US2289666A - Radio amplifier - Google Patents

Description

July 14, 1942. 1., MAGUIRE 2,289,666

I RADIO AMPLIFIER I Filed Nov. 16, 1938 5 Sheets-Sheet 1 c I, I o

i mvswron c 9 LLfM AGUlRE By 3 3 ATTY'S,

July 14, 1942. 6 l. L. MAGUIRE 7 2,289,666

RADIO AMPLIFIER Filed Nov 16, 1938 5 Sheets-Sheet 2 INVENTOR LL. MAGUIRE Q JI W A TTYS.

y 1942- I. L. MAGUIRE 2,289,666

RADIO AMPLIFIER Filed Nov. 16, 1938 5 ShetS-Slniet 3 INVENTOR I. L. MA GUIRE ATTYS.

J y 1942- I. L. MAGUIRE 2,289,666

RADIO AMPLIFIER Filed Nov. 16, 1938 5 Sheets-Sheet 4 lNVENTOR I. L. MAGUIRE ATTYS.

July 14,1942. L. {JIAGUIRE RADIO AMPLIFIER Filed Nov. 16, 1938 5 Sheets-Sheet 5 INVENTOR I.L.MAGU|RE YWPW) ATTYS.

Patented July 14, 1942,

- Irwin Leonard Maguire, Elwood, Victoria,

Australia Application November 16, 1938, Serial No. 240,823

. In Australia November 26, 1937 8 Claims. '(Cl. 179-171) This invention relates to radio amplifiersand' the like and it has been devised to enable signals to be faithfully reproduced over a wide range of frequency with a relatively high-signal gain and low current consumption.

The invention consists in the use of substan tially constant impedance and/or resistance networks for theflow oi current from the anode of one valve to the grid of its succeeding valve so that the voltages impressed on each of the valves may be proportional to,'or relatively proportional to, the input signal voltage on the input. grid of the first valve and so that the feed back coupling between thevalves may be reduced to an amount which is less than that due to the serial inter electrode capacity of the sum of the anode to 4 grid elements of the valve because the networks connecting the valves increase in impedance with frequency for feed back of energy from the grid of a valve to the anode of its preceding valve. The alternating output [voltage of the valves which is impressed. on the valve gridis arranged to be proportional to the signal input voltage by the use of constant impedance and/or resistance networks for the currentsandby arranging network output tappingpoints connected to the.

grids and/or loads so that the-potentials of these points may be directly proportional to the voltages impressed upon the networks.

The tapping point for the connection between I the anode load of a valveto the grid of a succeeding valve or to a load actuated 'by the amplifier is arranged between the condensers so as to isolate the relatively high voltages of the H. T.,

D. C., impressed on the anode load, from the grid and so that the voltage impressed upon the grid leg of this branch in either direction from this tapping point is a simple ratio'and nota surd quantity. I

The ba'ck coupling impedance, to prevent feed back of energy from a valve to its preceding valve, is increased by arranging the tapping point for the grid to be on the inductive branch of the grid leak network. The tapping point in the branch may have an inductance" and resistance leg on either side of the tapping point and while the impedance of a grid leak network is constant for energy flowing from the valve'through the grid leak network to the grid of the succeeding valve the impedance of the grid leak network to feed-back of energy in the reverse direction increases with frequency, The voltage impressed upon the grid leak resistance of the valve may be arranged to be proportional to the alternating voltage impressed upon the anode load by devising the constant impedance networks so that relatively small changes if any in the impedance of the anode load network are introduced by the connection to it of the grid leak network or resistance and by choosing a tapping point on the network for the grid and grid leak resistance or for the grid leak network connection where the potential is leak resistance may be proportional to the alter nating voltage impressed upon a valve which immediately'precedes-the grid leak resistance.

The impedance or resistance to feed-back ofv energy from a valve to its preceding valve may be increased to an extent greater than that provided by the small interelectrode capacity of the anode to grid of the valves, by providing suitable resistances and inductances in a network, hereinafter referred to as the grid leak network in which is the grid leak resistance.

