US2447040A - Modulation system - Google Patents

Description

g- 1948- E. K. STODOLA 2,447,040

MODULATION SYSTEM Original Filed Nov. 17, 1942 3 Sheets-Sheet l i F!G.6. MOD. i 3o 4 VOLTAGE FROM DATA IN RCA TUBE MANUAL RC I3 RP FOR scs TUBE EP=250 VOLTS ZIMMVM EDWIN K STODOLA Patented Aug. 17, 1948 UNITED STATES PATENT OFFICE 1W()D IIJ'IM YI.ION SYSTEM Edwin K. Stodola, Neptune, N. 3.

Original. application November 17, 1942, Serial No. 465,921. Divided and thisapplication March5, 1946, SerialNo.-652',210

(Granted under the act of .March 3, .1883, as amended April 30, 1928; 3'70 '0. G. "757) The invention described herein may be manufactured and used by or for the Government for governmental purposes, without the payment to me of any royalty thereon.

My present invention relates to modulation systems, particularly such as are suitable for modulation of carrier .waves. Thisapplication is a division of-my copending application forElectrical networks, Serial Number 465,921, filed November '17, 1942, which has matured into Patent No. 2,397,992, issued April 9, 1946.

In a signal system in which a source of energy is to be modulated in accordance with a signal, it is desirable to establish a linear relationship between the amplitude of the modulating signal andthe desired modulating effect. It is a principal object of my invention to devise a simple and efiective means for this purpose.

One object of my invention is to provide a modulation networkfor a source of carrier waves, said network including a nonlinear impedance adapted to be modulated and having-such characteristics that a desired modulation characteristicof-the potential across a portion of the network bears a linear relationship to the modulating potential.

Another object of my invention is to provide a network for producing a variable output from a source of input voltage in response to a control factor, variable over a predetermined range,-said network comprising at least a pair of elements connected in series across said source, one element being an impedance which is preferably fixed, the other element being a resistance adapted to be varied in accordance with variation of said control energy, said elements being so related that the ratio of the magnitude of a given characteristic of the voltage across said output circuit with respect to the magnitude of a like characteristic of the voltage across said network is a substantially linear function of said control factor Variation.

Another object of my invention is to provide a network for producing a variable phase output from a source of alternating current input voltage in response to a control'factor variable over a predetermined range, said network comprising at least a pair of elements connected in series across said source, one element being a reactance whichis preferablyfixed, the other element being a resistance adapted to be varied in accordance with said control factor, and ..an ..output circuit connected across one of said elements, said elements being so related .that the phase angle of 2 stantially linear function of .said control factor variation.

.In .accordance with my invention, a source of energy to be modulated is connected across ..a voltage divider including a pair of impedance elements, one of which may be fixed and the other-of which is variable by means of .a modulation potential, e. g., the space-current path of an electron .tube. The modulated output is derived across one of said elements, preferably the one which is variable. I have discovered that .by properly .proportioning the value of the fixedimpedance and properly choosing the relation between the impedance of said space-current path and its control factor, adesired characteristic of the modulated output with respect to the same characteristic of the input is a linear function of the modulating potential. If the input energy is constant, then the modulated characteristic of the output is a linear function of the modulating potential.

If the fixed impedance is resistive the network functions to produce linear amplitude modulation. If said fixedimpedance isreactive, linear phase modulation is obtained.

For a better understanding of the invention, together with other'objects thereof, reference is made to the following description, taken in connection with the accompanying drawings, wherein:

Figure 1 is a simplified diagram of an amplitude modulation circuit which forms the basis of my invention;

Figure 2 is a graph, illustrating the relationship between the grid voltage and the plate cathode resistance of a triode at present available on the market;

Figure 3is a more detailed diagram of the circuit shown in Fig. 1;

Figure 4 isa family of curves showing the relationshipbetweenthe grid voltage and the ratio between the tube voltage drop and total applied voltage across the .tube andresistor in Fig. 3;

Figure 5 is a simplified diagram of a phasemodulation circuit including a reactor in series with the tube;

Figure 6 is a more detailed diagram of the encuit shown in Fig. 5, and includes a tube and reactor in series to constitute a linear phase-modulation circuit; and

Figure '7 is a graph showing a family of curves directed to the relationship of .the phase angle between the voltage across the tube and the modulating potential on the grid in the circuitof the voltage across said output circuit is asub- Fig. 6.

