US2480842A - Automatic gain-control system - Google Patents

Description

N. B. FARNSWORTH AUTOMATIC GAIN-CONTROL SYSTEM sem9 6, 1949.

Filed March 6; 1945 no al Won/AL #MPL /F WM5 H l 2g j v y, Il

| l l I l INVENTOR A TORNEY Patented Sept. 6, 1949 AUTOMATIC GAIN -CONTRQL SYSTEM A Neil B. Farnsworth, Hempstead, N. Y., assignor to The Sperry Corporation, a corporation of Delaware Application March 6, 1945,'Serial No. 581,243

7 claims. (o1. 179-171)` This invention relates to improved automatic control systems, and more particularly to an automatic gain-control system especially adapted for use in a receiver for modulated carrier waves.v

' In the past, it has been common practice to provide an automatic gain-control system for modulated carrier Wave receivers'in which the rectified carrier wave is utilized to control the gain'of one or more amplifying vacuum tubes inversely in accordance with the strength of the input signal wave, so that an increasein the incoming signal strength 'causes a decrease in the ampliiication of the receiver and vice versa. Since such an automatic gain-control system is responsive to changes in amplitude of the rectified modulated carrier wave, filtering means must necessarilybe provided to prevent the automatic gain-control system from eectively removing the modulation component from the output of the receiver. In an ordinary radio broadcast receiver, for example, it is customary to provide a lter having a time constant whose valueis comparable to the period of the lowest audible 4frequency to which the receiver is intended to respond. Thisrlter is introduced in the automatic gain-control system in such a manner that changes in the output occurring at an audio-frequency rate have no appreciable effect upon the amplification of the receiver, while slow cy. In such receivers, phase shifts of the lmodulation envelope which maybe introduced due to the operation of the automatic gain-control system are relatively negligible importance. l

One form of vobject locating system radiates uniform short -pulses ofhigh-frequency .energy into space-With constantly varying directivity. The energy which strikes the object to be located is reflected toward thev receiver and provides information regarding the `direction Yof the object under observation.. Such a system may include a rotatable highly directional antenna, Iwhich is spun about an axis slightly different from the axis of its maximum directivity. If the object lies directly along the axis of rotation, the reflected Signal, which mi bereeived, by the Same @ritenz na, will have yconstraint amplitude, and its modulation envelope will have zero frequency.- When the object is displaced from the spin axis, however, the received pulses will have a modulation envelope having ai frequency corresponding to the speed of rotation ofthe. antenna. -The'phase of this modulation envelope, compared with that of a reference voltage developedV directly by the -rotation of the antenna, provides information as to the sense in which the object isdisplaced from the spin axis of the antenna. The amplitude of the modulation indicatesthey extent Lof such displacement. The average strength of the reflected pulses as-received varies between wide limits, dependent upon lthe range and the attitude of the object as well as other factors, allof which vary relatively slowlycompared with the spin period.

Ina practical-object locating system, an automatic gain -icontrol system must be provided which substantially compensates for the changes in average received signalystrength; Theproblem, as will be apparent vfrom the above brief description, is to provide an automatic gain-control system for thisrpurpose. which is also capable of operation withoutvthe introduction of appreciable phase shift at the modulation frequency and with arminimum effect upon the amplitude of the modulationfffrequency components of the received signal. The conventional type of automatic gaincontrol system has not proved to be satisfactory insuch cases. Y

It has been found in particular that phase shift of Vthe modulation envelope isV commonly encountered in amplifiers having a gain-control system incorporating the conventional type of resistance-capacitance low-pass filter. When the time :constant of such a filter is made sufficiently small so thatthe automatic gain-control system will effectively conpensate for relatively slow variations in input-signal strength, as for example those due to fading, a serious phase shift which is a function of thissignal strength is introduced.

