US2185367A - Thermionic valve amplifying circuit - Google Patents

Description

2, 1940- A. D. BLUMLEIN THERMIONIC VALVE AMPLIFYING CIRCUIT Filed June 24, 1937 5 Sheets-Sheet 1 INVENTOR A. D. BLUML E/N )r q V M ATTORNEY Jan. 2, 1940. A. D. BLUMLEIN THERHIONIC VALVE AMPLIFYING CIRCUIT Filed June 24, 1937 3 Sheets-Sheet 2 INVENTOF A. D- BLUMLE/N ATTORNEY Jam 1940- A. D. BLUMLEIN THERMIONIC VALVE AMPLIFYING CIRCUIT 5 Sheets-Sheet 3 Filed June 24, 193'! INVENTOR A. 0. BLUMLE/N ATTORNEY Patnted Jan. 2, 1940 a va an UNITED STATES THEBMIONIC VALVE Alan Dower Blumlein,

AMPLIFYING cracui'r Eaiing, London, England,

assignmto Electric 1; Musical Industries, Limited, Hayes, Mlddlesex, England, a British conim r v Application June 24,

1937. Serial No. 150,033

In Great Britain July 4, 1936 8 Claims.

According to the present invention, in a pushpull amplifier circuit arrangement, provision is made to attest discriminative amplification-between push-pull and push-push potentials supplied to the amplifier, said discrimination being in favor or push-pull potentials. The potentials are supplied to the amplifier through a coupling which is capable of transmitting equally pushpull or push-push potentials, and the amplification of the push-pull potentials is rendered greater than that of the push-push potentials by the use of negative feed back of the push-push potentials.

In a particular circuit arrangement according to the invention a pair of thermionic valves have 20 of the impedance element or combination of im pedance elements being such that the amplification of the push-push potentials is less than that of the push-pull potentials. A circuit embodying the invention may include twostages of amplification each comprising a pair of valves having a single impedance element or combination of impedance elements common to the anode cirsuits of both pairs of valves.

The impedance element or combination of impedance elements may be connected between the cathodes of a pair of valves and earth or a portion of the impedance element or the combination may be connected between the cathodes of a pair of valves and earth, while another portion is connected in a common conductor leading to the anodes of the pair of valves. The impedance element may consist of a resistance and inductance coil, or a combination of these.

This invention relates to thermionic valve ampliiying circuit arrangements and has particular reference to push-pull amplifiers.

Amplifiers of this type are well known, the expression push-pull being used to describe the condition of potentials on a pair of conductors in. phase opposition relative to earth. In contrast the expression push-push is used to describe the condition of potentials on two conductors in the same phase with respect to earth. Currents and voltages in push-pull are often de-v scribed as being balanced with respect to earth, and currents and voltages in push-push asbeing unbalanced with respect to earth or the amplifier valves are said to be operating in parallel. Transformers are usually used to couple the stages of push-pull amplifiers, this form of coupling giving amplification only or the push-pull potentials and not of those in push-push. Transformer coupling is undesirable in some applications of push-pull amplification on account of distortion which may occur due to, the iron circuit, or again, in the case of an amplifier handling a wide range of frequencies. If a. capacitative or direct form of coupling is employed instead of transformer coupling, any push-push potentials due to unbalances which may be set 10 up and any signals in push-push applied to the input of the amplifier will be amplified. In such cases, also, a phase reversing valve or similar device must be used in order to obtain push-pull signals from a single input potential. Such 1 phase reversing arrangements are shown in, for example, British Patent No. 325,833 which describes circuits enabling a push-pull output to be obtained irom an unbalanced input.

According to the present invention in a pushpull amplifier circuit arrangement, the coupling arrangements to the input and/or from the output of atleast one amplifying stage serve to transmit push-pull and push-push potentials 'without substantial discrimination, and negative feed-back is applied for the purpose of lessening the amplification of the push-push potentials compared with the push-pull potentials. According to a further feature of the invention in a push-pull amplifier circuit arrangement, negative feed-back is applied in such a manner as to decrease the amplification of push-push potentials to a substantially greater extent than the amplification of the push-pull potentials.

