|Publication number||US2429124 A|
|Publication date||14 Oct 1947|
|Filing date||12 Apr 1944|
|Priority date||12 Apr 1944|
|Publication number||US 2429124 A, US 2429124A, US-A-2429124, US2429124 A, US2429124A|
|Inventors||Cunningham Frederick W|
|Original Assignee||Arma Corp|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (7), Referenced by (9), Classifications (15)|
|External Links: USPTO, USPTO Assignment, Espacenet|
Oct. 14, 1947. v F. w. CUNNINGHAM ELECTRICAL AMPLIFIER Filed April 12, 1944 IN VEN TOR.
Patented Oct. 14, 1947 ELECTRICAL AMPLIFIER Frederick W. Cunningham, Brooklyn, N. Y., as-
signor to Arma Corporation, Brooklyn, N. Y., a corporation of New York Application April 12, 1944, Serial No. 530,589
This invention relates generally to electric amplifiers, has particular reference to booster type electronic tube amplifier units, and is especially adapted for supplying power to a low impedance load varying with a signal component which is available from a high impedance source, which source cannot of itself supply the electric power for energizing such load without being undesirably affected by the load.
Ordinarily, if a voltage from a high impedance source is applied to a low impedance load, the phase and magnitude of the supply voltage from the high impedance source are altered. In signalling where accurate signal transmission is required, as in computing apparatus, the result of such variations in phase and magnitude creates a source of error which must be compensated, but exact compensation is frequently difiicult to attain in cases where signal voltages vary widely and rapidly. If the voltage applied to a low impedance load could be balanced against the supply voltage from a high impedance source so that they have the same phase and magnitude at all times, the desirable operating characteristics of an amplifier for accurate signalling purposes would be attained, and it is the principal object of this invention to provide such amplifiers.
The invention comprises a semi-direct coupled electronic booster amplifier employing one hundred per cent inverse voltage feed-back, in which the signal voltage is amplified in an initial stage, supplied to a phase inverter through re" sistance coupling producing two signals 180 out of phase, which are fed through the aforementioned semi-direct coupling to a push-pull ampliier stage whose output is impressed upon the primary Winding of an output transformer, Whose secondary voltage is fed back to the input grid of the first stage to eiiect inverse voltage feed-back. Inverse voltage feed-back in the power output stage is employed to lower the output impedance of the amplifier by reducing the plate impedance of the last stage tubes, and high frequency oscillation is avoided by the use of condensers, principally in the output transformer primary circuit, whereas the hum level with zero signal is kept low by the use of filter networks in the initial* been obtained with the new amplifier. Also, high frequency oscillation has been eliminated for wide limits of load in the amplifier, which is also devoid of low frequency oscillation or bounce and still tolerates a high source impedance without becoming unstable. Other advantages will become apparent from the following description of the preferred construction of the amplifier and its field of use, taken in connection with the accompanying drawings, in which:
Figure l is a circuit diagram of the amplifier unit of this invention; and
Fig. 2 is a schematic diagram illustrating a method of application of the amplifier of this invention.
Referring to Fig. 1 of the drawings, the booster amplifier unitA 8 comprises an inputv circuit IIl and an output circuit I2 having a single common lead or conductor I4. The amplifier unit 8 also includes a pair of twin triode space discharge devices or tubes I6 and I8, such as of the GSC? type, a pair of beam power space discharge devices or tubes 20, such as of the 6V6 type, and a dual-secondary transformer 24. The tubes I6-2ii have cathode heating elements 26 provided with input terminals which are connected respectively to any suitable electricsupply source, not shown.
A suitable source of plate voltage, not shown, having a positive terminal B+, is connected between the respective anode or plate circuits of the four tubes, and a common negative conductor or ground. The plates 30 and 32 of the twin triode tube I6 are connected to the positive terminal B+ through a common circuit 34 including filter network resistors 36, 38 and 40 and divided circuits 42 and 44 including resistors 46 and 48, respectively. The plate 58 of the twin triode tube I8 is connected to the terminal B+ through an anode circuit 52 including a resistor 54, conductor 56, and the resistors 38 and 40. The other plate of the tube I8 is likewise connected to the terminal B+ through an anode circuit 6D including a resistor 62, the conductor 56, and resistors 38 and 46.
