US2244249A - Wave translation system - Google Patents

Description

G. GUANELLA WAVE TRANSLATION SYSTEM June 3, 1941.

Filed April 26, 1959 3 Sheets-Sheet l 4 1 qi 2 I 7 I 2 I i i B1 4 I: p p b s=c(r'q) /S 7" l c A;

F16 1 FIG. Z

INVENTOR.

ubi'a-uua-nella. s 9 4 ATTORNEY.

June 3, 1941.

QUANELLA WAVE TRANSLAT I OH SYSTEM Filed April 26, 1939 5 Sheets-Sheet 2 IN VENT OR gwsi'av' swnclla.

ATTORNEY.

June 3,- 1941. e. GUANELLA WAVE TRANSLATION SYSTEM 3 Sheets-Sheet 3 Filed April 26, 1959 FIG .8

F IG. 9

- IN VENTOR. susi'av guanella ATTORNEY.

Patented June 3, 1941 UNITED STATES Search Room PATENT OFFICE WAVE TRANSLATION SYSTEM Gustav Guanella, Zurich, Switzerland, assignor to Radio Patents Corporation, a corporation of New York Application April 26, 1939, Serial No. 270,125 In Switzerland December 31, 1938 17 Claims.

The present invention relates to improvements in feedback amplifiers and a method of operating the same and among the objects of the invention is to obtain and maintain a predetermined, such as a linear, input-output relationship in an electric translation or converting system.

It has already become known to maintain a linear relation between the input and output potentials of an electric amplifier serving for translating and/or converting variable signalling energy by the employment of inverse feedback or negative reaction, 1. e., by feeding back a portion of the amplified output energy to the input of the system. Such feedback in inverse phase relation to the input energy entails a reduction of amplification or loss of signal strength which makes it necessary in most cases to employ one or more additional amplifying stages. Furthermore, the elimination or compensation of distortion in feedback amplifiers of the heretofore known type to secure a desired input-output relation enables the attainment of this object to a limited degree only as it is possible merely to reduce but never to completely eliminate the undesirable distortion. Moreover, the oscillations fed back due to undesirable phase shifts by the intercoupling networks in the main amplifying and feedback channels for selected frequencies or freqency bands may be increased in amplitude, thereby upsetting the operation and eventually resulting in self-excitation and substantial impairment of the operational stability of the feedback system. These and other reasons have made it impossible in the past to make full and efilcient use of inverse feedback without the use of complicated and expensive circuit arrangements.

The above drawbacks and disadvantages are substantially overcome by the present invention. With this object in view the invention contemplates the provision of an auxiliary translation system or channel in addition to the main translaion system to be controlled or corrected which auxiliary system has equal or approximately equal propagation characteristics to the main translation system. There is applied to this auxiliary system the potential to be translated while a further correcting potential is impressed upon the input of both systems obtained by feeding back the difference of the output potentials of the systems through a suitable feedback channel or network. Due to the simultaneous impression of the correcting or feedback potential upon both systems, the new method will be designated as parallel inverse feedback for the purpose of this specification.

An advantage of the novel method is the fact that distortions of the controlled system may be suppressed entirely thereby eliminating the possibility of building up of self-excited oscillations. Further aspects and advantages of the invention will become more apparent from the following detailed description taken with reference to the accompanying drawings illustrating several practical embodiments of the invention and wherein,

Figures 1 and 2 are block diagrams showing the general layout and organization of an inverse feedback system according to the invention,

Figure 3 shows a theoretical diagram explanatory of the function and operation of the invention,

Figures 4, 5 and 6 are simple amplifier circuits embodying improved inverse feedback according to the invention,

Figures '7 and 8 show embodiments of the invention as applied to a radio transmitter to compensate distortions of the signal modulation, and

Figure 9 illustrates the invention as applied to a light control arrangement for photographic recording or like purposes.

Referring to Figure 1, there are shown a translating or converting system A such as an amplifier, etc., a parallel or auxiliary translation system B provided according to the invention, and a feedback system or channel C. The propagation factors of the three systems; that is, the real or complex ratio, usually being dependent upon frequency, between the input and output potentials are designated by the characters a, b, 0, respectively. The input magnitude or potential to be translated or converted without distortion by the entire arrangement is impressed at least upon the parallel system B or alternatively upon both the main translation system A and the parallel system B as shown in the drawings. The difference between the corrected output potential q of the system A and the output potential r .of the auxiliary system B is applied to the feedback channel C resulting in a combined correcting potential s impressed both upon the main translation system A and the auxiliary or parallel system B.

