US2385085A

US2385085A - Method of producing frequency modulated waves for radio transmission

Info

Publication number: US2385085A
Application number: US450596A
Authority: US
Inventors: Labin Edouard
Original assignee: Hartford National Bank and Trust Co
Current assignee: Hartford National Bank and Trust Co
Priority date: 1942-07-11
Filing date: 1942-07-11
Publication date: 1945-09-18
Anticipated expiration: 1962-09-18
Also published as: GB573406A

Description

p 8, 1945. E. LIABIN 2,385,085

METHOD OF PRODUCING FREQUENCY MODULATED WAVES FOR RADIO TRANSMISSION Filed July 11, 1942 2 Sheets-Sheet l su'lht cos? 7 cos (l t intensiiy modulator ffi. d :Diflerentihl Phase shifter Light Patented Sept. '18,

MODULATED MISSION WAVES FOR. RADIO TRANS- I Edouard Labin, Buenos Air-es, Argentina, assignor to Hartford National Bank and Trust Company, Hartford, Conn., as trustee Application July 11, 1942, Serial No. 450,596

10 Claims. (01. 179-1715) The present invention relates to radio transmission of the type in which the transmitted wave is; frequency modulated. More particularly, the

invention is concerned with the production of the frequency modulated wave by a method of direct phase modulation. 1

It is well known that the chief dimculty encountered in frequency modulated transmission, is that of ensuring the stability of the central or "carrier frequency (generally referred to by the symbol no) of the wave. The method of the present invention is based on the, idea of generatin the wave by taking the carrier term from a quartz pilot oscillator, whereby advantage may be taken of all the stability of standard transmitters.

It is known that a wave, the instantaneous i.requency of which is a function of the time flit). is written U sin! t) ,dt, where U is the amplitude of the wave. Now the expression for the wave it is desired to create, as deduced from theory is is the intelligence of crest A, It being the slope of the characteristic of modulation and the excursion of the frequency of the wave.

and therefore,

' =2r sin %sinat=sinat' (4a) The "index ofmodulation High uality gi fi s q in ca ons Quantity btosdoasting (mobile Y equipment) i Max. excursion given to instantaneous frequency of transmitted wave in p. s. B .m 70, 000 15, 000

in. frequency oiintelligence in p. s. (1min 35 300 ax, frequency of intelligence in mow vleoi'ssa""a;araztutime ligenoe, 00 0 Maximum amplitude of phase disls'cement, we. I 2.000 {50 inimum amplitude of phase displacement, his. in t 0. 02 0. 16 2 DynamicA-Wmu+\l'ai-. lmvmm well-lmown manner and on reasonable assumptions, as is usual in the art, and are intended to be chiefly illustrative.

The value of the dynamic A controls the problem of the integrating stage of the intelligence, at least in broadcasting, when it is very high. In-

deed, the tension 0 generated by the integrator must have the same range of variation as l', and

' this gives rise to a diillculty on the side of the measures the amplitude of the variable angle,

The following table gives values for certain of the principal quantities involved in frequency modulation which values have been calculated in 55 minimum values, since, if the maxima are arranged within the manageable domain of a fewvolts, the minima will fall within the domain of microvolts, where parasitic tensions reign. The

difllculty has been overcome by careful construction of the integrator. In mobile equipment for the transmission of conversations, where the dynamic reduces. to 350, the problem no longer arises.

The value of I'm controls the problem of the very generation of the frequency modulated wave.

Indeed, if Expression 1 for the frequency modulated wave is considered. it might be thought very simple to produce it with an no as stable as desired.

by Passing the output of a stabilised pilot-(no) through a phase shifter, the phase displacement of which will be I', and hence controlled by the in:- tegrated intelligence. But an prdinary phase shifter never gives phase rotations higher than a few 1,. in fact much less-about 2 5,if fidelity of the control by the intelligence is taken into account. Therefore, in order to secure the 2,000 radians in the case: of broadcasting, or even the formobile equipment, it would be necessary to multiply the argument generated by 500 or 10,000. This method has been, and still is, used, but it suffers from many complications and diiiiculties in design.

In order to avoid the need for neutralizing the amplifiers of the modulated wave and to bring the pilot frequency generator intothe domain in which its circuits and, above all, the quartz shall be more manageable, some frequency multiplications are known to be desirable. According to known practice. the carrier frequency in the gencrating stage and the total excursion may be made initially small and the requisite higher values obtained subsequently by ordinary frequency multiplication. In some instances, a still lower frequency may be. advantageous, in the generating stage, in which event the still lower frequency used may be increased to an intermediate frequency without multiplication, by frequency conversion (heterodyning) with a pure sinusoidal wave and the thus obtained intermediate frequency subsequently multiplied in the ordinary; way.

