US2251594A - Volume control for electrical signals - Google Patents

Description

5, 1941- w. H. MAYNE 2,251,594

-' VOLUME CONTROL FOR ELECTRICAL SIGNALS Filed larch a, 1940 .2 Sheets-Sheet 1 Aug 1941- w. H. MAYNE 2,251,594

VOLUME CONTROL FOR ELECTRICAL SIGNALS Filed March 8, 1940 v 2 She'ets-Sheet 2 Patented Aug. 5, 1941 1 wuuam Harry Mayne, san Antonio, Tom, a. signer to ()live S. Petty, San Antonio, Tex.

Application March 8, 1940, serial No. 323,009

1 Claim.

- This invention relates to improvements in methods of and. apparatus forcontrolling the amplitude of wave-form signals and has for its principal object the automatic control of the gain in an amplifier in response to variation in amplitude of the signals impressed thereon.

It is common practice in various fields to regulate amplifiergain-by means of a control circuit operable by and in response to the signal energy in such manner as to reduce the gain when the level of signal energy is high and to increase Circuits forv the gain when the level is low. achieving this result are widely used in radioreceiving systems in order to compensate in part for fading of signals from distant transmitters and in order to establish a generally uniformsignal level for the reception of stations of diiferent' but comparable field strength. ircuits functioning in the reverse sense, to vary the amplifier gain inversely to the amplitude of the received signal energy are also employed to a considerable extent in the reproduction of recorded music signals, and to a lesser extent in the reception of transmitted music signals, in order to expand the volume range. Thus in the preparation of recordings of the type employing a record groove, there is a limit to the intensity of signal which can be so recorded, since signals of high amplidegree of amplification by controlling thegrid bias of a tube in the amplifier. It can be shown,

howeventhat with such an arrangement it is difficult to reduce the gain to a low value without introducing distortion, since the usual thermionlc valve is constructed so as tooperate on the more nearly linear portion of its plate cur-1 rent-grid voltage curve only when thegain exceeds a predetermined minimum. Reduction of the gain to a low value results in operation of the valve on an arcuate portion of the curve,

and the plate current does not vary uniformly in response to change in the applied grid voltage.

It is therefore an object of the invention to avoid these difficulties and to provide an amplifying circuit in which the gainmay be varied from a. maximum-down to nearly zero without appreciable signal distortion.

In order to accomplish this result it is proposed to feed the signal energy through two paths, to combine the output of thesepaths in opposition, and to control the gain in each path .in such manner that when minimum gain is tude tend to cause the cutting stylus to break limits. Similar precautions are observed in the broadcasting of music signals to establish a range oi. volume which can be satisfactorily dealt with by the transmitter and by the ordinary receiving equipment. Thus by amplifying the signal in inverse ratio to the strength of the signal, an approximation of the original volume range may be established in the reproduction. Such devices may be referred to as volume, expanders.

In these and other applications of gain control, it has been difiicult heretofore to achieve as wide a variation of the amplification factor as is desirable. For example, a common type of gain control heretofore employed has regulated the required, the signal output of the two paths will substantially cancel. Preferably the arrangement is such that when minimum gain is'to be effected, the valves in both paths are functioning substantially on the linear portion of the I -E; curve, so that even for an extremely small or zero gain, the signals are undistorted. On the other hand, when maximum amplification is desired, the gain in one path is reduced to a minimum, whereby distortion of the signal in that path may result, but since the gain in the other path is concurrently the maximum, the combined energy output of the two paths is derived primarily from that path in which the valves are functioning on the linear portion of the Ip-Eg curve, so that distortion is not appreciable.

A further object of the invention is to provide, with a circuit of the character described affording two paths for signal energy, a grid bias control for the valves in thetwo paths which is responsive to and operable in accordance with the amplitude of the signal energy, so as automatically to establish the gain of the amplifier in accordance with the strength of the applied signal energy. I

A further object of the invention is to provide an automatic gain control operating over a wide range without appreciable distortion and without the use of specially designed and expensive valves and other equipment.

Further objects and features of the invention will be apparent from the following description taken in connection with the accompanying drawings, in which Figure 1 is a wiring diagram of a circuit embodying the principles of the invention;

Figures 2 and 3 are similar diagrams illustrating modified circuit arrangements;

'Figure 4 is a representation of an Ip-Eg curve of a thermionic valve such as may be employed in the practice of the invention; and

Figures 5 and 6 are wiring diagrams illustrating further modifications.

In order to facilitate an understanding of the invention, several preferred embodiments thereof are illustrated in the drawings and specific language is employed to describe the same. It will nevertheless be understood that no limitation of the scope of the invention is thereby intended, the several embodiments being merely representative of specific methods of carrying out the principles of the invention. Further modifications and alterations are contemplated such as would occur to one skilled in the art to which the invention relates.