`In addition to affecting the performance of the receiver for its intended purpose of receiving and comparing in .phase the modulated pulse signal, the phase shift in some cases may be so great as to cause cscillation of the receiver. The phase shift in an amplifier employing sucha filter may Vbe minimized-byincreasing thertimev constant o-f the lter, but such an expedient renders the response of the automatic gain-control system too slow for satisfactory compensation of changes in the input signal strength due to fading and other Similar phenomena. l

.In accordance with the present invention, the

problem of modulation phase shift which varies as a function of the input signal strength, in an amplifier having an automatic gain-control system, is effectively solved by providing a filter network in the gain-control system which has substantially zero phase shift at the modulation frequency and which, furthermore, has relatively low transmission at'this frequency. By the use of such an arrangement, it has been found that the component of modulation frequency which reaches the amplifier due to the automatic gaincontrol loop may be made to be of Such relative phase and magnitude as to have no appreciable effect upon the phase shiftnormally introduced by the amplifier upon the modulation', regardless of the strength of the input signal.

It is an object of the present invention, accordingly, to provide an automatic gain-control system having a response which varies in a predetermined manner with the frequency at which amplitude changes of the incoming signal occur.

Another object of the present invention is to provide an automatic 'gain-control system for a receiver of modulated carrier waves which causes no appreciable phase shift of the envelope of modulating signals lying -witl-iin a predetermined frequency band.

It is an additional object of the present invention to provide an improved automatic gaincontrol system for modulated carrier receivers which is suiliciently rapid in operation to correct for fading and other'relatively slow changes in the received signal strength, but which functions without introducing any appreciable phase shift of the modulation envelope as a function of the strength of the received signal.

Yet another object of the present invention is to provide an improved automatic gain-control system which is highly effective in counteracting relatively slow changes in the amplitude of the incoming signal, but which has a relatively negligible response to changes in the amplitude of the incomingl signal which occur-within a predetermined band` of frequencies.

Still a further object of the present invention is to provide an lautonfiati'c gain-control system incorporating' a frequency-selective filter network having no inductively reactive components.

In accordance with the present invention, there are providedv an amplifier having input and output terminals and electronic gain-control means, and means'for applying a modulated carrier wave to the input terminals. A rectifier is .connected to the output terminals, and a filter network is connectedbetween the rectifier and ,the gain-control means. This' network has such characteristics that the gain of the amplifier is varied as an inverse function ofv the strength of the carrier wave with no appreciable variation in the phase of its modulation envelope. In a Vpreferred embodiment of the invention, the filter network is of the non-inductive null type comprising impedance .elements arranged in a bridged-T configuration, In' a modified embodiment of the invention, the filter network may be of the parallel-T or twin-T configuration, in which the impedance elements. are also of the non-inductive type. lrreferably, the network'is S designed that'it has .substantially Zero phase shift and relatively low transmission at the modvlaton frequency, but basan effective time constant of the order of the period of the modulation frequency.

The invention in another of its aspects relates to novel features of the instrumentalities described herein for achieving the principal objects of the invention and to novel principles employed in those instrumentalities, whether or not these features and principles are used for the principal objects or in the same field.

The above and other objects and features of the invention will be better understood by reference to the followingdescription taken in connection with the accompanying drawing, in which like components are designated by like reference numerals and in which:

Fig. 1 is a circuit diagram, partly in block form, of an amplifier incorporating an automatic gaincontrol system in accordance with the present invention;

Fig. 2 shows, in graphical form, the performance of an arrangement in accordance with Fig. 1 compared with that employing a previously known type of filter network; and

Fig. 3 is a circuit diagram of a modified frequency-selective network which may be substituted for that employed in the gain-control system of Fig. 1.

Referring to Fig. 1 of the drawing, there is shown an amplifier I0 having input terminals I I andv output terminals I2. Amplifier I0 includes a vacuum tube I3 and, if desired, may include additional amplifying stages indicated by block I4.

Vacuum tube I3, which is preferably of the pentode type, has its` control electrode I5 coupled to one of input terminals Ir I' by means of capacitor I6. Cathode II of vacuum tube I3 is connected to ground through resistor I8 which is Icy-passed by capacitor I9. Screen-grid 20 of vacuum tube I3 is connected, at terminal 2|, to a source of moderately high positive potential with respect to ground. VSuppressor-gridY 22r is connected directly to cathode I'I. Plate 23 of vacuum tube I3 is coupled, by means of capacitor 24, to the input of block I4, the output of which is connected to output terminals'l'z of amplifier I0. It willbe understood that, if no additional amplifying stages are employed, capacitor 24 is connected directly to the high-potential one of output terminals I2. Plate 23 is also connected, through load resistor 25, to a terminal 26 which in turn is supplied with a relatively high positive potentialwith respect to ground.

