US2488606A - Frequency modulation receiver - Google Patents

Description

Nov 22, 1949 w. F. SANDS FREQUENCY MODULATION RECEIVER 2 shets-sheet 1 Filed May 2, 1945 .A KQ b, EN

.Nmmwn NOV 22, 1949 w. F. SANDS FREQUENCY MODULATION RECEIVER 2 SheetsfSheet 2 Filed May 2, 1945 ATTORNEY Patented Nov. 22, 1949 FREQUENCY MODULATION RECEIVER William F. Sands, Haddonfield, N. J., assigner to Radio Corporation of America, a corporation of Delaware Application May 2., 1945, Serial No. 591,484

(Cl. Z50-20) 2 Claims. 1

My present invention relates generally to frequency modulated (FM) carrier Wave receivers, and more specifically to an improved FM receiver of the type employing a frequency dividing locked-in oscillator prior to FM detection.

One of the problems encountered in all radio receivers, both amplitude modulation and fre-E quency modulation, is the presence of distortion in the audio frequency output of the receiver due to the presence at the receiver input terminals of an undesired signal at the same frequency, or very close to the same frequency, as that of the desired signal. In amplitude modulation (AM) reception on the present broadcast band there are a large number of channels each kilocycles wide, and, therefore, stations having the same frequency are either quite distant from each other, or else they do not, i-n general, operate during the same one-hour periods. Thus, the problem of common channel interference has usually not been Very serious, to date, in amplitude modulation reception.

However, the present assigned frequency modulation entertainment band extends only from 42 to 50 megacycles (mc.) With carrier separation of 200 kilocycles (kc.). This provides only 40 channels, but, in addition, the rst five channels are devoted to educational purposes leaving only 35 channels for entertainment purposes. Present plans call for FM broadcasting in the post-War period over a band .-of 88 to 108 mc. Whether IFMI broadcasting remains in the present frequency area or goes to the proposed area, it seems apparent that there will be many FM receivers located in regions in which sufficient signal energy will be received from at least .two FM transmitting stations on the same carrier frequency normally to operate the receiver.

In order to test the effect of such common channel interference, the outputs of two FM signal generators were fed tothe input terminals of an FM receiver employing cascaded amplitude modulation limiters. Both generators were tuned to 46 mc., as was the receiver. The desired signal was modulated i kc. at 400 cycles, and the undesired signal was modulated :i125 kc. yat 1000 cycles. It was found that for undesired signal strengths as low as 50% of that of the desired signal, the 400 cycle output of the FM receiver was decreased, and a considerable component of the 1000 cycle interference was .present in the output. When the signal strength of the two signals was the same, the 400 cycle and the 1000 cycle components Were, also, substantially equal. Further increase in the undesired signal strength caused it to become in effect the desired signal, and the 1000 cycle output predominated. Thus it was demonstrated that for an FM receiver with cascaded amplitude modulation limiters there was abroad region over which the audio output derived from the stronger FM signal suffered interference due to the weaker signal.

The FM receiver disclosed and claimed by George L. Beers in U. S. Patent No. 2,356,201, granted August 22, 1944, possesses a highly improved adjacent channel selectivity characteristic. The characteristic is due to the locked-in oscillator circuit feeding energy of divided frequency controlled by the received signal to the ,FM detector. The locked-in oscillator is constructed to be locked in With the applied FM wave over a range of frequency variations including, and substantially limited to, the range of frequencies of the modulated wave. I have discovered that there also exists a critical signal strength (developed at the signal grid of the locked-in oscillator tube) due to the signal at the receiver input terminals, at which an FM receiver of the Beers type also possesses a highly improved common channel interference suppression characteristic. Moreover, I have discovered that, in an entirely unexpected manner and at critical signal strength on the signal grid of the locked-in oscillator, a difference of only several percent in favor of the desired over the undesired signal enables the desired signal to take control, 4and to cause the audio output due to the weaker signal practically to disappear. My invention includes a solution of the problem of maintaining the desired common channel interference suppression characteristic over a Wide range of signal strengths at the signal collector.