As a result of the above connections, the output potential q may be expressed by the following equation:

The factor n is equal to 1 if the input potential p is applied to the system A and the system B with equal amplitude; in case of m=n, p is applied to A with 112 times the amplitude. If n=O, p is applied to the parallel system B only. The correcting potential s is obtained from the difference between the output potential (r-q) and in view of r=b(p+s) is represented by the following expression:

. a JEL' L s=c(1q)-c(bp+bsq) (2) by combining Equations 1 and 2 q is obtained as follows:

wherein for the sake of abbreviation:

d ac (4) In order to suppress distortion the constants b and c of the auxiliary system and of the feedback system should be chosen in such a manner that their product is equal to unity; that is,

in which case d disappears. This condition can be fulfilled by providing at least one amplifier in the systems B or C serving to compensate the attenuation in the other system. In this case the output potential q is obtained as follows:

From this latter expression it follows that the output potential q is free from all distortions and interference having its origin in the main translation system A and that its dependence upon the input potential p is determined solely by the propagation constant b of the auxiliary or parallel system B, while the propagation constant a of the system A which may be subject to substantial variations has no influence whatsoever upon the output potential.

The constants a, b, c may be chosen in such a manner that in addition to condition (5) approximately the following condition exists:

in which case a reduction of the amplitude of the output potential q compared with the output potential without feedback is avoided. An incorrect balance of the systems B and C whereby the factor d does not disappear entirely will only slightly affect the total propagation factor due to the denominator (1+d) in Equation 3 and in view of the fact that in the equation the distortion appears multiplied with the factor d.

The relative distortion in the system A without feedback may be represented as follows:

This distortion factor is reduced by employing parallel inverse feedback according to the invention as follows:

The latter, in case of fulfillment of Equation 7, corresponds approximately to the factor (1. The

distortions may also be considerably reduced if condition (5) is not completely fulfilled. In general, the factor d may be kept very small whereby with sufficient approximation the output potential will be as follows:

In designing the system consideration should be given to the fact that the propagation characteristics in most cases are dependent upon frequency. The condition (5) should be fulfilled as far as possible for the frequency range of the distortions a: to be suppressed. If only a single disturbing oscillation of definite frequency is to be suppressed it may suffice to comply with condition (5) for this frequency only. In many cases the oscillations to be transmitted and the distortions to be eliminated include a larger frequency range. A dependence of the factor 1) upon frequency which under circumstances cannot be avoided may be avoided at least partly by the design of the feedback circuit C to have a suitable frequency response as regards its propagation factor 0 to comply with condition (5) as far as possible.

From expressions 1, 2, 5 and 7 the output potential of the parallel system B is obtained as follows:

1': (2-n) bp-a: (12) This potential is a minimum; that is, the power transmitted by the parallel system B is a minimum for n=2 i. e., by applying the input potential p to the system A to be corrected with an amplitude twice the value of the potential applied to the auxiliary system B. In this case r=a: so that the potentials translated by By represent only the distortions, whereby not only the power but also the amplitudes in the parallel system B are small.

The correcting potential s considering Equations 6 and 12 may be expressed as follows:

ments.

In practice at least one of the systems B and C should contain an amplifier. By the proper choice of the factor n the amplitudes within this amplifier may be maintained as small as possible.

The system A may consist of several individual sections in series or cascade such as shown in Figure 2 wherein the propagation constants of the systems A1 and A: are designated by or and 0.2, respectively and whereby the total propagation factor of the systems is represented as follows:

The output potential q: of A: is again determined by Equations 3, 6 and 11. If A: is a system free from distortion and interference, the output potential of A1 taking into consideration Equations 6 and 5 will be as follows:

that is. the potential derived from the output of the system A1 is independent of interference and distortion produced in this first translation system. The employment of the second system A2 is advisable if the output potential of A1 occurs in a form unsuited for combination with r or for the feedback through the system C. Thus, for instance, p, r, s and q: may be electric currents or potentials while qi may have the form of electro-magnetic or mechanical oscillatory energy converted into electric currents or potentials by the system As.

