Патент USA US2406019код для вставки
Aug; 20, 1946. E. LABIN 2,406,019 PULSE MODULATION SYSTEM Filed April l. 1941 4 Sheets-Sheet 1 .n . . Wì| k ATTORNEY Aug. 20, 1946. E. LABIN PULSE uonULATIo? SYSTEM Filed April 1, 1941 2,406,019 . 4 Sheets-Sheet 2 LV .NÓEY INVENTOR ATTORNEY l Allg. 20, 1946. _ E, LABlN 2,406,019 -PULSE IODÚLATION SYSTEM - Filed April 1, 1941 ' 4 Sheets-Sheet 3 FIGA; lNVENTOR BY ATTCRN l Aug. 20, 1,946. 2,406,01 9 E. LABIN PULSE MODÚLATION SYSTEM vFiled April 1, 1941 4 Sheets-Sheet 4 MW„aM6 w amava/.Aw ¿wn-calml l . reared' 20» 1946 amig e! È. 2,409,019 PULSE MODULATÍON SYSTEM Emile Labin, New York, N. Y., assignor to Inter ' national Standard Electric Corporation, New Y York; N. Y., a corporation of Delaware application April i, i941, serial »10,386,282 i2 claims. (ci. 25o-»6) l , This invention relates to improvements in pulse modulation systems. It is a primary objective of the invention to provide an improved pulse modulating system. Another object is to provide more interference free system of radio communication. ' A further object is to provide an improved sys tem of radio transmission which may function despite deliberate attempts to jam the transmis sion frequency. ~ io It is also an object to provide an improved and 2 (i. e., from the beginning to the end of the modu lated pulse) to be intelligible as in the former case. It might also be observed that neither of the above-indicated systems is free of phase mod ulation, as will later be clear. In accordance with features of my improved form of pulse modulation system only one ex tremely short pulse is needed to identify a. signal, no matter what the degree of modulation. This improved method contemplates the time-dis placement of a single pulse one side or the other about a given norm in accordance with the de simplified pulse modulation reception system. Still another object is to provide an improved system of secret radio transmission. A further object is to provide an improved form of pulse modulation system in which a. single, gree of modulation; hereinafter alternately re ferred to as time-modulation. Since the maxi mum time-displacement of the impulses is pref erably small in comparison with the time-inter constant-width impulse may suiilce for indication val between pulses. a large part of the time is not of intelligence. used for transmission; and suitable receiving ap paratus may include a sharply selective circuit It is also an object to provide a pulse modula tion system adaptable to relatively simple receiver 20 tuned to the pulse frequency and blocking means energized by said selective circuit for excluding means. substantially all noise but the displaced impulses. Another object resides in provision of a pulse Appropriate displacement responsive means may modulation system in which there is no phase modulation of, the impulses. then detect out the audio` signal in a known man Other objects and various further features of novelty and invention will hereinafter be pointed’ I consider it preferable to a more ready under out or will occur to those skilled in the art from standing of my invention ñrst to indicate gen a reading of the following specification in con-, erally how I treat signals in accordance with fea junction with the drawings included herewith. tures of this invention, and then to describe pos 30 sible means for giving such treatment. Accord In said drawings Fig. l is a graphical showing of the effect of ingly, vreference will first be made to Fig. 1 which represents graphically the general treatment of signal treatment in accordance with features of my invention; pulses in accordance with features of my modu Fig. 2 is a block diagram of a pulse transmitter according to my invention; Fig. 3 is a block diagram of a pulse receiver incorporating features o-f my invention; Fig. 4 is a schematic circuit diagram of a por 35 lating system. . The pulses modulated as indicated correspond to a spectrum of frequencies with bands located near the harmonics of the pulse repetition fre quency. Each harmonic may be modulated in tion of the transmitter of Fig. 2; and 40 amplitude by time modulation of the pulses. If Fig. 5 is a graphical showing of successive any particular harmonic of the transmitted en changes in wave shape in portions of the circuit ergy be selected for detection or demodulation, of Fig. 4. the original signal may be restored by 'relatively simple means forming a, feature of this invention. . Pulse modulation systems are known in which the amplitude of the pulse is always maintained 45 In order to avoid combined amplitude and constant and the duration of the pulse vmade to phase modulation of harmonics and increase the vary in accordance with speech or other modu percentage of amplitude modulation of harmon lating signals. It is also known that the whole ics, I propose to displace alternate impulses sym pulse-so modulated need not be