The ungrounded output terminal I2 of amplifier Il) is coupled, by means of capacitor 21, to the anode or plate 28 of a vacuum tube 29, which functions as a diode. Cathode 30 of vacuum tube 29 is connected to the movable arm 3| of' a potentiometer 32, and is by-passed to groundl by means of capacitor 33. The ungrounded terminal of potentiometer 32 is connected through resistor 33 to a terminal 34 to whichv is applied a -relatively high positive potentialwith respect to grounded, as shown at 54.

The output terminal of lter network 36 is connected to the grid 4I of a vacuum tube 42, and is also connected to ground by means of resistor 43 which may be shuntedby a capacitor 44' if desired, The cathode 454 of vacuum tube 42 is connected through resistors lltV andI 41 in series to a terminals I I.

terminal 48, to which is applied a relatively high negative potential-with respect to ground. The plate 49 of vacuum tube `42 is connected at terminal 50 to a moderately high source of positive potential with respect to ground. The junction 5I of resistors 46and 4l -is-connected, through coil 52, to control-grid I5 of vacuum tube I3. Junction 5I is also connected, by lead 53, to block I4.

In operation, a modulatedcarrier wave is applied to input terminalsf-II. Assuming for the moment that junction 5I `is substantially at ground potential, a suitable bias voltage for control-grid I5 of vacuum tube I3 is developed across cathode resistor I8 due to its normal plate current, so that this tube provides substantially maximum amplication of the input signal wave. The signal wave is subjected to further amplication in unit I4, it being understood that the provision of such additional amplifying stages is optional. rIhe amplified signal wave appears at output terminals I2 and isA applied to anode 28 of vacuum tube 29,` which operates as a diode rectifier to demodulate the modulated carrier wave.

Assuming for the moment that movable arm 3l is at the grounded terminal of potentiometer 32, a voltage having a negativegdirect-current component corresponding to the rectified carrier wave and an alternating component which is a function of the modulation -is supplied to lter network 36. The purpose of potentiometer 32 is to provide an adjustable threshold for the automatic gain-control system. If movable arm 3| is moved to the left, a positive potential is applied to cathode 30 of vacuum tube 29, and hence no rectication or demodulationrtakesplace until the signal voltage applied to anode 28' exceeds this threshold voltage. VT-he provision of ysuch an ad- .lustable threshold is optional, and is'in no sense to be considered an essential element of the present invention.V

Due to the action of ,iilter network 36, the alternating component of the @output of rectifier 29 due to the modulation is greatly attenuated, so that the voltageV developed across resistor 43 and applied to grid 4I of vacuum tube 42 is substantially direct and of a magnitude depending upon the strength of the signal wave applied to input The'polarity of this voltage is such that grid 4I becomes increasingly negative with respect to ground as the input signal strength increases. Vacuum tube 42 functions as a cathode follower. The potentials applied to terminals 48 and 50, together with the values of resistors 46 and 41, are normally so chosen that, when no voltage is present atthe output of lter network 36, junction 5I is substantially at ground potential. When there is an output voltage from filter network 36, however, junction 5I becomes negativeY with respect to ground, to an extent dictated by the value of the voltage applied to grid 4I. 1 i

Thus it will be seen that, whenever the modulated carrier wave applied to input terminals II exceeds a predetermined value, a direct-current control voltage having a small ripple component of modulation frequency will be developed at the output of lter network 36 and utilized by means of vacuum tube 42, to provide a control-grid voltage forvacuum tube I3 of a polarity which tends to reduce the amplification of this tube as the Ysignal strength increases. If unit I4 is used, the

the amplitude of the output'sign'al at terminals I2 will be maintained substantially constant-in spite of wideuctuations inthe strengthof :the input signal. VFurthermore, due to the characteristics of filter network 36 as explained more fully below, the automatic Vgain-control system will have substantially no effect upon the phase 'of the modulation as the signal wave passes through amplifier I0, and relatively little effect upon the amplitude of the modulation.V