The stability conditions of a system according to the invention are explained by reference to Figure 3. The propagation factors of the systems A, B, C are in general dependent upon frequency. The product 41.0 of the constants of the systems A and C may'be represented in a known manner by a vector whose point for variable frequencies describes a polar curve as shown in Figure 3, while at the same time the point of a further vector he moves along a second curve. For the frequencies to be translated and especially the disturbing frequencies to be suppressed the vector 11.0, in accordance with Equation 5, should coincide with the abscissa point +1 as nearly as possible. The same requirement applies to (1.0, if Equation 7 is to be fulfilled. The difference vector g=b.c--a.c also being dependent on frequency represents the total propagation factor of the entire system for a potential 3 from the input of the systems A and B to the output of the system C. Since the output of the latter is applied to both inputs through the feedback circuit building up of oscillations or self-excitation according to the investigations by Nyquist (see Bell System Technical Journal 1932, page 126 et seq.) is possible only if the polar curve g encloses the abscissa point +I.

From the foregoing it is seen that the factor Q being the difierence of two vectors whose absolute values are in most cases smaller than unity and may exceed unity only under the most unfavorable conditions may be kept within the admissible limits by most simple means to insure operational stability of the system. By adapting the phase shifting circuits in the system A to those of the system B it is furthermore possible to obtain a coincidence of the polar curves ac and 17.0 to such an extent that g can never exexcitation may therefore be maintained in the new system in a more simple and efficient manner than is possible in the known inverse feedback amplifier in which latter the product of the transmission factors of the main system and of the feedback system on the one hand should assume large negative values and on the other ha; d should not exceed unity.

s will be evident from the foregoing the new system is especially suited for suppression of interference and distortions in translating systems by employing a definite relation (Equation 5) between the propagation constants of the parallel and feedback systems, respectively. From Equation 3 it is seen that in case of negative values of the factor d, the polarity of the distortions will be reversed. Thus, for instance, for d=-2 there is obtained the following:

-q=(b2an)p2a: (16) By an over-correction of this type of the distor- $earch Room tions in the individual translation systems it is possible to compensate the distortions in an associate non-corrected system.

Thus, in case of telephone lines with intermediate amplifiers a distortion 2.1: may be produced in two successive amplifier sections. In a third amplifier wherein the distortions are over-compensated with a factor d=2 there are thus produced distortions -2:c, whereby the total distortion is eliminated by employing similar over-correction in every third line amplifier. This however presupposes that the effect of non-linear elements are equal in all amplifiers and that the distortions x are small compared with the signals being transmitted. In addition, the phase shifts in the individual amplifiers should be proportional to frequency to prevent a, change of the signals by the line transmission excepting a delay and amplitude reduction in dependence upon frequency. When using a different degree of over-correction; that is, by selecting a diflerent value of the factor 11 the requisite number of over-corrected amplifiers may be varied. A negative value of the factor d may be obtained while maintaining the stability conditions discussed with reference to Figure 3 in a simple manner by choosing the propagation constants c and b on the one hand and the propagation constant c on the other hand to be of definite polarity.

In Figures 4 to 9 there are shown several exemplifications of inverse feedback systems embodying the improvements according to the invention.