At least one of the amplifying stages in a pushpull amplifier circuit may according to the invention consist of two valves or electron paths and an impedance associated with both valves or electron paths is connected so as to provide substantial negative feed back for the purpose of reducing the amplification of push-push potentials by the two valves or electron paths compared with the push-pull potentials. In particular circuit arrangements according to the invention, an impedance may be connected between 5 the cathodes of a pair of valves or electron paths operating in push-pull and a point at earth potential or an impedance may be connected in the anode circuit of a pair of valves or electron paths operating in push-pull, said latter im- 60 'pedance being'eiiectively included in the control circuit of said valves or electron paths. The impedance may consist of a resistance, an inductance or a combination of these.

In a push-pull amplifier circuit arrangement 66 embodying the invention and provided with D. C. couplings, a positive potential applied to the grids of the two valves of an amplifying stage is compensated for relatively to the cathodes of said valves by a resistance connected between the said cathodes and earth, said resistance serving to provide a suitable positive potential for said cathodes, and also to provide a negative feed-back which reduces the amplification of push-push potentials without reducing to the same extent the amplification of the push-pull potentials. Again, D. C. coupling may be obtained by connecting the grids of amplifier valves to tapping points on potentiometers bridged between a point at earth potential or a source of bias potential, and the anodes of preceding valves and the cathodes of said amplifying valves may be connected to a point at earth potential through a resistance which serves to make the cathodes of said amplifying valves more positive in potential than their grids, and simultaneously to provide negative feed-back for the purpose of reducing the amplification of push-push potentials without reducing to the same extent the amplification of push-pull potentials.

The invention may be applied to push-pull amplifier circuit arrangements including amplifying stages provided with means for neutralizing inter-electrode capacity, and again the invention may be applied to arrangements employing transformer couplings.

An amplifying circuit arrangement embodying the invention may for example be employed in converting a single input potential into a substantially push-pull output for supplying the input to a balanced cable. Again, an amplifier circuit embodying the invention may be used to amplify a source of push-pull potential having superimposed unwanted push-push potentials, such a source being constituted for example by a balanced cable subject to interference.

Further applications of the invention will be referred to hereinafter.

In order that the invention may be more clearly understood and readily carried into effect, some amplifying circuit arrangementsembodying the invention will now be described by way of example with reference to the drawings, in which Figures 1 through 6- show alternative forms of circuit embodying the invention.

Referring to Fig. 1 of the drawings, the input terminals l and 2 feed the control grids of valves 3 and 4. If these valves are similar and the input is purely push-pull no alternating current component will flow through a resistance 5 connected between earth and the cathodes of the valves 3 and l which are connected together. If the input potentials are not in perfect pushpull relationship, or if the valves '3 and 4 are not similar, an alternating current component will flow through the resistance 5 tending to reduce the push-push currents by negative feedback. If for example, the valves 3 and 4 have a magnification slope of three milliamps. per volt this will be effective for push-pull amplification, but if the resistance 5 has a value of ten thousand ohms, the slope of the two valves in parallel will only be about 0.1 milliamp. per volt for push-push signals. If the anode resistances 6 are of two thousand ohms in the case of each valve, the push-pull gain will be 6 as against 0.1 for the push-push channel, the effective anode resistance for push-push potentials being one thousand ohms. The resistance 5 need not have the relatively high value of ten thousand ohms in order that amplification of push-push signals may be reduced and a resistance having a value of the order of two thousand ohms will also be effective. The two valves 3 and 4 are coupled to two further valves 1 and 8 which also have a common cathode resistance 8 which reduces the push-push gain. In the latter pair of valves, resistances I0 and H are introduced separately in the cathode circuits. These resistances serve to straighten the characteristic curves of the valves 1 and 8 by providing negative feed back, and the effect of inserting these resistances is to reduce the amplification of signals in both the push-pull and push-push senses. It will thus be seen that while producing a negative feed back, the resistances l0 and H do not act in the same manner as the resistance 9. This resistance 8 is effective to produce negative feed back, which reduces amplification for push-push signals without materially affecting the amplification for push-pull signals. Instead of arranging the three resistances 9, l0 and il as shown in the drawings, they may be replaced by three resistances in delta formation, that is to say a resistance of comparatively small value may be connected between the two cathodes, and from each cathode to earth a comparatively high resistance may be connected. In such an arrangement it cannot be said that any one of the three resistances depreciates the push-push amplification without depreciating similarly the push-pull amplification. The two high resistances connected between the cathodes and earth have the effect of producing the same amplification or attenuation of both push-push and push-pull signals. The resistance connected between the two cathodes, however, removes the greater part of this negative feed back for the push-pull signals without removing the feed-back for the pushpush signals, so that the same ultimate effect is obtained. The anodes of the valves 1 and 8 are connected to output terminals I2 and I 3 which may be connected for example, to a cable through suitable coupling condensers. The usual coupling condensers and resistances have been shown in Fig. 1 and it will be understood that in this and in the subsequent figures to be described, the arrow heads at the ends of the resistances shown diagrammatically indicate that these resistances are connected to appropriate sources of high tension or grid biasing potentials as required.

