US 2282714 A
Description (OCR text may contain errors)
May 12, 1942. J. FAGOT R AMPLIFICATION 0F.
METHOD AND MEANS. FOR THE LINEAR TRANSMISSION AN AMPLITUDE-MODULATED CARRIER WAVE Filed Dec, 1, 1959 3 Sheets-Sheet l "FOUL/7717B 8 m & Sousa: 0F o'sc' IL 4 Ana/vs I 2 Z'mnentor A Jae up: ill/gm; BB 7 Gttorneg May 12, 1942. J. FAGOT 2,282,714
METHOD ANDMEANS FOR THE LINEAR TRANSMISSION OF. AMPLIFICATION OF AN AMPLITUDE-MODULATED CARRIER WAVE Filed Dec. 1, 1939 g 3 Sheets-Sheet 2 ISnvenfor Gttorneg May 12, 1942. J. FAGOT 2,282,714
METHOD AND MEANS FOR THE LINEAR TRANSMISSION OR AMPLIFICATION 0F 'AN'AMPLITUDE-MODULATED- CARRIER WAVE Filed Dec. 1, 1939 3 Sheets-Sheet 3 M MODl/LA'TOE 3nventor 0 "SOURCE OF OSCILL/lT/ONS BB v Gttomeg Patented May 12, 1942 TRANSMISSION AMPLIFICATION OF AMPLITUDE M o n U L A T E n CARRIER WAVES Jacques Fagot, Paris, France, assignor to Compagnie Generale de Telegraphic Sans Fil, a corporation of France Application December 1, 1939, Serial No. 307,022 In France December 2,1938
4 Claims. The transmitter system of the present application has for its object to provide a modulated wave amplifying means and method applicable to radio-telephonic transmission.
The present invention involves on the one hand the system of modulation through phase shift, which is described in older patents (see especially United States Patents Nos. 1,882,119, 1,892,383, 1,946,308 and 2,009,080. See, also, Proceedings I. R. E., November, 1935, A Communication from H. Chireix).
The present invention involves on the other hand operation similar to that of an ordinary modulation system in which variation of the amplitude of the carrier wave is accomplished.
The present invention thus can in a certain.
sense be considered as improvement of the two preceding methods.
One of the objects of the invention is to provide a method 'and means for high efiiciency operation for radio-telephonic transmitters. 'Another object of the invention is to provide a method and means for amplification and trans- The essential arrangement of the invention resides in feeding to the end or power tubes or to an intermediate amplifier stage of a transmitter connected along the lines of a modulation system predicated upon outphasing, amplitudemodulated potentials originating from a single input amplification chain or cascade, and in the use of reverse or negative feedback in the load circuit with a view to insuring modulation based upon amplitude variation only for the points of the modulation cycle which in the load circuit correspond to an instantaneous current value below the effective or root means square value of the'carrier wave, and modulation based both on amplitude variation and outphasing for points of the modulation cycle which correspond to larger load.
In certain embodiments of the essential arrangement as hereinbefore outlined, a transmit-' ter comprises only a last symmetric stage or preceding stages that are also symmetric to which the modulated potentials may be impressed either directly or else by the intermediary of auxiliary amplifiers.
The circuit organization, moreover, includes compensating means whereby the distortingaction of the modulation cycle may be diminished, this means being'efiective especially when distortion is due to cuits.
In describing my invention, reference will be made to the attached drawings wherein:
Figure 1 illustrates a modulated wave amplifier system wherein two radio-frequency waves are fed to a common load in'phase-displaced relation to produce a resilient radio-frequency wave the amplitude of which depends on the phase relation which, in turn, is controlled by the signals.
Figure 2 illustrates a modulator and modulation amplifier arranged in accordance with the present invention. In this system I make use of modulation by phase shift and by carrier wave amplitude variation;
Figures 3 and 4 are voltage vectors used in illustrating the manner of operation of my system as illustrated in Fig. 2;
Figure 5 shows means for compensating nonlinearities occurring in the transmitter of Figure 2; while Figures 6 and 7 are modifications of thearrangement of Figure 2. r It is known in fact that the two modulation methods referred to hereinbefore differ essentially by the manner in which there is obtained the modulation of antenna currents under the action of the speech currents, this modulation producing in the two cases, amplitude variations of the carrier wave.
In the customary modulation system, this amplitude variation of the carrier wave is obtained by applying to the antenna a high frequency potential, already modulated in amplitude by the potentials, which are substantially equal, form between each other a rather wide angle for instance). Under the action of a phase modulation in the reverse sense of these two elementary potentials, their relative phase shift varies, for example, 'between 120 and and this has the effect that the instantaneous current amplitude the saturation of the load .cir-
in opposition to the potential Us.
in the antenna likewise varies between two extreme I and I1 which, for a deep modulation of the transmitter, may be in the neighborhood of zero for the one value, while the other is twice the effective value of the carrier Im.