Fig. 1 shows a voltage-divider circuit including two resistances 2 and 3 connected in series and energized by a voltage E from an input circuit 4. An output circuit 5 is connected across the terminals of the resistance 3. If resistances 2 and 3 are fixed, the voltage drop across each will always maintain the same proportion relative to the input voltage E, regardless of the variation Where the resistance 2 is kept at constant value and the resistance 3 is made variable, the drop in potential across resistance 3 will not vary directly in proportion to the variation of said resistance, but will vary according to the ratio of the resistance 3 to the total resistance of the voltage-dividing circuit. It will be obvious, therefore, that if resistance 3 was linearly modulated in accordance with a signal potential, the modulated output voltage would not be linear with respect to the signal potential.

In accordance with one aspect of my invention, linear amplitude modulation is obtained by varying resistance 3 in accordance with a modulating signal in such a manner that the ratio of said resistance with respect to the sum of both resistances 2 and 3 is always proportional to the instantaneous voltage of the modulating signal. Simple calculations show that the required rate of variation of resistance 3 with respect to the modulating voltage to obtain linear modulation is given by the following equation:

Where:

R3 is the resistance of resistance 3,

R2 is the resistance of resistance 2,

Em is the modulating factor which varies resistance 3., and

K is a constant of proportionality which establishes the extent of the modulating function necessary to produce a given efiect.

I have found that at least a portion of the dynamic plate-resistance characteristic of an ordinary triode closely approximates the requirements of the above equation. Hence, by making resistance 3 the space-current path of a gridcontrolled electron tube, the impedance of which varies nonlinearly with the modulation voltage, the requirements of the equation can be realized. The curve in Fig. 2 shows the variation of plate resistance with respect to grid potential of one type of commercially available triode, generally designated as the type 6C5, with a D. C. plate potential of 250 volts. By choosing the proper operating portion of this characteristic and a proper value for fixed resistance 2, it is possible to obtain linear amplitude modulation. A more detailed embodiment of the elementary circuit of Fig. 1 is shown in Fig. 3, wherein the plate-cathode circuit of the tube 3a corresponds to the resistance 3 in Fig. 1. The input circuit 4 is shown connected to the voltage-dividing circuit, which includes the resistor 2 and the triode 3a, through coupling condensers 8 and 9 of low impedance to the impressed frequency of input potential E, which represents a source of carrier frequency currents.

The triode 3a is shown as a typical heatertype triode, in which the heater element is shown as energized from a source of heating energy, as a battery ll, connected to the heater element through two R. F. chokes l2 and I3. The cathode is connected to one conductor of the input circuit fi through the condenser 9. The grid is provided -with a biasing potential derived from a source indicatedas a battery I 4, connected to the grid through an R. F. choke 23 and a high series resistance 30. A modulation voltage is applied across resistance 30 and battery I to vary the grid bias. Blocking condenser 30 is of low impedance to the modulating currents. Shunt F condenser 30 is of high impedance to the modulating currents but of low impedance to any residual radio frequency currents.

In order to balance out the interelectrode capacitance of the triode a parallel shunt, including an inductance H and a capacitance I8, is connected across the tube 3a. The constants of the shunt elements I? and I8 are such, that, at the operating radio frequency, the tube capacity and the shunt are resonant to said frequency and constitute a substantially infinite impedance, non-reactive circuit, whereby the shunt resistance is ver high at all times compared to the plate-cathode resistance. The plate voltage for the triode 3a, which in the present example operates at 250 volts, is provided by a source indicated as a battery IS, the negative terminal of which is grounded, and the positive terminal connected through an R. F. choke 2| to the plate. The cathode D. C. return path is provided by an R. F. choke 22. The grid biasing battery I4 is grounded at the positive terminal. The plate of the tube is connected to the resistor 2 through a blocking condenser 15 having negligible impedance at the operating radio frequency to prevent short circuiting of the plate battery l9 by the inductance l1.

, The grid of triode 3a is shown provided with a by-pass condenser 25 to the cathode, to provide a low resistance path to any appreciable radio frequency voltage between the grid and the oathode while, at the same time, permitting the application of direct current biasing voltage and modulation voltage between the cathode and the grid.

