US2159731A

US2159731A - Neutralizing high-frequency pushpull amplifiers

Info

Publication number: US2159731A
Application number: US104634A
Authority: US
Inventors: Buschbeck Werner; Prost Hans
Original assignee: Telefunken AG
Current assignee: Telefunken AG
Priority date: 1935-09-13
Filing date: 1936-10-08
Publication date: 1939-05-23
Anticipated expiration: 1956-05-23

Description

May 23, 1939. w. BUSCH BECK ET AL 2,159,731

NEUTRALIZING HIGH-FREQUENCY.PUSH-PULL AMPLIFIERS Filed oct. s, 1936 2 Sheets-Sheet 1 cur/160E INVENTORS W. BUSC H BECK BY H. FROST I ATTORNEY May 23, 1939. w; B USCHBECK ET AL 2.15%131 NEUTRALIZING HIGH-FREQUENCY PUSH PULL AMPLIFIERS Filed Oct. 8, 1936 2 Sheets-Sheet 2 INVENTOR I w. BUSCH BECK BY H. PROST A QRNEY Patented May 23, 1939 PATENT OFFICE NEUTRALIZING HIGH-FREQUENCY PUSH- PULL AMPLIFIERS Werner Buschbeck and Hans Prost, Berlin, Germany, assignors to Telefunken Gesellschaft fiir Drahtlose Telegraphie m. b. H., Berlin, Germany, a corporation of Germany Application. October 8, 1936, Serial No. 104,634 In Germany September 13, 1935 3 Claims. (Cl. 179-471) the neutralization of high frequency amplifiers can be rendered independent of frequency in that not only the gridanode capacity be neutralized, but also the cathode-grid capacity and the cathode-anode capacity. According to applicant's proposal, this neutralization circuit is improved to" obtain a still better independence of frequency, in that not only the capacities of the tube, but also the inductances of the electrode lead-ins are reproduced in the neutralization branches. 1

For a better understanding of the invention, reference is made to the accompanying drawings, 151 in which:

Figs. 1,2 and are circuit diagrams of a pushfl p p ull amplifier neutralized in accordance with our 3 invention, v

,d Fig. 3 is a portion of the bridge circuit of Fig. 2 arranged to showcertain voltage relations,

I i Fig. 4 is aportion of the bridge circuit of Fig. 2

l l l It is known from German Patent 532,515 that arranged to show other voltage relations between A certain points and Fig. 6 isv a portion of the bridgecircuit of Fig. 5

showing certain voltage relations. I

, An example of application to the push-pull circuit is shown in Fig. 1 in which only the tubes and the neutralizing branches are shown. This reproduction is carried out in the simplest manner in that a cold tube is used in the neutralizing,

or a geometrically exact imitation with or with-.

out glass bulb.

It was found that especially in case of short waves, considerable difiiculties are encountered 36 owing to thepassing of an electron current be- H tween cathode and grid and between cathode and "anode. This electron current in fact, changes on -the one hand the capacities between cathode and x grid and between cathode and anode, owing to its 4 :jspace charge, and on the otherhand, it produces a damping of the grid-cathode path and a negative damping of the anode-cathode path. .These conditions produce the following actions V uponneutralizatiomThe amplifier may be for example neutralized in that 'a high-frequency alternating potential is applied to the input terminals while the potential at the output is reduced to zero. If the tube is a non-heated tube, and a neutralizing-condenser forms an exact 're- 5o production of the tube, the output potential is equal to zero, 1. e., the stage is neutralized. Now

if the tube is heated, so that a direct grid current flows,the output potential is no longer zero. However, neutralization can be reestablished by 56; increasing accordingly the grid-cathode capacity by the neutralizing-condenser. At larger grid currents, a matching of the reactances is no longer sufficient, and it would be necessary to provide corresponding damping for the grid-cathode path of the neutralizing-condenser, in order to reproduce in the neutralizing-branch the phase displacement produced by the grid current. This matching would however only be valid for a single operating condition.

Wholly equivalent conditions appear, if for carrying out the neutralization an alternating plate potential is applied to the output terminals, and the alternating grid potential at the input is suppressed.

Hitherto these influences of the electron current to the grid have been reduced in that sufliciently high capacities were connected in parallel to the electrode capacities so as to reduce the influence of the electrode capacities. This method is useless as such in case of ultra-short waves, since on the one hand the capacities must be maintained as low as possible, while on the other hand, the additional parallel capacities can be connected only across a considerable line section.

This invention relates to a means and method by which the detrimental influences of the gridcathode path andanode-cathode path can be eliminated without prejudicing the operation especially at ultra-short waves. In accordance with the invention, the inductances of the electrode lead-ins and the corresponding inductances of the neutralizing-branches are so chosen that a for the alternating grid potential there are formedother equi-potential points besides the output terminals (bridge diagonals) as well as all anodecathode paths and the corresponding paths of the neutralizing branches. For the alternating plate potential, besides the input terminals (other bridge diagonals) also all grid-cathode paths and. the corresponding paths of the neutralizingbranches are formed to effect equi-potential points, so that variations in these tube paths, cannot disturb the equilibrium of the bridge.

tube to the electrodes. Although the inductances of the lead-ins situated within the tube could as such be dimensioned in accordance with the data according to the invention as discussed, froma practical point of view, the correct conditions will haveto be assured by the connection of inductances outside the glass bulb, since for reasons connected with the manufacturing methods of the glass bulb, the electrode lead-ins will not always have the dimensions herein required.

