US2485821A

US2485821A - Translation of duration modulated code pulses into equal length code pulses

Info

Publication number: US2485821A
Application number: US90252A
Authority: US
Inventors: Gloess Paul Francois Marie; Libois Louis Joseph
Original assignee: Individual
Current assignee: Individual
Priority date: 1948-08-05
Filing date: 1949-04-28
Publication date: 1949-10-25
Anticipated expiration: 1966-10-25
Also published as: BE490359A; CH278421A; GB659020A; DE824067C; NL82618C; FR970030A

Description

Oct. 25, 1949. P. F. M. GLOESS ETAL 2,485,321

TRANSLATION OF DURATION MODULATED CODE PULSES INTO EQUAL LENGTH CODE PULSES Filed April 28, 1949 4 Sheets-Sheet 1 22 (ODE PULSE /wxR cm".

(MP/1R. com /m cot u'r. 0E1 HY 0am DELHY n coMP/m. 1 GET) 7 7 71 I2 15 c 19 c r U h u M U K 4 H I2 I H 4:1;- n' H9 i 13 If 55mm ((71 (SECOND (or 1 room 2" 0005!? 05 PULSE MIXER (fr ZNVENTORS.

1? f? M. G206. Z. J. Zjbm Oct. 25, 1949. P. F. M. GLoEss ETAL 2,485,821

TRANSLATION OF DURATION MODULATED CODE PULSES INTO EQUAL LENGTH CODE PULSES Filed April 2 8, 1949 4 Sheets-Sheet 2 PULSE HGENT- Oct. 25, 1949. P. F. M. sLoass ETAL 2,485,821

TRANSLATION OF DURATION MODULATED CODE PULSES INTO EQUAL LENGTH CODE PULSES Filed April 28. 1949 4 Sheets-Sheet 3 a t I INVENTORS I 620655 g- 5 z. J: 0m

Oct. 25, 1949. P LoEss AL 2,485,821

.. TRANSLATION OF DURATION MODULATED CODE PULSES INTO EQUAL LENGTH CODE PULSES Filed April 28, 1949 4 Sheets-Sheet-4 [/7 VIM/7728.5.

21'? M. 62065.5 L J [Mums Patented Oct. 25, 1949 TRANSLATION F DURATION 'MODULATED :CODEPULSES IN TO EQUALLENGTH CODE BULSES Paul FrancoisMarie Gloessand Louis Joseph -Libois, Paris, France Application April28, 1949, Serial- No. 90,252 'In France. Augustwj1'948 -11Claims. (Cl.*1'77--'-380) 5 The present invention relates to improvements in electrical codificationsystemsaand more particularly in the translation of duration modulated pulsesintolcode pulses.

It is known-thatthe modulation ofelectric signals by codepulses maylbeeiiected as. follows:

The low frequency signal to .be transmitted modulates, in alknownmanner; pulses repeated at a .suitab1e.rhythm or repetition frequency, modifying, according tothe amplitude voi the signal, one of the characteristic parameters of the pulses used, this parameter being generally -.the amplitude, the .duration or theposition'in time.

The pulses thus modulated, from a number of communication channels are .thengroupedas is well known, inthe multiplex pulse transmission systems.

These modulated pulses are transmitted to a coding device which converts each one .of them into a train of n codepulses, :each of which can assume each .of two valuesaccording .to the importance of themodulation of the.-,pulse of the corresponding channel, -insuch .a manner as to define .2 different levels .for the modulation parameter of the channel ,pulses, .It..islknown that, practically, one of the twoamplitude values of the codepulses is zero,.i..enan -absenceloflthe corresponding pulsathe other value having. generallya predetermined constant level.

The present invention -.has, as an object, a coding device enabling the conversion, into .code pulses, of pulses modulated in .duration Joya dis placement of their trailingedges.

Accordingto one characteristicofthe invention, the coding device successively compares duration modulated pulses with .,predetermined intervals, each one of which is half the. previous one, said time intervals offering a a simple relationship- [with the maximum duration of the initial pulses.

According to. another. characteristic, the coding device comprises .a number of successive stages equal to the number ofeelements of the codeconsidered, each stage beingconnected to'the next one by two: difierentcircuits, and, in eachvstagc, means, controlled v by the 'signal :applied ".to said stage, to direct the signal throughaone tor the other f said circuits, according to its-duration referred to a time interval characteristic o'i the stage 1 considered and, simultaneously, to render inoperative the circuit not utilized.

