US1988609A - Synchronizing system - Google Patents

Description

Jan. 22, 1935. A. H. RIEYEVES SYNCHRONIZING SYSTEM Filed July 21, 1931 3 Sheets-Sheet 1 JNVENTOR A. H REEVES A TTORNEV Jan. 22, 1935. A. H. REEVES ,6

" SYNCHRONIZING 'SY-STEM Filed July 21, 1951 'sheets-sheet 5 INVENTOR V AHREEVES T/h wwa A TTORNE Y Patented Jan. 22, 1935 UNITED STATES 1,988,609 smcnnomzmc SYSTEM Alec Harley Reeves, Paris, France, minor to Western Electric Company, Incorporated. New York, N. Y., a corporation of New York Application July 2 1, 1931, Serial No. 552,179 In Great Britain August 1, 1930 3 Claims. (Cl. 250-36) This invention relates to non-mechanical means for controlling one or more characteristics of a network according to the value of a current or potential difference in an associated circuit.

An object of the invention is to provide an arrangement whereby a constant frequency difference may be maintained between two oscillators remote fromeach other.

Another object of the invention is to provide a variable reactance without mechanical control.

In accordance'with a feature of the invention, an arrangement is provided in which the effective reactance of a network is varied in accord- .15 ance with the variation of an associated impedance, said impedance being located in said ,network or in another network coupled thereto. The variable impedance may be the space discharge path of a vacuum tube, impedance variation being obtained by varying the biasing potential.

According to a further feature of the invention ,a predetermined relation between the frequencies of two or more oscillators may be obtained by controlling the conditions of oscillations of one or more oscillators, independently of the conditions of oscillation of one or more 01' the remaining oscillators.

Another feature of the invention resides in a high frequency signaling system comprising two oscillators, one at a transmitting station and the other at a receiving station. A portion of the output from the transmitting oscillator is transmit ed to the receiving station where two tuned circuits with overlapping resonance curves act differentially on a variable impedance. When the currents in the tuned circuits are not equal, variations in the impedance modify the tuning of the receiving oscillator,.thus keeping the frequency of the two oscillators difierent from each other by a substantially constant amount.

These and other features of the invention will be described with reference to the accompanying drawings, in which: 7

Fig. 1 shows a circuit arrangement in which an inductance is adapted to vary depending on the variations of a resistance;

Fig. 2 shows an arrangement of the invention in which an inductance is-adapted to vary as a function of the variations of the biasing potential applied to the grid or. control element of a three element vacuum tube; I

Fig. 3 shows a slight variation of Fig. 2;

of the circuit as one. The grid voltage being then varied. the

Fig; 4 shows an improved frequency multiplier device of the invention involving the principle illustrated by Fig. 3;

Fig. 5 shows a frequency divider device of the invention based on the same principle as the 5 multiplier of Fig. 4;

Fig. 6 shows an arrangement for synchronizing a local oscillator to the average frequency of a ource whose frequency is varying;

Fig. 7 shows a resonance curve;

Fig. 8 illustrates the application ofan automatic gain control arrangement;

Fig. 9 illustrates an alternative way of using a receiver embodying features of the invention; and.

Fig. 10 illustrates a circuit which is effectively a combination of the circuits of Figs. 6 and 8, that is, an adaptation of the circuit of Fig. 6 to include the automatic gain control arrangement of Fig. 8.

' Fig. 1 discloses a network comprising inductances L1 and L2 and a resistance R1 in series and having terminals A and B. Shunting inductance L2 is a resistance R: whose variation will modify the inductance between the termi-' nals A and B. This will appear from the consideration of the total impedance of the circuit which is given by the following equation:

Otherwise stated above, resistance R: changes the total effective reactance of the circuit. If the resistance R: be replaced by a three-electrode 35 vacuum tube T as shown in Fig. 2 and if the platev impedance of this tube be varied by changing the grid bias, for example, the effective in'ductance of the whole circuit from A to B would be changed and therefore if a condenser C is provided in parallel between A and B as shown in Fig. 2 the resonance frequency of the circuit thus formed will be changed in accordance with variation of the biasing potential of the vacuum tube. It should be noted that the above method of controlling reactance by means of a vacuum tube is not limited to its use in conjunction with resonant or anti-resonant circuits, but the application to resonant circuits is merely given for illustrative purposes.

