|Publication number||US2492004 A|
|Publication date||Dec 20, 1949|
|Filing date||Jul 13, 1946|
|Priority date||Aug 1, 1945|
|Publication number||US 2492004 A, US 2492004A, US-A-2492004, US2492004 A, US2492004A|
|Original Assignee||Fr Sadir Carpentier Soc|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (1), Referenced by (5), Classifications (11)|
|External Links: USPTO, USPTO Assignment, Espacenet|
Dec. 20, 1949 G. POTIER 2,492,004
PULSE MODULATING SYSTEM Filed July 13, 1946 5 Sheets-Sheet 1 Voltage across a l Rcsis l'l ta Grwnd I Vglhge nross diode I6 am i Gaston Pofier Dec. 20,- 1949 c. POWER PULSE MODULATING SYSTEM s Sheets-Sheet 2 Filed July 13, 1946 Gus'ron Po'rier Dec. 20, 1949 G. POTIER PULSE MODULATING SYSTEM 3 Sheets-Sheet 3 Filed July 13, 1946 I Dales Lina JL I n I l l Milli Pilo+ Pulse Generafor Gaston Potier Mod u lei-w Patented Dec. 20, 1949 PULSE MODULATING SYSTEM Gaston Potier, Versailles, France, assignor to Societe francaise Sadir-Carpenti'er',
France, a corporation of France Application July 13, 194.6,. Serial No.. 683,299 In France July 31,1945
The present invention relates to electric signalling by means of width modulated pulses and more particularly to certain improved means for generating modulated pulses the length of which is av linear function of the instantaneous amplitude of a modulating. signal.
Fig. 1 is a diagrammatic. showing of width modulated saw tooth. pulses;
Fig. 2 is another diagrammatic showing of Width modulated saw tooth pulses;
Fig. 3 shows a circuit diagram illustrating an embodiment of the invention;
Fig. 4 shows the pulse diagram corresponding to said circuit embodiment;
Fig. 5 shows a circuit diagram illustrating. another embodiment of this invention;
Figs. 6, '7 and 8 show the pulse diagrams cor-- responding to embodiment shown in Fig. 5.;
Fig. 9 shows a conventional diagram. of a modulation system for pulse multiplex transmission, using time division;
, Fig. 10 shows the pulse distribution in a system such as shown inlFigr 9;
Fig. 11 shows a. general diagram of a modulation system for multiplex transmission, according to the general process of the present invention; and
Figs. 12 and. 13 show two examples of application of the invention to multiplex transmission systems, employing the circuits of Figs. 3 and 5 respectively.
Heretofore, a generally utilised process for generating. width modulated pulses consisted as diagrammatically shown inFig. l of the annexed drawings, of superimpressing on the modulating signal (curve I.) of. saw-tooth shaped signals 2, 3, of prerequisite recurrence frequency, constant amplitude and slope. A. suitable device, such. as a clipper stage associated with. a gating circuit, converts that portion of the saw-tooth which is above. a predetermined level indicated by line 4 into a rectangular pulse. the width of which equals that. of the saw-tooth atv said. level. This width is a function. of the instantaneous amplitude of curve. I as the saw-teethare'of invariant shape, hence the required result is. obtained.
In the application. of such a process, it is very difiicult to obtain. a substantially linear relation between. the length (or. duration) of the pulses and the instantaneous amplitude of signal II, which results in a noticeably high degree of distortion; furthermore, the following disadvantage occurs: in the. case of an overmod'ulation, i. e. when the amplitude of the. modulating. signal. exceeds the level of'line 4,.the device delivers a pulse 2 of awidth equal to AB as shown in Fig. 1 whether or not a saw tooth such as 5 does exist. This width AB may be of much greater value than the maximum width or the concerned saw'-tooth. This may give an idea of the ensuing trouble.
According to certain features of the invention, such trouble is eradicated by providing first an amplitude modulation of the pulses and then converting said amplitude-modulated pulses into corresponding length-modulated pulses. Other- Wise the pulses of the recurrence frequency equal to that of' the final pulses to be generated are first derived from the modulating signal I (Fig; 2), such pulses being of a fixed width equal to the maximum width admitted for such final pulses, and having an amplitude equal or linearly related to the instantaneous value of said signal I. To the pulses 6, illustrated as hatched areas, are then added saw-tooth pulses l as in the" above described generating process, the bottom width of which equals that of'pulses 6, the following operations being such as summarized with respect to Fig. 1.