Figure 1 illustrates a control system which is operable automatically to reduce the gain on reception of signals of high amplitude and to increase the gain when signals of lower amplitude are received. Thus the signal energy, which may be amplified or otherwise treated in any desired manner before delivery to the control circuit, is fed to the input terminals l and thence through the primary winding ll of a transformer l2. 'At this point the signal energy is divided into two paths by the use of separate secondary windings l4 and I5. One end of the winding [4 is connected to the grid I8 of a thermionic valve 19, the other end of the winding being connected to one end of a resistance 20. A center tap 2| on the resistance leads through a source of grid bias voltage 24 to the cathode 25 of the valve l9.

Similarly, one end of the secondary winding I is connected to the grid 28 of a thermionic valve 29. The other end of the winding i5 is connected through a source 3| of grid bias voltage to the opposite end of the resistance 20, the grid circuit for the valve 29 being completed through the cen ter tap 2! of the resistance and the grid bias source 24 to the cathode 32 of the valve.

The outputs of the valves [9 and 29 are combined in opposition, the plates 34 and 35, respectively, of the valves being connected to opposite ends of the primary winding 36 of an output transformer 38. The main portion of the signal energy is applied to output terminals 4| by the secondary winding 42 of the transformer 38. Plate voltage for the valves is supplied from a source 40 which is connected between the cathodes of the valves and a center tap on the winding 36 of the transformer 38.

The transformer 38 includes an additional secondary winding 44 which is used to apply a portion of the output signal energy to" control the gain of valves l9 and 29. Thus one end of the winding 44 is connected through a. voltage source 45 to the anode 46 of a diode 41. The cathode 49 of the diode 41 is connected to one end of the resistance 20 and the opposite end of the secondary winding 44 is connected to the opposite end of this resistance.

The source of grid bias voltage 24 operates trough the respective parts of the resistance 20 supply substantially equal negative grid bias to .mwalves i9 and 29, but the grid 28 of valve 29 issupplied with an additional negative bias from the source II. Thus the valve l9 will normally be operated on the linear portion of the I -E curve to deliver its normal output, whereas the strongly negative bias on the valve 29 will cause the latter to operate on the lower or arcuate portion of the curve so as to afford greatly reduced gain. Consequently, although the outputs of the valves l9 and 29 are combined in opposed phase, the output energy will be derived-principally from the valve l9, which is operating on the linear portion of the curve, and distortionwhich might be occasioned by the excessive negative bias applied to the valve 29 will be minimized.

The anode 46 of the diode 41 is initially provided with a negative bias by the source 45, so

that no current flows in the diode circuit. However, when the signal energy in the output transformer 38 exceeds apredetermined minimum, the potential developed in the winding 44 will exceed the potential established by the source 45, the anode 46 of the diode valve will become positive, and current will flow in the diode circuit. This will in turn develop potential differences across the two parts of the resistance 20, with the result that the bias voltage applied to grid 28 of valve 29 will. become less negative, whereas the bias voltage applied to the grid l8 of the valve I 9 will .become more negative, the extent of this change in grid bias on the two valves being determined by the amplitude of output signal energy of the circuit. Thus the gain in the valve 29 will be increased and the gain in the valve 1 9 will be decreased, so that the output signal energy resulting from opposing the energy from these two paths will be decreased. The constants of the circuits may obviously be so selected that as the amplitude of received signal energy increases, a

point is reached at which the outputs of the valves 19 and 29 are substantially-equal, and the gain of the whole circuit will be reduced to an absolute minimum or to zero. Since both valves are then operating on a substantially linear portion of the IpEg curve, there will still be no appreciable distortion of the signal.

It will be obvious that the relative amounts of signal energy flowing in the two paths will determine the extent of gain in each path required for proper operation. For example, if the amount of signal energy applied to the valve 29 is substantially less than that applied to the valve I9, greater gain must be effected in the valve 29 than in the valve IS in order that the outputs of the two valves may nearly cancel at high levels of input signal energy to reduce the amplification factor of the whole system to a minimum. It is therefore intended, when reference is made herein to substantially equal gain in the two paths for the p rpose of producing a minimum gain in the whole circuit, to include within the scope of the language an arrangement employing a relative gain producing the desired result when difierent amounts of energy are caused to fiow in the two paths.

In order to employ the circuit shown in Figure 1 as a so-called volume expander, it is only necessary to delete the source 3| of grid bias voltage for the valve 29, so that valves l9 and 29 normally operate at nearly the same point on the Ip-Eg curve so that the gain of the whole system is very small. If, then, the input signal energy increases so as to cause current to fiow in the circuit of the diode valve 41, the operating points of the two valves l9 and 29 will be displaced along the Ip-Eg curve to afford increased gain as hereinbefore explained. In other words, when the amplitude of the signal energy in the transformer 39 exceeds a certain predetermined minimum value, the gain represented by the difference in amplincation or the two paths will be correspondingly the potential applied to the grid 28 or the valve 29 by the source it be represented as Er, and the potential applied to the grids of both valves l9 and 29 by the source 24 be represented asEz, the plate current of the valve 29 may be represented at 11 and the plate current of the valve l9 may be represented at I2. The output or the whole circuit will be represented by the difierence between these values of plate current and will be quite lage; since distortion occurs principally in the valve 29, byreason of operation on the lower portion of the curve, the distortion will be small since the output of the valve is small.