A modification of the arrangement of Fig. 2 is shown in Fig. 3. In this case the inductance L2 instead of being in series with the main inductance is coupled to it to a degree less than and the frequency of the wave total eifective inductance will also vary. The resistance component of the total impedance will also be changed to some extent, but in the practical application of this device, this resistance variation is unimportant in its effect.

It should be noted that the element La shown in Figs. 1 and 2 may alternatively .be a condenser while maintaining the essential principles of the system.

. In the case of Fig. 3 the coupling between the two coils may be performed by means of either an air core or of a core of magnetic material.

Fig. 4 shows an application of a non-mechanical reactance control in a frequency multiplier system In Fig. 4, S represents a standard of low frequency such'as a tuning-fork, or a magnetostriction, oscillator. Supposing that it is desired to obtain in the output of the frequency multiplier a frequency corresponding to a high harmonic of the standard frequency, for instance, the harmonic corresponding to the low frequency when multiplied by 8192. For this purpose an oscillator 1, of any well known type, is adjusted to a frequency within two or three per cent of the desired harmonic and its output frequency is divided by 8192 by means of an aperiodic frequency divider diagrammatically indicated by F which may be, for example, like the one disclosed in British Patent No. 296,827, both the frequency divider output and the output of the standard low frequency source being connected to the grid of a rectifier tube 2.

The voltage drop across the plate resistance R of rectifier 2 is used as the grid bias of the low impedance control tube 3 which has a plate coil D coupled to the oscillatory circuit of oscillator 1.

The operation of the frequency multiplier system will now be apparent. The voltage drop across resistance R will vary from a small to a high value, at the beat frequency between the outputs of divider F and standard oscillator S. Let us consider a fewcycles of the wave from oscillator S during which the output wave of divider F lags behind that of oscillator S. If the frequency of the wave from oscillator 1, for any reason, becomes slightly greater thanthe required harmonic of the wave from oscillator S, then the two low frequency voltages on the grid of tube 2 will have a smaller phase difference than before, which will cause the voltage drop across R to increase.

The resulting increase in the plate resistance of control tube 3 will cause the coil D to have less shunting eifect than before on the inductance with which it is coupled so as to effectively increase such inductance and cause the frequency of oscillator 1 to be lowered, i. e., the two low frequency outputs will tend to become 90 apart as before.

If, for any reason, the frequency of the wave from oscillator 1 tends to be reduced below that giving zero beat between the low frequency outputs, exactly the reverse action takesplace, the two low frequency voltages becoming more out of phase, the grid bias of tube 3 being reduced from oscillator l correspondingly raised.

The result is an automatic synchronizing action of the standard low frequency source on the high frequency oscillator 1, such that the frequency of the latter is exactly that of the required harmonic, the order of the harmonic being determined by the frequency divider F. This scheme has the distinct advantages that only one tuning control is required namely the main oscillator tuning condenser.

In Fig. 5 there is illustrated a frequency divider of the invention adapted to be used in measuring a relatively high frequency. The device operates on the same principle as the multiplier of Fig. 4 although in a reverse manner. The frequency to be measured, of the order for instance of 20 megacycles, is applied to the grid of the screened grid tube E. This tube amplifies the unknown frequency and at the same time prevents back coupling on to the source of this frequency. Oscillator l and control tube 3 are of the same design as the similarly identified element in Fig. 4. The oscillator output is however now appliedto the harmonic generator M. The wave from oscillator '1 is in the region of 1 megacycle, that is, the limit of the working frequency of the aperiodic divider. The circuits of M and 1 are adjusted so that a particular harmonic, for example the twentieth, is at a frequency within 2 or 3% of the frequency to be measured. The output of devices M and of E are applied together to the rectifier tube 2 which is of the' same design as the similarly identified tube in Fig. 4 and performs a similar function.