Referring to Figure 4, the system A1 to be corrected comprises a two-stage amplifier including the electron amplifier valves V1. V3. The output of valve V1 is impressed upon the input of the valve V3 through a. resistance coupling network of known construction comprising an anode or load resistance R1, coupling condenser C1, and a grid-leak resistance Re. The valve V3 is connected to the output terminal 3 through a load resistance R5 and a coupling condenser Ca. the other output terminal 4 being connected to ground or cathode in a manner well known. The input potential p to be amplified is impressed across the terminals |-2 through a portion of an inductance coil L simultaneously upon the input grid of the valve V1 of the main amplifier system A1 and upon the input grid of valve V: forming part of the auxiliary or parallel system B. The output of the valve V2 is derived through a network comprising anode load resistance R2, 9. coupling condenser C2 and a. grid-leak resistance R4. By suitable design of R2, C: and R4, the propagation characteristics of the parallel system B may be adapted as far as required to the characteristics of the main amplifier A1 thereby fulfilling the conditions for stability discussed with reference to Figure 3. The output terminals of the coupling condensers C: and C3 inthe auxiliary and main translation systems, respectively, are connected through a potential divider W1. A tap point on the latter is connected to the lower end of the induction coil L whereby the latter represents the feedback system C according to Figures 1 and 2. The tap 32 of the potential di vider W1 which latter correspondsto section A: in Figure 2 is chosen in such a manner that the amplified output potential (11 provides a component q: of the potential derived from the tap 32 which component is approximately equal and opposite to the component 1' supplied at point I! by the output of the auxiliary system B (valve V2) thereby fulfilling the requirements according to Equation 7. The feedback system C as pointed out is provided by the inductance L forming an auto-transformer for the potential r-q: to be impressed upon the inputs of the main and auxiliary translation systems. The ratio of transformation is adjusted by selecting the tap point 22 in such a manner that a definite potential variation at the control grid of valve V: will. cause a like potential variation at the upper terminal 23 of the induction coil L, thus fulfilling the requirements according to Equation 5. By neglecting a slight increase of the input potential p by the auto-transformer L, the output potential :01 will be as expressed by Equation 15; that is, it will be independent of interfering potentials originating in and distortions produced by the amplifier V1-V3.

In Figure 5, there is shown a modified amplifier employing parallel inverse feedback, wherein the parallel system B and the system A1 to be corrected each consist of an amplifier valve unit V4 and V5, respectively, arranged in a common envelope. The input potential p supplied at terminals I-2 is impressed upon the control grid of both amplifying units through the center tap of the secondary of the feedback transformer L44. The difference r-q1 between the output potentials of both systems is applied to the primary of the transformer L44 constituting the feedback system, whereby the input or grid potential impressed upon both amplifiers is increased by an amount s=c(rq). R9 and R10 are anode load impedances and C is a coupling condenser for deriving amplified output potential.

There are further provided a pair of impedances R1 and Rs connected between the end of the primary of the transformer L44 on the one hand and between ground or cathode lead on the other hand. If these impedances are equal, a reaction of the fed back potential upon the preceding circuits connected to the terminals I and 2 is prevented. By equal design of both amplifier units V4 and V5 Condition '7 may be fulfilled. Condition 5 can be maintained by a suitable choice of the transformer L44.

In the foregoing arrangement, the output potential q derived from a relatively large coupling condenser C5 will be dependent on the input potential p as determined solely by the amplifying characteristics of the amplifier unit V4, inde pendently of the load connected to the output terminals 34. This arrangement is therefore specially suited in all cases where constant voltage transmission conditions are required independently of load variations both with regard to amplitude and frequency of the load current.

In the circuits according to Figures 4 and 5 the input potential p is applied to the amplifier to be corrected and to the parallel system with equal amplitude; that is, the constant 1: according to Equations 3, 12 and 13 is equal to unity.

In Figure 6 there is shown an simplifying circuit wherein the input potential 1: is applied to the amplifier to be corrected with twice the amplitude; that is, with the factor n=2. There is applied through the. transformer Lu a potential 272 to a two-stage amplifier comprising the valve Va and V1. The latter are connected through a resistance coupling network comprising a coupling condenser Cs and a grid-leak resistance Rm. The output of valve V1 is applied to the terminals 3-4 through an output or load transformer Lac. Input potential 9 derived from a tap point of the secgndary of the transformer Lu is simultaneously applied to the control grid of a further valve V: forming part of the auxil-v iary system through a coupling condenser Cr and grid leak resistance R13. Condenser C1 and resistance R13 may be designed in such a manner as to adapt or balance any phase shifts produced by the coupling elements Cs and R1: in the main amplifier to insure operational stability for extreme frequencies as discussed in connection with Figure 3. The anode of valve Va which is directly connected to the pole of the high potential source is connected further to the ground or cathode end of the primary of the output transformer Lac through a resistance Ru. The cathode leads of valves V7 and Va include resistances R14 and Rm, respectively, and there is connected to each of the cathodes a further re sistance R15 and R11, respectively. The open terminals of the last mentioned resistances are led to a common Junction and the latter is connected to the center tap 8 of the secondary of the input transformer L55. The cathode lead of the valves V6, V7, Vs are returned to the junction of a pair of equalresistances R11 and R1: connected in series and across the secondary of the input transformer Lss.