The two condensers 8 and 9 in the input circuit, and two similar condensers 2G and 21' in the output circuit, connected to the terminals of the triode, are provided to isolate the direct current voltage within the voltage-divider assembly of Fig. 3, when that assembly is connected to the rest of a circuit. These condensers all have a negligible reactance to the impressed frequency.

In the arrangement shown in Fig. 3, as the grid voltage is varied,,the resistance of the tube varies. As above pointed out, the ratio of the output voltage, representing the drop across the tube, to the input voltage, as applied to the voltage-dividing circuit, depends upon the value of the resistor 2, and the relationship of that value to the resistance characteristics of the tube 3a. Since the resistance of the tube 3a between the cathode and the anode, will vary with the excitation of the grid, the ratio of the output voltage to the input voltage will also vary with the grid excitation.

In Fig. 4, several curves 28a to 28c are shown, representing the relationship between the ratio of output voltage to input voltage and the corresponding grid voltage for different values of resistance in the series resistor 2 in Fig. 3.

The value of the resistor 2 corresponding to each curve 28a to 2819 is shown at the upper end of each curve. The middle curve 280 corresponding to a resistance value of 15,000 ohms in series with the tube, shows a linear relationship be tween the ratio of output voltage to input voltage EO/E and the grid voltage, between values of Eo/E=0.'795 and Eo/E=0.47. Thus, with a circuit including a tube and a resistor having the proper resistance values corresponding to the characteristics of the tube, a direct linear voltage relationship may be established between modulating voltage on the grid of the tube and the output voltage over an extended range, provided that the input voltage is held constant.

While inthe above circuit a tube has been, used as a non-linear resistance, it is evident that any other known type of element which has a nonlinear relation between its resistance or impedance and the controlling factor applied thereto can be used. In addition, all or part of resistance 2 in Figs. 1 and 3 can be constituted by the internal resistance of the source .of voltage E,

Thus, by making the impressed voltage E, in Fig. 3, a source of carrier potential and placing a source of modulating potential in series with battery M, the output Eu will be an amplitudemodulated carrier which will bear a linear relationship to the modulating voltage over a range depending upon the constants of the circuit.

Another object of my invention is the use of the principles above described to obtain linear phase modulation. It is well known that when an alternating voltage is applied to a circuit including a resistor and a reactance in series, the Voltage across the reactance is phase displaced with respect to the voltage across the resistor, and that each such voltage is phase displaced with respect to the applied voltage. Such circuit arrangement is shown, for example, in Fig. 5, in which the applied voltage from an input circuit 4, similar to that of Fig. 2, is applied to a circuit including the resistor 3 and reactance 29. The circuit is similar to that shown in Fig. 1, except that either an inductive or capacitative reactance 29 is substituted for resistor 2. It will be evident that the output voltage E0 will be phase displaced with respect to the input voltage E, the direction of such phase displacement depending on whether reactance 29 is capacitative or inductive. If the applied voltage E and the reactance 29 are held constant while the resistor 3 is varied in accordance with a modulation voltage, the phase displacement of output voltage E0 will be modulated.

The required law of variation of resistance 3 with respect to the modulating voltage function to obtain linear phase modulation of the output E0 with respect to the modulating voltage is given by the following equation:

tan KE Where X is the reactance of reactor 29 and the other symbols are similar to those in Equation I.

As is the case with Equation I, the requirements of Equation II can also be closely approximated by making the resistance 3 the space-current path of an electron tube. Such a circuit is shown in Fig. 6 in which the space-current path of tube 3a, which may be of the 605 type, corresponds to resistance 3 of Fig. 5. In other respects the circuit is substantially similar to Fig. 3.

R3 (II) The modulation characteristics of the circuit in Fig. 6 are illustrated by the curves in Fi 7.. Curves 3llaw3fldshow the relationship between the grid voltage on tube 3a and the phase angle between the applied voltage E and the output voltage E0 across the spaceecurrent path of the triode.

The curve 301;, corresponding to the case where a reactor 29 having 15,000 ohms reactance is used, shows a linear relationship between the grid voltage and the phase angle between the applied voltage and the output voltage for the range of values of the phase angle between 15 and 55. Thus, by use of proper circuit constants in a. voltage-dividing system such as shown in Fig. 6, a triode may be employed as a variable resistor, and an output voltage derived across the triode, whose phase relationship to the input voltage will be a linear function of the amplitude of the modulation voltage applied to the tube over a substantial range of that voltage.