Reverting now to the main problem, namely, the production of the frequency modulated wave, it has been shown (above) that that wave may be expressed as u=U sin (iht+P) where I is a term representing the integrated intelligence and is given by 2rk fs(t) dt. It has also been pointed out that the integration of the intelligence is comparatively simple. Now, if it were possible to produce, from the intelligence s(t) an oscillation tension v=V cos I, where the value of the angle 1' is kfs(t)dt, with an amplitude of this angle as large as desired, then evidently the problem would be solved, since it would suffice to mix with an L. F. tension v=V cos I, a sinusoidal H. F. oscillation u=Uo sin Oct, of the carrier frequency Sic, taken from a pilot wave source of perfect stability, to obtain an oscillation u of the type refrom the circle, by producing tensions of the type quired, which is directly proportional to the sum of the angles representing the carrier frequency and the integrated intelligence, and is written, as above, u==U sin (nct+-l which is the type required.

At this point, it is convenient to study more in detail the process of mixture, on the assumption that a L. F. oscillation varying as the sine or cosine of the integrated intelligence, is available. If the two oscillations cos P and sin Oct are impressed on a modulator, amongst other terms the product cos l sin Oct is obtained, which still differs fundamentally from the desired oscillation sin (nct+-l'). Since the product contains this oscillation plus an oscillation in sin (act-T) it might be though easy to isolate one of these expressions by means of a filter; but deeper reflectlon will show that this would have no meaning. It is necessary to employ the following artifice; Take two radio-frequency oscillations in quadrature and of equal amplitudes:

u U sin fl t u;-' U2 cos G t} (5) and two L. F. oscillations, also in quadrature and of equal amplitude:

11 V cos v V? sin 5 (6) Thereupon it would be necessary to multiply or modulate m by m, and u: by m, in thesame degree m, thus:

mum U sin il t. cos p (7) ""1403: U cos Shtsin t and then add the two products, giving, by elementary trigonometry:

mu1vi+mu=m=U sin-('tat- -w) (a) On the assumption that an oscillation cos -1 can be produced, it is obvious that the complementary sin I could, be produced would be ideal for obtaining the frequency modulated waves, the central wave of which will have all the stability of the classical pilots, is obvious from the very constitution of such waves, and was recognised from theearliest times by several authors. But what v always appeared to be an insuperable difliculty was. precisely the generation of tensions of. the type V0 cos I' or V0 sin -I Tosay that the variable anglel ==kfs(t)dt is the argument of the tension .to be generated, is to say that that ten sion hasalready been frequency modulated by the intelligence s(t). If this modulation is effected by standard methods, it always appears around a central frequency 520' and, by following the described process, it will readily be seen that a wave would be generated having a central frequency 90:90, so that it would be necessary to stabilise 90. Obviously this is a vicious circle.

Latterly, attempts have been made to escape cos 1' or sin -1 by direct and unusual methods which avoid thepassage through a pre-modulation having a central frequency. D. Sabarof has published, in Communications, September 1941, a diagrammatic description of a method which utilises the trace of a sine wave on black and white paper, a system of illumination, photo-electric cells, rotary mirrors driven by the modulating signal, and systems of slits and optical projections. Such a'method is obviously cumbersome and costly, but it represents the only practical suggestion madeprior to the present invention.

It has now been found that a very simple and efiicient method of obtaining the desired result can be devised by making use of a photoelectric system including in the path of the exciting beam of light a pair of rectilinear polarising means arranged in quadrature and between said polarising means, a medium having voltage responsive double refraction properties, and so adjusting the operating conditions, that the rectilinearly polarised beam entering the medium from the first polarising means has impressed on it during its passage therethrough, a phase displacement which is a function of the intelligence, whereby said beam is enabled to pass through the second rectilinear polarising means and impinge on the photo-electric means, from which a current may then be derived which will contain the desired term.

It is, therefore, a principal object of the present invention to provide a novel and improvedmethod for generating, in connection with frequency modulated radio transmission, an electrical quantity varying as a sinusoidal function of an angle proportional to the integrated intelligence.