The voltages impressed upon the grid of the valve may be made proportional to the alternating voltage'impressed on the anode load by tapaping a point in the network where the potential is always proportional to the voltage across the grid leak network by placing the tapping point in a branch of the network where the ratio of the impedance of the branch to the impedance of the simply proportional to the voltage impressed upon the anode load'to which it is connected, in

' lieu of, a tapping point where its potential is related to the voltage impressed on the network by'some surd quantity which varies with frequency. This may be done by making the impedance or resistance of the grid leak network or resistance large compared to the impedance or resistance of the branch of the constant impedance or resistance anode load to which it is connected. Alternatively the grid leak network or resistance may be connected between two condensers in a type of constant impedance anode load where there is reactance only in a branch of the network, one of the condensers maybe large compared to the other and the relatively large condenser is connected to the end of the branch of the network across which the impedance of the network is constant, thus the potentialof the grid leak network or resistance connected to the relatively large condenser may bepractically the same as the potential at the end of the constant impedance network to which the relatively large condenser is connected. By connecting the grid leak resistance in this manner the potentiometer efiect of connecting a valve grid between a condenser and grid leak resistance is practically eliminated.

drawings wherein:

Fig. 1 shows a simple circuit whose voltage output is independent of frequency, which is suitable where the amplification and range of frequencies is not too great.

Fig. 2 shows a circuit using a more complicated coupling network, suitable for greater amplification and range of frequencies.

Fig. 3 shows a modified form of coupling grid network whose impedance is independent of frequency if proper values are chosen, having one branch containing resistance and capacity in series and another parallel branch carrying the output lead with series inductance and resistance on each side of the lead.

Fig. 4 shows another modified form of network of similar characteristic, having one branch containing resistance and inductance in series and another parallel branch carrying the output lead with series capacity and resistance on each side of the lead.

Fig. 5 shows a further form of network of similar characteristic having one branch with two resistances in series and a second parallel branch with capacity and inductance in series with capacity cross connection between points intermediate the elements in each branch.

Fig. 5a shows an amplifier circuit using a network unit of a type similar to that of Fig. 7 succeeded by a network similar to Fig. 3 with the addition of cross connections from spaced points in one branch to the cathode lead.

Fig. 6 shows a two-unit network whose impedance is independent of frequency, if suitable values are chosen, with each unit having two parallel branches.

Fig. 7 shows a network with three branches one of which contains only capacity, whose impedance is independent of frequency if suitable values are chosen, and which constitutes a part of the circuit of Fig. 1.

Fig. 8 shows an amplifier coupled by the network of Fig. 4 succeeded by the network of Fig. 3.

Fig. 9 shows an amplifier coupled by the network of Fig. 4 succeeded by pure resistance.

Fig. 10 shows a push-pull amplifier circuit using a network unit of a type similar to that of Fig. 7 succeeded by a network of the type of Fig. 3.

Fig. 11 shows a push-pull amplifier circuit using a network unit of a type similar to that of Fig. 7 succeeded by pure resistance.

Fig. 12 shows the equivalent circuit through which back coupling would take place for the circuit of Fig. 8.

Fig. 13 shows the equivalent circuit through which back coupling would take place for the circuit of Fig. 9.

Fig. 14 shows the equivalent circuit through which back coupling would take place for the circuit of Fig. 2, and

Fig. 15 shows an arrangement alternative to the network of Fig. 1, the inductances being arranged in series additively.

In practical embodiments of the invention there may be provided the types of constant impedance networks shown in Figs. 3, 4, 5, 6 and 7 and marked 4, 5, 6, I and 8 respectively. These networks may be used in combination with one another or with resistances and any particular one of the circuits of Figs. 3, 4, 5, 6 and 7 may contain within itself any one or more of these circuits for use as a constant impedance which is equivalent to useof an ohmic resistance.

It will be evident from an inspection of Fig. 3, that if k equals ratio of potential difference between points A and B to potential difference between B and D and if this ratio is a simple numeral then Len equals kL and Run equals kit and LAB equals (1k)L and RA]; equals (1k)R where R equals the resistance in series with C and RP equals L/C.

Similarly with the circuit shown in Fig. 4 if potential difference between points B and D is to be some simple proportion of the potential V i. e. the potential between points A and D, then r: equals kit and and also r1 equals (1k)R and where R equals resistance in series with the inductance and C equals It is also apparent that the circuit shown in Fig. 5 is characterised by being a constant impedance if and that, without altering the constant impedance characteristic of the circuit, that connection may be made at D and A or D and g.