transmitted, but metrically and in opposed senses about their nor rather two pulses, each marking one determinant 50 mal unmodulated position. In other words, the end of the modulated pulse. This latter method impulses are displaced to new, modulated posi-_ has advantages over the former, in that still tions which are in time successively before and greater-signal-to-noise ratios may be obtained, after` the normal instants at which unmodulated but has the disadvantage that a signal so trans impulses would occur. In such a modulation sys mitted requires just as much of a time-interval 55 tem the impulses may be considered as made up g 2,406,019 3 Y lof two superimposed trains of impulses, each o! normal, unmodulated period T-in one of which time-modulation occurs generally in onev direc tion along the time axis, and in the other of which ` as will be clear. After substituting Equations 4 and 5 in expression (3), we obtain ` time-modulation occurs generally inthe oppOsi tion sense'. ` Referring to the diagram in Fig.- 1, odd impulses Now, if we assume that the time-displacement 0v is always very small, such that 0=-_Ktu, - I, 3, 5 kare shown schematically with respect to time and as being time-modulated in a sense op (O<K<1), then, calling posed to that of the modulation of even impulses 2, 4, Ii.` It wil1 be clear that if to represents the 15 ' maximum amplitude of time displacement for the Y impulses, one set of impulses I, 3, 5 etc. will be ' displaced :2te with respect tothe other (2, 4, 6 2”,_‘„"t"=mwt„<1 . a distortion factor e may be described as etc.) for maximum modulation, depending upon the magnitude of modulation. ` As stated above, the series of pulses shown in 20 e=2mTm<<i simplifications for- our purposes: Fig, 1 may be considered as formed of two identi cal series, each having a period T, which series cos mw0==11 we shall designate as P1 (for pulses I, 3, 5 . . .) and ’ ` 25 and P2 (for pulses 2, 4, 6 . . .), respectively. <8) Such an assumption will permit the following tall mw9=mw0 By appropriate choice of time origin the center Wherefrom an appropriate expression for Ap be of the ñrst pulse I of series P1 may occurat time _ comes _ AP: Em( 01, and the amplitude A of pulse I may be de scribed as- *_ Ap being the sum of amplitudes at frequencies ’"/r each of which is amplitude modulated bythe ‘ AP1=ZMAW cos mou-61) signal 0, as will be clear. Now, assuming the simplified case of 0 sinus oidally varying with a period a, We may write (l) in which , l and m is any harmonic. It will be noted that no sine terms appear in this expression because the function is symmetrical. v(10) once be observed that the depth of' modulation (2) Where 62 represents the time at which the center » mao. tan mwb=K.e. tan mob It will be recalled that K is always equal to or y less than 1, and that e is much less than 1. 45 Therefore. in order to obtain large orders of modulation Ke. `tan mwb should approximate unity, or tan mwb should equal of impulse 2 occurs with respect to the time , origin. 0=K.ío. Sin (at-Ho) ` 40 for the mth harmonic is--- In the same Way, the amplitude of a pulse in the series P2 may be written as_ Ap2=2„,.Á,„.-cos mw(t--02) ` Upon examination of expression (9), it will at „ By combining expressions (1) and (2) it will be clear that the combined eiîect of both trains 50 of impulses will be described by If `the time origin has been appropriately se lected, a simpliñed solution may result. Assum l Ke In determining what phase relation one series of impulses should have to the other, care must Vbe taken that at no time, with or without modu lation, do the pulses of one series pass beyond 55 precisely the mid-point in time between two im-. pulses of the other series. Such occurrence, it will be clear, would give rise to the undesirable 1 effect of doubling the impulse frequency. ing, then, a time-displacement of both series P1 In order, then, to avoid a doubling of the period-T, and P2 in opposite. directions by a time 0, We may 60 characteristic of each seriesl of impulses, we may arbitrarily describe 61 and 62 as define b as T 01 :î- b- 9 . T i Hand , _ , b=E--to (11) 65 in which case, for m odd, 92 b -|- Ú tan mwb= tan e =-1e in which 2b arbitrarily represents the distance between the axes of modulation of the ilrst two and cos mwb=sin e=e pulses after the time origin, and from which 70 and, for m even, and ' -Y cos mob-:cos e=1 75 The general term of series (3) may now be writ 2,406,019 5 . ten in two forms, depending on whether m is odd a or even; thus, for m odd 2Am.e[1-K sin (at-l-qnl cos met desired and interfering signals may be expressed (12) and, for m even - assuming an interfering signal substantially _uni 2Amll-Ke2 sin (at-l-qù] cos mot (13) form with respect to time. Of courseffor best From expressions (12) and (13) it will be noted reception, free as possible from interference. that the presence of e in the amplitude term will