The plates 62 of the beam power tubes 26 are connected through conductors 66, 66, respectively, to opposite terminals of the primary winding 10 of the transformer 24, the Winding 'I0 having a center tap 6B connected to the positive plate Voltage supply terminal B+ by a conductor l2.
The twin triode I6 is used as the first amplifier stage in the amplifier unit 8. The control grid 14 of such triode is connected to a signal supply circuit through a lead 'I6 provided with a shield 1l, which may be grounded through a suitable resistance, or as is shown, connected by a conductor 'I8 to the common input-output lead I4, which is grounded through a high resistor 'I9 across the output winding I9 of transformer 24. Capacitance between the input grid circuit 'I6 and ground is thus avoided. If not avoided, such capacitance would be the cause of high frequency oscillation, which is highly undesirable. The other grid 88 of the twin triode I6 is connected directly to ground through lead 8|. The common cathode 82 of twin triode IB is connected to ground through a suitable cathode resistor 83.
The lter network, comprising resistors 36, 38 and 40, also includes suitable condensers 84, 86 and 88 connected between the common circuit 34 and ground, the arrangement being such that such filter network functions, in the plate circuits of the twin triodes I6 and I8, to maintain a low hum level with zero signal. No ltering is needed in the push-pull power output stage.
The control grid of twin triode I8 is connected to the output circuit 44 of twin triode I5 through an input circuit |32 containing a condenser I34, the grid |80 being connected to ground through a resistor |06. The common cathode |88 of such tube I8 is likewise connected to ground through a resistor I I8. The output plate 58 of the first triode amplier stage of the tube I8 is connected through a condenser ||4 to resistors IIG and IIS, the control grid i 20 of the second triode of tube I8 being connected by a lead |22 to the same at a point between the resistors IIS and I i8. The tube I8 thus acts as a phase inverter by virtue of such resistance coupling. For balanced push-pull operation of the amplifier, the resistance value of resistor IIS should be equal to the sum of the resistance values of resistors ||6 and ||8 divided by the gain of the rst triode of the twin triode I8. Two amplified signals 180 out of phase are thus obtained in the output circuits 52 and 68.
The output circuits 52 and 60 are coupled to the input circuits |24 and |26 of the beam power tubes 28 according to the invention by novel semi-direct couplings, each including a coupling condenser |28 and a resistor |30 shunted across such condenser |28. Such semi-direct coupling |28, |38, among other advantages, minimizes low frequency oscillation.
Balancing resistors |32 and |34 are connected between the input circuits |24 and |26, respectively, and ground. A series circuit containing a resistor |36 and a condenser |38 is connected from the input circuit |24 to ground. The impedance values of resistor |38 and condenser |38 are such as to cause a reduction in the gain for relatively high signal frequency values without phase shifting of the translated signal, in the same manner as the semi-direct coupling |30, |32 functions at relatively low values of signal frequency.
The beam power tubes 28, 20 are connected in push-pull relation, the control electrodes or grids |40 of such tubes being connected to the input leads |24 and |25. The anodes 62 are connected to opposite terminals of the transformer primary l0 through conductors 55, as pointed out above, while the cathodes |48, |43 are connected through conductors |48, |48 containing resistors |50, I 50 t0 opposite terminals of the transformer secondary winding |I. The secondary winding |5| is provided Iwith a center tap which is grounded through a resistor I 52.
A circuit |54 containing condensers |56 and |58 is connected across the primary winding 'I0 of the transformer 24 for the purpose of avoiding high frequency oscillation with the screen grids |60, |60 of the tubes 20 connected to a conductor I 62 which is connected between the circuit |54 and the circuit '|2, which supplies the screen grids |60 with the full B-I- voltage direct from the source.
The secondary winding |64 of the transformer is included in the output circuit I2 of the amplifier unit 8, and the output voltage from such winding is fed back to the grid 'I4 of the twin triode I6, 180 out of phase with the input required voltage. The booster amplifier unit 8 employs inverse voltage feed-back.