As hereinafter indicated, it is an important object of my invention to provide a method of, and means for, maintaining a predetermined optimum desired signal strength at the locked-in oscillator input electrode despite Wide variations in the signal at the received input terminals thereby to obtain and retain the aforesaid critical, highly desirable common channel interference suppression characteristic.

In carrying `out this object of my invention, I provide a system of AVC (automatic volume control) for an FM receiver of the ,aforesaid Beers type, wherein the intensity level of the I. F. (in-v termediate frequency) signals at Vthe .signal grid ,of the locked-inoscillator is maintained substanf tially flat over a very Wide range of signal levels Vat the receiver input terminals. More specifically, I have replaced tne VC Ysystem of the aforesaid Beers patent by a parallel channel AVC system adapted to apply full AVC bias to prior amplifier stages and partial AVC bias to later stages thereby to keep the desired signal strength at the locked-in oscillator input grid at a constant value despite very wide changes in the strength of the signal at the receiver input terminals.

Still other objects of my invention are to improve generally an FM receiving system of the type using a locked-in oscillator as a frequency divider, and more especially to provide a system with improved common channel selectivity in a reliable, eiiicient and economical manner.

Wherever the term frequency modulation is used throughout the specification and claims, it should be understood to refer to any modulation wherein the instantaneous frequency of the transmitted waves is varied by the application of modulating Voltage of an alternating character such as music or speech. There are many possible functional relations between the instantaneous wave frequency and the modulating voltage, which are, or can be, used. For example, if the instantaneous frequency is caused to shift in direct proportion to the instantaneous amplitude of the modulating voltage, there results one common form of frequency modulation; or if the instantaneous frequency is caused to vary as the time integral of the modulating voltage there results a type of frequency variation, which is usually called phase modulation because it is with equal correctness, and somewhat more simply, definable as a modulation which causes the phase of the transmitted waves to shift in direct proportion to the instantaneous amplitude of the modulating voltage.

In other words, the term phase modulation and frequency modulation are tied together by the fact that a changing frequency necessitates a changing phase, and vice versa, Furthermore, in many practical systems going under the name of frequency modulation (FM), the instantaneous frequency does not vary directly as the voltage of the modulating voltage, nor yet as its integral, but in some intermediate fashion. Regardless of the exact nature of the functional relation mentioned above, however, the system, of the present invention can be employed, and, hence, such terms as frequency modulationf. frequency modulated, and the like should be taken in the broad sense here defined.

Other features of my invention will best be understood by reference to the following illustrative description taken in connection with the drawings in which I have indicated diagrammatically several circuit organizations whereby my invention may be carried into effect.

In the drawings:

Fig. 1 is a schematic circuit diagram which illustrates one form of an FM receiving system embodying the invention;

Fig. 2 is a schematic circuit diagram illustrating another form of FM receiving system incorporating my invention and providing certain performance advantages over the system shown in Fig. 1; and

Fig. 3 shows graphs illustrating certain operating characteristics of the invention.

Referring to the accompanying drawings, wherein like reference characters in the different figures designate similar circuit elements, the signal input circuits of my receiving system may be of any suitable type. To simplify the drawings, block diagrams are employed to represent the radio frequency amplifier 5, first detector 6, local oscillator 1, and intermediate frequency (I. F.) amplifiers 3 and 8 of the usual superheterodyne type of vacuum tube receiver circuit. A conventional dipole antenna is indicated at 9 as a suitable collector of signals for the radio frequency amplifier 5.