According to another-characteristic, one-'o'f the circuits of each stage passes only thesignals having a duration longer than the-reference time interval of the stageconsidere'd andproduces, un-

der" the control 'ofthese through-going signals, a codapulse-together with a signal rendering the other circuit of the same stage inoperative, and sup-pliesthe following stage with a signal whose duration is decreased by the reference interval of the stage "considered and which is delayed by thisduration.

According to another characteristic, the other circuit receives the signals which have a duration smallerthan the reference time interval of the stage considered, delays these signals by this time interval and'transmits them to the next stage without decreasing their duration.

According to another characteristic of the invention, the last stage of the coding device comprises only onecircuit and means are provided in this circuitto pass only signals having a duration greater thanithe shortest reference interval and toilpro'duce then .a code pulse.

Finally, according to another characteristic, the firstc'of'the circuits of each stage preceding the last one comprises an electronic device with two stability positions, controlled by a signal derived from the signal applied to the stage when said applied signal has a duration greater than the reference interval of the stage, said electronic "device controlling, when operated, the emission :of a code pulse and further creating a blocking signal'for the second circuit, this blocking signal having a duration substantially double that o'f'the reference interval of the stage considered.

The inventionwill now be described indetail, with reference to a preferred example of embodiment, which will illustrate exactly its purpose, characteristicsand advantages. This description will be made with reference to the appended drawings wherein:

"Figure 1 shows, schematically, a coder in accordance with the invention;

Figure 2 shows in greater detail, the electrical circuits of the two first stages of a coder in .accordance with the invention;

Figures '3 and 4 are diagrams of pulses occurring at various points of the coding chain; and

Figurefi shows an example of embodiment of the circuitsof the last stage of a coder in accordance with the invention.

For "greater simplicity, Figure 1 shows, schematicallyga coder giving three codepulse positions and-giving, consequently, a definition corresponding to 2 different levels, but it will be obvious, upon reading the description, 'thatthe codwmightcomprise any larger number of elements and that, in such a case, a number of stages would be utilized equal to the number of these elements, each stage being similar to those whose functions will now be defined in connection with Figure 1.

In the description, T will designate the maximum duration selected for the modulation of the incoming pulses and t the duration of any pulse.

In Figure 1, the pulses I, modulated in duration and which are negative, for instance, are applied to a delay device 2, delaying said pulses by a time T/2, and delivering to a comparator circuit 3 a signal 4 whose shape will be defined in connection with Figures 2 and 3. The signal 4 will be positive, for instance. cuit 5 receives, on one hand, the initial pulse I and on the other hand the signal 4.

If the duration 2? of pulse l is larger than T/2, the circuit 3 delivers to a second delay circuit 8 an impulse l delayed by T/2 and having a duration (t-T/Z) further, it sends out a code pulse 8 which is transmitted to a mixing circuit 9 and, finally, it applies to device 5 a blocking signal l whose duration will be substantially equal to T so as to prevent this device from transmitting any signal to the circuit 6.

On the contrary, if the duration of pulse I is smaller than T/2, the device 3 does not operate and consequently does not supply any of the signals I, 8 and I 0. The device 5 therefore is not blocked and it transmits to the delaying circuit 6 a pulse ll corresponding to pulse 1, i. e. whose duration is 75, but it delays this pulse H by T/2 with respect to pulse I.

The device 6 subjects the signals it receives to a delay T/4 and, as previously, the signal 12 produced is applied simultaneously to a comparator device l3 and to a device [4 which also receives the pulse 1 or H. The devices l3 and 14 act, respectively, in a manner similar to that of the

devices

3 and 5 of the first stage: if the pulse 1 or H entering the second stage of the coder has a duration greater than T/4, the device [3 projects a code pulse towards the mixer 9 and it also supplies a pulse l5 whose duration is decreased by T/4 with respect to that of signal 1 or II and which is delayed by T/4 with respect to said signal, and it sends to circuit I4 a blocking signal [6 whose duration will be about T/Z, this duration being sufficient to prevent the circuit I4 from operating while the pulse 1 or II is applied to it, since these signals 1 or II can only have a duration equal to or smaller than T/2 by virtue of the adjustment of the first stage of the coder.