Another object of the present inventionis to provide a novel and improved method for generating, in connection with frequency modulated transmission, one electrical quantity varying as a sinusoidal function of an angle proportional to the integrated intelligence, and another electrical quantity having a sinusoidal variation proportional to the central frequency, and combining said two quantities to give a phase modulated component of the wave to be transmitted.

a} object of 'this invention is to provide animprove'd method of producing a frequency modulated ,wave for radio transmission bygeneratin'g a carrier frequency wavein a stabilised pilot; said wave'simultaneou'sly through two parallel shifting zones one of which is in "quadrature with the other, generating an intellito two parallel phase modulation zones one w ch is in quadrature with the other, one of phase modulation zones corresponding to one ot the phase shifting zones, and the other phase odulation zone corresponding to the other phase or light to pass through each phase modulai tion Qne whereby its polarisation is modified in rsliifting izone, causing a rectilinearly polarised rice and applying said intelligence simultanefjresponse to variations in the intelligence, passing latedj carrier frequency emerging from one phase Izone-with the electrical quantity derived fromthe corresponding phase modulation zone, combining the modulated carrier frequency emerging-from the other phaseshifting zone with theelectrical quantity derived from the other phase'imodulation zone, and simultaneously ,ap-

Dlyingjthetwo combinations to a common point of a transmitting means.

JA'stilli'urther-object of the present invention is supreme an improved method of producing a frequency modulated wave for radio transmission bygenerating a carrier frequency, passing said frequency simultaneously through two parallel'phase' shifting zones one of which is in quad- J raturewith' 'the other, generating an intelligence andapplying said intelligence simultaneously to ,jtwoparallel phase modulation zones, one of which isQin qua-dratu e with the other, one of the phase zones corresponding to one of the jphaseshiftingzones and the other phase modulation zone-corresponding to the otherphase shift- 'ing e,;, ;generatin8 .two beams of rectilinear light, said beams through respective preliminary-polarisation zones, passing one of said a first intensity modulation zone while applying to said zone the phase shifted car- ,rier frequency issuing from oneof said phase shifting zones, simultaneously passing the other of a second intensity mod- ;ulation;;.zone.--in-.-parallel with the flrstintensity qdul ti noz ner while applyi g to said second I intensity modulation zone the phase. shifted car'- rlerfrequencyissuingfrom the other phase shifting'zoneypassing each intensity modulated beam 1 throughankintermediate rectilinear polarisation izonezin quadrature with therespective first shifting *zonei throughthe corresponding phase "modulation Y zone, o. simultaneously passing the other varying intermediate 'beam through the other phase modulation zone, passing each phase modulated beam through a final rectilinear polarisation zone in quadrature'with the respective intermediate rectilinear polarisation zone whereby to produce "a varying issuing beam from each phasemodulation zone, the variations of sinusoidal phase modulation of the present invention. p

Fig. 2a is a diagram of the circuit connections of a differential intensity modulator.

Fig. 2b is a diagram of circuit connections for a phase shifter. I

Fig. 3 is a diagram illustrative of the operation of 9. Kerr cell. 1

F 4 is a diagram illustrating anothermethod' of operating the Kerr cell.

Fig. 5 is a diagram illustrative of the compounding of the sinusoldal factors corresponding to the carrier frequency and the intelligence.

Fig. 6 is agraph illustrative of the luminous intensityof the Kerr cell, and

Fig. 7 is a diagram illustrating a modified.

tion, namely, the generating of a sinusoidal phase modulation by utilizing a Kerrcell, will best be understood by reference to Fig.1, in which a source of light S is shown sending a beam through a lens L1 and then through a Nicol prism P1, from which the beam issues as a ray rectilinear-1y polarised in a certain plane. Said ray thenpasses through 9. Kerr cell K, which provides a path I cm s. long between two spaced plates G. which are d cms. apart, said path being filled with pure nitrobenzene'. It is known that if a potential V be applied to the plates G, this liquid has the property of becoming doubly refractive with an index of double refraction proportional to the square of the electric field V/d applied perpendicularly to the passing ray. The index of double refraction is also proportional to the length of path 1 through-the field. When V=0 there is no double refraction and the polarisation of the ray remains, at exit from the cell, in the same planeas at entry, so that if on leaving the Kerr cell, the ray is compelled to pass through a second Nicol prism :Pa'the plane of polarisation of which is perpendicular to that of said first prism,

But as soon as V no'light will pass beyond P2. has some definite value, the liquid becomes doubly refractive. and causes the ray to leave the cell with a polarisation difi'ering from that impressed thereon by the first Nicol P1. portion of light will pass through and beyond the second Nicol prism P2, to be focussed by lens Ln onto the sensitive surface of a photo-electric cell F. As a net result, an electric current is obtained at the output of the'photoelectric cell I", which is controlled :by the double refraction of Hence, a definite the liquid and it has been shown that the value of the current is exactly where B is the Kerr constant of the liquid and I is the value of the current corresponding to the total intensity of the luminous ray used if it fell directly on the photoelectric cell F.