The intermediate frequency amplifier 8 is provided with an output transformer l0 having primary and secondary circuits resonant to the intermediate frequency. As illustrated, the primary and secondary circuits of each of transformers I0 I0 are suitably coupled to give each of them a resonance curve substantially as depicted above transformer I0. In the present example, the intermediate frequency may be assumed to be 4.3 mc., and the transformers I0 and I0 are designed to pass with suitable tolerances the frequency band occupied by the modulated carrier wave with modulation. For the frequency modulation broadcasting standards in use at the present time, which provide for kc. maximum frequency swing of the transmitted signal wave, the input circuits of the receiving system, including the intermediate frequency transformers, are constructed to respond eiciently over a frequency band somewhat greater than 150 kc. to allow for frequency drift of local oscillator i, over-modulation at the transmitter, etc.

It will be understood by those skilled in the art that the stages 5, 6 and 'I are provided with the usual variable tuning means. The input circuits of amplifier 5 and first detector 6 are for best reception each tuned to the mean or center frequency of the desired station channel by respective tuning devices 5 and 6', while the oscillator 'I is tuned by device 'I to an oscillation frequency differing from said mean frequency by the value of the desired intermediate frequency. A common control I actuates the adjustable elements of tuning condensers 5', 6 and l.

The output circuit II is followed by a lockedin oscillator I2 and a suitable frequency selective detector network by which the applied signal is converted to amplitude modulation. The signal is detected by suitable rectifiers. The audio frequency, or modulation, output connection lead 2l is coupled to the detector circuit I4 through a suitable capacitor 22. The FM detector may be of the form shown in the said Beers patent or, alternatively, of the form shown by S. W. Seeley in U. S. Patent No. 2,121,103, granted June 21, 1938. If desired, the FM detector circuit of Conrad U. S. Patent No. 2,057,640, granted October 13, 1936, may be employed.

The oscillator I2, which immediately precedes the discriminator network, produces oscillations of frequencies less than the intermediate frequency, and having a frequency swing less than that of the intermediate frequency wave. The oscillator tube envelope contains a grid 35 connected to the intermediate frequency secondary circuit Il, which, in turn, is connected to ground through a by-pass condenser I2. In the present example, the tank circuit I3 connected to the anode I0 of oscillator I2 comprises a variable magnetic core inductance 3l and a shunt tuning capacitor 32, and is tuned to a sub-multiple frequency of the intermediate frequency signal applied to the grid 35 to cause the tube I2 to produce oscillations of such sub-multiple frequency.

The circuit I3 is grounded through a by-pass condenser 30, and is caused to oscillate at the predetermined sub-multiple frequency by feedback through inductive couplings between the inductance SI and an inductance indicated at 42. The latter has a shunt tuning capacitor 43 included in a circuit 4l, and is connected between a `second control grid 44 and ground, or cathode, through a suitable grid resistor 45 shunted by a grid capacitor 6. The normal oscillator frequency is primarily determined by the tuning of circuit I3. The tube I'2 may be a pentagrid tube and may further be of the 6SA7 type, and is preferably locked up Vby the impression of the 4.3 mc. signal on the first grid of the tube, which is the grid 35 in the present example. The cathode 36 of tube I2 is grounded. Screen Vgrids and a suppressor grid are arranged as shown on the drawing in Fig. 1.

Coupling connection between the locked-in oscillator and the FM Adetector is provided through a lead 26, a resistor 25, and a coupling capacitor 2l connected to the anode 40 of the e oscillator. Since the oscillator I2 applies its output voltage to the discriminator input circuit whether a signal is being received or not, and since both the frequency and the frequency swing of the voltages which it produces are different from the intermediate frequency signal which is applied to the grid 35 it is apparent that the radio frequency signal waves collected at the signal collector device, such as the usual antenna, do not themselves pass through the receiving system to the discriminator, but operate to control the frequencies of the voltages produced by the oscillator. I

By proper choice of the circuit constants of the oscillator I2 and in particular of the coupling between the circuits I3 and 41 and the operating potentials applied to the tube electrodes, the oscillator voltage applied to the discriminator can be made substantially independent of the strengthv of the intermediate frequency signal.