If, on the other hand, the pulse 1 or II has a duration smaller than T/4, the comparator circuit 13 cannot operate and the device l4 delivers a pulse I! which is identical but delayed by T/4.

One or the other of pulses l5 or I! reaches the third stage of the coder which is the last Further, a cirstage in the example considered. It is applied to a delay circuit l8 causing a delay T/B (T/2 in the case of a coder with 12 pulse positions) and delivering a signal l9 applied to a last comparator circuit 28 which delivers a code pulse 2! if the signal I5 or I! has a duration greater than T/8 but delivers no signal in the opposite case.

For this last stage, it is obviously unnecessary to provide a second circuit for the direct transmission of the pulses since signals of a duration smaller than T/2 do not give rise to the production of any code pulse.

The code pulses eventually delivered by the circuits 3, l3 and 28 are suitably distributed by the mixer 9 which may also, if desired, give them the characteristics suitable for transmission, and they are received at 22.

Figure 2 shows an example of embodiment of a coder according to the invention. For greater simplicity, only two stages of this coder have been shown, since all the stages are similar. However, an example of embodiment of the last stage is shown in Figure 5.

It will be noticed that certain electron tubes used have been represented under the shape of triodes or pentodes, but it is obvious that in practice tubes can be used comprising a different number of electrodes, and that such tubes may offer operating characteristics which make it possible to obtain in better conditions the desired results. The circuit modifications to be efiected in such a case will be obvious to technicians.

It will be assumed in the description that the duration of the incoming pulse considered is between T/2 and (T/2+T/4) With this assumption, the first stage of the coder gives a code pulse but the second stage does not give any and this makes it possible to explain completely the operation of the system in connection with the two stages shown.

It will further be assumed that the incoming pulses 23 are modulated in duration by displacement of their trailing edges and that they are of a negative polarity, as shown.

The description of the operation of the circuits of Figure 2 will be made in connection with the diagrams of Figures 3 and 4. Figure 3 shows schematically the shapes of the signals produced by the circuits corresponding respectively to the

assemblies

2, 3, 8 and 13 in Figure 1. Figure 4 shows the shapes of the signals procluced by the circuits corresponding to the devices 5 and M of said Figure 1.

The incoming pulse 23 is applied to the control electrode of an electron tube 24 whose anode circuit comprises particularly a delay line 25, open at its end 26 and closed at its input 21 on its characteristic impedance so as to reflect the signals applied to it. The back and forth transmission time on line 25 has been selected equal to T/2. The pulse 28, appearing on the anode of the tube 24 under the action of the pulse 23 is shown on line a of Figure 3.

This pulse 28 is transmitted simultaneously to the input 21 of the delay line 25 and over connections 29 and 30. These connections also transmit the pulse 3|, shown on line b of Figure 3, caused by the reflection of pulse 28 in line 25 and delayed therefore by T/2 with respect to this pulse 28. The combination of

pulses

28 and 3| produces the signal 32 shown at c, Figure 3; this signal offers a crest 33 of duration (tT/2) caused by the partial superposition of the

pulses

28 and 3|. The signal 32 is applied to the control electrode of a second electron tube 34, normally biassed beyond cut-off voltage by means, for instance, of a voltage divider 3536 designed in such a manner that the cut-off voltage, shown at 31 on line c of Figure 3 allows the tube to pass only signals having an amplitude higher than that of the pulse 28. Thus, only the crest 33 of the signal 32 goes through the tube 34 which delivers on its anode a negative signal 38 shown at d on Figure 3 and whose duration is that of the crest 33, i. e. (t-T/Z) The said signal is further transmitted, through lead 30 to the control electrode of an electron tube 39, such as a pentode, to the suppressor .gridof which is applied the incoming pulse 23 which thus suppresses during atime t the anode currentof this tube, as has been shown on lines a to c.0f :Figure l. In the anode circuit-oi tube 39 is inserted an inverter-transformer 40 at the outputof whichthe positive signal 4! .is'received, whose duration is T/ilasshown by line 0 of Figure 4: and which isdelayed by t with respect vto the incoming pulse 23. Thusthe electron tube 42 controlled by the tube -39 through the vtransformer 48 cannot send out any signal pending the duration 12 of the incoming pulse, as will appear below.

