the magnitude of said fixed resistance being sufficient' to prevent substantial changes in impedance'of said'loadelement from affecting the time constant of the load'impedance connected with said second anode, said load impedances having similar'time constants to prevent the production of self-sustained oscillations.

19. In an amplifier, the combination of an electron discharge device having fiveelectrod'es includinga cathode, a first control electrode, a first; anode, a second" control electrode and a second anode.lying serially'along a single electron discharge" path, means for impressing a signal potentialbetween said. first control electrode and cathode, meansfor' producing amplified signal potential on" said first anode and'for' impressing such amplifiedsignal potential on saidsecond control electrode, andmeans for developing amplified signal potential on said second anode, said second control. electrode having a static voltage amplification. factor sufiiciently large that ourrent in said second anode remains substantially I cut off with said second control electrode near that potential at which current begins to flow in said second control electrode as the continuous potential of a source to which said second anode is connected rises from zero voltage to a substantial positive voltage.

20. In an amplifier, the combination of an'electron discharge device including a first control electrode, a first anode, a second control electrode and a second anode lying serially adjacent each other along a single electron d scharge path, said device having separate connections to said anodes, means for impressing a signal potential, on said first control electrode, means for producing amplified signal potential on said first anode and for impressing said ampl fied signal potential on said'second control electrode, and means for developing amplified signal potential on said second anode, means maintaining said first and second anodes at substantially the same continuous potential, said second control electrode being so constructed that, with substantially equal potentials on said anodes and the potential of said second control electrode near that potential at which electron current begins to fiow'therein, substantial changes are effected in the electron current distribution between said anodes by small potential changes of said second control electrode.

21. In an amplifier, the combination of an electron discharge device including a first control electrode, a first anode, a second control electrode and a second anode lying serially adjacent each other along a single electron discharge path, said device having separate connections to said anodes, means for impressing a, signal potential on said first control electrode, means for producing amplified signal potentialon said first anode and for impressing said amplified signal potential on said second control electrode, and means for developing amplified signal potential on said second anode, said second control electrode being so constructed that, with substantially equal potentials on said anodes and the potential of said second control electrode near that potential at which electron current begins to fiow therein, substantial changes are effected in the electron current distribution between said anodes by small potential changes of said second control electrode, said signal producing means and said signal developing means having substantially the same time constant throughout the range of frequencies of said signal potential.

22. In combination, an electron discharge device having a first control electrode, a first anode, a second control electrode and a second anode lying serially along a single electron discharge path, means for impressing a signal potential on said first control electrode, means for transferring amplified signal potential from said first anode to said second control electrode, said transferring means comprising a load impedance connected with said first anode, and means for developing amplified signal potential on said second anode, said developing means comprising a second load impedance, the ratio of said second load impedance to said first load impedance being as small as possible, allowing for the tolerance of elements available to form said load im pedances, that saiddevice does not produce selfsustained oscillations.

7 23. In combination-,an electron discharge device having a first control electrode, a first anode,

a second control electrode and a second anode lying serially along a single electron discharge path, means for impressing a signal potential on said first control electrode, means for, transferring amplified signal potential from said first an- Ode to said second control electrode, said transferring means comprising a load impedance connected with said first anode, and means for developing amplified signal potential on said second anode, said developing means comprising a second load impedance, the ratio of said second load impedance to said first load impedance being at least as small as the static voltage amplification factor of said second control electrode with respect to said second anode but not so small as to produce self-sustained oscillation.

24. In combination, an electron discharge device having a first control electrode, a first anode, a second control electrode and a second anode lying serially along a single electron discharge path, means for impressing a signal potential on said first control electrode, means for transferring amplified signal potential from said first anode to said second control electrode, said transferring means comprising a load impedance connected with said first anode, and means for developing amplified signal potential on said second anode, said developing means comprising a second load impedance, said load impedances being so adjusted that, allowing for the tolerances of elements available to form said load impedances, the gain of said electron discharge device is as high as possible Without the production of selfsustained oscillations, the ratio of said second 7 load impedance to said first load impedance being at least as small as the static voltage amplification factor of said second control electrode with respect to said second anode.

25. In combination, an electron discharge device including a first control electrode, a first anode, a second control electrode and a second anode all lying serially adjacent along the same electron discharge path, a first impedance Z1 connected with said first anode, a second impedance Z2 connected with said second anode, means for impressing a signal potential on said first control electrode, and means for impressing amplified signal potential from said first anode upon said second control electrode whereby amplified signal potential appears on said second anode, said impedances being of such magnitude that at any frequency is greater than RzGM in which inequality, R1 is the internal resistance of said first anode, R2 is the internal resistance of said secondanode and GM is the static transconductance from said second control electrode to said first anode.

26. In combination, an electron discharge device including a first control electrode, a first anode, a second control electrode and a second anode all lying serially adjacent along the same electron discharge path, a first impedance Z1 connected with said first anode, a second impedance Z2 connected with said second anode, means for impressing a signal potential on said first control electrode, and means for impressing amplified signal potential from said first anode upon said second control electrode whereby amplified signal quency

Images