Since the locked-in oscillator I2 operates at a sub-multiple of the instantaneous frequency of the modulated carrier impressed on grid 35, frequency division is obtained in which not only the mean carrier frequency is reduced, but also the frequency swing is correspondingly reduced. In the present example with .a mean intermediate frequency of 4.3 mc. supplied by the circuit Il, the tank circuit I'3 may be tuned to a submultiple frequency of 860 kc. therebyproviding a 5:1 frequency reduction and a corresponding reduction in the frequency swing. For example, if at any instant, the received signal has been heterodyned to an intermediate frequency of, for example, 4.375 mc (up 75 kc. from the mean or center frequency of 4.3 ma), the frequency of the locked-in oscillator will then be 875 kc. (up kc. from the mean sub-multiple frequency of 860 ko), preserving the 5:1 ratio with the intermediate frequency on grid 35. The FM detector circuit I4 will also be responsive to a band of frequencies, at least kc. wide and centering at 860 kc.

As heretofore indicated, the sharp common channel interference suppression characteristic of the system of Fig. 1 depends upon an optimum and critical value of signal attheY grid 35. In Fig. 3 I have shown actual performance curves taken on an FM receiver of the Beers type and as shown in Fig. 1, with signal inputs provided by a pair of FM signal generators. Each of the signal generators. impressed, on radio frequency amplifier 5 frequency modulated:V energy having a mean frequency of 46 mc, The tuning .control vI was adjusted to tune the receiver to 46 mc. A desired signal was modulated- $25 kc. at 400 cycles, and an undesired signal was modulated :1:25 kc. at 1000 cycles.

Fig. 3 shows three pairs of curves, d, e and f, secured for particular levels of the desired signal. The abscissae are microvolts of interfering signal, while the ordinates are audio millivolts. The set of curves d shows the relation between the desired signal curve (shown as a dash line) and the interfering signal curve (shown as a solid line) for a desired signal of 6 microvolts. The set of curves e shows the relation between the same curves at a desired signal of 100 microvolts. The curves f show the relation at a desired signal level of 6000 microvolts. It will be seen from Fig. 3 that the relation existing at e is desirable because the area of overlap of the desired and undesired signals is very small, it being seen that even at 90 microvolts very little audio response to the undesired signal was obtained. This is the interference suppression characteristic desired due to the performance of the locked-in oscillator with critical signal strength on its control grid 35. The overlap at d permits substantial response to interfering signals at strengths well under 6 microvolts. It may be pointed out that an FM receiver with cascaded limiters cannot exceed the selectivity illustrated at d.

It is to be noted that a change of only several percent in the ratio of desired signal level to undesired signal level at e is sufficient to cause either one of the signals to take control and the audio output due to the weaker signal practically to disappear. It may, also, be seen that the magnitude of the audio output of the stronger signal is not decreased until the respective strengths of the two signals differ by less than 10%. Further increase in the intensity level of the desired signal changes the character of the curves until at 6000 microvolts level, for example, the relation at f exists. While the f curves are superior to those at d, they are decidedly inferior to the curves shown at e.

I have invented a relatively simple and effective method of maintaining the relation adapted to produce the two-signal, common channel interference curves illustrated at e of Fig. 3 despite a very wide range of signal input levels at signal collector 9. In general, I utilize a parallel channel AVC system in conjunction with the lockedin oscillator circuit. Such a system will maintain the level of the I. F. signal voltage at the grid 35 substantially flat over a very wide range of signal levels at collector 9. The at AVC operation may be secured by proper adjustment of the various AVC circuit constants, and will depend somewhat upon the design of the remainder of the FM receiver.

Referring again to Fig. 1, the I. F. signal voltagefor the AVC channel is derived from the secondary winding of the transformer i0. Prior to rectification of the I. F. voltage, the latter is applied through condenser IE to the input electrodes of an I. F. amplifier 4I. The latter may be of any suitable construction, and may be constructed in a manner similar to the stages 8 and 8. The amplied output of amplifier 4I is applied todiode rectifier I'I. The I. F. transformer I8, whose primary and secondary circuits are each tuned to the operating I. F. value of 4.3 mc., has flat selectivity over the nominal range of deviation of the transmitting .stations (i. e. at least :':75 kc.) in order that the AVC action will not change for various amounts of modulation at the FM transmitter.