There is developed by the cathode resistance Rm of valve Vs a potential drop r corresponding approximately to the grid potential. A further potential drop being proportional to the output potential qr is developed by the cathode resistance R14 of the valve V1. The resistance-Ru is large compared with the resistance R11 whereby there is produced at the junction between these resistances a feedback potential s=c(r-g a wherein c=1 and qz=az.q1. This feedback potential is applied through the center tap I of the transformer secondary Lee to the valve Vs on the one hand and to the valve V; on the other hand. The resistances R11 and R1: serve to maintain symmetrical conditions to prevent reaction of the fed back potential upon the circuits connected to the input terminals l2. Due to the fact that n=2, the amplitudes in the parallel system according to Equation 12 are at a minimum, whereby an effective equalization or elimination of distortion is ensured when using a valve Vs with a non-linear operating characteristic. In contrast to the circuits shown in Figures 4 and 5, the circuit according to Figure 6 functions by current control by maintaining a definite ratio between the input potential and the output current independently of the load connected to the terminals 3-4.

Referring to Figure '7, there is shown an exemplification of the invention as embodied in a radio transmitter for preventing distortion of the signal modulation. The transmitter representing the system A1 includes a push-pull modulator stage comprising a pair of electron valves Va and V10 having applied thereto the high fre-- quency carrier 71 from a suitable source across input terminals 5-! and through a coupling transformer Let. The amplitudes of the high frequency carrier oscillations are modulated in accordance with the low frequency control potential (p+s) impressed upon a different grid electrode of both valves Va and V1'o. Ln repre-'- sents a common inductive load inductance connecting the anodes of the valves Va, V10 and having its center tap connected to the pole of a high potential source in a manner well known in" the art. The amplitude modulated carrier oscillations are fed through coupling condensers Cu and Cu to a high frequency amplifier E and U11! bll serve to energize an antenna for radiation in the form of an electro-magnetic wave having instantaneous amplitudes varying according to the potential or any other magnitude (11.

A small fraction of the radiated energy is absorbed by, an auxiliary receiver constituting the system A2 and including a tunable circuit comprised of an inductance coil L shunted by a condenser C10. The received oscillations are rectified such as by a pair of diode rectifiers V11 and V12 and the rectified potential derived from the diode load resistance R21 shunted by a condenser C11 in accordance with well known practice. In this manner, a low frequency potential of desired amplitude q2=a2.q1 derived from a suitable tap point of the resistance R21 is obtained varying in proportion to the instantaneous amplitude of the high frequency oscillations m. The low frequency potential p+s besides being applied to the transmitter is simultaneously impressed upon the auxiliary system or network B whose propagation characteristics may be adapted by special means such as a. series capacity C12 and shunt resistance R22 shown to the characteristics of the main translation channel comprising the systems A1 and A2. The tap point on the diode load resistor R21 is connected to a suitable tap point of the resistance R22 through an impedance such as a choke coil In, the center tap of which is connected to the control grid of an electron valve V13 constituting part of the feedback system C. In this manner there is produced in the output of valve V13 a feedback potential s=c- (r-q2). The input potential 11 supplied at terminals |--2 is impressed upon the transformer Lee through the center tap 9 and together with the feedback potential s supplied through the secondary of the transformer Les results in a corrected input potential p+s developed by the resistance R2: and applied both to the transmitter A1 and to the auxiliary system B. The lower end of the primary of the transformer Lee is connected to the input terminal 2 through a compensating resistance R24. If both resistances R2: and R24 are of equal value, reaction of the fed back energy upon the preceding circuits connected to the terminals |--2 is prevented. By adjusting the potentiometer R22 in the auxiliary system B it is possible to vary the propagation factor b to fulfill the requirements according to Equation 5. As is understood from Equation 16 an arrangement of the type described which substantially corresponds to the principle circuit according to Figure 2 will result in a complete compensation or elimination of inherent distortions of the transmitter A1 provided no additional distortions are produced in the receiver A2.