make for odd harmonics of relatively small ampli 10 tude, while even harmonics, not being associated .should be made small as possible; and in the case with the factor e, will he relatively large. On the of ` other hand, odd harmonics are highly modulated, whereas there is hardly any modulation of even harmonics. ` It will further be noted from these same ex It will be seen that, for any given value YofV e, p pressions that, inasmuch as ai and 02 vary in is inversely proportional to m, and that, there opposite directions, there are no phase modula fore the higher the harmonic chosen for detec tion terms, and that, therefore, all modulation >is in amplitude. That such relative variation of 20 tion, the greater will be freedom from inter ference. Since e may be selected as desired, it 6i and 02 is responsible for'elimination of phase will also be appreciated that _to should be as small modulation may be seen if we consider one series as possible, both in order to obtain less prob of impulses as unmodulated while modulating ability of interference and to see that modula- the other. In such an assumed case, for m odd, tion (see expression (7)) follows a substantially we would obtain, instead, of expression (12) 25 linear4 rather than a tangential law. In review ZAWeIÍII-â-ç sin (at-Hä] cos mw(t----g-) (14) then, as high a harmonic as possible should so be chosen that, when considered in view of the smallest possible to and d, the term e is not so lation is necessarily only one half that of the pre 30 large that linear modulation is impaired. In ferred case of expression (l2), and that phase any case, the final value chosen will be a com promise between the frequency band available modulation is present. It may be observed, how and freedom from interference. ever, in this connection, that the phase modula In the drawings I show a preferred possible tion present in the case of expression (14) is not very harmful to speech transmission; but, of 35 circuit for obtaining the above-indicated de sired results in accordance withfeatures of the course, for utmost fidelity purposes, it would be invention. In Fig. 2, which shows a transmitter considered undesirable. Y In this case it will be noted that depth of modu Returning again to the preferred system in _in block diagram form, impulse energy may be derived from a pulse generator 20 and then sup which both series of impulses are modulated in opposite directions, let us consider reception and 40 plied to a converter circuit 2|, which preferably includes a multivibrator, as will later be clear. detection of signal impulses. Since each odd Converter circuit 2| may simultaneously be fed harmonic, as will be clear from expression (12), ~ by speech or other modulating energy by known is purely amplitude-modulated, a circuit tuned . means 2‘2. The appropriately modulated im to some such odd harmonic of l/T would per pulse energy output from converter 2| may mit the amplitude modulated term to be removed thereafter be treated in accordance with known and thus result in detection of the signal. Fur means, which are suitably labeled on the rest of ther utilizing received energy in accordance with -the transmitter diagram. features of the invention, any Veven harmonic Fig. 4 shows a possible circuit for converter term, which as we have seen from expression ' (13) may be very large, may be employed, after 50 2| schematically and in some detail. In a pre ferred form the circuit is designed to set up the suitable shaping. to obtain a generally square shaped blocking signal for blocking out recep tion of energy for substantially the interval .be tween pulses. By making the even harmonic cir two series of pulses l, 3, 5 . . . and 2, d, 6 . . . so that, when speech or other signals are applied, time-modulation of one set of pulses will be op cuit as selective as possible, it will be clear that 55 posed to time-modulation of the other. The cir cuit of Fig. 4 may more readily be understood parasitic disturbances may be reduced to a min when viewed jointly with the wave forms of Fig. imum, and that substantially only the signals 5. First, we shall consider the case of no modu themselves will be admitted to the detection lation, tracing wave treatment throughout the circuits. A further desirable feature of my invention 60 circuit of Fig. 4, as depicted, in an exaggerated `manner for purposes of clarity, by waves a may be seen in its use as an- improved secrecy through g of Fig. 5. A source (not shown) of system. Since the steady component existing continuous variations preferably supplies input in the signal is modulated but very slightly, ordi terminals 23, 24 with sinusoidal variations (see nary detection methods will only be able to de tect a small unintelligible disturbance, and com 65 Fig. 5a) in synchronism with impulses later to be superimposed thereon, that is, such impulses mon circuit and other noise will probably be large preferably occur whenever these 'sinusoidal enough compared with such very slight modula variations pass through zero. For purposes that tion that the signal may be completely unintel-y will later become apparent I next prefer to shift ligible, as will be clear. This brings us to a consideration of the signal 70 the phase of this input voltage so that it is sub stantially in quadrature with the impulses, that to-noise ratio and related factors. If we assume is. so that the impulses later to be superimposed that the receiver is blocked continuously except occur at instants of time corresponding to posi during the pulse and modulation interval tive and negative maxima of the input voltage. 