The diode I1 includes a load resistor I9 in circuit between the grounded cathode and the low potential side of the secondary circuit 20. A positive direct current voltage is applied to the ungrounded end of resistor I 9 through a resistor 53, the resistor being bypassed by condenser 23 for I. F. currents. There will be developed across resistor I9 a direct current voltage whose magnitude is directly proportional to the magnitude of the I. F. signal voltage at the high potential side of the secondary circuit of transformer I8. 'I'he positive voltage, which is an AVC delay voltage preferably applied through a large resistor 24, functions to prevent the performance for weaker signals.

It will be understood that the positive delay voltage may be derived from any suitable positive voltage point in the direct current supply of the receiver. The anode end of resistor I9 is connected to ground through a resistor 28, and I preferably adjust slider 29 to a predetermined intermediate point on resistor 28 such that only a portion of the AVC voltage across resistor I9 is applied to the signal grid of the I. F. amplifier 8. This is accomplished by connecting the low potential side of the secondary circuit of transformer I to slider 28 through a filter resistor 48. The lower end of resistor 48 is connected to ground through a filter condenser 49. In Fig. 1 the full AVC voltage is applied to the signal grids of all the preceding stages 5, 6 and 3 by connecting the signal grids thereof through respective filter resistors 50, I and line 52 to the upper end 54 of resistor 28. Of course, the AVC filter resistors and condensers will be chosen to prevent any pulsating components of the rectified voltage output of diode I1 from being applied to the signal grids.

The positive potential E is derived from a suitable point in the receiver such as a tap on the common high voltage supply or by utilization of the voltage across the cathode resistor of the audio output tube. If, for a moment, consideration of the effect of the controlled grids upon the potential at point 54 may be postponed, then the potential at point 54 due to E may be determined. Now, if, by way of example, we let E=11.0 volts, resistor 24 be 8.2 megohms, resistor 53 be 820,000 ohms, resistor 28 be 1.2 megohms, and resistor I9 be 120,000 ohms and the measured contact potential at point 55 is found to be 0.5 volt, then the voltage from point 54 to ground will be +1.() volt. Now, with no signal present, the potential from 54 to ground would actually not be -l-l.0 volt because the grids of the various controlled tubes draw current upon application of the positive potential, and thus effectively reduce the voltage at point 54 to zero voltage, or even to a slightly negative value. When the signal is applied, and gradually increased in value, the controlled grids no longer draw current. The current through resistor 53 due to voltage E is in opposite sense to the current due to the developed voltage at point 55, and the AVC voltage is, therefore, delayed" for weak signals.

The slider 29 is adjusted to a point on resistor 28 such that the AVC bias applied to the controlled stage 8' is less than that applied to the preceding amplifiers. The amount of AVC voltage used on the last I. F. stage should be just enough to make the voltage across resistor I2" flat. The desired setting of slider 28 is readily determined by this means. In order to secure the superior receiver performance in maintaining the critical desired signal intensity at the locked-in oscillator input grid, it is only necessary to adjust slider 29 as directed. The adjustment of the slider will be such as to distribute the AVC bias so as to maintain the relation shown by the curves e in Fig. 3. Of course, the required level of I. F. signal voltage at the locked-in oscillator grid will depend upon the particular type of tube I2 used for the locked-in oscillator, upon the Q and L./C. ratios of the locked-in oscillator circuits, and, also, upon the self-bias arrangements of the grids 35 and 44. Having chosen the proper constants for the locked-in oscillator network, for which reference may be made to the aforesaid Beers patent, the combination of full and partial AVC will maintain the desired I. F. level at the critical value satisfying the requirements of e in Fig. 3.