In Fig. 7 the circuit is characterised by being of constant impedance if and includes a blocking condenser Ca, which, if made large in comparison to C, renders the circuit for practical purposes of constant impedance and the potential differences between either side of Ca and the point A are the same. This point will be better appreciated if we assume a case for Fig. 7 where C equals 7 mmfd. and Ca=1 mf., the ratio of the reactance of C to Ca equals approximately.

Figs. 3, 4, 5, 6 and 7 show preferred constant impedance networks but to those skilled in the art it will be understood that where no direct current component is present the blocking condensers C", C and C and C in Figs. 5, 6 and 7 respectively may be omitted.

Applications of the invention in the circuits are hereinafter described with reference to the diagrammatic drawings accompanying this specification.

Where the amplification and the range of frequencies of signals is not too great the circuit shown in Fig. 1 may be used. The circuit is self explanatory- A constant impedance circuit I, see Fig. 7, is used in series with a resistance H for the anode load. G is a grid leak resistance connected at point B of Fig. 7.

Where the range of frequencies of signals and the amplification is greater than in Fig. 1, the circuit shown as Fig. 2 may be employed. Fig. 2 shows alternative arrangements for the grid leak network and for the anode loads. The circuit for the valve on the left utilises the anode load as at 8 Fig. 7, in series with a resistance H and uses a grid leak network consisting of circuits shown as 4 see Fig. 3 in series with a resistance G and on the right is shown a circuit employing a grid leak network as shown in Fig.. 3, the resistance G being omitted. The circuit shown on the left enables the use of a smaller total inductance in the grid leak network than is required for the valve on the right but its impedance to feed back of energy between valves is not so great. Fig. 14 shows how the resistance to feed back of energy for circuit shown in Fig. 2 is increased by an amount which may be greater than that due to the capacity of condensers Cga which represent the interelectrode capacity between anode and grid of a valve. This may be better appreciated if it is remembered that the value of G is .2 to 1 megohm, and H is .2 megohm in the valve circuit. On the right of Fig. 2, the resistances of Fig. 3 are about 1 megohm and in 8 shown in Fig. 7 the resistance His .2 megohm, the value given being suppositious.

, Other embodiments are shown in Figs. a, 8, 9, 10 and 11. c

Any of the circuit arrangements described may be arranged for push pull connection of valves, as is usual the circuits being symmetrically duplicated' upon either side of the negative of the H. '17., lead and the anode loads of the valves being joined together symmetrically to the positive H. T. lead. As examples see Figs. 10 and 11 which are preferred embodiments.

It is apparent that if the resistance of the grid leak network 4 (see Fig. 8) is large compared to the anode load circuit to which it is connected, that the branch of, this network, previously described with reference to Fig. 3 and consisting of a resistance R be omitted.

It will be apparent that in any of theinductive branches of the constant impedance networks that if the inductances are connected so as to be 4 series aiding then the sizes of the winding for the chokes may be reduced since in this case the resistance of the network may be where L is the inductance of each winding and M is the mutual inductance of the windings and C is the capacity of the condensers oi the network. As an example of this type of circuitsee Fig. 15.

in series with a capacity C, may

although the coupling arrangements have been described with reference to circuit arrangements for amplifying electric oscillations including the coupling of the anode of one tube to the grid of the next, the inductive and/or capacitative load of the next succeeding device, as for example the voice coil of aloud speaker, may be combined with suitable resistances and/or inductances and/or capacities to form a constant impedance anode load for the output tube of the amplifier. In the following claims, it is to be understood that the expression constant impedance" network includes constant resistance networks and networks of constant impedance and resistance.

I claim: c

1. A circuit arrangement for amplifying electric oscillations, comprising a plurality of therm-, ionic discharge tubes, means for producing a constant impedance load, said means comprising a pair of networks of constant impedance coupling the anode and grid of related tubes, each network consisting of two branches and having an output tapping point dividing one of said branches in each networkinto a pair of branch legs, each of said tapping points being located on its respective branch where the ratio of impedance in that branch to the impedance of the leg of that branch in either direction from said point is a simple ratio, one of said branches including an inductance and resistance leg on either side of its said tapping point.

2. A circuit arrangement for amplifying electric oscillations, comprising a plurality of thermionic discharge tubes, a grid leak for at least one of said tubes, means for producing a constant impedance load, said means comprising a pair of networks of constant impedance coupling the connected to a grid leak connected to one of said tubes and inwhich the impedance of the lastmentioned network is large compared with the impedance of the branch of the anode coupled constant impedance network connected tothat anode load.