As is seen, there is, in the variable part of the photoelectric current I, the magnitude in cos I which is wanted, and the angle -1 which is desired to create is here given by readily be shown by calculating limiting values based on considerations of practical dimensions that only the electrical efllciency is of interest, that is to say that the sensitivity of the response to the modulation, and the luminous losses will be of no importance, since they do not call for the use of unreasonably large lamps.

In order to ascertain this last mentioned point, the worst case for the light, that is to say, when the cross-section of the luminous beam has been reduced to the last limit, will now be considered. Such reduction is a natural tendency because, in order to improve the electrical sensitivity, it is always convenient, according to Equation 10, to reduce the thickness d of the Kerr cell, which necessarily forms part of the dimensions of the luminous beam. If the thickness of the Kerr cell be reduced to the lowest desirable workable limit, any trouble with respect to the small value of the thickness may be compensated by a much greater increase in width of the beam. This beam would be defined by the opening in a diaphragm probably the actual condenser of the Kerr cell would already play this part by reason of the lateral aperture it offers to the lightand'such diaphragm would be illuminated by a beam, the cross section of which would be determined by the conveniences of optical construction of the lens L1 of Fig, 1. The lenses here concerned need no offer refined optical qualities, and there would be no difliculty in finding and using ,very small lenses with a large relative aperture.

The efficiency of thephotoelectric cell F in microamperes of output current per lumen of light input must now be considered. It is already possible to work comfortably with currents the amplitude of which may be indifierently 10, 1, or 0.1 microamperes. Moreover, it is always possible to use photoelectric cells with internal current amplification by secondary emission, and such amplification nowadays reaches, in commercial types, into the millions. This means that it only the output current (not thelight for projection purposes) is used, it is possible today to utilize infinitesimally small light fluxes.

Theoretical consideration would seem to show that with a current of 1 A with which to work.

and a sensitivity of the photoelectric cell of 200 A/lumen, the power required would be 50 watts. and this value can be still further reduced by choosing a slightly smaller value for the current, or a cell with a somewhat greater sensitivity.

The electrical sensitivity may be increased by the liquid used (nitrobenzene) would appear to be a further means of influencing the sensitivity.

The reduction of the separation between the plates and the increasevof the length of path through the Kerr cell cannot be carried beyond certain limits in practice, for constructional and travels impressed on the beam by two lateral mirrors. This type of Kerr cell is known but it was introduced into television for the purpose of increasing the optical efficiency.

Obviously, the increase in electrical sensitivity was also utilized, but the development was towards a multiple cell structure which could accommodate a beam of light the cross-section of which was enormously extended in a direction perpendicular to that of the necessarily reduced separation d of the plates G. This led to quite complicated structures which entailed such an increase in the capacity of the Kerr condenser that, hearing in mind the very high frequencies used in television, the impedance offered to the voltage V dropped until it formed a highly objectionable load for that voltage in so far as the power to be developed was concerned. In the present case, the structure may be very much simpler: two mirrors, or better still, two silver deposits on the glass walls which enclose laterally the Kerr cell, are sufllcient, as maybe seen from Figure 3. If the light is caused to fall on said mirrors with a small angle of incidence, as many to and fro passes may be made as desired, with a very small width of plate. With a width of 2 cms. for the 5 cms. length which is found admissible, and with a thickness of 0.01 cm., and bearing in mind the dielectric constant of nitrobenzene, the capacity C becomes approximately 1,000 [LII-F- At the highest frequency amax which the applied tension V will contain, that is to say, 100, p. p. s., this capacity represents an impedance of 15,000 ohms. It can be shown that, with such an arrangement the impedance constituted by the cell not only represents a completely negligible load for the tension V of the order of magnitude to be used (as hereinbelow defined), but is alsoadvantageous for immediately allowing V to reproduce faithfully the grid voltage of the valve from which it issues, independently of the circuit associated with the plate. In fact, it suflices to employ for that valve a triode with a suitable internal resistance say of the order of 5,000 ohms, in order that the Kerr cell shall always represent an open circuit, for the output. This will represent the optimum conditions for the, process of integration as known In the second manner or method, the

from the theory of this operation.

Calculations following the lines indicated above, lead to a possible maximum value for the Constant a This shows that the electrical sensitivity also.

need not be. regarded as a troublesome problem, and so much is this so, that in the following, it

is possible directly to eliminate all eonsidera-.

tions as to the construction of the Kerr cell, since there is the foregone assurance that it can be adapted to all classes of specification with an almost inexhaustible field of play in all the mechanical and optical dimensions.