In considering the operation of the amplier unit 8, a signal voltage Vs is impressed on the grid I4 of twin triode where it is amplified. In order to get maximum gain at this point a unique hook-up of the twin triode I6 is used. The entering signal on the grid 'I4 causes current to flow between cathode 82 and plate 30, and since cathode 82 is connected to ground not directly but through a resistance 83, grid bias is developed proportional to the signal. But this resistor 83 unfortunately also acts to reduce the gain of the tube. For this reason the other triode of the twin triode is brought into play. Here, the grid voltage being always zero (due to grid 83 being directly grounded) the eect of the cathode voltage swing is to decrease current now in the robber circuit 42 and thereby increase current in the main circuit 44. By this novel means the degeneration due to the cathode resistance 83 is neutralized and the fullest possible gain of the triode 82, 14, 3U is secured.
The amplied output of tube I6 is next fed into twin triode I8, which acts as phase inverter through resistance coupling IIB-H8. The resistance values of the resistors ||6, ||8 are such that balanced push-pull operation of the amplier is obtained. The two signal components, which are out of phase, are fed through the semi-direct coupling |28, |38 to the control grids |40 of the beam power amplifier tubes 20 which are connected in push-pull relation. The semidirect coupling consisting of resistors |30 shunted across coupling condensers |28, minimizes low frequency oscillation.
If at a given signal frequency the combined phase shift of an amplier and feed-back circuit is 180 or greater, the product of the amplifier gain and the vectorial transmission between the output and input circuits should be less than unity, since, if otherwise, the resultant reversal of phase might supply suiicient output voltage to sustain oscillations begun by some perturbation of the loop. This trouble is overcome according to the present invention in which semidirect coupling is employed between successive amplifier stages of the amplifier unit. Such semi-direct coupling functions to maintain to a minimum value the phase shift at very low frequency values of the signal due to coupling without appreciably affecting the magnitude of the output voltage. Stability for such frequencies is obtained by keeping the total phase shift of the amplifier to less than 180 for low frequencies. `Compared to direct-coupling, the advantage of semi-direct coupling includes a loss of gain at low frequencies that makes it easier to get the overall drop in gain that is required for stability at low frequencies, and much less dependence on tube characteristics. Compared to resistance-capacity coupling, the big advantage of semi-direct coupling is the decided decrease ln phase shift at low frequencies, without appreciably affecting the phase shift or magnitude of the translated signal component at higher frequencies.
The output of the push-pull power stage is fed into the primary of the output transformer 24. Inverse voltage feed-back is accomplished by feeding the output voltage from the secondary |64 of transformer 24 back to the grid 14 Of twin triode I6. The inverse voltage feed-back in the power output is accomplished by feeding the voltage from secondary |5| of the transformer 24 to the cathodes |46 of the beam power amplifier tubes 20. The beneficial effects of feed-back are:
1. Reduction of harmonic distortion.
2. Greater stability, including constance of characteristics with changes in tubes or input voltage Vs.
3. Reduction of phase distortion.
4. Reduction of hum level.
5. Reduction in the internal impedance of the amplifier.
This last factor is important in that the output impedance of the amplifier is kept to a minimum.
The following advantages have been derived in booster amplifier unit 8:
1. Freedom from bounce or low frequency oscillation.
2. High frequency oscillation has been eliminated for wide limits of load.
3. A high source impedance can be tolerated.
4. The hum level is less than 0.004% of the maximum full signal input voltage.
High frequency oscillation is avoided by condensers |56 and |58 across the output transformer primary 1D and by a condenser |38 in series with a resistor |36 on the input |24 of the third stage. The values of resistor |36 and condenser |38 are such as to cause a reduction in the gain for high frequencies without shifting the phase, in the same manner as the semi-direct coupling effect on 10W frequencies. y
Although a high -source 'Impedance can be tolerated, care must be observed to avoid capacitance from the input grid circuit to ground, thereby causing high frequency oscillation. This is readily taken care of by connecting the input circuit shield 11 to ground through a suitable resistance or by connecting the shield 11 to the input-output common lead I4 with conductor 18.
The hum level with zero signal is kept low by proper filter networks in the plate circuits of twin triodes I6 and |8, comprising resistors 36, 38, 40 and condensers 84, 86, 88. No filtering is needed in the push-pull power output stage 60.