If the signal amplitude now increases to the extent necessary to cause current flow in the circuit .containing the diode 41, the operating points of the two valves will be displaced toward. each other on the curye, and as the signal continues to increase. these points of operation will eventually coincide at the point represented by E3, at which point the output energy will be a minimum or zero. It will be observed .that

- will become positive with when this condition is established, the valves are both operating underifavorable conditions so that distortion will be minimized,

Figure 2 illustrates a further embodiment of the invention in which the control bias for the valves is supplied in inverse relation to the amplitude of the signal energy. Thus the incoming energy is delivered to the terminals 59 and thence through the primary winding 5| of a transformer 52. Two secondary windings 54 and 55 are employed, the winding 54 delivering energy to the valve 59 and the winding 55 delivering energy to the valve 50 in precisely the manner described with reference to Figure 1..- The output of valves 59 and 5.9 isfed in opposed relation to the primary winding 54 of a transformer 5! or which the secondary winding 52 delivers energy to the output terminals 53 while the secondary winding 65 delivers energy to a control circuit hereinafter described. Plate voltage for the valves is supplied from a source 65 connected between the valve cathodes and a center tap on the transformer primary 54. The initial grid bias voltages are supplied from a source 69 .of which the positive terminal is connected to the cathodes of the valves 59 and 50 and the negative terminal is connected to the center tap I0 of a resistance II, the ends of which are connected to the grids of the valves through the respective secondary windings 54 and 55. It will be appreciated that the circuit thus far described is identical with that shown in Figure l with the exception that the grid bias source .3! is omitted, the valves being adjusted to normally operate at or substantially at the same point on the I -E curve, to afiord minimum gain as hereinbefore explained.

One end of the secondary control winding 55 is connected through a source of bias voltage 12 to the anode I3 of a diode-triode valve 11, the cathode 14 ofwhich is returned through a resistance 15 to the opposite end of the winding 65.

current flows in the diode circuit until the energy developed in the winding excee sa predetermined minimum, at which time the bias aflorded by thesource I! will be exceeded, the anode I9 respect to the cathode 14, and current will flow in the diode circuit,

with the result that. a potential difference will I be established across the resistance I5. The resistance I5 is also connected to the'grid the triode circuit, and the bias on the grid 19 becomes more negative as the potential across the resistance I5 increases. thus regulating. the

flow or current in the circuit including the cathode 14, the plate 80, the plate voltage source 9|, and the resistance II, tovary the potential difference across the latter. A by-pass 0011-? denser 92 aiifords apath for pulsating current in the triode circuit.

It will be appreciated that when no current flows in the diode circuit between the cathode I4 and the anode 19, as will be the case when the,

received'signal amplitude is low, the associated triode will deliver maximum, plate current. Maximum potential will then be developed across the resistance 1|, and the operating points of the valves 59 and 89 will be displaced widely'on the Ip-Eg curve, so as to afford maximum gain as hereinbefore described. when, however, the level of received signal energy increases, current 11 flow in the diode circuit, the grid 19 oi. the triode will be rendered more negative to reduce the potential developed across the resistance II, and the difference in grid bias applied to the valves 59 and 50 will be correspondingly reduced, so that the points of operation of these valves will be moved toward each other on the Ir-Eg curve to reduce the gain eflected thereby. Eventually, as the input signal energy continues to increase, the points of operation 01 the valves In Figure 3 is shown an embodiment of the.

invention which is in some respects simpler than those previously described. Thus the energy to be amplified is supplied to the input terminals 95 and thence through the primary winding 86 of a transformer 81, one end of the secondary winding 89 being connected to the grid 89 of a valve 90, and the other end being connected to oneend of a resistance 92. The center tap 93 of this resistance is connected through a source of grid bias voltage 94 to the cathode 96 of a valve 91 and the cathode 98 of the valve 90. The other end of the resistance 92 is connected through a resistance IOI to the grid I92 of valve 91.

The operating voltage for the grid 192 is derived from the combined outputs of the valves 90 and 91. Thus the plates I05 and M5 of the respective valves are connected to each other and As in the'previously described embodiment, no

condenser I I6 and the opposite ends of the resistance are fed with control voltage from terminals M8 by control circuits which may be constructed as described with reference to Figures 1 and 2.