By means of the potentiometer R21 in the receiver A2, the propagation factor a: and as a result the output amplitude q1 of the transmitter A1 may be adjusted as follows from Equation 15. For a definite adjustment of R21 0. will be equal to a1-a2=b; that is, the output amplitude qr will be equal to the amplitude of a transmitter without correction or equalization.

Referring to Figure 8, there is shown an arrangement similar to Figure 7 but operating with suppression of the carrier and equalization of distortionsmbyniyl lfseiieedhacklhaccordanse, .witli. 0

high frequency potential induced in the output transformer L14 will disappear and if the valves have non-linear operating characteristics the induced high frequency potential will vary in accordance with the low frequency potential impressed upon the grids in opposite phase through the modulating transformer L13. The high frequency signal with suppressed carrier is amplifled in E and radiated in the form of an electromagnetic wave qi similar as in the preceding figure. The receiver A2 comprises a balanced demodulator controlled by an auxiliary carrier h supplied from the transformer L17. The input transformer L15 of the balanced modulator has its secondary terminals connected each to one of the grids of the modulating valves, the center tap of the transformer secondary being connected to the cathodes in a manner well known. There is thus produced between the anodes of the valves v18, V11 a low frequency potential varying in proportion to the instantaneous amplitudes of the high frequency potential qr both in sign and. magnitude. This low frequency potential is transmitted through the transformer Lia, amplifier V12 to voltage divider R28 to serve as control potential 112 of the auxiliary receiver A2 applied to the feedback system C. The latter contrary to the preceding figure contains no amplifier. On the other hand, an amplifier V18 is provided in the auxiliary system B controlled by an input transformer L11 having its primary connected in series with the input terminals l-2 and a portion of an impedance L12 in the feedback system C and a portion of the output resistance Rat of the valve V13. tion coil L12 is connected to a tap of the voltage divider R25. The feedback potential 3 derived from the tap of the induction coil L1: is thus applied together with the input potential 11 to both the input transformers L11 and L1: of the aux iliary system B and of the transmitter A1. .As a result, distortions occurring in the transmitter are equalized in a manner understood from the foregoing.

In Figure 9 there is shown an. embodiment of the invention for controlling the intensity of a beam of light in accordance with a low frequency magnitude or potential for use in photographic sound recording or like arrangements. The arrangement shown comprises an electrically controlled light source A2 forming the system to be corrected which may be a gas discharge lamp of known construction. The intensity qz of the light flux radiated by the source A2 varies in accordance with the output potential m of the amplifier A1 supplying control energy to the device A2. A: is a photo-electric cell receiving a small portion of the light flux radiated and serving to convert the received light into a corresponding electrical potential as which is amplified by means of an amplifier A4 to obtain an amplified potential (14 impressed upon the feedback system C. The parallel system in the example shown comprises a pair of amplifying valves V21 and V22 for amplifying, respectively, the input potential p and the correcting potential 3 in such a manner as to obtain a differential potential r=b(p+s) between the anodes of the valves. The differential potential is applied through a transformer L1: to the voltage divider R2; in the feedback system C. From the latter there is derived the correcting potential s=c('r-q4) by suitable tapping which potential is in turn applied to the input of the amplifier A1 and of the parallel resistance B in a manner readily understood from the fore- The other end of the induc- HUUH going. Sinc the input; potential p is only impressed upon the parallel system p, the factor n in Equations 1, 3, 10, 12, 13 is equal to zero. By complying with the condition according to Equation it is seen from Equation 3 that in an arrangement of this type a required relation between the input potential 1) and the output potential qr of the entire translation system A1A2A3A4 may be obtained. If the systems A3 and A4 operate without distortion, there is furthermore obtained a desired relationship between the input potential 1; and the light flux qz; that is, all distortions of and interference produced in the amplifiers A1 and the light source are eliminated.

In the arrangements described a special adjustment of the parallel system B and of the feedback system C is required to fulfill the condition (5) in order to insure an efficient distortion elimination. This adjustment may be effected in an easy manner by increasing the factor b or c with the rest of the circuit being disconnected by increasing the amplification or decreasing the attenuation to a point where self-excitation sets in. Sure self-excitation by feedback e is possible only if he exceeds unity value, it is undertsood that the condition according to Equation 5 is just fufilled at the point of starting of the oscillations. If then the potential is applied, the oscillations will again disappear, as this potential opposes the control potential 1'.