2me-d, d rheling the length of the pulse, the prob ability of interference for equal amplitudes of 75 A circuit for eiïecting the desired phase displace 2,406,01 9 8 , ment may. for example, include series capacitance to tube V4 momentarily drives thatidischarge de-> means Ca and shunt resistance means Re, as shown.` 'I'he input sinusoidal voltage, so shifted in phase may then be substantially squared on by well-known limiter means including an over vice below cut-oiï, thus permittingv a sharp in crease in plate voltage. This relatively positive potential on the plateìis transmitted through excited amplifier tube V1 and appropriate cir cuit elements. Output energy from tube V1 may Thereupon, because' of the conducting state ‘of condenser C, to render tube V5 conducting. tube Vt, a resultant drop in cathode resistor R shifts cathodes of both multi-vibrator tubes posi preferably divided for application to tubes V2 tively. This shift has the effect of making the and V3 in push-pull. The outputs of tubes V2 10 control grid of tube V4 so negative with respect and Va may be connected in parallel as shown so to the cathode that the cut-off condition set up that a unidirectional output signal may be ob by the momentary impulse is maintained a little longer. v tained. For purposes hereinafter to be indi cated the t'ap of potentiometer 25 is so adjusted The cut-oiï state of tube V4 will in every in that the push-pull control potentials applied to 15 stance be maintained for a period of time de tubes V2 and V3 are unbalanced; that is, one is pending upon the magnitude of negative impulse substantially- greater than the other. As a re applied. For example, pulse 21 being more nega sult of the limiting eii’ect, adjustment of `poten tive than pulse 30 will retain tube V4 in the cut tiometer 25, and the parallel output connection oiî state longer than pulse 30 will. This will be of‘the tubes V2 and V3, an output wave form 20 clearly seen from the fact that condenser C1 is somewhat conforming to Fig'. `5b is obtained. variously charged depending upon the magni At this Ystage òf the sinusoidal input signal tude of the positive pulse output of tube V4 when treatment impulse energy (shown in Fig. 5c to the latter’s conductivity is first cut off. This be substantially in quadrature with the sinus positive potential in the form of a charge -on oi'clal voltage, that is, characterized by impulses` 25 condenser Ci will continue to be _applied to the occurring as the sinusoidal input passes through control grid of tube V5 (thus keeping tube Va zero) may be superimposed upon the wave form conducting) until the leakage effect of resistance of Fig. 5b by appropriate grid control means 26, Riso reduces the magnitude of the charge that 21 shown associated with tubes^V2 and Va', re the output current of tube V5 is insuiiicient to spectively. The resultant wave so obtained will 30 maintain the drop in cathode resistor R necessary be of the general form shown in Fig. 5d as will to keep tube V4 below cut-oil. When the drop be clear. in resistor R gets that low, then, tube V4 will In accordance with features of my invention, I again conduct and tube _Vs will be non-conduct next apply the Wave of Fig. 5d to a multivibrator ing. This is the point at which the operation of circuit of known form including tubes V4 and V5. 35 the multi-vibrator was assumed to commence, These multivibrator tubes are preferably so ex and so a complete cycle of multi-vibrator action cited that an impulse, say 21 (Fig. 5d) of a cer has been described. tain magnitude will energize one of the tubes V4 It has just been pointed out that a longer time and V5, and a drop below a certain exciting po will elapse before decay to a predetermined trip tential, say 28 (Fig. 5e), will cause operationof 40 level 28 for larger applied pulse magnitudes than the other multivibrator tube. The circuit of for lesser applied pulse magnitudes. An at these tubes (V4 and V5) preferably includes ap tempt has been made Ato show this phenomenon propriate time-constant decay circuit means for in the curve of Fig. 5e._ It will be obserbed that deriving an output from the multivibrator of the pulse 21 and pulse 38 of Fig. .5d represent re general