The 'performance of the arrangement of Fig. 1, compared with that of previously known systems, xwill become more apparent by reference to Fig. 2 of the drawing. In this figure, the ordinate axis represents the relative amplitude of the control :voltage developed at the output of lter network 36, as for example the voltage developed across resistor 43, compared with the zero-frequency value of this voltage, it being assumed that the output voltage from rectifierY 29 remains unchanged. The ordinate axis also represents rela- 'tive phase between the modulation voltage appearing at output terminals vI2 and the ripple 'voltage developed across resistor 43.A The abscissa 'axis of Fig. 2 represents frequency incycles per second. l

Curve 60 in Fig. 2 represents the relative am- Curve 6I shows the 'phase shift'which occurs in the system of Fig'. 1.l As shown by thiscurve,

the phase shift isvsubstantially zero at 30 cycles per second, and Vhas a maximum deviation from fzero of plus or minus 10 degrees as the frequency 'varies from 25 to 35 cycles per second. The effective time constant of thisV system is approximately 0.053 second.

ACurves 62 and 63 show the performanceofa system similar to that of Fig. 1, except that filter `network 36 was replaced by a conventional resistance-capacitance network. In order to provide comparable performance .in the low-frequency region, the latter network was so designed thatV its half-power point also occurred at,3icycles per second, a time constant of 0.053 second. At 30 cycles, as shown by curve 62, this system has a relative amplitude :of approximately 0.1. Such an arrangement does not provide satisfactory operation, however, in view of its extremelylarge phase shift which, as shown by curve 63, is greater "than 80 degrees at the modulation frequency.

In a further attempt to duplicate, by means of a conventional filter, the improved performance realized by the arrangement of the present invention, the resistance-capacitance network was redesigned to have substantially zerotransmission at 30 cycles. As 'shown by curve 64, this Vresult was obtained but, as also clearly shown by 65"' in the low-frequency region.

the curve, at the expense of-proper performance The half-power point in this case occurs at approximately 0.16 cycle per second, and the 1.0-second time constant of such a design is much too long to provide satisfactory operation in a 30-'cycle system.

It will be apparent from Fig. 2 thatV neither the expedient of employing a resistance-capacitance network having a time constant so long that its transmission at 30 cyclesper secondis substantially zero, nor theuse of a resistancecapacitance network having a time constant substantially equal to the effective time constant of filter network 36 but with a Very appreciable phase shift at 30 cycles, provides performance in any way comparable with that achieved in accordance with the present invention.. To state this advantage of the present invention in another way, it provides a filter network whichhas no appreciable effect upon the phase and which is approximately twenty times faster Vthan would bev a resistance-capacitance filter having'no appreciable effect upon the phase of the modulation voltage. Thus the arrangement of the present invention is capable of following and correcting for relatively rapid changes in the strength of the input wave at input terminals Il without at the same time appreciably affecting the phase of the modulation carried by such Signals.

' Fig. 3 of the'drawing shows a lter network 65` which may be substituted for filter network 36 of the arrangement of Fig. 1. Filter network 65 comprises series'resistors 66 and 5l together with shunt capacitor 68 in one 'branch of a parallel-T or twin-T configuration, and series ca pacitors 69 and 1D associated with shunt resistor 1I in the other branch thereof. Since the transfer impedance of such a network varies relatively rapidly withchanges in frequency near the frequency at which it is designedfto operate, the filter network of Fig.'3 is especially adapted for use in a system in which the modulation frequency remains substantially constant. Under such a condition, the network of Fig. 3 is capable of providing substantially zero transmission as well as substantially zero phase shift at the modulation frequency.

In one successful embodiment of the invention in accordance with: l, which was designed for 30-cycle modulation, vacuum tube I3 was a type 6AK5 and vacuum tubes 29 ande/lf2 were each a type 6C4, The following values of re- Vsistance and capacitance wereemployed:

Resistor I8 ohms 300 Resistor do 1,000 Potentiometer 32 1 1 do 25,000 Resistor 33 do, 620,000 Resistors and 43 megohms. 1.5 Resistors 31 and 38 ohms 240,000 Resistor IGli dog--- '750 Resistor 41 do 60,000 Capacitors I6 and 24 micromicrofarads r 100 Capacitors l!!V and 44 microfarad 0.001 Capacitor 33 do 0.01 Capacitor 2'L do 0.0039 Capacitor 39 do 0.1 Capacitor 40 do 0.015

The potentials applied to terminals 2l, 26, 34, 4-8 and 50 were respectively +100,Y +250, +250,

1 +250' and +120 volts.