By the designation electrode lead-ins there is meant the connections inside and outside the to these paths.

While in the anode-neutralizing circuit solely the grid-cathode capacity and in the grid neutralizing circuit solely the anode-capacity can be formed into additional equi-potential points, both conditions can also be fulfilled at the same time in case of the push-pull circuit. For this reason the full result can be obtained only when applying the invention to the push-pull circuit.

It should also be pointed out that the change in the capacities by the space charge is not entirely without influence upon the neutralization, however in applying the invention its influence will be considerably reduced;

For a clearer explanation, the push-pull circuit according to Fig. l is redrawn in Fig. 2 in connection with the conditions of the bridge. The capacities are the electrode capacities assumed to be concentrated in the center of the electrodes, and the corresponding capacities of the neutralizing-branches. ignated by letters such as LA, LG and LR: are assumed to be concentrated in the electrode leadins (not especially shown in Fig. 1). As in all lines (in contrast to coils), they are not determined by the lines proper, but by the entire closed circuit. When interrupting the cathode connections of all tubes, and neutro-branches, and when short-circuiting all capacities, CAG the circuit starting at G1, Fig. 2, across A1, G2, A2 and back to G1 furnishes a resultant inductance, namely, 4:X(LA+LG). A closed circuit starting from A1 across the anode of tube I, the short-circuited capacity CAK of tube I, point K and anode a of the neutro-branch l and back, furnishes an inductance 2 (LA+LK). Acorresponding closed circuit from G1 across K results in 2 (LG+LK)- The three inductances LA, Lo and L1; thus are exactly defined.

In accordance with the invention, the induc-- tances LA, LG and LK and the corresponding inductances of the neutro-branches are so chosen that when applying a potential to a bridge diagonal, not only the other bridge diagonals but also the electrode paths lying in the direction of these other diagonals will have zero voltage. When applying an alternating grid potential to the points G1 and G2 an ordinary neutralizing circuit provides zero voltage between A1, K and A2. Through proper choice of the inductances according to the invention, the voltages at all anode-cathode paths likewise become zero. The apparent resistance of the anode-cathode paths thus is of no, importance to neutralization. This will be readily understood, when considering that voltages (and hencealso resistances) are applied These alternating potentials have no effect at the terminals G1 and G2.

On the other hand, when applying an alterable subscript, alongside the arrows.

In Fig. 3, one of the two tube systems is shown again. When applying an alternating grid potential Uebetween K and G1, the voltage between K and A1 is zero owing to the equilibrium of the bridge, so that the same alternating grid The inductances des- 1 potential UG lies between A1 and G1, and which for A1-G1 and K-G1 will be found as follows:

. CAK shall have zero voltage, so that therefore This value inserted in the third equation gives,

Between G1-A1 there exists the same voltage 1 j 2 Jc(.l o wCAG Ccx will be without voltage when: r

The two conditions furnish:

which gives the proper dimensioning for this example in accordance with the invention. In pratctice these relations will be attained through tes These relationships assume another form in case of tubes with concentric electrode lead-ins for the purpose of reducing the lead-in induction and for other reasons. pushgpull circuit with such tubes is shown in Fig.

L ZL IL C KZ AK: AG

On the one side of the tube, the lines leading to grid and cathode and on the other side, the lines leading to anode and cathode are brought out concentrically arranged to each other. The

principal difference as compared with the case,

shown in Fig. 2, resides in that the concentric lead-ins produce magnetic coupling between the electrodes. concentric tubings is equal to theself induction of the outer tubing. Fig. 6 shows one of the two tube systems of Fig. 5. In place of self inductions and capacities, the designations of the reactances are immediately indicated.

The bridge circuit of a- The counterinduction between two I 2,159,731 When applying an alternating grid potential,

the points A1, K1 and K2 have the same potential owing to the equilibrium of the bridge, In Fig. 6,

CAK and Can, means connecting together the endsof the inductances LK of the tube networks remote from the cathodes and the corresponding It follows from the above that;

J3 Kl M1)( K2 XM2) By eliminating the currents and voltages there is:

The conditions for the zero voltage of the gridcathode'path at appliedalternating plate potential are obtained by interchanging the indices 2' and 1 and A with G.

the values Xe, XMl, XA, XK2 any two values thereof can be selected and their value so chosen that the two conditions are fulfilled thereby. In

this connection the lead-in inductances may also become negative, i. e., they may become capacities.