form shown in Fig. 5e. The multivi 45 spectively pulses _one of greater magnitude than bratorcîrcuit including tubes V4 and V5 is one the other. Pulse 21, being relatively larger, causes capable of controlling the instant of transition a relatively large output voltage 29, with the from one condition to another, and operates . result that decay from this voltage to the level spontaneously without external control to return 28 takes a relatively long time t'. Pulse 30, being to the ñrst condition. The multi-vibrator ac 50 xsomewhat smaller in magnitude than 21, conse tion is attained by virtue of a conventional for quently causes a lesser output voltage 3l, with ward coupling from tube V4 to tube V5 over a the result that decay from yvoltage 3| to level resistance capacitance arrangement R7, C1, R1, 28 takes a correspondingly smaller time interval together with a back coupling from the output of t” than t’ in the case of pulse 21. tube V5 to the input of the tube V4 by virtue of a 55 It will now be clear that the respective widths common cathode resistor R. The forward cou t’ and t" each recur regularly with a periodicity pling through R7, C1, R1 is like an ordinary inter T. In_accordance with features of my inven-stage coupling except that the values of> con tion I employ the relative position of the instant denser C1 and resistance R1'are such that con at which the decaying voltages reach level 28, denser Ci may be charged to a substantial value 60 that is, when the drop across cathode resistor R in a few microseconds. due to conduction of tube V5 decreases to such a In considering operation of the multi-vibrator magnitude that the control grid of tube V4 is circuit, assume as an initial condition an instant ` no longer effectively biased beyond cut-oil?, to de be taken from across a potentiometer 25 and is or so before occurrence of impulse 2‘1 (Fig. 5d), termine the two series of impulses I, 3, 5 . . . and when tube V4 is carrying relatively high plate 65 2, 4, 6 . . . of period T. current; thusmaking its plate far less positive To obtain the two series of pulses from the than the potential of plate supply. This reduc curve of Fig. 5e I prefer to employ a Well-known tion in the plate potential of tube V4 has the resistance-capacitance derivative circuit, which effect of applying a'negative potential on the may‘include capacitance C2, resistance R2 and control grid of tube V5 after a delay, instituted 70 an over-biased tube Vs, in the output of the by condenser C1 and resistor R1, suflicient to multi-vibrator. Output from the derivative cir allow the negative potential to be substantially cuit, as obtained across R3 may be of the form of attenuated and permit tube V5 to draw grid cur rent. Thisy is the condition of tubes V4> and Vs Fig. 5f, as will be clear. It will further be clear that peaks 32, 33, 34, etc. so obtained represent when pulse 21 appears. Application of pulse 21 75 the desired. appropriately displaced two series 2,406,0_1 9 of impulses. Since the other peaks 35, 36, 31, etc. increaslngthe retardation or decay time. Con versely; a decrease in R, or decrease in bias, will result in an opposite displacement eñect. Rela tive displacement of the two series> of pulses, one with respect to the other, may be varied by ad obtained from the derivative circuit are all regu larly spaced with period T/2 andare independent of the magnitudes of pulses 2l, 3d, they are of no use to a transmitter in accordance with my in vention, and may therefore be suppressed by justment of the ground tap on potentiometer 25, well-known means (not shown), as I have indi which, as we have seen, controls the »relative cated schematically in Fig. 5g. The unmodu iated signal of Fig. 5g is now fully prepared for Fie. 5b. amplitudes of adjacent lobes of the curve of v modulating potential, shown in Fïg. 5h. It will be clear that the voltage of Fig. 5h may represent speech or other intelligence to be transmitted.. While a form of converter circuit 'shown has been described in particular detail, it is to be understood, of course, that such description is merely by way of example to illustrate how the desired form of impulse treatment may be ef fected. A possible and preferred form of receiver for detecting impulses, transmitted as outlined above, is shown schematically in Fig. 3. Impulse In a preferred form this speech may be applied energy received by antenna t5 may ñrst be its transmission by any desired means. The case of unmodulated pulses has first been taken for discussion because it is obviously simpler than when the impulses are modulated. Figs. 5h through 5l illustrate this latter more complicated case for an assumed sinusoidal 10 to modulate the pulse signals at the push-pull 20 treated in a customary manner in radio and in termediate frequency ampliñer stages,f shown rectider stage V2, V3 of the converter of Fig. 4. generally as a block B6. Thereupon