The anode load of tube3 l1consists, in. addition to resistor 43 at whose terminals appears the negative signal 38, of a transformer i l which re verses and differentiates thepulsetfl, producing a signal 45 as shown by line-e of Figure 3. This signal 'comprisesa short positive pulse 4*,(30119- spondiugto-thefirst edge of the inverted signal from the pulse 38, anda short'ncgative pulse. il, caused by the second edge of this inverted signal.

The signal 45 energizes-an electronic device havingzone position of permanent equilibrium and capable of assuming, during. apredetermined time intervahanother position of equilibrium under the control of a suitable signal. Such devices are well known and the one which is represented by way of example in Figure 2 consists of a double triodc tube 48, connected in a known manner.

The left hand element of this tube is normally non-conductingbut it is made conducting by the positive potential applied to its control electrode by the pulse 46. The potential at the terminals of the load resistor 49 ofthis element thus dcreases suddenly when this pulse 46 appears, which gives a negative pulse 5.". shown on line e of Figure 4. The time constant of the circuit formed by the resistor 49 and by thevcoupling capacitor 5! is determined in such a manner that the tube 48 comes back to its position of permanent equilibrium by unblocking of its right-hand element, only after a timeinterval T, i. e. at a timei3T/2 afterthe beginning of the pulse 23;. An examination of Figure 4 shows that this duration is suflicient to include the duration of the pulse 4! produced by the tubeEEl even if the initial pulse 23 had the maximum duration T.

Of coursaknown devices (not shown) comprising, for instance, a detector circuit will be uti- U lized so that the negative pulse-47 of the signal 45 .does not cause the tube -48 to come back untimely to its position of permanent equilibrium.

Through the medium of a suitable capacity :52 and resistance'53, the negativepulse is transmitted to the control electrode of the tube 42 which thus receives the signal shown on line 2 of Figure l. The circuit elements are proportioned in such a manner that the portion 55 of signal 54 caused by the appearance of signal 4| has an amplitude smaller than that ofthe cut-off voltage of the tube 42, a voltage which is indicated by the horizontal line 56; thus the tube 42 remains blocked in spite of the existence of the positive pulse 4 l.

The incoming pulse'having a duration greater than 172, the device 48 also controls the emission of a code pulse produced as follows:

When the pulse Gil modifies provisionally the equilibrium condition of thertube '48, the right hand element of this tube is blocked and a positive pulse 57 appears atthe terminals of the load resistor 58; this pulseiS represented on line f of Figure 3. It is differentiated by a capacitor '59, whose values have been chosen sufiiciently small,

andgivesa signal fill (gzomfigurefi) ,sconsistin of .a short positivejpulse :6 I and .a-. short snegative pulse 62. .The-signal-tfl :is applied at TBSi-tol-a mixer circuit 64 whoseiterminalscare connected to a detector -65 which -;suppresses the negative pulse 62 and thusproduces asignal JEE -shown on line h'rof Figure .3. As this figureashows, the signal v(56 is produced by the leadingedgeof the initial pulse 23, delayed by T/2; it thus occupies a fixed position with respect to this leadingedge, i. e., a fixed position in the iterative period :of the pulse 23.

.The pulsefit,isrtransmitted to avshapingcde- .vicetl, for thepurposeof giving the codesignals suitable characteristics, and which delivers, r-for instance, at its output terminals 68 thesignal-iGQ shown on line i of Figure 3.

Further, the pulse 38 whose duration, (t-T/Z), by hypothesis, issmaller than TM, is applied to the control-electrode of an electron tube T0 is theinput tube to thesecond .coderstageand which corresponds in that stage to the tuber-34.

The tube 1!] vconverts this pulse into acpositive pulse 1 l, shownon line 1" of-Figure 3zand whichris appliedito thevinput of a-delay line l2 similar-to line 25. Like the latter,the line-l2 comprises an open end l3 but is closed-at its input IA-.onits characteristic impedance so aslto reflect. at ,tone of its endsthe signals appl-iedt it. This line 12 has abackand .iorthtransmission time equal :to T/4 or T/22. Similarly, the delay line of any stage p of the coderhasaiback and forth-transmission time equal to T/Z The pulsell is further transmittedthrough'aa lead 15 vto the controlielectr-odeaof anxclectron tube 16 andthrough a lead "it is alsoapplied to the control electrode of another-electron tube 18 such as a pentode tube.