From the explanation given in the previous section it will now be clear that the action of the device is to synchronize automatically the frequency of the wave from oscillator 1 (which is of the order to be measured by the aperiodic frequency divider) so that a particular known harmonic of that frequency has exactly the same frequency as that to be measured, in other words,

so that the frequency of. the wave from oscillator 1 is a definite sub-multiple of the unknown frenary. stable oscillator to re-supply the carrier' frequency. When working in the region of 20 megacycles, however, this method is very difficult, even when the best quartz crystal oscillators are used, as the local carrier must not differ by more than 20 cycles from theoriginal suppressed carrier. Up to the present date crystal oscillators have not been developed sufliciently to give the required stability under commercial conditions, i. e. a maximum frequency difference of one part in a million (which is a very severe requirement). An alternative solution is to synchronize automatically the oscillator at the receiver with the original carrierfrequency by means of some synchronizing signal transmitted over the circuit. The power in the latter signal may be quite small compared with the side band power, and therefore, the increased efliciency given by the single side band method is not appreciably impaired.

"A method which may conveniently be used in a short wave single side band receiver is to transmit a small amount of ordinary carrier frequency as well as an inverted single side band. At the inverter at the same time as matically at, for instance, 4000 cycles away from the original carrier. As in the transmitting inverter equipment the speech side band is displaced upwards by 1000 cycles as well as being inverted, the resulting beat frequencies at the receiver give ordinary straight speech, in other words the present single side. band receiver acts as a receiving it performs its normalfunctions. 1

One of the chief difficulties to overcome isdue to thefact that when using short waves, the carrier or any particular frequency fades out completely for short intervals, so that if a single frequency is used as a synchronizing signal the resulting synchronizing action is intermittent. Therefore it is necessary to use the average and not the instantaneous effect of the synchronizing signal. The method by which the functions described in this and the precedingparagraplis have been accomplished is shown in Fig. 6.

The screened grid amplifier tube 4 (Fig. 6) has the control grid connected to the source of frequency which is to be used as a synchronizing signal (i. e. the partially suppressed carrier of the distant station, stepped down to intermediate frequency).

The output of 4 and also of the local oscillator 5, which is connected to the local receiving circuit (not shown) by circuit 12, are connected to the gridof rectifier 6 in the plate circuit of which the beat frequency between the local oscillator and the synchronizing signal appears.

until this heat note is approximately the frequency required and the output voltage at this frequency the resonance points are spaced equally on opposite sides of the desired beat frequency as shown in Fig. "l, and the decrements adjusted, for ex-' ample,-so that, at exactly the frequency required, each resonance curve is 6 decibels down from its peak value. When this condition occurs, it is evident that the direct current outputs in rectifiers H and K will be equal and therefore thatthere will be no voltage difference between the points 9and 10, which are, respectively, at the alternating potentials of the cathode and control electrode of the control tube L, that is the total grid bias of the control tube L will only consist of the normal value given by battery 11, just sufflcient to bring the tube to the middle of the lower bend of its characteristic.

The circuit is arranged so that if, for any reason, the beat note between the local oscillator and the synchronizing signal departs from the desired value the resulting inequality in outputs of H and K gives a voltage difference between 9 and 10 in such a direction as to cause the resulting change in impedance of L to change the frequency of 5 torestore the original beat frequency.

As it is desired to use, not the instantaneous but the average effect of the synchronizing signal, the high value resistances N and P and condenser Q are added to give the required time constant to the frequency changing device.

This time constant is adjusted to such a value that "when the synchronizing signal disappears for its maximum length of time the local oscillator B can never drift more than the allowable number of cycles from its desired frequency. The output at the frequency required is then taken at point- Fig. 8 illustrates an arrangement somewhat analogous to that of Fig. 6 but provided with au- The frequency ofthe wave from oscillator 5 is adjusted.

tomatic gain control means. The reference characters used in this figure correspond to those used in Fig. 6, that is, like parts are designated by the same reference characters.

The signals (waves from distant transmitter and local oscillator) are applied on the input of rectifier 6 as in Fig. 6, and we will consider below what happens when the signals change in amplitude as may occur under fading conditions.