In the case of the resistances GI and 62 in Figure 1, it may be necessary to connect their free ends each to a source of positive potential, because, in order that substantial feed back discriminating against push-push potentials may be produced, the values of the resistances 5 and 9 will have to be higher than the values required to provide normal grid bias potentials for the valves 3 and 4. 1

The circuit described may for example be used grease? 3 valves 1 and l, a portion may be included in the anode circuit. This can be effected by joining the upper ends of the anode resistances of the valves 1 and I to the upper ends or the resistances l in the anode leads of the valves 3 and 4, the common junction being connected to the high tension supply through a further resistanoe. The last mentioned resistance will, in such a case, produce negative feed back for phantom currents or push-push currents in the valves 1 and 3, since an increase of current in. these two valves produces a negative drop at the lower end of the added resistance, thus producing a negative potential on the grids of the valves 1 1g and 3 and tending to prevent the increase of current.

Such a modified arrangement is of advantage if the output from the valves I and 3 is taken from the cathodes instead of the anodes, and

such a further modified arrangement is shown in Fig. 2 of the drawings. In this drawing it will be seen that the separate anode resistances of the valves 1 and 3 of Fig. l are omitted, a common anode resistance 46 having been inserted.

23 Separate resistances 41 and 43 are, however, inserted in the cathode leads to earth, and the output terminals 12 and I3 are connected to the upper ends of these resistances.

Since the load is taken from the cathodes, the

common resistance in the cathode circuits would increase the efllciency for push-push signals, but the common load resistance 43 inserted in the anode circuit feeds back push-push voltages directly on to the anode circuits of the valves 3 33 and 4. It will be seen that triode valves are shown in place of the tetrode valves shown in Fig. 1 of the drawings.

In Fig. 3 01' the drawings an amplifier circuit particularly adapted for terminating a cable is shown. Input terminals 14 and ii are connected to the two conductors of a balanced cable. The wanted balanced signals arrive at the terminals l4 and IS in push-pull and are amplified by valves l6 and I1. Interfering voltages arriving 5 in push-push at the terminals i4 and ii are attenuated due to the insertion of a resistance It connected. in a similar manner to the resistance 5 shown in Fig. l of the drawings. Further attenuation of push-push potentials may be effected by adding further stages as in that figure, or as shown, the valves l6 and I1 may be coupled to a phase reversing valve is which serves to reverse the potentials from the valve l6 and adds them to the potentials from l1. The valve is has a resistance 20 in its cathode lead, this resistance having a value such that the gain of the valve i3 is unity. The anode of the valve is is coupled to an output valve 2| which may conveniently be of the cathode follower type 00 described in patent application No. 448,421. The

grid bias potentials for the valves l6, l1 and I9 have been shown as derived from the tapping points in the resistances l8 and 20 and although this is a convenient arrangement, it will be understood that this method is by no means essential. In the case in which the bias potential is derived from the resistance l8 for example,

a tapping is necessary since, in order to obtain substantial reduction of push-push amplification compared with push-pull amplification, the resistance l8 as a whole has to have a value greater than that required to bias the valve.