it may be According to this embodiment, balanced, that is equal, speech signals are applied by means of detectedl and some disturbing effects removed an input transformer 33 to appropriate control by suitable limiter and detector means dl. At grids of rectiñers Ve and Va. 25 this point, in accordance with a feature of my The edect of such application of speech sig invention I provide means, responsive to the out put of detector t?, for blocking out reception of nais may be to vary the curve of Fig. 5d as shown any signals substantially for the period between in Fig. 5i, that is, to make pulses corresponding successive impulses. Such blocking means pref to 2l and 3d of either greater-or lesser relative magnitude in accordance with the modulation. 30 erably includes a sharp selective circuit 68, tuned Thus. pulse 39, being at an unmodulated point, twice to the pulse frequency (i. e., of period T) is unmodulated, and hence of the same magni and a square-wave generator t9 controlled there by. By suitable adjustment means (not shown) tude as pulse 217. However, pulse 40, being dis placed by the rising portion of the modulating the square-wave output of generatorßil may be potential, is of lesser magnitude than pulse 3i). 35 made to produce a blocking signal occurring only Recalling the discussion in connection with Fig. when no impulses are due for reception. 'I‘his blocking signal when applied to the earlier stages 5e, it will be seen that pulse 39 will cause a multi vibrator output signal of duration t'; Whereas 46 or ßl of the receiver may effectively cut out pulse dd, being of lesser magnitude than pulse substantially any and all parasitic disturbances 3d, will cause a multivibrator output signal of 40 which would otherwise interfere with reception of the impulses. still shorter duration (tm) than t", as will be clear. A relatively simple blocking signal of uniformly The next succeeding impulse ¿il of series 39, recurring duration may be derived by well-known el, d2, which may correspond to series i, 3, 5 . . . of Fig. l, _has been so modulated that its peak is of considerably greater magnitude than those of 2t or 39. Consequently, voltage decay from means utilizing, say, the second harmonic of the impulse period T. Such a signal could, for ex ample, be obtained by half-wave rectification of the second harmonic and adjusting generator 49 the multivibrator output voltage corresponding to give as large a constant-width blocking sig to pulse di may be of proportionally greater nal as possible without interfering with impulse duration tm' than t', for pulses 2l and 39. Con 50 reception. Suitable oparatus for obtaining a sidering this phenomenon as to its effect on the signal of the indicated nature having a desired l displaced peaks of Fig. 5k, it will be apparent that, as the modulating signal is, say, increas `width has been disclosed in the copending ap plication of H. G. Busignies, Ser. No. 380,186, ` ing, the pulses (3W di) of one series (39, di, d2) filed February 24, 1941, and entitled "Radio are time-modulated or displaced in one direc 55 electric impulse systems." Such a blocking sig tion (i. e.. retarded in time) from their 'unmodu nal is shown schematically by the dash-dot-dot lated relative positions, and the pulses (say tii’) line 50 in Fig. l, and is designated as a simple of the other series (fili, 43, M) are time-modu blocking signal. It will be observed, however, lated or displaced in the opposite sense (i. e., that such a blocking signal is not effective com advanced in time) from their unmodulated rela 60 pletely to block out reception between pulses in tive positions. When prepared for transmission treatment, the modulated pulses, corresponding to a modulating signal of Fig. 5h, may then re semble the curve of Fig. 5e, as will be clear. that one series of pulses, as we have seen, is other than exactly in phase opposition to the other series of impulses, and, furthermore, modula tion of one series is opposite in sense to that of In the circuit of Fig. 4 it will be appreciated 65 the other. What is needed, therefore. for much that many adjustments may be made to vary greater blocking efñciency, is some sort of com the relative displacement of the two series of plex signal which will block reception for both pulses, width of pulses, modulation depth, etc. the short interval of time between, say, pulses i ~ etc. For instance, by changing the value of re and 2, 3 and ß, 5 and 6, etc. and for the longer sistor R,- or varying the bias on tube V4, the width 70 interval between pulses 2 and 3, d and 5, etc., as of pulse in Figs. 5e or 5i may be controlled,»in will be clear. that ampliñcation, thus controlled, willoccasion A possible arrangement for this odd propor-` diiferent pulse widths t’ or t", etc. Thus, an tioning of blocking signals could include appa increase of resistance R, or increase in the nega ratus similar to that outlined