The greatly simplified nature of such a system can be represented as is indicated in the Figure 1, attached herewith. In this figure, two tubes I and 2 having their anodes 5 and iconnected to circuits tuned substantially to the same carrier frequency feed into a common load circuit, represented in a highly schematical fashion in the form of a resistor R which is the equivalent, for instance, of the total antenna resistance which would be coupled in practice in a suitable manner with the two anode circuits L1 Cl and L2 C2.
According to the method of modulation by phase shift which has just been recalled summarily, the grids g1 g2 of the two tubes I and 2 are excited respectively by two potentials in and r us obtained from the same high-frequency oscillator 0 through two separate amplification chains C and C, these two potentials beingmodulated in modulators in mi and m'; by opposite phase variations. The resultant current in the load is correspondingly modulated in amplitude.
The amplifying system forming the subject matter of the present invention is represented schematically in Figure 2. It differs from the preceding one by the mode of excitation of the grids g1 and g2 of. the tubes I and 2, the platecircuits of these tubes being the same and being controlled in the same manner.
The modulated alternating potentials applied to the grids g1 and ya in my system result from the composition of four elementary high frequency potentials U1, U2, Ua and Us, as will be explained hereinafter.
The potentials U1, U2 and Ub-coming from the high-frequency source 0 across a single amplification chain C, permits use of a modulator M of any type of means of which the amplitudes of said potentials are varied in accordance with the amplitudeof the speech currents.
The potentials U1, U2 which are substantially equal to each other are applied respectively to each of the grids g1 ya but in opposite phasewith respect to each other,while the potential Us is applied substantially cophasally to the two grids with a phase in quadrature relative to the two preceding potentials V1 and V2.
The potential Us finally is furnished by the resistor Rs of theload circuit (antenna circuit for instance). Vs is likewise a high-frequency potential modulated in amplitude and the arrangement is such that this potential Us is also The poten-- tial Us can thus be considered as in inverse reaction.
The resultant alternating potentials which are applied to the grids g1 and Q: are represented by the diagram of Figure 3.
In this figure there is drawn from the point 0 corresponding for instance to the ground potential (point 0 of Figure 2), a vector 0A equal to the potential Us arising in the impedance Rs, then in the opposite direction a vector AB equal to Us, and finally two vectors BG1 and BGs equal respectively to U1 and to Us but displaced in phase each by 90 relative to the preceding vectors, the one in advanced phase and the other one lagging in phase. The resultant potentials applied to the grids g1 an of the two tubes 1 and 2 (Fig. 2) are then represented respectively by the two vectors 0G1 and 0G: which form between each other an angle of 1802a.
In practice; the circuits will be controlled such that in the absence of modulation the angle a has a small value,.20, for instance.
Thus, it is seen that in the absence of modulation, the alternating potentials which are applied to the grids g1 gz of the two tubes are similar to those which they would receive in accordance with the usual scheme of modulation by phase shift as illustrated in Fig. 1.
For reasons of economy, it is known, moreover, that it becomes expedient in general to control the transmitter in such a manner that this operation (non-modulated carrier wave) corresponds substantially to the appearance of saturation phenomena in the output circuits of the two power tubes 1 and 2.
When examining thus the performance taking place at modulation it is found that for all points of a modulation cycle for which the amplitude in the load circuit is smaller than that which would correspond to the carrier alone, the potentials U1 U2 U and U vary (in amplitude only) in a linear fashion, 1. e., the diagram of the potentials applied to the grid (Fig. 3) becomes larger or diminishes but without its form being varied,'i e., by remaining homothetical to itself.
The final stage thus behaves like a simple amplification stage modulated in amplitude and with inverse reaction.
In fact, this inverse reaction acts in accordance with the phase but the latter plays hereby but aii intermediate role.
The output of the power stage for all these points is by the way the more favorable the more the amplitude approaches the value of the carrier. It becomes poor only for the smallest amplitudes but since the power which presents itself is then low, the influence of said. amplitudes upon the overall output is of slight importance.
However, for the points of a modulation cycle which correspond to larger amplitudes than the means value of the carrier,- the potential Us owing to the saturation of the load circuits has the tendency of not increasing according to a linear law, while the potentials U1, U2 and Us continue to increase proportionally. The result is that, at the same time that the diagram becomes larger, it also undergoes a deformation. The angle a which at. first was 20", for instance, will be increased and for the points of the cycle which correspond to the maximum amplitude there will be obtained, for instance, the diagram shown in Fig. 4. The potentials U'i U'2 and U's are equal respectively to 2 U1, 2 U2 and 2 Us while U is less than 2 Ua. It follows for instance that U'b-U'a has not become equal to 2 (UbUa) but equal to (Ub-Ua) and that the angle a has become equal to 2d namely equal to 40.
When assuming, for instance,'that Fig. 3. corresponds to a point of the modulation cycle such that the instantaneous power-of the transmitter be equal to the means power of the carrier wave, Fig. 4 will then correspond to the crest point of the same cycle.