This booster amplifier 8 operates over a signal voltage range of 200 to 1 with an accuracy of better than i0.1%.
Such a booster amplifier unit B may be used in conjunction with a resolver |68. In the application of an electrical resolver, which is illustrated and described in detail in copending application Serial No. 346,183, filed July 18, 1940, it is highly desirable that the phase of the secondary voltage and its magnitude when set at the maximum position be the same as those of the signal voltage. It is clear that if the primary impedance of the resolver were constant a simple phase shifting network might be employed between the booster amplifier and the resolver to correct its phase. However, the resolver has a ferro-magnetic core and its main circuit resistance is a function of the flux density and so of the applied voltage. For this reason, a com- 6. pensator |10 is used as'shown in Fig. 2.- The leakage reactance of the resolver is due principally to iiux which has a long path .in air, and the leakage resistance is due to the copper Windf ing. The ratio of reactance to resistance of the compensator primary |12 and resistor |14 is made the same as the ratio of reactance to resistance of the leakage impedance of the resolver. Then, since the same current flows through both of these impedances the voltage across the primary coil |12 of the compensator will .be proportional to the loss of voltage in the leakage impedance.
The turns ratio of secondary |16 which is closely coupled to the primary |12, to that coil is made such that the voltage across the -secondary |16 is equal to the voltage across the leakage impedance of the resolver. This voltage is then introduced into the input of the booster amplifier unit 8 in the direction required to neutralize the leakage impedance of the resolver |68. If freedom from temperature effects is required the resistor |14 should be made of copper.
It is possible to vary the ratio of input signal voltage Vs to output signal voltage Vo, or the effective transformation ratio of the resolver |68 when used in conjunction with the booster amplifier unit 8 and compensator |16, by means of a shunt resistor |18. If the resistance value of resistork |18, or leakage resistance, is decreased, by shunting with the resistor |14 the ratio Vo/V is increased, as will the ratio VO/VS.
Although a preferred embodiment of the invention and one use thereof are illustrated and described herein, it is to be understood that the invention is not limited thereto, but the invention is susceptible of various uses as well changes in form and detail within the scope of the appended claims.
1. The combination with a signal supply circuit having a relatively high value of effective impedance, and a load circuit having a relatively low value of effective impedance, of an amplifier circuit responsive to a signal component derived from said signal supply circuit comprising a twin triode amplifier tube, one o-f said triodes thereof increasing the gain of the other triode, for applying a similar component of substantially the same phase and magnitude as such signal component to said load circuit.
2. The combination with a signal supply circuit having a relatively high value of effective impedance, and a load circuit having a relatively low value of effective impedance, of an amplifier circuit responsive to a signal component derived from said signal supply circuit comprising a twin triode amplifier tube, one of said triodes thereof increasing the gain of the other triode, for applying a similar component of substantially the same phase and magnitude as such signal component to said load circuit, said amplifier circuit having a value of effective internal impedance with respect to that of said signal circuit such that the signal supply circuit is not affected when a signal component is derived therefrom by said amplifier circuit and applied to said load circuit.
3. The combination with a signal supply circuit having a relatively high value of effective impedance, and a load circuit having a relatively low value of effective impedance, of an amplifier circuit responsive to a signal component derived from said signal supply circuit for applying a similar component of substantially the same phase and magnitude as such signal component to said load circuit, said amplifier circuit comprising electronic means adapted to amplify such signal component, a phase inverter circuit responsive to such ampliiied signal component for producing two signal components which are 180 out of phase with each other, a pair of power ampliiier space discharge devices connected in push-pull relation, and semidirect means coupling the output circuits of said phase inverter circuit to the input circuits of said power ampliiier space discharge devices.