above in connec tive bias of tube V4, will displace all pulses by 75 tion with the simple blocking signal (for the 2,406,019 . . , y 11 _ - shorter intervals) and an additional superim posed blocking signal suitably phased with re understood that many additions, omissions, sub stitutions, and modifications may be made within spect to the first-mentioned blocking signal in the scope of the invention. What I claim is: accordance with teachings in the above men - tioned H. G. Busignies application and recurring with a period T for substantially> blocking out the longer intervals. It will be clear that known wave-shaping methods may bel employed for uti lizing preferably the second harmonic to obtain 1. An impulse communication system, impulse generator means for supplying a plurality of relatively short impulses, means for time-modu lating said impulses substantially without phase modulation of odd harmonics of the impulse fre the additional superimposed, as well as the first quency, and a receiver including means sharply indicated blocking'signal. The complex blocking selecting frequencies in the neighborhood of an . signal 5| so obtained has been designated as such in Fig. 1 and is shown as a dotted line. «by it is possible to demodulate by ordinary ampli-, After the above-indicated treatment o_f re ceived signals, it will be seen that output tothe rest of the receiver vial line 52 will include only frequency. date the modulating frequencies either side of the harmonic selected. Thereafter, simple detector 54, audio ampliñer 55, and speaker 3. _A receiver for an impulse-modulation com munication system, said receiver'including ñlter with theoretically perfect ildelty. as will be ap impulses, amplitude-detecting means, 'and block preciated. ing means responsive to an even harmonic of odd harmonic of the' impulse frequency, where tude detection of an odd harmonic of the impulse 2. A radio communication system including a transmitter and receiver, said transmitter in the pulses themselves with a very slight amount cluding impulse generator means for generating of adjacent parasitic noise, which may com two series of impulses of the same impulse fre pletely have been eliminated in the limiter cir cuit 41. It will be recalled from the above dis 20 quency, and means for time-modulating one of said series of impulses with respect to th'e other cussion of theory involved that any odd harmonic in line 52 will include amplitudeé-modulated of said series, said receiver including filter means passing substantially only a band of frequencies pulses; see expression (12). Thusfall that now is needed is an appropriate band-pass filter 53 for immediately about an odd harmonic of the im selecting the desired odd harmonic, the pass 25 pulse frequency of one of said series, and ampli band being, of course, wide enough to accommo tude-detection means. ' means passing a band of frequencies including an , means 56 may restore the original input signals 30 oddv harmonic of the frequency of the received said frequency of the received impulses, said -It might here be observed that theoretically any amplitude-modulation receiving set ought blocking means including a generator responsive to be able to listen-in on pulse-modulated sig-. 35 to said even harmonic for supplying energy to cut out reception of signals during substantially the nals` in accordance with features of this inven period between received impulses. tion. Practically, however, it seems unlikely that such will be possible. for it would be neces 4. In a receiver for receiving and detecting time-modulated impulsev energy, blocking means sary to have a set including an extremely stable and critical local oscillator for receiving only 40 for suppressing reception during the-interval be ordinary set would not be provided with my novel tween impulses of the received impulse energy, said blocking means including a sharp selective blocking means, so that the average signal-to circuit attuned to an even harmonic of the fre . one particular odd harmonic. Furthermore, an quency of said received impulses, wave-shaping noise ratio might, and would in all probability, be so great'as not torpermit reception at all. 45 means for generating a generally square-shaped signal to occur substantially during the'interval Thus, it will be seen that transmission and ap between said impulses, and means responsive propriate reception in accordance with my in to the output of said last-mentioned means for vention may even take place without suspicion suppressing effective operation of said receiver in by anyone that any such communication is tak 50 accordance with said generally square-shaped ing place. N signal. In accordance with ,still another feature of the 5. An impulse time-modulation transmitter invention, still better signal-to-noise ratios may according to claim 11, in which the impulses gen be obtained. I propose to utilize the