There are now available three courses for com- 7 pleting the task involved in the present invention, said courses being three manners of modulating the I Angle 21B1V proportional to the square root amplification can be avoided,'by dividing the tension V into a fixed portion V0 and a portion 0 which will reproduce the integrated intelligence but which will always remain small integrated intelligence; I In the more obvious first manner or method, the integrated intelligence v=2wkfs(t)dt, is-

passed through a specially designed amplificationstage which supplies an output tension 3') proportional to the tension 22:

V= MI V =bv=21rkfs(i)dt (11 v Such a design is already known by the technique which allows all kinds of response curves to be made with electric waves. See, for example, M. Ziegler; The Construction of a Hot Wire Anemometer with Linear Scale, Verhandelin'gen der Akademie Amsterdam, Deel xv, No. 1. If the output voltage of the square root stage is applied to the Kerr cell, then -I'=Icfs(t) dt will be ob-- tained directly as desired. The two signals in cos I and sin I will be obviously taken from two cells, of which the pairs of Nicol prisms (511i, P2) will be disposed perpendicularly to each 0 er. r

The electrical design is now very easy. As a matter of fact, the presence of the square root stage makes it possible to work at its input with a large maximum voltage 11. Hence, the situation of the intelligence integrating stage which sup plies v is considerably relieved with respect to current practice.

If the values of Table 1 are introduced, it follows that the conditions may be fulfilled with the following values:

Broad- Mobile 7 casting equipment V. in volts 100 50 2% 0. 0125 0.0004

square root of the .input and therefore cos a(Vo=+v') =1. It will be'shown below how the upsetting of such a cholse by t!!! variations in the term v may be prevented. Then cos l =cos 2111700 (141) and it is'seen that the angle I! has now become obtained:

sin I'=sin ZaVo'v (14s) The values V0 and Vo'may lie very closely together, since it is sufilcient if and it can be shown that values of V0, V0 can be used which are sufilciently big so that the difference of I shall correspond to difierences less than 1% in V0. Then the two cells will give exactly the two desired signals.

It is, of course, necessary to limit 0 so that the angle as shall never-exceed a small value. From this it follows that the first condition of the problem will therefore be of the type:

awm=a25 (15) The second condition will be, as'always, that the useful maximum angle 2aVuvmu shall correspond to the desired total excursion. According to Table I, and accepting a normal frequency multiplication, we have a condition of the type:

Following out the calculations with these conditions, we arrive at values of the following type:

Broad- Mobile casting equipment a 0. 05 0; VL 350 QC") Finally, it should be noted that the present tegrating amplifier of the modulation, which change does not always follow immediately, since the operating point of such an amplifier is, as is well known, quite critical. I

The third manner in which the process of phase modulation can be practiced, is developed as follows: The use of multiple passes of the ray,

which can as easily reach high as well as lowv values, suggest the utilisation of the same parameter of length l for controlling the angle in Considering Fig. 4 in which'the electrical field V is suppose-d applied perpendicularly to the plane of the paper, it is seen that when the mirror E is turned, on which the ray coming from S is reflected before being admitted into the cell, the length of path may vary enormously, from one to and fro movement to tens of such movements. Obviously the ray will leave the cell from one or other side thereof, and though this cannot be controlled, it is not serious since it is sufllcient to locate outside of this cell and to one side thereof, a reflecting system having a, field large enough to send everything to one and the same side. A more serious matter would be an irregularity in the length l of the path due to there not always being a whole number of to and fro passes. But on making accurate calculations, it is readily seen that such irregularity does not exist and that the length is always exactly:

if 0 is the angle of incidence of the ray at entrance and h the height of the lateral mirrors.

The angle 0 is controlled .by the rotation 8 oi the mirror E. If n reflections are used between the rotary mirror properly so called, and the entrance into the cell, then 0=oo+2n6 (l9) where 00 is the minimum incidence giving the greatest length of path liner. Therotation 5 may be controlled in its turn in several known ways, by the current i, which reproduces the integrated intelligence. For example, the rotary mirror may be suspended from a coil through which said current i passes and which is located in a fixed magnetic field H, as in moving coil ammeters; or else the mirror may be attached to a magnetic needle, the current then circulating in amagnetising coil of field H =hi. By adding to the system various fixed magnetic fields, duly arranged, in amplitude and direction, it is readily possible to obtain a law of rotation is as a function of i: i

such that, when introduced into Equation 18, with the help of (19), a practically linear control of the length lby the current i is obtained, within the limits of variation of i and 6:

= A Bi (20 bis) graphs, since it must respond to variations having up to 10,000 periods per second.