It will be seen that this circuit permits the use oi'resistance coupling and of an input transformer having a single winding, and is to this extent a simplification of the circuits hereinbefore described. Except that the driving voltage applied to the grid I02 of valve 91 is supplied from the subtractive outputs of valves. Hand 80, so that there .must always be a residual signal to excite the grid, the operation is quite similar to that described in connection with the aforementioned figures, bias voltages on the valves being varied in accordance with variation in amplitude of the received signal so as to displace the operating points of the valves on the Ip-Eg curve and thus automatically regulate the gain.

Turning now to Figure 5, it will be observed that the arrangement of the circuit elements is generally similar to that shown in Figure 2, and to simplify the description, the same reference numerals are applied as in Figure 2 to elements functioning similarly. There is one minor differenc in the circuit arrangement, the input energy being delivered to the grids of the valves 59 and 60 in" opposed phase relation, and the outputs of the valves being combined by directly connecting the anodes of the valves. .It will be appreciated that this alteration is more or less obvious and is an arrangement which may be employed in any of the circuits described herein. With either arrangement the outputs of the valves may be described as opposed, this language being used in the generic sense to signify in Figure the control energy is taken from the input transformer 52 by a secondary winding 51, in lieu of the secondary winding 65 of the form of the invention shown in Figure 2. The control circuit including the valve 11, is otherwise identical with that shown in Figure 2 and the grid bias on the valves 59 and 60 is similarly varied.

Certain advantages are secured with the modifled arr ement shown in Figure 5. Thus by the use of the variation in amplitude level of the incoming signals to control the gain of the amplifier, greater controlling effect can be obiained, since this eflect is independent of gain. In other words, when the control voltage is derived. from the output of the valves 59 and 60, such voltage decreases as the gain decreases, and consequently as the output approaches the normal IeVeLlittIe-or no control voltage is available. It is of course highly desirable, in the event the control voltage is derived from the incoming signal, to so adjust the constants of the circuit that the gain varies inversely as a linear function of the signal.

In the further modification illustrated in Figure 6, the general principle is the same as that involved in the operation of the circuits shown in Figures 2 and 5, but the control voltage is derived both from the incoming signal and the output 01' the main valves, so that the advantages of both systems are obtained. This involves alteration only of the control circuit including the control valve, and the reference numerals used in Figures 2 and 5 to designate corresponding elements of the main circuit are employed in Figure 6. a

The control valve indicated at I22 is constructed to afford in effect two diodes. having a single cathode I23 and anodes I25 and I28. The anode I25 is connected through a source of anode voltage I28 and a secondary winding I29 associated with the transformer 52 to the tap I30 of a resistor I3I, and thence through one end of the resistor to the cathode I23, by-pass condensers I and I38. being associated with the resistor I3I. The anode I25 is connected through a source of anode voltage I31, the sccondary winding 62 of the transformer 6i, and the resistor Hi to the cathode I23. The grid I39 of the valve I22 is connected to that end of the v resistor I3I which is remote from the cathode,

and the output anode II, supplied with voltage from a source 8|, varies the potential diflerence across the resistor II to vary the grid bias of the valves 59 and 50 as in the form of the invention shown in Figure 2.

It will be apparent from a comparison of the circuit shown in Figure 6 with those previously described that the control voltage which serves to vary the grid bias of the main valves is a function. of both the input and the output signal energy, and adjustment ofthe gain is therefore correspondingly improved, control voltage being available from the incoming signal even though the combined energy output of the valves 59 and 60 is quite low, so that the proper gain P value is promptly stablished.

whether the energy delivered to the control circuit is derived from the input or the output side of the amplifying means, or from both, it is obviously appropriate to state that the control circuit is operableby and in response to variation in amplitude level of the signal energy.

It will be appreciated that the invention is especially applicable to signaling devices in which a very wide range of gain control is desirable. Such is the field of radio receivers for use in automobiles, in which the received signal varies widely in strength while the vehicle is in motion. Another important field is geophysical exploration, including seismic surveying .and the like, in which it is desirable to maintain a generally uniform amplitude level in a record of seismic impulses where the amplitude of the incoming signal may vary in the ratio of 600:1 or more. will be appreciated that the circuit described herein is applicable to amplification control of signal energy at both radio and audio frequen cies, and that the circuits illustrated in the -awlngs can in general be introduced at any convenient point in any amplifying system for wave-form energy and may be used with or without additional amplification as desired.

Having thus described the invention, what is claimed as new and desired to be secured by Letters Patent is:

In a gain control for wave-form signal amplifiers, the combination with thermionic valve amplifying means affording two paths for signal energy, means combining the outputs of said valve means, means feeding back a portion of the combined output energy to one of said valve means, and means controlling the relative amplification efl'ected by said valve means to cause the outputs thereof to approach or recede from a common amplitude level.

WILLIAM HARRY MAYNE.

Images