interval be erated by said impulse generator are relatively tween blocking impulses (in the receiver), which interval is relatively short compared to the time 55 short with respect to the interval between succes sive impulses, and in which, for the same degree between pulses, momentarily to increase the gain, ’ of modulation, time retardation of impulses of say, of amplifier 55. Although the means for one of said series is of substantially the same so improving reception has not been shown, it magnitude as the time advance of impulses of the will be appreciated that it may include suitable wave-shaping means for applying a momentary 60 other of said series. ' abnormally high voltage to the anode or, say, an 6. In an impulse time-modulating transmitter, accelerating electrode in an amplifier tube of cir cuit 55. While sustained application of such a an impulse generator for supplying a series of regularly spaced impulses, circuit means syn chronously related to said impulses for supplying voltage might ordinarily harm this tube, a momentary application thereof conceivably will 65 a full-wave rectified alternating wave of a fre quency one-half the recurrence frequency of said not. At the same time, it will be clear that impulses and having oddv maxima greater than amplification and efliciency of the tube may be even maxima, means for superposing said im materially increased for this, the desired and pulses and said rectified alternating wave, multi useful instant of time. It will be seen that I have provided a rela 70 vibrator means responsive to such superposed energy and includingtime-constant decay means tively simple radio transmission apparatus of for delayingalternative operation of said multi greatly improved efllciency and having many vibrator means in accordance with the magnitude other desirable features. While the invention of said impulses so superposed, and derivative has been described in particular detail and pre ferred forms illustrated, it is, of course, to be 75 circuit means responsive to the output of said 2,406,019 13 multivibrator means for deriving impulse peaks corresponding to alternative action of said multi 11. In an impulse time-modulation transmit ter, an impulse generator for generating a ñrst vibrator means. series of regularly spaced impulses and another 7. A transmitter according to claim 6, in which additional circuit means responsive to modulat ing energy is provided for modulating the magni series of regularly spaced impulses of the same frequency of recurrence _as said first mentioned series, modulating means for time-modulating impulses of said first mentioned series of im# _ tude of successive maxima of said Wave. pulses in one sense and for time-modulating im 8. A transmitter according to claim 6, in which pulses of said other series in an opposed sense, additional circuit; means responsive to modulat ing energy is provided for modulating the 10 and means displacing impulses of said ñrst series in time with respect to those of said other series, greater maximaof said alternating wave in one said displacing means being eiîective to displace sense concurrently with modulation of the lesser said iirst and said other series respectively to an maxima of said alternating wave in a sense oppo extent that an impulse of said first series always site the said first-mentioned sense. 9. A receiver for an impulse-modulation com 15 precedes an impulse of said other series with an interval of time greater than that by which an munication system, said receiver including filter ' impulse of said ñrst series follows animpulse of means passing a band of frequencies including said other series, whereby succeeding impulses a harmonic of the frequency of the received im alternately determine an interval greater than pulses, amplitude-detection means, and blocking means responsive to a harmonic of said frequency 20 one-half the periodic recurrence of impulses of one of said series and a succeeding interval lessof the received impulses, said blocking means in ' than said one-half. cluding a generator responsive to said last-men 12. An impulse time-modulator transmitter tioned harmonic for supplying» energy to cut out according to claim 11, in which said displacing reception of signals during substantially the pe means is eiïective to displace said ñrst and said riod between received impulses. 25 other series respectively to an extent that at l0. A receiver for detecting time modulated maximum modulation of impulses defining said impulse energy, odd impulses of said energy re succeeding interval, said succeeding interval will curring with a given periodicity, said receiver not exceed one-half the periodicity of recurrence comprising filter1 means passing substantially ‘ only a band of frequencies immediately about 30 of impulses of one of said series. an odd harmonic of said periodicity and ampli EMILE LABIN. tude detection means.