Considering first what takes place when there is a change of amplitude in the received signals, it will be seen that:

l. A change of amplitude will change the output of the rectifier tube 6 which after amplification in the amplifier LFP is applied through a transformerTRtoihegrids of a balanced detector,

comprising two'tubes- H and-K." Supposing that the amplitudeof the receivedsignals increases then the plate currents of the two tubes' Ha'ndK' will also increase thus increasing the voltage drop in resistances R1 and Re and so changing the potential of point 13 with respect to ground. It will be seen therefore that the potential of points. 13 will vary depending upon the change of amplitude of the received signals applied at the input of rectifier 6. The potential of point 13 is applied through a biming battery and through a choke coil CK to the grid of the rectifier 6 the adjustment being such that the change of the potential of point 13 restores substantially the output of rectifier 6 to its original value.

2. When the frequency of the received signals mains substantially constant in spite of change within certain limits in the frequency of current applied to tube 6 since, when different potentials I are set up across resistances R1 and R6 due to change of frequency, the average of these potentials will determine the potential of 13 with respect to ground. It will be clear from an examination of the resonance curves shown in Fig. 7

that this average potential will be substantially constant since, by symmetry ordinates A'B and A"B" are equal and the sum of either and the corresponding ordinate of the other curve will constantly equal value A3 or twice the ordinate at the crossing point.

It will be seen that the arrangement illustrated in Fig. 8 fulfills two functions:

(a) to maintain approximately constant the amplitude of the output of the rectifier 6;

(b) to provide a variation in the plate current of a tube L depending upon a variation of the frequencies applied at the input of the rectifier 6. This variation of plate current of the tube L may be used to restore the resulting output frequency to its originalvalue in the manner above de scribed.

Fig. 9 is a block schematic diagram of the receiving equipment of a high frequency signaling system embodying features of the present invention. This system is a high frequency signaling system of the single side band partly suppressed carrier type.

In this figure 14 is a device on which is applied odyne receiver which may, for instance, be with-' in the range of 500 to 503 kilocycles for one received speech side band. The output of '14 is applied to a frequency changer FG, the output of which (within the range 20 to 23 kilocycles) is applied to a repeater 15 followed by a filter 16 of pass range equal to 20 to 23 kilocycles. The output of the filter 16 is applied to a balanced demodulator BDM associated with a local carrier oscillator COSC adapted to give a frequency of 20 kilocycles. Finally the original signals, for instance, speech, are restored in the terminal output equipment S0.

A synchronized oscillator SOSC of say 520 kilocycles is associated. with the frequency changer FG and with a gain control and rectifying device GR similar to device of Fig. 8 which is associated with an oscillator 080 of 16 kilocycles and with a repeater 17 whose output passes through low frequency tuned circuits LF andthence to a balanced detector BD. The balanced detector BD and the gain control device GR are associated through a gain control lead 18. Finally the output of the balanced detector ED is applied to a reactance control device RCC which adjusts the synchronized oscillator 3080, through a synchronizing lead 19.

When the signals are transmitted over a path including a wireless path subjected to bad fading conditions, it has been found necessary in a system of the above type to limit the amplitude of the beat note (for example, 4000 cycles) before applying it to the tuned circuits 7 and 8 in Fig. 6. Without such a device it was found during heavy fading that the controlled frequency depended to some extent on the amplitude of the incoming signals as well as on its frequency thus preventing synchronization within the required limits.

In order to avoid'these drawbacks means are provided, as explained above, whereby the grid bias of the rectifier giving the beat note is obtained from the average voltage drop in two resist-ances associated with .a balanced detector. As long as the beat note is within the working limits therefore, the above mentioned grid bias depends only on the amplitude of the received signals and not on its frequency. The increased bias lowers the gain of the detector when the signal amplitude increases so that the resulting beat frequency voltage is substantially constant within wide limits of signal voltage.