In this method or manner, it will be seen that it is possible to use a single Kerr cell for gen erating the two complementary signals, by separating the paths of the two rays and giving the one, by an arrangement of mirrors, a supplementary. travel corresponding to an angle of the angle being concisely given by which in practice approximates and is assumed equal to 'the intelligence.s(t)-if necessary already integrated-and makes of it the two signals in quadrature of Equation 6. The said phase modulator is precisely the element which comprises essentially the Kerr cell, as hereinabove explained, and in which the phase, modulation is impressed on the intelligence according to the method which is the principal object of the present invention. The multiplication of each L. F. oscillation with the corresponding H. F. oscillation is carried out in two classical modulators Mi, M2, which have been designated intensity modulators" to distinguish them from the phase modulator. To add the two products thus obtained,

it is sufiicient to apply the outputs of the intensity modulators simultaneously to a common point of a transmitter, as indicated at T in the figure.

An ordinary intensity modulation generates not only the products ('7), but at the plate of the modulating tube, there also appear currents which reproduce the two individual oscillations applied. The L. F. oscillation will not give an output voltage, because the plate circuit will be closed through a. filter capable of rejecting it completely. But such a filter cannot reject the current in the carrier frequency 0c, located right in the middle of the band which it must transmit. However, a differential modulator of the classical type shown in Figure 2:: can be used to eliminate the carrier by a process 0! internal opposition. Furthermore, it should be mentioned that the presence of a certain portion of pure carrier bee yond the modulating system, and hence up to within the aerial, does not do much damage to the reception of the useful wave, and so long as such portion does not exceed a value of some tenths (in power) of the useful wave. In fact, it is known that the reception oi .trequency modulated wave offers a very good protection against interference signals the amplitude of which is a few times lower than that o! the desired signal, wherever the frequency of the. for-- mer is located with respect to that of the latter. The same may be said of any lack of symmetry Therefore matters must be so arranged that the,

between the two paths, which is translated by the persistence of a lateral band at the output, alongside the desired wave. It is thus seen that the arrangement of the system of mixture is not critical. Fig. 2b illustrates one of many. possible types of phase shifter.

Direct intensity modulation for application to frequency modulation may be achieved as follows. The multiplication of a L. F. tension v1==-V cos I by a H. F. voltage u1= Uo sin not, may be effected without the differential modulators of Fig. 5, whenever the intensity V of the L. F. volt- 8 can be controlled independently. Let it be supposed that the phase modulator. which provides V cos I from s(t) (or from I =kfs(t) dt) operates without perturbations whatever may be the amplitude V, and, more particularly, even if the amplitude V varies in time with a H. F.

rhythm. Then it is purposely controlled by the piloted frequency, that is to say, it is made to be of the type V= Vo sin (kt, so that from the phase modulator the output is directly the product Va sin 9st. cos I The complementary phase modulation path, which generates sin I', would be modulated in intensity by the piloted oscilla-' tion in quadrature cos not. The combination at a common point of the outputs of the two paths effects the addition required to give the complete wave. In this second method of intensity modulation, two differential modulators as well as two separate outputs for the two phase modulation paths are suppressed, and the necessary devices are added to control the amplitude of the L. F. oscillations in accordance with the rhythm of the H. F. oscillation. The net result of this is always a simplification of the system.

It will now be shown how the amplitude of the magnitude created in the phase modulation Kerr cell may itself be modified. As shown above, a Kerr cell supplies an electric current -31 cos @11 being the intensity corresponding to the total intensity Jo of the ray of light entering the cell. Now, it is known that the cell has an inertia not greater than 10- sec., which means that the cell will perform the previously stated work without change in spite of any concomitant variation in Io, provided that such variation is no more rapid than 10' times a second. It is therefore seen to be possible, with the frequency Qcr of several megacycles, to utilise the foregoing principle, that is to say, to suppress the separate intensity modulators and directly modulate the luminous intensity Jo at the rhythm of the piloted oscillation sin Slot. The subscript r will indicate the magnitudes before frequency modulation. V

The modulation of luminous intensity offers no great difficulty; another Kerr cell between crossed Nicol prisms would he used, and such cell will now be briefly described.

angle aU always varies in the neighbourhood where cissmall. The arrangement in Figure '6 in which the curverepresentsxthe normal response Jo=f(U) 'of a cell as a function of U (theangle of aU correspondingto 11 being shown beside U on the horizontal axis), and

to the piloted Voltage u. but of the type I providedwith a mean portion. the V which will issue from the photoelectric cell be: s 1

iO-cos W)+I7(A sin c e -cos w 22) v The second part of this is desired, the first ,But since the useful part is of high frequency,-

Since the luminous response of a cell as a function of the tension U energising the plates, is J=' i(1 cos aU2) response of the cell vary not as cos a, but as u.