The way in which pulses such as 6 are obtained from signal I is well-known in itself. One may see that, in such a case as provided by the invention, any trouble from an overmodulation is avoided as, in all cases, the width of the finally generated pulse cannot exceed that of pulses 6, at whatever level the clipping may occur.
Other features of the invention relate to certain particular means for generating pulses 6 and I as well as for providing circuits therefor which comprise only a small number of vacuum tubes, thus. enablingv a great manufacturing. econ.- omy of. the apparatus embodying such features.
Other. features of the invention relate also to certain particular manners of applying such process and means as provided above, to multiplex transmissions in order to obtain an important reduction of tubes and associated equipment.
Referring to Fig. 3. in the circuit shown, the pentod'e tube 81 is used for generating amplitudemodulated pulses; for this purpose, its. control grid receives, from generator 9,, through a resistance-capacitance connection, rectangular positive pulses of. constantampl'itude and width and of the required recurrence frequency. A modulating signal; is supplied from generator Ill to the suppressor grid of tube 8;. a. suppressor bias being provided by D. C. source. H. such that the operating point of the tube Si is placed at the mid-point of the linear portion of said suppressor characteristic. Thus,intheplate-circuit of. tube 8 negative amplitude-modulated pulses are generated and said pulses are applied through a network comprising resistance l2 and capacitance l3 to the terminals of a grounded resistor M. The generator 9 is connected to the control grid of said tube 8 through a resistance Zil-capacitance 2| network. The resistance 20 is of high value and tube 8 is blocked between the pulses delivered by generator 9, according to the well-known rectifying process occurring in the tube gride cathode space, which ensures a definite and accurate conversion of signal 1 into pulses t (Fig. 2).
Across the terminals of resistor i l, a condenser l and a diode (or other rectifying element) it, the cathode of which is grounded, are seriesconnected. A co-nstant current generator ll, of any suitable kind has its "positive pole connected to the plate of said diode. The purpose of such a combination l ll5itll is to generate sawtooth waves and superimpose them upon the afore-mentioned amplitude-modulated pulses. Its operation will be explained in relation to the curves shown in Fig. 4 in which voltages are plotted against time for three different heights of the pulse delivered by tube 8.
With no negative pulses supplied by the tube 8, the current of generator ll? passes through diode it, the internal resistance of which may be considered as negligible. Curves a indicate the voltages with respect to the ground across resistor l4, which are developed by the transmitted pulses.
As soon as a negative pulse appears, the voltage at points C and D, Fig. 3, becomes negative, the time constant of the network iii-l6 being negligible, as a result, diode if; is blocked. Hence the 'constant current output of generator ii only flows through the i--l 5 network, which creates a conduration of the charge of condenser is is at most equal to the duration of any pulse 6 of Fig. 2. Curves 1) indicate voltage across point D and ground, in the shape of saw-teeth which begin at points E1E2E3, determined by the amplitude of the corresponding pulse a. The current generated by I! being of :constant value, points F1F2F3 where a saw-tooth meets the zero voltage axis is a function of the amplitude of the corresponding pulse.
In other words, the spacing G1F1 (or G2 F2 or G3 F3) is linearly related to the amplitude of the corresponding :1. pulse, which is obviously the wanted result. When pulse a ceases, the voltage at point C is suddenly changed to the ground voltage, that at point D is suddenly raised, but this condition does not last, condenser quickly discharging through diode l6. Thence, a positive peak P1 (or P2 or P3) appears on curves 1), but this departure does not introduce any trouble since it is suppressed by the unit is following tube I8, (for which a triode is shown but a pentode is of more advantageous use). Upon the control grid of tube It is applied the voltage between point D and ground. The tube It is permanently operated except for the periods in which the saw-tooth negative pulses are received. The circuit constants are such that, even for the smallest amplitudes as G3 E3 of the sawtooth pulses, the plate current of [8 is partially suppressed, thus giving curves 0 of Fig. 4 which show the voltages appearing across the plate load resistance of tube I8.
These curves are then converted into rectangular pulses d the width of which is a time function of the signal amplitude from It, such a conversion taking place in the unit [9 in any well known manner. It shall be noted that the saw teeth b are obtained from the beginning of the charge of condenser 15, so as to use the desired linearity of charge characteristic, and that the discharge characteristic, which always presents a curvature to some degree, is not used as in other usual modulation systems.