In Fig. 2 I have shown a modification of the system of Fig. 1, wherein the I. F. signals for ampliiier 4I are derived from the primary circuit of transformer IS. This provides a broader selectivity curve (substantially as shown above the primary) than at the secondary circuit. Under certain conditions this added broadness at the input to the parallel AVC channel may be needed. In addition, the delay voltage (secured from a suitable point of positive potential in the receiver) is applied to the cathode of the AVC rectifier II. A resistor 33 (whose Value is low in comparison with that of resistor 28) is connected from the cathode to ground, the resistor 33 is bypassed by the capacitor 34. The condenser 31 connects the ungrounded end of resistor 28 to the positive end of resistor 33. The AVC voltage is not developed across the resistor 33, but is actually developed across the resistor 28. The resistor 33, which is of low value compared to the resistor 28, serves to provide a direct current path to ground for the cathode, and thus completes the diode circuit.

The capacitor 34 is a bypass for the I. F. voltage, which would otherwise be developed across the resistor 33. The capacitor 31 is the AVC iilter capacitor. The operation is such that the diode does not conduct until the peak value of the I. F. signal across circuit 20 exceeds the positive direct current potential at the cathode of the diode I1. The AVC bias is taken oli from across resistor 28. The full AVC lead 52 is connected to the anode end of resistor 28, and applies the AVC bias to the signal grids of amplifiers 5 and 8. The slider 29 is adjusted to apply a portion of the AVC bias to the respective signal grids of the first detector 8 and the amplifier 8 through suitable filter resistors 48. The reason that partial AVC is shown on both the 1st detector and the last I. F. stage is as follows: Application of full AVC is sometimes not advisable on converter tubes in which both the mixer and oscillator functions are performed in the same tube structure (i. e., self-excited converters). The effect of full AVC on such tubes is that the oscillator frequency is shifted by changes in the strength of the received signal. The use of partial AVC obviously causes less shift in frequency.

While I have indicated and described several systems for carrying my invention into effect, it will be apparent to one skilled in the art that my invention is by no means limited to the particular organization shown and described, but that many modifications may be made without departing from the scope of my invention.

What I claim' is:

1. In a frequency modulation receiver of the type comprising a plurality of cascaded selective amplifiers, a frequency divider of the locked-in oscillator type having input electrodes coupled to the last of said selective amplifiers, said oscillator having a relatively sharp rejection characteristic for interfering signals of the same carrier frequency b-ut of relatively weaker intensity, an automatic gain control circuit, responsive to variations of desired signal amplitude at an amplifier preceding said frequency divider, for controlling the gain of said amplifiers so as to maintain the value of signal at the oscillator input electrodes substantially constant regardless of wide variations in the ratio of desired to interfering signals at the receiver input terminals, said gain control circuit comprising a rectifier, means for delaying rectification by the rectifier, means for applying the full rectifier output voltage to at least the first of said cascaded amplifiers, means for applying a part of said rectifier output voltage to a later one of the amplifiers, and means for adjusting the value of the partial Voltage.

2. In a frequency modulation receiver of the type comprising a plurality of cascaded selective amplifiers, a frequency divider of the locked-in oscillator type having input electrodes coupled to the last of said selective amplifiers, said oscillator having a relatively sharp rejection characteristic for interfering signals of the same carrier frequency but of relatively weaker intensity, an automatic gain control circuit, responsive to variations of desired signal amplitude at an amplifier preceding said frequency divider, for controlling the gain of said amplifiers so as to maintain the value of signal at the oscillator input electrodes substantially constant regardless of Wide variations in the ratio of desired to interfering ignals at the receiver input terminals, said gain control circuit comprising a rectifier, means for delaying rectification by the rectifier, means for applying the full rectier output voltage to at least the first of said cascaded amplifiers, and means for applying a part of said rectifier output voltage to a later one of the amplifiers.

WILLIAM F. SANDS.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 2,144,304 Braden Jan. 1'7, 1939 2,356,201 Burs Aug. 22, 1944

Images