Alternatively the synchronized oscillator has its frequency shifted about 20 kilocycles away from the frequency of the incoming, partly suppressed, carrier frequency and acts as a beating oscillator; the resulting side band (on the 20 kilocycle range) is then put through a filter having sharp cut off on each side of the received side band. An ordinary stable oscillator at the required frequency (about 20 kilocycles) is then used to re-supply the carrier frequency into a balanced demodulator.

Fig. 10 illustrates what amounts, effectively, to a combination of the features illustrated in Figs. 6 and 8. The lettering is the same as that which has been used in said Figs. 6 and 8, these indicating the functions of the respective elements in such a way as, with the like physical arrangement of elements in the figure, to obviate the necessity of explaining the figure in detail. In part Fig. 10 is a reproduction of Fig. 8 to include the circuit and structure anterior to the 76 combining device 6, this inclusion therefore cov ering tubes 4 and 5 and their immediately associated circuits. Then, in order to make the system complete, the output circuit of tubeL of Fig. 8 has been connected to the oscillator 5 of Fig. 6, which it controls, this connection being shown exactly as in Fig. 6. It is evident therefore that Fig. 10 differs from the circuit of Fig. 6 substantially only in the addition of the gain control feature of Fig. 8, Fig. 10 therefore disclosing both the gain control feature of Fig. 8

and'the frequency averaging feature of Fig. 8.

It will be seen from the above that the present invention may be embodied'in widely different structures and is particularly useful in high frequency communication systems particularly in those of the carrier suppressed or partly suppressed types.

It will be readily apparent that instead of employing resonantcircuits with overlapping resonance curves, filter circuits may be employed having overlapping attenuation curves. By this ,means, a greater degree of control of the fre quency range over which synchronizing occurs is possible. For instance,.by employing low pass filters of which the attenuation curves slope in opposite directions, synchronizing will occur over the width of the frequency band included within the two slopes, and in general, this band can be made wider than that included within the two resonance peaks shown in Fig. 7.

What is claimed is:

1. A high frequency signaling system comprising a controlling wave source and a controlled oscillator, a resonant frequency determining circuit in said controlled oscillator, means for combining the outputs of said wave source and oscillator to obtain a beat frequency, two circuits having overlapping resonance curves each in energy flow relation with said combining means,

means comprising a variable shunt circuit whereby the differential current from said two circuits act differentially upon a variable tuning impedance in the frequency determining circuit of the controlled oscillator in such a manner that the frequency of the wave from said controlled oscillator when varied relatively to that of the wave from said controlling wave source is modified to compensatorily restore the initial frequency relation between currents from said wave source and oscillator, and means comprising a series resistance and a shunt condenser in the input path of said variable shunt circuit whereby the tuning of the controlled oscillator varies as a function of the average frequency of the wave from the controlling wave source over a given interval of time.

2. A high frequency signaling system comprising a controlling wave source and a controlled oscillator, a resonant frequency determining circuit in said controlled oscillator, means for combining the outputs of said wave source and oscillator to obtain a beat frequency, two circuits having overlapping resonance curves each in energy flow relation with said combining means,

means whereby the differential current from said' two circuits act differentially upon a variable tuning impedance in the frequency determining wave source and oscillator, and a time constant circuit embodied in said last mentioned means whereby the controlled oscillator is given such a time constant that-its average frequency of oscillation maintains substantially its initial value it the controlling waves cease for an interval of time dependent on said time constant.

3. A high frequency signaling system comprising a controlling wave source and a controlled oscillator, a resonant frequency determining circuit in said controlled oscillator, means for combining the output of said wave source and oscillator to obtain abeat frequency, two circuits having overlapping resonance curves each in energy flow relation with said combining means, meanswhereby the differential currents in said two circuits act differentially upon a variable impedance in the frequency determining circuit of the controlled oscillator in such a manner that the frequency of the wave from said controlled oscillator when varied relatively to that of the wave from said controlling wave source is modilied to compensatorily restore the initial frequency relation between currents from said wave source and oscillator and a gain control means directly associated with said last mentioned means tor rendering the frequency control function of the system substantially immune from the eflects of amplitude variations of the waves received from the controlling oscillator.

ALEC HARLEY REEVES.

Images