4. The combination with a signal supply circuit having a relatively high value of eiiective impedance, and a load circuit having a relatively low value of eiiective impedance, of an amplifier circuit responsive to a signal component derived from said signal supply circuit for applying a similar component of substantially the same phase and magnitude as such signal component to said load circuit, said amplifier circuit comprising electronic means adapted to amplify such signal component, a phase inverter circuit responsive to such amplified signal component for producing two signal components which are 180 out of phase with each other, a pair of power ampliiier space discharge devices connected in push-pull relation, and semi-direct means coupling the output circuits of said phase inverter circuit to the input circuits of said power ampliiier space discharge devices, said semi-direct coupling means comprising a coupling condenser between said input and output circuits, and resistors in shunt circuit relation with said coupling condensers, for minimizing low frequency oscillation in the amplifier circuit.
5. In combination, a relatively high impedance signal supply circuit, an untuned current amplifier circuit the input circuit of which is coupled to said signal supply circuit, and a relatively low impedance load circuit coupled to the output circuit of said untuned current ampliiier circuit, said untuned ampliiier circuit including an amplifier, a phase inverter, a pair of space discharge devices connected in push-pull circuit relation, and a semi-direct coupling between said amplifier and phase inverter, respectively, and said space discharge devices, said coupling consisting of condensers in series with the control grids of said space discharge devices and the output circuits of said ampliiier and phase inverter, and resistors connected across said condensers, whereby said untuned current amplifier circuit is adapted to apply to said load circuit a voltage component of substantially the same phase and magnitude as that received by its input circuit from said original supply circuit.
6. In a inverse feed-back booster amplier unit of the class described, a plurality of ampliiier stages comprising electronic tubes, at least two amplifier tubes of which are connected in push-pull relation, and semi-direct coupling means between the input circuits of such tubes and the output circuits of the preceding stage, consisting of capacitance and resistance in parallel, the impedance values of which are proportioned so as to inhibit inherent phase displacement due to a change of frequency of a signal component translated by the unit.
FREDERICK W. CUNNINGHAM.
REFERENCES CITED The following references are of record in the iile of this patent:
UNITED STATES PATENTS Number Name Date 2,200,055 Burnett May 7, 1940 2,324,279 Clark July 13, 1943 2,306,749 Potter Dec 29, 1942 2,162,878 Brailsford June 20, 1939 2,276,565 Crosby Mar. 17, 1942 2,068,112 Rust Jan, 19, 1937 2,215,796 Rust et al Sept, 24, 1940
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US2068112 *||20 Aug 1935||19 Jan 1937||Rca Corp||Amplification and selectivity control circuit|
|US2162878 *||26 Jan 1937||20 Jun 1939||Rca Corp||Automatic gain control circuits|
|US2200055 *||23 Feb 1938||7 May 1940||Rca Corp||High impedance attenuator|
|US2215796 *||26 Apr 1937||24 Sep 1940||Rca Corp||High frequency circuit arrangement|
|US2276565 *||23 May 1939||17 Mar 1942||Rca Corp||Limiting amplifier|
|US2306749 *||22 May 1941||29 Dec 1942||Gen Electric||Amplifying system|
|US2324279 *||29 Nov 1941||13 Jul 1943||Rca Corp||Amplifier|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US2595443 *||14 Mar 1946||6 May 1952||Becker Harry W||High fidelity amplifier|
|US2595444 *||26 Jun 1946||6 May 1952||Becker Harry W||Amplifier|
|US2614227 *||6 Aug 1949||14 Oct 1952||Moorc Electronic Lab Inc||Cathode follower photoelectric direct current amplifier circuit|
|US2635224 *||18 Oct 1946||14 Apr 1953||Arma Corp||Calculating instrument|
|US2646467 *||13 Jul 1949||21 Jul 1953||Mcintosh Frank H||Wide band amplifier|
|US2706245 *||24 Feb 1950||12 Apr 1955||Miller Joseph L||Electromagnetic transducer-detector|
|US2744971 *||12 Jul 1951||8 May 1956||Llopes John D||High fidelity audio frequency amplifier|
|US2934713 *||9 Dec 1955||26 Apr 1960||Itt||Anode-follower amplifier|
|US3416088 *||27 Oct 1964||10 Dec 1968||Rank Bush Murphy Ltd||Electrical signal amplifier|
|U.S. Classification||330/81, 330/87, 330/117, 322/45, 330/121, 330/182, 330/122, 330/98, 330/91, 330/107, 330/100|
|International Classification||H03F1/36, H03F1/34|