It will be understood that in the arrangement shown in Fig. 3 the push-pull output potentials to of the cable could be obtained to the exclusion or the push-push potentials by the use of a single valve such as the valve 13. However, the particular arrangement shown in Fig. 3 has the advantage that the relative attenuation of the push-push potentials given by the valves l3 and I 11, reduces the critical nature or the adiustment required for the valve I! for a given degree of balance. Again, the valves l3 and 11 are capable of handling comparatively large push-push potential inputs without overloading due to the fact 10 that for push-push potential inputs, their cathodes follow the potentials of their grids. It very large low frequency interfering input potentials are expected at terminals 14 and IS, a high inductance choke may be inserted at the lower end ll of the resistance 18 in order to raise this impedance for audible frequencies. Similarly, chokes may be employed in place of any of the resistances I, 3, I3, 20 and 43, although it the circuit is to handle a wide range of frequencies, I

for example, or the order employed in television systems, series resistances will also be required it a low degree of amplification of the push-push potentials isrequired over the whole range of frequencies.

In Fig. 4 of the drawings the invention is shown applied to a circuit arrangement suitable for supplying the scanning potentials to a cathode ray tube in a case in which it is required to apply substantially push-pull potentials to the Q.

deflecting plates. The anodes of two tetrode valves 22 and 23 are connected to two potentiometers 24 and 25 connected across the source of high tension supply for a cathode raytube, one pair of deflecting plates 26 and 21 of which are I. shown. The potentiometers 24 and 25 serve also to constitute the anode loads of the valves 22 and 23. In a particular example, if the source of high tension supply is four thousand volts,

the potentiometers 24 and 25 may each be of U eight megohms. The anodes of the valves 22 and 23 are supplied from tapping points 28 and 29 half way along the potentiometer resistances. If each anode resistance draws 0.5 milliamp. of

current, the anode potentials in each case will u,

be one thousand volts and the effective anode load will be two megohms. The screening grids of the valves 22 and 23 are supplied from a separate source not shown in the diagram. The cathodes of the valves 22 and 23 are connected 3 together and have a large common resistance 30 which may be of one hundred thousand ohms. Thus, if the screening grids also take about 0.5 milliamp. each, this will give a cathode potential of two hundred volts. Thecontrol grids of the valves 22 and 23 are fed by condensers and leak resistances, the latter being fed from tapping points at slightly less than two hundred volts in the potentiometer resistances 24 and 25. Should either of the valves 22 and 23 tend to take. less steady anode current than the other, the grid potential of that valve will rise and tend to equalize the steady currents. The grid condenser 3! of the valve 23 is earthed, and the grid condenser 32 of the valve 22 is fed with a scan- 55 ning potential wave such as a frame frequency saw-tooth. Theresistance 30 reduces the slopes of the valves in push-push to about 0.01 milliamp. per volt, as against for example, 0.2 milliamp. per volt for each valve in push-pull. The 70 resultant output on the scanning plates 26 and 21 is chiefly push-pull since the push-pull gain will be forty times the push-push gain. The scanning plates 26 and 21 are fed through condensers 33 and 34 with very high leak resistances 7 85 and 38 connected to the high tension source of supply.

In Fig. 5 of the drawings, two triode valves of a power amplifier connected together in accordance with the invention are shown. A resistance 31 serves to reduce the amplification of the pushpush potentials applied to input terminals '38 and 38. In order to neutralize the anode to grid capacities of the triode valves 40 and 4|, condensers 42 and 43, equal in value to the anode grid capacities are connected as shown. Given equal anode leads for the valves 40 and 4| the condensers 42 and 43 serve to balance out the effects of the anode to grid capacities for pushpull operation. In the case of push-push operation, however, the added condensers only increase the chances of instability occurring at high frequenoies due to stray inductance of anode load impedances. The introduction of the resistance 31, decreases the push-push gain and thus increases the stability. The resistance 31 also tends to keep the valves in balance and so improves the neutralizing balance if the valves are not similar. The output potentials from the valves 40 and 4| are fed to output terminals 44 and 45.