,It should be noted that curve 206 in Fig. 7 shows that amplitude modulation will accompany the phase modulation when reactance 29 has a value of 15,000 ohms. If this is considered undesirable it can be eliminated by conventional amplitude limiting or automatic volume control methods.

As is the case with Figs. 1 and 3, any other type of non-linear impedance can be used instead of tube 3a in Figs. 5 and 6. Also, all or part of reactance 29 can be constituted by the internal reactance of voltage source E.

Although in the above described circuits the linear relationships are functions of the grid voltage, the circuits can be arranged to make said relationships a function of the voltage impressed on another electrode or combination of electrodes.

The phase modulators above described can be used as components of conventional frequency modulation systems. It hasbeen pointed out by Armstrong and others that pure frequency modulation is equivalent to phase modulation in which the maximum phase shift for a given modulating voltage is inversely proportional to the modulating frequency. Hence, any device which produces phase modulation will produce frequency modulation by making the modulating voltage inversely proportional to the modulating frequency. of course, at any single modulating frequency there is no distinction between the two. Hence, in this application reference has been made to phase modulation with the understanding that by properly correcting the modulating voltage the same devices and methods will produce frequency modulation.

The specific circuits and constants above described are to be considered as illustrative of the principles of the invention. Numerous modifications can be made without departing from the spirit of the invention as set forth in the appended claims.

I claim:

1. A network for providing a variable-phase output from a source of alternating-current input voltage in response to a control factor variable over a predetermined range comprising: a reactance adapted to be coupled to said source in series with the space-current path of an electron tube, an output circuit coupled solely across said path, and means adapted to be controlled solely by said control factor to vary the resistance of said path, the resistance of said path over said range of variation of said control factor being a non-linear function of said variation, the magnitude of said reactance being so related to said non-linear function that the phase angle between the voltage across said output circuit and the voltage across said network is a substantially linear function of said control factor variation.

2. A network for providing a variable-phase output from a source of fixed-frequenc input voltage in response to a control voltage variable over a predetermined range comprising: a fixed reactance adapted to be connected across said source in series with the anode-cathode path of an electron tube, an output circuit coupled solely across said path, and a control grid in said electron tube adapted to be coupled solely to said control voltage to vary the resistance of said path, the resistance of said path over said range of variation of control voltage being a non-linear function of said control voltage, the magnitude of said fixed reactance being so related to the average resistance of said path over said range that the variation of the phase angle of the Voltage across said output circuit is a substantially linear function of said control voltage variation over said range.

3. A network for varying the phase angle of the voltage of a source of alternating current in response to a control factor variable over a predetermined range, said network comprising an input circuit including a pair of impedances con nected in series across said source, an output circuit coupled solely across one of said impedances, one impedance being predominantly reactive, the other impedance being predominantly resistive, the resistance of which is adapted to be controlled solely by said control factor, the variation of said resistance being a non-linear function of the variation of said control factor over said range, the magnitude of said reactive impedance being so related to the average resistance of said resistive impedance over said range that the ratio of the phase angle of the voltage across said input circuit relative to the phase angle of the voltage across said output circuit is a substantially linear function of the magnitude of said control factor over said range.

4. A network for varying the phase angle of the voltage of a source of alternating current in response to a control voltage variable over a predetermined range, said network comprisin an input circuit including a air of impedances connected in series across said source, an output circuit coupled solely across one of said impedances, one of said impedances being fixed and predominantly reactive, the other of said impedances being predominantly resistive and comprising the space-current path of an electron tube, a control grid in said electron tube adapted to be controlled solely by said control voltage, the resistance of said path being a non-linear function of the variation of said voltage over said range, the magnitude of said reactive impedance being so related to the average resistance of said path over said range that the ratio of the phase angle of the voltage across said input circuit relative to the phase angle of the voltage across said output circuit is a substantially linear function of the magnitude of said control voltage over said range.

5. A network as set forth in claim 4, including means to minimize the effect of the interelectrode capacitance of said electron tube.

6. A network as set forth in claim 5, wherein said means includes reactance shunted across said interelectrode path to form therewith a circuit which is parallel resonant to the frequency of said source.

EDWIN K. STODOLA.

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

UNITED STATES PATENTS Number Name Date 2,259,393 Van Roberts Oct. 24, 1941 2,371,285 Crosb Mar. 31, 1945 2,383,848 Crosby Aug. 28, 1945

Images