From an illustrative point of view, a circuit designed as above has given modulated pulses of a maximum width of 1.5 as of a mean width of l p. and of a periodicity of recurrence equal to 3.2 s, the linear distortion degree being about 2%, without special care involved.
It is also to be noted that in the described system, the fronts of the pulses remain fixed (i. e. coincident with the fronts of the signals supplied by generator 9) the rear faces of said pulses moving in accordance with the modulating signal, which may be taken advantage of, in some cases.
An alternative, which may be useful for some applications, consists of the omission of diode [6, the function of which may beaccomplished by the grid-cathode space of tube 18, when the internal resistance of said space is low with respect to the value of resistor 14.
The circuit shown in Fig. 5 makes use of the charge, under constant current, of a condenser 28, until it reaches a value fixed by the instantaneous voltage of the modulating signal, after which the charge current is by-passed through a secondary circuit comprising a resistance 34 across the terminals of which the width-modulated pulse may be picked up.
In Fig. 5, a generator 9 generates rectangular pulses at an uniform recurrence frequency (that which is wanted for the final output pulses) and of suitable fixed amplitude and width. These pulses, of positive sense, are applied to the control grids of two tubes 22 and 23 (pentode tubes, for example, the screen and suppressor grids of tube 22 having been omitted for the sake of clearness in the drawing) by means of resistance capacitance networks 24-45 and 26-27. These tubes are unblocked during the pulses and blocked in the intervals between the pulses due to the charges picked-up by condensers 24 and 26. The plate circuit of tube 22 comprises a condenser 28 across which is connected a pentode'or triode tube 29. The plate circuit of tube 29 comprises a resistor 30 the lower end of which is connected to the grid of tube 29. The screen-grid of tube 23 is supplied, in a well known manner, by a network constituted by resistors 3l-32 and condenser 33. The plate circuit of tube 29 also comprises a resistor 34 of a relatively low value, for a purpose which will be further indicated.
The modulation voltage from generator 35 is fed to the suppressor grid of tube 23 by means of a network resistance 31 and capacitance 35. A suitable bias is supplied to this grid from battery 38. This tube 23, operates as described for tube 8, in Fig. 3 and thus delivers at the terminals of its load resistance 30, amplitude-modulated pulses of the same width as those delivered by generator 9. An output pulse of constant amplitude from generator 9 is shown at a, Fig. 6, and at b is shown the pulse which synchronously appears across resistor 3B, but is of varying amplitude.
As soon as tube 22 is unblocked, condenser 2'5 begins to charge and this charge lasts until tube 29 is unblocked, which as may be easily seen,
occurs when the voltage capacitor 28 equals the voltage across resistor 36 plus the volt! age of out-oft, which is practically negligible.
- Fromthis moment, the plate current. of tube 22 flows through tube 29; resistance 34 being of low magnitude, introduces only a negligible voltage drop; hence condenser 28 retains. its. charge as long as the positive pulse applied on tube 22 lasts. As soon as said pulse ends, tubes 22 and 23 are blocked. Condenser 28 discharges through resistor 34 and tube 2% and the voltage across this condenser 28- varies as shown in curve 0, Figs. 6 7 and e. The voltage across resistor 34 is such as-sho-wn in curve at, Figs. d, T and 8. It is clear, from the, above, that. equal widths of the flat portions of curves c and d are linear functions of the amplitudes of pulses 1)., Figs. 6 7 and 8, since the charge oicondenser 28 is oftected with a constant: current condition through pentode tube 22.
The resulting widthamodulated pulses are collected across resistance 34. It will be noted that, in the circuit shown in Fig. 5 the higher the modulation voltage, the smaller the width of the generated pulse. This indicates that the linear relation between the modulation voltage and the pulse width is, a decreasing one. Moreover, it is now the rear edge of the pulse and not the front edge, that. is fixed, and this rear edge coincides with the rear edge of the pulse delivered by generator 9.
It shall be noted that; in addition togenerator 9, Fig. 3, the circuit utilises three tubes only, which leads to an economical; design.
The invention also provides for the application of such process and meansas here above described to multiplex transmissions with a view of greatly reducing, the required number of tubes and associated elements.