Thus, it is seen that in my novel system for the high values of the modulation cycle operation is similar to that of the system of modulation through phase shift, with the difference, however, that in my system at these high values of the modulation cycle the amplitudes of the excitation potentials on the grids of the amplifier tubes are also increased.
When considering finally the overall operation of the transmitter in any modulation, it is seen that the output which is higher than 60% for the carrier and above the latter and which is only low for the very small loads, remains very favorable in the arrangement.
In order to give an illustration there is chosen in Fig. 3 Ub=l (Uh-U8.) Then, it can be seen that there suflices a relative drop of of Us at the crest value in order to double the relative value of UbUa and, therefore, to have the angle a pass substantially from to 40, 1. e.. for doubling substantially the antenna current and the circuits of the plate of thelast stage is adapted for this as in the system of modulation through phase shift.
The amplitude curve will under these conditions indicate a relative drop of 10% in the crest which value does not appear exaggerated.
It is clear 'on the other hand that this drop will be lower if for Us and for Us higher values relative to their difierence are chosen respectively.
Moreover, in accordance with a further feature of the invention, this drop can be overcome in various ways.
For instance, instead of applying to the grids the frequency in accordance with a law of inverse rotation.
Utilized for modulating the input of the main transmitter, this potential, will, therefore, compensate in full or in part, its phase rotations.
The action of such a compensator could be increased further if necessary by repeating it several times by means of several compensators placed in series.
Returning to the general functioning of the system in accordance with the invention and,
more particularly, to Fig. 2, it should be understood that in practice the reaction potential could relatively to Fig. 2.
be applied to one of the preceding stages of is shown at l and 2' two tubes of a preceding g1 ya the potential Us. proper or a potential which is proportional thereto, a potential may be applied which tends itself to fall towards the crest in accordance with the same law.
In accordance with another variation of the invention as illustrated in Fig. 5 this correction may further be eflfected by means of an auxiliary transmitter of very low power or simple amplifier G possessing the same distortion characteristic as the main transmitter Em.
Negative feedback is added to this auxiliary transmitter or to this amplifier serving as a compensator.
In designating by. Ee the input potential and by. E8 the output potential of the compensator, the efiective potential applied between grid and cathode will be equal to EeK Es wherein K is the coefficient of tlie inverse reaction. This potential will contain aside from the fundamental term all the harmonics of the principal transmitter Em (since the distortions of these two transmitters are similar) but in phase opposition Q with respect to the fundamental term.
In order to compensate the phase lag of the principal transmitter, the phase of the small I transmitter could thus be made to lag to a lesser degree in order to be able to apply to the latter transmitter the inverse reaction. The potential between grid and cathode of the input tube will thus have its phase which will advance with stage. The potentials U1 U2 Ua and Ub obtained respectively from the chain C and from .the resistance Rs are then applied to the grids g'l and g: of the two tubes l and 2'.
It may also be .of advantage to apply directly to the grids of the tubes of theend stagethe two potentials U1 and U2 and to apply to them on the contrary the potential Ub-Ua across one or several amplifier tubes. Such a circuit is shown in Figure 7 indicating in 3 such an amplifier tube. Here the voltage Uh is applied to the grid 9 of tube 3 and the resulting amplified voltage appearing in the tuned circuit ll connected with anode i3 is fed to the grids g1 and g: i
by blocking condenser I5.
It should be understood finally that the circuits shown canbe provided or modified; without departing from the scope of the invention, in accordance with any known arrangement. Thus, instead of utilizing for the power stage of the transmitter two tubesonly, for this stage also four tubes could be used which operate in pairs in syrmnetrical circuits or even a larger number could be utilized.
What is claimed is:
1. In a modulated carrier wave amplifier, a load circuit, a source of modulated wave energy, a pair of electron discharge devices each having input and output electrodes, means for applying voltages from said source in phase displaced re= lation on the input electrodes bf said tubes, means for applying voltage from said source in phase on the input electrodes of said tubes, 8. load impedance connected with said output electrodes, and means for impressing voltages which vary in accordance with variations of the amplitude of the potentials developed in said load impedance on said input electrodes in phase opposition to said second named voltage impressed in phase on said input electrodes.
2. The method of high efficiency linear relaying of an amplitude modulated wave which in cludes the steps of, transmitting substantially components to produce a resultant dependent upon the components amplitude and phase relation as combined, and
modifying the phase relatransmission is substantially linear, combining said transmitted components to produce a resultant, producing voltages representative or substantial departures or said resultant from linof said voltages to augment the resultant and to thereby maintain linearity over the entire modulation range during said relaying process.
4. In a modulated carrier wave amplifier, a
load circuit, a source of modulated wave energy,
a pair of electron discharge devices, each having input and output electrodesfmeans for applying voltages from said source in phase opposition on the input electrodes of said tubes, means for applying voltage from said source in phase on the input electrodes 01' said tubes, a load connected with said output electrodes, and means for impressing voltages from said load on said input electrodes in phase opposition to said second earity, and reducing the phase displacement of a said components being transmitted'as a function named voltage impressed in phase on said input electrodes.