Having described our invention, what we claim as novel and desire to secure by Letters Patent is:

1. An amplifying stage comprising a pair of input terminals and a pair of output terminals, a pair of tube networks forming a push-pull circuit each of said tubes including an anode, a cathode and a grid electrode, inherent capacity between thetube elements comprising CAe between said anode and grid, CAK between said anode and cathode and CGK between said grid and cathode, said tube networks including leads for connecting said tube elements in the circuit, inductance LK I in the lead from each cathode, inductance LA in the lead from each anode and inductance LG in the lead from eachgrid, said inductances being related to said capacities substantially in accordance with the proportion LA:LK.LG=CGK.CAG.'CAK

a pair of balancing networks composed of capacities and inductances substantially identical both as to arrangement and magnitude with said tube network capacities and inductances,

whereby each of said balancing networks includes inductances Lx, LA and La, and capacities CAG,

of said one tube network remote from the grid and the corresponding end of inductance LG of said other balancing network together and to an input terminal and means for connecting the In view of the fact that the tube capacities reend of inductance LG remote from the grid of said other tube network and the corresponding end of inductance LG of said one balancing network together and to the other input terminal, whereby voltage impressed between the output terminals produces no difference of potential either between the input terminals or across CGK of either tube and conversely voltage impressed between said input terminals produces no difference of potential either between the output terminals or across CAK of either tube in spite of the existence of inherent inductance in said leads.

2. An amplifying stage comp-rising a pair of input terminals and a pair of output terminals, a pair of tube networks forming a push-pull circuit each of said tubes including an anode, a cathode and a grid electrode, inherent capacity between the tube elements comprising CAG between said anode and grid, CAK between said anode and cathode and Cox between said grid and cathode, said tube networks including leads for connecting said tube elements in the circuit, inductance LK in the lead from each cathode, inductance LA in the lead from each anode and inductance Le in the lead from each grid, and wherein the following proportion issatisfied LK:LG=CAG:CAK, a pair of balancing networks composed of capacities and inductances substantially identical both as to theory and magnitude with said tube network capacities and inductances, whereby each of said balancing networks includes inductances LK, LA and LG, and capacities CAG, Cm and Cox,

means connecting together the ends of the inductances Lx of the tube networks remote from the cathodes and the corresponding ends of the inductances Lx of the balancing networks, means for connecting the end of the inductance LA remote from the anode of one tube network and the corresponding end of inductance LA of one balancing network together and to one of said output terminals, means for connecting the end of inductance LA remote from the anode of the other tube network and the corresponding end of inductance LA of the other balancing network together and to the other output terminal, means for connecting the end of inductance L of said one tube network remote from the grid and the corresponding end of inductance LG of said other balancing network together and to an input terminal and means for connecting the end of inductance LG remote from the grid of said other tube network and the corresponding end of inductance Le of said one balancing network together and to the other input terminal, whereby voltage impressed between the output terminals produces no difference of potential either between the input terminals or across CGK of either tube in spite of the existence of inherent inductance in said leads.

3. An amplifying stage comprising a pair of input terminals and a pair of output terminals, a pair of tube networks forming a push-pull circuit each of said tubes including an anode, a cathode and a grid electrode, inherent capacity between the tube elements comprising CAG between said anode and grid, CAK between said anode and cathode and Cox between said grid and cathode, said tube networks including leads for connecting said tube elements in the circuit, inductance L1; in the lead from each cathode, in-

ductance LA in thelead from each anode and inductance LG in the lead from each grid, and wherein the following proportion is satisfied LA:LK=CGK:CAG, a pair of balancing networks composed of capacities and inductances substan: tially identical both as to theory and magnitude with said tube network capacities and inductances, whereby each of said balancing networks includes inductances LK, LA and LG, and capacities CAG, CAK and Cox, means connecting together the ends of the inductances LK of the tube networks remote from the cathodes and the corresponding ends of the inductances Lx of the balancing networks, means for connecting the end of the inductance LA remote from the anode of one tube network and the corresponding end of inductance LA of one balancing network together and to one of said output terminals, means for connecting the end of inductance LA remote from the anode of the. other tube network and the corresponding end of inductance LA of the other balancing network together and to the other output terminal, means for connecting the end of inductance LG of said one tube network remote from the grid and the corresponding end of inductance LG of said other balancing network together and to an input terminal and means for connecting the end of inductance Lo remote from the grid of said other tube network and the corresponding end of inductance LG of said one balancing network together and to the other input terminal, whereby voltageimpressed between said input terminals produces no difference of potential either between the output terminals or across CAK of either tube in spite of the existence of inherent inductance in said leads.

WERNER BUSCHBECK. HANS PROST.

D l S C L A l M E R 2,159,731.Werner Buschbeck and Hans Prost, Berlin, Germany. NEUTRALIZING HIGH-FREQUENCY PUSH-PULL AMPLIFIERS. Patent dated May 23, 1939. Disclaimer filed December 1, 1939, by the assignee, Telefunlcen Gesellschajt fair Dmhtlose Telegraphie, m. b. H. Hereby enters this disclaimer to the three claims in said patent.

[Ofiicial GazetteDecember 26, 1939.]