This invention, while providing a means of discriminating against push-push amplification, also provides a means of facilitating D. C. coupling in push-pull amplifiers. As is well-known, when providing D. C. couplings, it is necessary to provide some means for correcting for ,the positive potential which is applied to the grid of a valve when this grid is connected directly to the anode of a preceding valve.

The application of the invention to a direct coupled push-pull amplifier may be illustrated for example with reference to Figure 1 of the drawings. Thus, in this case, the coupling condensers in the leads to the control grids of the valves 3 and 4 and 1 and 8, and also the grid leak resistances of the valves 1 and 8 are omitted. In consequence, a large positive voltage is applied to the grids of the valves 1 and 8 and it is necessary to raise the cathodes of the valves 1 and 8 to a slightly higher potential so that the control grids will be biased negatively with respect to these cathodes. This can be done by making resistance 9 of a sufilciently high value that at normal values of applied voltage, the cathodes of the valves 1 and B are more positive than the'anodes of the valves 3 and 4, unless it is required to run the valves 7 and 8 into grid current which is practicable with the direct connection to the anodes of the valves 1 and 8. In this latter case the cathodes of the valves 1 and 8 may have about the same potential as the anodes of the valves 3 and 4. The resistance 9 then serves the dual purpose of providing positive cathode potential and negative feed-back for push-push potentials.

With these arrangements it is necessary to supply a high voltage for the valves 1 and 8, since the voltage applied to the anode and screen grids must be selected with regard to the cathode potential which is already as high as that of the anodes of the valves 3 and 4. The direct coupling arrangements may be applied to a power amplifier in which the high voltage for the successive stages rises due to the use of larger tubes in the successive stages. An arrangement for avoiding excessive high tension voltages while still maintaining the D. C. couplings will now be described with reference to Figure 6 of the acaccompanying drawings. In this case it is as sumed that an unbalanced D. C. input is applied at the terminal I and the input terminal 2 is connected to earth through a condenser 49 and a leak resistance 50, to a suitable source of potential corresponding to a reference potential at the terminal I. Decoupling condensers a and decoupling resistances b are inserted in the anode leads of the first pair of valves 8 and 4. These valves are coupled with valves I and 8 through condensers c shunted by resistances d and grid leak resistances e. These components are proportioned in such a manner that the ratio a d e is the same as the ratio 0 b 6, so that a true'D. C. coupling is obtained in the manner described in the specification of British Patent No. 456,450.

Owing to the insertion of the cathode resistance 9 the cathodes of the valves 1 and 8 are of sufiicient high positive potential to enable the D. C. coupling to be effected in spite of the positive potential applied to the grids of these valves through the D. C. potentiometer formed by resistances d and c.

Any difiioulty experienced in obtaining the correct operating conditions can be overcome by applying suitable positive or negative bias potentials to the lower end of each of the resistances e. Such bias potentials may be made slightly different for the two valves so that the feeds may be balanced for a given mean working position. Decoupling resistances SI and condensers 52 are included in the anode leads to the valves 1 and 8 so that the output terminals I 2 and 13 may be connected directly through direct coupling circuits to further valves.

The arrangement described with reference to Fig. 2 of the drawings, or such an arrangement having additional stages, forms a very convenient limiting amplifier. Thus, for example, assuming that a badly shaped square topped wave-form is applied to the terminal I, and if the input is of sufiicient amplitude, the grid of the valve 3 will be driven negatively during the downward swing to such an extent as to render the valve 3 non-conductive, the wave-form on one side being made more truly square. Similarly, the positive swing on the grid of the valve 3 may be sufficient to enable this valve to pass the whole current flowing through the cathode of resistance 5, in which case the valve 4 will be rendered non-conductive, tending to render the other side of the wave-form more square. The amplified signals passing to the valves 1 and 8 will operate similarly on those valves, thus limiting the wave-form still further.