If the propagation factors of the several systerms are adjustable by the provision of special regulating organs, the conditions according to Equation 5 and/or 7 or any other condition may be automatically maintained by an operative mechanical connection between the regulating organs independently of the adjustment made. In order to .prevent undesired variations of the propagation characteristic of the parallel and feedback system due to fluctuation of the con stants of the amplifiers embodied in these systems, these amplifiers may be individually corrected to have linear input-output relation by means of any of the known means.

As will be evident from the foregoing the invention is not limited to the specific arrangement as shown and described for illustration, but that the underlying principle and concept disclosed are susceptible of numerous variations and embodiments difiering from the exemplification shown herein and coming within the broad scope and spirit of the invention as defined in the appended claims. Thus, the invention is not to be interpreted limitatively as applying to electric wave energy only but may be employed with equal advantage in connection with arrangements for translating and converting other forms of oscillatory or vibratory energy such as sounds, mechanical vibrations, etc. The specification and drawings are accordingly to be regarded in an illustrative rather than a limiting sense.

I claim:

1. In a wave translation system, a main wave path and an auxiliary wave path, means for impressing wave energy to be translated upon at least the input of said auxiliary wave path, a load circuit connected to the output of said main wave path, means for combining a portion of the energy of said load circuit in inverse phase with the output energy of said auxiliary wave path to produce differential wave energy, a feedback path, means for impressing said differential energy upon the input of said feedback path, and

further means for impressing the output energy of said feedback path upon the inputs of both said main and auxiliary wave paths.

2. In a wave translation system, a main wave path and an auxiliary wave path, means for mpressing wave energy to be translated upon at least the input of said auxiliary wave path, a load circuit connected to the output of said main wave path, means for combining a p rtion of the energy of said load circuit in inverse phase with the output energy of said auxiliary wave @path to produce differential wavey energy, a feedback path, means for impressing said differential energy upon the input of said feedback path, and further means for impressing the output energy of said feedback path upon the inputs of both said main and auxiliary wave paths, the product of the propagation constants of said auxiliary and feedback paths being substantially equal to unity.

.3. In a wave translation system, a main transmission channel, an auxiliary transmission channel, means for impressing input wave energy upon at least said auxiliary transmission channel, the propagation constant of said auxiliary transmission channel being equal to the propagation, constant of said main transmission channel, further means for combining output energies of said channels in inverse phase relation, a feedback channel for impressing the combined energy simultaneously upon the inputs of both said main auxiliary transmission channels, and a utilization circuit coupled to the output of said main transmission channel.

4. In a wave translation system, a main wave path and an auxiliary wave path, means for impressing wave energy to be translated upon the inputs of both said wave paths in predetermined amplitude ratio, a load circuit connected to the output of said main wave path, means for combining a portion of the energy of said load circuit in inverse phase with the output energy of said auxiliary wave path to produce differential wave energy, a feedback path, means for impressing said differential energy upon the input of said feedback path, and further means for impressing the output energy of said feedback path upon the in zfigs of both said main and auxiliary wave pa 5. In a wave translation system, a main wave path and an auxiliary wave path, means for impressing wave energy to be translated with equal amplitude upon the inputs ofboth said wave paths, 9. load circuit connected to the output of said main wave path, means for combining a. portion of the energy of said load circuit in inverse phase with the output energy of said auxiliary wave path to produce differential wave energy, a feedback path, means for impressing said differential energy upon the input of said feedback path, and further means for impressing the output energy of said feedback path upon the inputs of both said main and auxiliary wave paths.

6. In a wave translation system, a main transmission channel, an auxiliary transmission channel, means for impressing input wave energy simultaneously upon both said channels, the energy impressed upon said main transmission channel having twice the amplitude of theenergy impressed upon said auxiliary transmission channel, further means for combining output energies of said channels in inverse phase relation,

a feedback channel for impressing the combined energy upon the inputs of both said main and auxiliary transmission channels, and a utilization circuit coupled to the output of said main transmission channel.