The processgenerally uitjlfied' for obtaining width-modulated pulses is shown in Fig. 9. A pilot pulse generator 41 feeds a delay line 42 comprising a number of output terminals n equal to the number; oi channels in the concerned system. At these terminals, appear pulses, regularly shi-fted in time which are-brought to nmodulators 43 controlled by the modulating voltages from the various channels. Width-modulated pulses are thus collected at the output of said modulators and fed to a mixer circuit 44 also supplied through a connection 45 with a pilot pulse (used, at the far end, for the purpose of selecting the various channels). The mixer 44 comprises, for instance, an impedance common to the plate output circuit-s of the last stage tubes of circuits 4| and 43.
Fig. 10 shows at a the pulses supplied by generator 4| and pulses b and c represent those fed by modulator 43, to the mixer 44 and pulses 11 represent the complete train at the output of mixer 44.
The modulators 43 may consist of a minimum of three tubes although generally four tubes are employed not considering the tubes in the genorator 5| associated with each of the units 43.
The number of tubes N may be appreciably decreased if, according to some features of the present invention, the operation process is made as follows:
(1) Amplitude-modulated pulses are generated for every channel;
(2) Said pulses are mixed;
(3) Those mixed amplitude-modulated pulses are converted by a single element into Widthmodulated pulses;
6 (4): Width-modulated pulses. are mixed with the pilot. pulse train.
Such an arrangement is shown in Fig. 11, in which 41f. a. pulse generator,v 42. is a delay line, 46 are modulators delivering amplitude-modul' d p lse 41 is a mixer for pulses fed from modulators 46., 4B is a pulse converter from amplitude-modulated into width-modulated pulses and 49 is a second mixer adding pilot pulses fed by conductor to said width-modulated, pulses. As. the. mixers do not require any tube and a single tube per channel is only necessary for generating, amplitude-modulated, pulses, the total number of tubes is then:
in which K is the number of tubes required for the converter circuit. As K=3 or 4 and n from 6*- to 2-0 or higher, it is seen that N is much less than N. Hence. a substantial economy of tubes and associated elements, and a decrease in capacity of supply sources is effected by this invention.
Fig. 12 is a schematic diagram of; a multiplex circuit in which the, modulators 46. mixer ll and pulse converter 48- are designed according to the embodiment of Fig; 3, the parts. of said, Fig. 3, unnecessary for an understanding of said circuit, being omitted. The pulse generator 4|. and the delay line '42, are connected, as in. Fig. 11,... The generation of amplitude-modulated pulses is. offected by means of a pentode tube 50. (tube 8 of Fig. 3) the suppressor grid, of which is. supplied byhe o p t i na of a generator 5!, of the channel concerned. Tube 50 taken, along with generator El and associated elements= constitutes a modulator indicated as 51. There are n such modulators provided in the, system. The plate circuits, of the n tubes 50 are connected. by a resistance-capacitance network to a resistance 52;. These plate circuits may. comprise a common resistance, Which connected. to the resistor 52 by a single coupling, condenser... Across; the resistance 52 are connected a. condenser 53. and a diode or other rectifier 5,4, supplied with the output of a constant current generator 55; Then. a clipping circuit, 56. Suppresscsjthe, top. oi all: signals to a predetermined level and a second mixer adds the pilot signal to the pulses thus shaped. All elements 52-46 and 59 constitute a single apparatus 58 common for all channels, the operation of which may be easily understood from the preceding description.
Fig. 13 shows an embodiment making use of the circuit of Fig. 5. Here also superfluous details are not shown; only the generator 4|, delay line '42, modulator tube 50 and generator 5|, as described above.
Moreover, the terminals of the line 32 are connected through coupling condensers 59 and. separating resistances 67, of high value with respect to the characteristic impedance of the line 42, to a common conductor 60 connected to the grid of tube 6| (tube 22 in Fig. 5). A leak resistance 66 determines the potential of the grid of the tube 6| as well as the amplitude ratios of the signal applied to this grid and the potential of the various signals at the terminals of the delay line. The plate circuits of tubes 50 have a common load resistance 52 (3B in Fig. 5) controlling the grid potential of tube 63 (29 in Fig. 5) with a plate resistor 64 across which width-modulated pulses are delivered, which are amplitude-limited at 56 and mixed with pilot pulses in 49. Here again, the elements 50-5| constitutes a channel modulator 51 (one for each channel) and the elements 7 6! to 64 a single unit 65 common to all channels.