While thereV has been described what is at Ypresent considered the preferred embodiment of the invention, it will be obvious to those skilled lin the art that various changes and modificartions may be made therein without departing :plying a modulated carrier'wave to said input terminals, a rectier connected to said output terminals, and a filter network connected between said rectifier and said control electrode, said network comprising a iirst branch including series resistive elements and a shunt capacitive element, and a second branch including a capacitive element connected in shunt with the rst branch, the latter capacitive element being of such value as to eliminate substantially phase shift of the modulation envelope at the output of the filter network.

2.y In combinatioman amplifier having input and output terminals and including an electron tube having a control electrode, means for applying a modulated carrier Wave to said input terminals, a rectifier connected to said output terminals, and a filter Vnetwork connected between said rectifier and said control electrode, said network comprising Vimpedance elements arranged in a T configuration and a capacitive element connected in shunt therewith, said capacitive element being of such value as to eliminate substantially phase shift of the modulation envelope at the output of the filter network, whereby the gain of said amplifier is varied as an inverse function of the strength of said carrier wave with no appreciable variation in the phase of its modulationl envelope. v

3. In combination, an amplifier for e radar scanners and the like having input and output terminals and including an electron Ytube having a control electrode, means for applying a carrier wave to said input terminals modulated according to scanner spin frequency, a rectifier connected to Ysaid output terminals, and a filter network designed to reject spin frequency components connected between said rectifier and saidV control electrode, said network comprising impcdance elements arranged in a parallel-T configuration, whereby the gain of said amplifieris varied as an inverse function of the strength of ksaid carrier wave with no appreciable variation .scanner spin frequency, a rectifier 'connected to said output terminals, and a non-inductive filter network `Connected between said rectifier and said control electrode forr rejecting modulation components, said network comprising impedance elements arranged in a bridge-T configuration,

whereby the gain of said amplifier is varied as an inverse function of the strength of said carrier wave with no appreciable variation in the phase of its modulation envelope.

5. In combination, an amplifier for radar scanners land the like having input and output-terminals and including an electron tube having a control electrode, means for applying a carrier wave to said input terminals modulated according to thespin frequency of the scanner, a rectifier connected to said output terminals, and a non-inductive filter network connected between said rectifier and said control electrode, said network comprising impedance elements arranged in a parallel-T' configuration so proportioned as to eliminatev modulation components, whereby the gain of said amplifier is varied as an inverse function of the strength of said carrier wave with no appreciable variation in the phase of its modulation envelope.

esas-12 6. In an automatic gain-control system for radar scanners and the like, in combination, an amplier having input and output terminals and including an electron tube having a control electrode, means for applying a carrier wave modulated ac-cording to the spin frequency of the scanner to said input terminals, a rectifier connected to said output terminals, and a lter network comprising resistance and capacitor elements arranged in parallel-T configuration connected between said rectifier and said control electrode, said network having substantially zero vphase shift and relatively low transmission at said modulation frequency, whereby the gain of said famplier is varied as an inverse function of the strength of said carrier wave with no appreciable variation in the phase of its modulation envelope.

7. In combination, an amplifier for radar scanner systems and the like having input and output terminals and including an electron tube having a control electrode, means for applying a carrier Wave modulated according to scanner spin frequency to said input terminals, a gain-control circuit including a rectifier connected to said output terminals, and a lter network of the non-inductive null type connected between said rectier and said control electrode, said network com- REFERENCES CITED VThe following references are of record in the me of this patent:

UNITED STATES PATENTS Number Name Date 1,836,556 Schelleng Dec. 15, 1931 2,106,785 Augustadt Feb. 1, 1938 2,240,533 Wilson May 6, 1941 2,284,102 Rosencrans May 26, 1942 2,302,866 Hunt Nov. 24, 1942 2,307,308 Sorensen Jan. 5, 1943 2,350,803 Newcomb June 6, 1944 2,382,097 Purington Aug. 14, 1945

Images