The delay line 12 causes the; appearance oina pulse 79 shown-on line lc of vFigure 3, identical with the! pulse H but having with respect toithat pulse ll a delay T/4. Since, by hypothesis, the duration of the pulse H is smaller than T/ -i, the total signal-180 transmitted over leads l5 and" consists of two identical successive pulses as shown on line. I of Figure3.

The tube 16 is .biassed beyond-cutsofi voltage, like the corresponding tube 34 of the firststage, by means, for instance, of a voltage :dividerdl H82, so as topass only-.signalshaving an :amplitude greater than that of the pulse 1 I. 'Thusthisatube does nottransmit the signaltfl, sothat no control signal is sent through the differentiating transformer 83 to the device fl lysimilar/to the device 48 of the first stage and, likeit, shown in the form .of a double triode tube. This tube thus remains .at its permanent equilibrium position when the signal applied tothe-stage has'a 'duration smaller than the'back and forth propagation time of the delay line ofthi stage.

It was seen that the signal is applied; on theother hand, through theilead TI to the control electrodev of thetube 18. On thesuppressor grid of this tubeis also applied the negative signal 38 which, having given rise to'the signal-H coincides in time with this signal. The signal 38 is again shown'on'line h of Figure 4. The tube 78 is blocked pending the duration of the signal 38 and thus passes only the first pulse H of the signal 80. Only the pulse 19 goes through this tube and after inversion by the transformer 85 supplies a pulse-86 :delayed'by T/4 withrespect to the pulse H and which is appliedto the control electrode of an electron tube 81.

:A circuit isimilarvto that 'ofithe'tube fl of 'the first stage also connects the control electrode of the tube 8'! with the anode of the left hand element of the tube 84; but since the latter tube does not leave its permanent equilibrium position, it delivers no signal capable of preventing the tube 87 from operating. Therefore, there is received at the terminals of the load resistor 88 of tube 81 a negative pulse 89 shown on line a of Figure 4.

This pulse 89, delayed by T/4 with respect to the pulse 38 is also applied to the control electrode of the input tube 90 of the third stage, and also, eventually, to a pentode tube not shown, such as 39 or 18 of the third coder stage, through the lead 9|, unless this third stage is the last coder stage as will be discussed in the description of this last stage in connection with Figure 5.

The third stage, like the previous ones, comprises a delay line 92, open at its end 93 and whose back and forth propagation time is T/2 =T/8.

It may be noted that the load resistor 94 of the left hand element of the tube 84, and the capacitor 95 for connecting the control electrode of the right hand element of this tube must be chosen so as to offer a time constant equal substantially to T/2. When the tube 84 leaves its equilibrium position under the control of a pulse such as 38, whose duration is greater than T/ i, it thus delivers a negative blocking pulse 95, with a duration T/2 sufiicient for blocking the tube 8'1 until the end of the signal transmitted in this case by the tube I8 since, in any case, the signal from the second coder stage cannot have a duration greater than T/2. On the other hand, the blocking pulse 96 must not have a duration substantially greater than T/2 in order not to hinder the operation of the coder when it receives the following pulse, which occurs substantially at a time T/2 after the appearance of the pulse 96.

Always in case the pulse such as 38 has a duration greater than T/4, the right hand element of the tube 84 delivers a positive pulse which is differentiated by the capacitor 9! and applied to the mixer device 64 as was seen during the description of the operation of the first stage.

' The pulse 89 delivers, at the output of the tube 90 a positive pulse 98 represented on line is of Figure 4, a pulse which is transmitted to the delay line 92 and eventually through the lead 99 to the control electrode of the tube of the third stage, such as previous tube 39 or I8.

As concerns the last coder stage, the nth stage, it is quite obvious that it does not have to transmit signals whose duration is greater than T/Z Its circuits are therefore simpler than those of the previous stages, as shown by Figure 5.

We shall assume, for explanation purposes, that the last but one coder stage delivers a signal I whose duration is greater than T/2 This signal is applied to the control electrode of an electron tube IN, and there is received at the terminals of the load resistor I02 of this tube a positive pulse I93, which is applied to the input of a delay line IB S. As in the preceding stages, this delay line is open at its end I95 and closed at its end I06 on its characteristic impedance and its back and forth propagation time is taken equal to T/Z.