3. A circuit arrangement for amplifying electric oscillations, comprising a plurality of thermionic discharge tubes, means for producing a constant impedance load, said means comprising a pair of networks of constant impedance coupling the anode and grid of related tubes, each of said Fig. 15 shows an alternative arrangement to the network 8 in Fig. 1, the inductances being arranged so as to be series aiding.

Fig. 12 shows back coupling circuit for Fig. 8; Fig. 13 showsback coupling circuit for Fig. 9; Fig. 14 shows back coupling circuit for Fig. 2, and Fig. 5a illustrates a circuit incorporating the network of Fig. 5.

From the foregoing it will be apparent, that points located therein, said inductance branch in one of said networks being connected to a grid leak connected to one of the tubes.

4. In an interstage amplifier circuit, an input source connected to the anode of a first tube, an

output load connected to'a grid of a succeeding tube, an input network unit and an output net- I work unit interconnected and each being of constant impedance at all frequencies and each comprising two branches in shunt respectively connected across the terminals of said input source and output load, a first branch of the said output unit being connected to said grid of said succeeding tube and comprising inductance and resistance and further comprising the grid leak of said succeeding tube, said grid being connected to a tapping point in said first branch, said input unit being connected to said anode and having a tapping point in one of its branches whereby it is connected to a common point of the branches of said output unit, each said tapping point being located on its respective branch where the ratio of total impedance of said branch to the impedance of each leg of its branch in either direction from said tapping point is a simple ratio.

5. A circuit according to claim 4, the imped-' ance of said output network unit being large compared to the impedance of the branch of the input network unit to which said output network unit is connected.

6. In an interstage amplifier circuit, an input source connected to the anode of a first tube, an output load connected to a grid of a succeeding tube, an input network unit and an output net-' work unit interconnected and each being of constant impedance at all frequencies and each comprising two branches in shunt respectively connected across the terminals of said input source and output load, a first branch of the said output unit being connected to said grid of said succeeding tube and comprising inductance and resistance and further comprising the grid leak of said succeeding tube, said grid being connected to a tapping point in said first branch, said input unit being connected to said anode and having a tapping point in one of its branches whereby it is connected to a common point of the branches of said output unit, each said tapping point being located on its respective branch where the ratio of the total impedance of its entire branch to the impedance of each leg of its branch in either direction from said tapping point is a simple ratio, at least one of said units having a first branch consisting of resistance and of a first reactance of one given sign connected in series, and further having a second branch comprising two legs in series separated by the tapping point of said unit, each of said legs comprising a resistance in series with a reactance of sign opposite to said first reactance, the said resistance in said first branch being in value equal to the square root of the ratio of the total inductance in one of said branches to the total capacity in the other of said branches.

'1. In an interstage amplifier circuit, an input source connected to the anode of a first tube, an output load connected to a grid of a succeeding tube, an input network unit and an output network unit interconnected and each being of constant impedance at all frequencies and each further comprising two principal branches in shunt respectively connected across the terminals of said input source and output load, a first leg of said output unit being connected to said grid of said succeeding tube and comprising inductance and resistance and further comprising the grid leak of said succeeding tube, said grid being connected to a tapping point in said first branch, said input unit being connected to said anode at a first common point of its two principal branches and having a tapping point in one of its principal branches, said input network unit further comprising an auxiliary branch extending between the second common point of the principal branches of said input unit and said tapping point of said input unit and comprising two condensers connected in series, the said condenser connected to the said second common point having large capacity relative to the other said condenser, a connection between a point in said auxiliary branch intermediate said condensers and a common point of the branches of said output unit, each of said tapping points being located on its respective branch where the ratio of the total impedance of its entire branch to the impedance of each leg of its branch in either direction from said tapping point is a. simple ratio.

8. A circuit according to claim 7, at least one of said units having a first branch consisting of resistance and of a first reactance of one given sign connected in series, and further having a second branch comprising two legs in series separated by the tapping point of said unit, each of said legs comprising a resistance in series with a reactance of sign opposite to said first reactance, the said resistance in said first branch being in value equal to the square root of the ratio of the total inductance in one of said branches to the total capacity in the other of said branches.

IRWIN LEONARD MAGUIRE.

Images