'7. In a wave translation system, a main wave path and a first auxiliary wave path, means for impressing wave energy to be translated upon at least said auxiliary wave path, load means connected to the output of said main wave path, means for combining a portion of the load energy in inverse phase with the output energy of said auxiliary wave path to produce differential wave energy, a further auxiliary wave path, means for impressing said differential energy upon the input of said further auxiliary wave path, further means for impressing the output of said further auxiliary wave path upon the inputs of both said main and auxiliary wave paths, and an amplifier included in at least one of said auxiliary wave paths, whereby the gain of one auxiliary wave path is substantially equal to the attenuation of the other auxiliary wave path.

8. In a wave translation system, a main transmission channel subject to distortion, an auxiliary transmission channel having a substantial distortionless input-output characteristic, means for impressing input wave energy upon at least said auxiliary channel, further means for differentially combining output energies of said main and auxiliary transmission channels, a distortion-free feedback channel with means for applying the combined diflerential energy to the input thereof, means for applying the output of said feedback channel to the inputs of both said main and auxiliary transmission channels, and a utilization circuit connected to the output of said main transmission channel.

9. A wave translation system as claimed in claim 8 wherein the propagation constants of said main and auxiliary transmission channels on the one hand and of said feedback channel on the other hand are of opposite polarity to obtain a desired over-compensation of the distortion component produced by said main transmission channel.

10. A wave translation system as claimed in claim 8 wherein input energy is impressed simultaneously upon both said main and auxiliary channels.

11. A wave translation system as claimed in claim 8 wherein input energy is impressed simultaneously upon said main and auxiliary transmission channels with an amplitude ratio of 2:1 and wherein both channels have equal propagation constants.

12. A wave translation system as claimed in claim 8 including phase shifting means in said auxiliary channel to equalize its phase characteristic with that of said main transmission channel.

13. In a wave translating system, a main translation channel subject to distortion and adapter to convert electrical input energy into corresponding output energy of different character, a distortion-free pick-up system for reconverting a portion of the output energy of said translation channel into corresponding electrical.

energy, a distortion-free auxiliary transmission channel, means for applying electrical input energy to at least said auxiliary transmission channel, means for differentially combining the out- UUui uu put energies of said pick-up system and said auxiliary transmission channel, a distortion-free feedback channel with means for impressing thereon the combined difierential energy, and means for applying the output of said feedback channel to the inputs of both said main translation and auxiliary transmission channels, said auxiliary channel and said feedback channel having propagation constants the product of which is substantially equal to unity.

14. In a wave translation system, a main amplifier subject to distortion, a converting device fed by the output of said amplifier for converting the electrical energy into energy of different character, a distortion-free pick-up system for reconverting a portion of the converted energy into corresponding electric energy, a distortionfree auxiliary amplifier, means for applying input energy to be converted to at least said auxiliary amplifier, further means for differentially combining the output energies of said pick-up system and said auxiliary amplifier, a distortionfree feedback channel with means for impressing thereon the combined diflferential energy, said auxiliary amplifier and said feedback channel having propagation constants the product of which is equal to unity, and means for applying the outputs of said feedback channel to the inputs of both said main and auxiliary amplifiers.

15. In a wave translation system as claimed in claim 13 wherein said main and auxiliary amplifiers possess equal gain and frequency characteristics.

16. In a wave translation system, a main transmission channel having imperfect transmission properties, an auxiliary transmission channel having substantially perfect transmission properties, means for impressing wave energy upon the input of at least said auxiliary transmission channel, a feedback channel having substantially perfect transmission properties, a load circuit coupled to the output of said main transmission channel, means for differentially combining a portion of the energy in said load circuitwith the output energy of said auxiliary transmission channel, means for impressing the combined diflerential energy upon the input of said feedback channel, and further means for impressing the output of said feedback channel upon the input of both said main and auxiliary transmission channels.

17. In a wave translation system, a main transmission channel having imperfect transmission properties, an auxiliary transmission channel having substantially perfect transmission properties, means for impressing input wave energy simultaneously upon both said channels in predetermined amplitude relation, a feedback channel having substantially perfect transmission properties, a load circuit coupled to the output of said main transmission channel, means for differentially combining a portion of the en ergy in said load circuit with the output energy of said auxiliary transmission channel, means for impressing the combined difl'erential energy upon the input of said feedback channel, and further means for impressing the output energy of said feedback channel upon the inputs of both said main and auxiliary transmission channels.

GUSTAV GUANEILA.

Images