Having now described and ascertained my invention, I declare that what I claim is:
1. A process for generating width-modulated electrical pulses, by which a substantially linear relation is ensured between the width of each pulse and the corresponding instantaneous amplitude of the modulating signal, comprising the steps of generating amplitude modulated pulses, by generating from a modulating signal pulses of a recurrence frequency equal to that of the final width-modulated pulses wanted, of fixed Width equal to the maximum width required for said final pulses and of varying amplitude equal or proportional to the instantaneous amplitude of the modulating signal, the further step of converting said amplitude modulated pulses into width modulated pulses by adding to said amplitude modulated pulses, saw toothed pulses, the height of which is at least equal to the maximum amplitude variation of said modulating signal and the Width of which is equal at the base to that of the fixed width of said amplitude-modulated pulses, the further step of clipping said combined pulses at a constant inferior level, suitably determined about the maximum amplitude of said constant-width pulses.
2. A combination for the generation of widthmodulated electrical pulses comprising, means for generating substantially rectangular pulses of substantially constant amplitude, width and recurrence frequency, means for generating a modulating signal, a pentode tube, means for applying said pulses and modulating signal to electrodes of said pentode tube, means for biasing said pentode to cause it to deliver pulses of constant width and recurrence frequency having amplitudes that are in linear relation to the corresponding linear amplitudes of said modulating signal, means for generating from a constant current source, and under the control of said amplitude-modulated pulses, saw-toothed pulses of the same width as said amplitude-modulated pulses of the same recurrence frequency and of a height at least equal to the maximum'variation of amplitude of said amplitude-modulated pulses, means for adding said amplitude-modulated pulses and said saw-toothed pulses; and means for applying the resulting pulses to a gating circuit which suppresses the portions of said combined pulses lower than a predetermined level substantially adjusted to the maximum amplitude of said amplitudemodulated pulses, in order to obtain pulses of a varying Width in a substantially linear relation to the amplitude of said amplitude-modulated pulses.
3. The combination for the generation of width modulated electrical pulses comprising, means for generating substantially rectangular pulses of constant amplitude, width and recurrence frequency, a vacuum tube, means for generating a modulating signal, means for applying said pulses and modulating signal to electrodes of said vacuum tube, means for biasing said vacuum tube to cause it to deliver pulses of constant width and recurrence frequency, the amplitude of which are in linear relation to the corresponding linear amplitudes of said modulating signal, means for generating saw-toothed pulses from said rectangular constant amplitude pulses, said saw-toothed pulses being of the same width and recurrence frequency as said amplitude-modulated pulses and of a height at least equal to the maximum variation of amplitude of said amplitude-modulated pulses, means for adding said amplitude modulated pulses and said saw-toothed pulses and means for applyingthe resulting pulses to a gating circuit which suppresses the portions of said combined pulses lower than a predetermined level substantially adjusted to the maximum amplitude of said amplitude-modulated pulses, in order to obtain pulses of a varying width in a substantially linear relation to the amplitude of said amplitude-modulated pulses.
' GASTON POTIER.
REFERENCES CITED UNITED STATES PATENTS Name Date Labin Sept. 24, 1946 Number
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US2408077 *||Aug 25, 1944||Sep 24, 1946||Standard Telephones Cables Ltd||Multichannel system|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US2551816 *||Feb 18, 1948||May 8, 1951||Hartford Nat Bank & Trust Co||Multiplex transmitting device|
|US2641699 *||Mar 25, 1949||Jun 9, 1953||Joseph Libois Louis||Multiplex pulse time modulation system|
|US4613974 *||Mar 16, 1984||Sep 23, 1986||Vokac Peter R||Method and system for modulating a carrier signal|
|US4989219 *||Sep 22, 1986||Jan 29, 1991||Gerdes Richard C||Midlevel carrier modulation and demodulation techniques|
|WO1988002199A1 *||Sep 22, 1986||Mar 24, 1988||Vokac Peter R||Method and system for modulating a carrier signal|
|U.S. Classification||332/111, 370/212|
|International Classification||H04J3/04, H03K7/00, H03K7/08|
|Cooperative Classification||H04J3/042, H03K7/00, H03K7/08|
|European Classification||H04J3/04B, H03K7/08, H03K7/00|