It is a characteristic of the amplifier described that it can be arranged that overloading does not cause grid current to flow. The discriminative amplification in favor of push-pull signals only continues while both valves of a pair are conductive. As soonas one of the valves ceases to conduct due to an excessivenegative potential swing on the grid, the other valve is placed in series with the following push-push feed back resistance which is operative to reduce the pushpull amplification so that as soon as one valve ceases to conduct, there is a very little push-pull amplification. Thus, by operating the valves at an average position on their characteristic curves so that the valves can pass more than double the normal input without grid current being caused to fiow, it is possible to overloadthe amplifier for each direction of potential swing without grid current flowing, the overloading being produced by one or other of the valves being rendered nonconductive, with the consequence that a large negative feed back is applied to the valve which is still conducting, thus limiting the push-pull amplification.

The D. C. coupling amplifier described may thus be used with advantage for correcting a badly shaped square topped wave-form, or if the input signals are intermittent or very unsymmetrical about the mean A. C. value.

The arrangement described with reference to Fig. 1 of the drawings is suitable for use with a square topped wave-form symmetrical about a mean A. C. value.

The arrangement of. Figure 6 of the accompanying drawings also may be used conveniently for limiting the amplitude of signals amplified.

The D. C. couplings ensure that the limiting amplitudes bear a fixed relationship to the absolute input potential.

The overloading feature may be used for limiting the amplitude of spurious signals or atmospheric disturbances. For example, an incoming radio signal subject to strong atmospheric disturbances of brief duration may conveniently be passed through an amplifier to limit the amplitude of the disturbances without any risk of grid current being produced thereby.

A further application, lying in the television field, consists in limiting the amplitude of spurious pulses set up when the electron beam of a cathode ray tube is returned from one end of a line to the beginning of another. It is of advantage to be able to limit the amplitude of. these pulses during the amplification of the vision signals, and by passing the signals through an ampiifier of, the type described with reference to Fig. 1 of the drawings, limitation may be effected.

Various other alterations and modifications may be made in the present invention without departing from the spirit and scope thereoi, and it is desired that any and all such modifications be considered within the purview of the present invention except as limited by the hereinafter appended claims.

Having now particularly described and ascertained the nature of my said invention, and in what manner the same is to be performed, I declare that what I claim is:

1. An amplifier circuit comprising a pair of thermionic discharge tubes, each including a cathode, a control electrode and an anode, means including a resistance for connecting the oathodes to a negative source of potential, means for maintaining the control electrodes at a predetermined potential, means including a resistance connected to each anode for maintaining the anodes. positive with respect to said cathodes, a condenser connected between the anode of each tube and the control electrode of. the other tube, said condensers being of a value equal to the magnitude of the anode-control electrode ca pacity of the tubes, means for applying an input signal between the control electrode and another electrode of one of the tubes, and an output circuit connected to the other tube.

2. An amplifier circuit comprising a pair of electronic discharge tubes each including at least a cathode, a control electrode and an anode, means including a resistance for connecting the cathodes to the negative terminal of a source of current, means including a resistance connected to each anode for maintaining said anodes positive with respect to said cathodes, a pair of series resistances connected between each anode and the negative terminal of the source of current, means including a resistance for connecting the control electrode of each tube to the junction of: the associated pair of series resistances, means including a condenser for connecting the control electrode of one of said tubes to earth, means for connecting the control electrode of the other tube to a signal source, and an output circuit connected to the anodes of said tubes.

3. An amplifier circuit comprising a pair of electronic discharge tubes each including at least a cathode, a control electrode and an anode, means including a resistance for connecting the cathodes to the negative terminal of a source of current, means including a resistance connected to each anode for maintaining said anodes positive with respect to said cathodes, a resistance connected between each anode and the negative terminal of the source of current, means including a resistance for connecting the control electrode of each tube to a point along the assoelated last named resistances, means including a condenser for connecting th control electrode of one of said tubes to earth, means for connecting the control electrode of the other tube to a signal source, and an output circuit connected to the anodes of said tubes.

ALAN BLUMLEIN.

Images