theflrst step must be the suppression of the ef- Y whereas the useless part is of low frequency, it is easy to prevent it from being troublesome by pro viding at the output, instead of a resistance, a circuit tuned to ncr (with a band passing in)! In any event it is always necessary to providea tuned circuit at a point where it is desired to take? oil a high frequency voltage. 1

The diagram of Fig. 7 illustrates the elements included in twomodulation paths according to the present method. The light J00 generated'by the lamp S passes through the lenses L1, L1", two

Nicol prisms P1, P1.", and then through the in.

tensity modulation cells K1, K1". Theserhave applied to them the same mean voltage U0, and the two tensions in quadrature of frequency. Qcrz 3 or 6 me, supplied by phase shifters P8. connected to the output of the pilot p. The Nicol;

Prisms P2, P2", through which the two rayspass on issuing from the cells K1, K1", act simultaneously as prims of entry to the phase modular: tion cells K2, K2", and as complements to the prisms P1, P1". The last mentioned cells are. controlled by the intelligence according to one of, the three methods already described above. The; device is completed by the two prisms Pa, Pa", and the lenses L2, L2", which direct the two rays; together onto a common photoelectric, cell F, the

output circuit of which is provided with a circuit;

tuned to (in.

According to the foregoing descriptionand the values which have been found by typical calculations, it will be appreciated that in the practi cal application of the invention, the whole of the. parts which effect the modulation, both as to intensity and as to "phase, can. always be genclosed in cases the dimensions of whichwill not; go beyond declmeters in the case of mobile equip-7 ment, nor be very much biggerin-the case of broadcasting. Such cases .will have. onlythreeg external terminals, to one of which the piloted:

oscillation would he applied, another receiving-the (integrated) intelligence, while, from the third units which will adapt themselves very readily to all the usual types of transmitter.

Although in the course of the foregoing description reference has been made to certain preferred values, modes of operation and forms of embodiment of apparatus, the present invention is not intended to be limited thereby, and it is understood that sundry modifications may be made therein without departing from the spirit and scope thereof as defined in the accompanying claims.

I claim: I

l. A method of generating an electrical quantity varying as a sinusoidal function of the sum of two angles, comprising the steps of generating a wave having a stabilized frequency, generating a varying and oscillating electrical quantity, de-

riving from said wave two oscillations in phase quadrature, rectilinearly polarizing two light beams, the polarization of one beam being in quadrature with that of the other, phase modulating each polarized beam by subjecting the same to voltage excited double refraction, varying the which is a sinusoidal function of said varying and oscillating electrical quantity and the function in one beam is in quadrature with that in the other beam, converting each beam into a corresponding varying electrical quantity, combining the electrical quantity derived from one of the phase modulated beams with one of the oscillations derived from the stabilized wave, combining the electrical quantity derived from the other of the phase modulated beams with the other of the oscillations derived from the stabilized wave, and electrically combinin the so-derived combinations.

2. A method of generating an electrical quantity according to claim 1, in which said control quantity is directly proportional to the square root of the sum of the useful oscillating electrical quantity and a mean quantity equal at least to the maximum of said oscillatingelectrical quan-, tity.

3. A method of generating an electrical quanv tity according to claim 1, in whichsaid control quantity is directly proportional to the squareroot of the sum of a mean portion and the useful and oscillating electrical quantity and said oscillating electrical quantity has such a value with respect to the mean portion that the square root relatrajectory in response to a control quantity havmg a relationship with said oscillating electrical quantity, rectilinearly polarizing the doubly refracted beam in quadrature with the initial polarization whereby the resultant intensity of the beam is caused to vary as a sinusoidal function of a useful angle depending on said control quantity, adjusting the relationship between the control quantity and the useful oscillating electrical quantity so that said useful angle shall be directly proportional to the useful oscillating electrical quantity, and converting the beam into a similarly varying electrical quantity.

5. A method of generating an electrical quan-' tity varying as a sinusoidal function of the sum of two angles, comprising the steps of generating a wave having a stabilized frequency, generating 'a varying and oscillating electrical quantity, de-

riving fromsaid wave two oscillations in phase quadrature, rectilinearly polarizing two light :beams the polarization of one beam being in quadrature with that of the other, phase modulating each polarized beam by subjecting the same to voltage excited double refraction over a por- I of whereby the resultant intensity of eachbeam is caused to vary as a sinusoidal function of an angle depending on said control quantity, adjusting the relationship between said control quantity and the varying and oscillating electrical quantity so that said angle shall be proportional to said varying and oscillating electrical quantity and each beam shall have impressed on it a phase shift which is a sinusoidal function of said varying and oscillating electrical quantity and the function in the one beam is in quad-.-

prature with that in the other, converting each beam into a correspondingly varying electrical quantity, combining the electrical quantity derived from one of the phase modulated beams with one of the oscillations derived from the stabilized wave, combining the electrical quantity derived from the other of the phase modulated beams with the other of the oscillations derived from the stabilized wave, and electrically combining the so-derived combinations.

. 6. In the art of radio transmission, the method of generating an electrical quantity according to claim 4, in which the length of path of the beam in the portion of its trajectory subjected to voltage excited double refraction is varied by impressing on the beam an angular change in direction prior to the entry thereof into said'tralecto y portion, subjecting the beam to a plurality of alternating opposed transverse reflections while traveling through said portion of its trajectory, the number of said reflections being dependent on the degree of said angular change of direction,

and controlling the degree or said angular change of direction by said control quantity.

\ 7. A method of generating an electrical quantity according to claim 5, in which the length of path of each beam in the portion of its trajectory subjected to voltage excited double refraction is varied by impressing on the beam an generating a varying and oscillating electrical quantity, deriving from said wave two oscillations in phase quadrature, rectilinearly polarizing two beams of light the polarization of'one beam being in quadrature with that of the other, subjecting each beam to voltage excited double refraction while controlling the degree of refraction of one beam by one of the oscillations derived from said 'wave and the degree of refraction of the other beam by the other of the oscillations derived from said wave to impress on each beam an intensity modulation corresponding to the respective oscillations, rectilinearly polarizing each intensity modulated beam in quadrature with the respective initial polarization whereby the intensities of the twice polarized beams corresponds directly to a sinusoidal function of one of said angles one function being in quadrature with of another angle depending on said control quantity, adjusting the relationship between the control quantity and the varying andoscillating electrical quantity so that said other angle shall be proportional to said v rying and oscillating electrical quantity and each intensity modulated beam shall have impressed on it a phase shift which is a sinusoidal function of said variable and oscillating electrical quantity the motion in one Ibeam being in quadrature with that in the other, converting each beam into a similarly varying electrical quantity and electrically combining the so-derlved combinations.

9. A method of generating an electrical quantity according to claim 1', comprising the step of controlling the variation of the electrical quantity by adjusting the mean portion of the varying and oscillating quantity.

10. A method of generating an electrical quantity varying as asinusoidal function of the sum or two angles comprising rectilinear-1y polarizing a light beam, subjecting the polarized light beam to voltage excited double refraction proportional to variations of a control quantity, rectilinearly polarizing the doubly refracted beam in quadrature with the initial rectilinear polarization, subjecting the polarized doubly refracted beam to voltage excited double refraction proportional to variations of a second control quantity, rectilinearly polarizing the said beam inphase with the initial rectilinear polarization, r'ectilinearly polarizing a second light beam, subjecting said second the other, subjecting each polarized intensity I modulated beam to voltage excited double refraction, controlling the degree of double refraction in response to a control quantity having a relationship with said varying and oscillating electrical quantity, subsequently rectilinearly polarizing each of said beams in quadrature with the respective second polarization thereof whereby the resultant intensity of each emerging beam varies as a combination of the respective sinusoidal functions of said one of the angles and light beam to voltage excited double refraction proportional to variations of a control quantity in quadrature with the first of the control quantities applied to the first light beam, rectilinear ly polarizing the doubly refracted second beam in quadrature with the initial polarization thereof, subjecting the so polarized doubly refracted beam to voltage excitedz double refraction proportional to variations of. a second control quantity in quadrature with the other of the control quantities applied to the first light beam, rectilinearly polarizing the second light beam in phase with the initial rectilinear polarization thereof, converting the so treated beams into corresponding electrical quantities, and combining the so derived electrical quantities.

i EDOUARD LABIN.

mte e? @errafion Patent No. 2,385,085,. Y

September 18, 1945.

EDOUARD LABIN.

It is hereby certified that errors appear in the printed s ecification of the above numbered patent requiring correction as follows: Page 2, t column, line 51, for I the word though read thought; page 3, second column, line 30, for fsinusoldal? read sinusoidal; page 4, first column, linen-14' and-15, page 5, first column, lines 6, '22 and 69, same page, second column, line 8, and age 6, second column, line 13, for that portion of the equations reading B1 read Z; page 4, first column, line 12, after "which second occurrence, insert the word it; page 5, first column, line 43, for waves read valves; and second coll, line 20, for cholse read choice; and that the said Letters Patent should he read with these corrections therein that the same may con form to therecord oi the case in the Patent @fice.

Signed and eeeled this 25 day of December, A.- D. 1945.

LESLIE FR,

First Assistant Commisee" 1: ej Patenteo