With the assumption made, the line I04 thus delivers a signal Ill! built up by the partial overlapping of the pulse I03 and of the delayed reilected pulse, delayed by 172. The signal I01 is applied to the control electrode of an electron tube I08 biassed beyond cut-off voltage by means, for example, of a voltage divider I09I I 0 in such a manner as to pass only signals having an amplitude greater than that of the pulse I03. There is thus received, for example, on the cathode of the tube I08, a positive pulse III delivered by the peak of the signal IN. This pulse III is differentiated by a capacitor I I2 and. thus produces a signal II3 applied to the mixer device 64, for instance, on its terminal II4. Owing to the detector 65, there is received on the terminal 63 a positive signal II5, transmitted to the shaping device 61 (Figure 2) to produce a code pulse.

Of course, the device described in connection with Figure 5 has been given by way of a purely illustrative example and might be made up in any suitable manner.

It may be noted that the system described in connection with Figures 2 to 5 delivers code pulses at fixed times with respect to the position in time of the leading edge of the initial pulses 23. As a matter of fact, under the action of the delay lines of the successive stages, and due to the clipping caused by the comparator tubes, these pulses, when they do exist, appear with respect to this leading edge, with respective delays T/2, (T/2+T/4), (T2+T/4+T/8) It is known that it is advantageous in practice, to distribute regularly, in time, these coded pulses so as to facilitate transmission.

According to one characteristic of the invention, this result may be obtained very easily by means of a mixer circuit 64 consisting, as shown on Figures 2 and 5, of a delay line closed at its two ends on its characteristic impedance. If it is assumed, for instance, that the code comprises four elements, the corresponding pulses may be spaced by a regular time interval equal to T/4. As the two first code pulses occur with a shift of T/ i as mentioned, they will be received successively on the input terminal 63 of the mixer circuit. The third pulse oiTers a delay T/8 with respect to the second one; it will be applied to a terminal I IS (Figure 2) located at a distance from the terminal 63 corresponding to a propagation time T/B so as to have a total delay (T/8+T/8) with respect to the second pulse. Similarly, the fourth one will be received on a terminal I I? such that the propagation time from this terminal to terminal 63 is equal to 3T/IB, etc. the code signal being derived from the terminal 53 to be transmitted to the shaping device 81. As already mentioned, the pulse from the last stage of the system may be collected on the output terminal H t of the circuit and the total propagation time will then be equal to (T/8+T/I6+T/32+. .+T/2

Of course, the time interval between two consecutive code pulse positions may be different from the one just indicated by way of example, especially if the code comprises more than four pulse positions. A simple calculation will allow the determination of the positions of the various terminals of the circuit 94 capable of giving the desired recurrence frequency.

It is quite obvious that the system described has been indicated only by way of illustrative example; numerous modifications to the circuits described will be obvious to those skilled in the art, but it must be understood that any modifications incorporating characteristics of the present invention are within its scope.

What is claimed is:

1. Method for translating recurrent electric pulses modulated in duration by time displacement of their trailing edge into coded pulses, according to a chosen code, whichcomprises comparing successively, onthe one hand, the duration of the initial pulses or the signals. successively derived as modifications thereof with, on the other hand, a series of reference time intervals based on' the maximum modulation duration and themselves serially inter-related by a com mon factor, and producing a code signal each time a comparison determines that the initial pulse or its signal modification exceeds in duration the corresponding reference interval.

2. Method according to claim 1, including staggering the produced code pulses in time in a predetermined order and with predetermined time intervals.

3. Method according to claim 1, wherein each comparing operation includes for the next comparison either diminishing the duration of the modulated pulse or of its signal modification proportionally with the change in the corresponding reference interval if said duration exceeds said interval, or retaining its duration in full if it does not exceed said interval.

4. A device for translating recurrent electric pulses, modulated in duration by time displacement of their trailing edges, into coded pulses, comprising a cascade of coupled coding stages equal in number to the code elements, means in each stage whereby the stages compare successively the duration of said modulated pulses with predetermined reference time intervals one for each stage and successively related in value by a common multiplier, the first one of said time intervals having a predetermined value amounting to a substantial part of the maximum modulation duration, and each stage also having means to transmit a code pulse corresponding to that stage in response to excess value of the modulation duration of the incoming pulse over the reference interval of that stage.

5. A translating device according to claim 4, wherein the coupling between stages comprises two signal transmission circuits, and the means to compare includes devices responsive to incoming signals to direct the signal selectively over one of said circuits in dependence on its duration respecting the reference interval of the directing stage, and to block the other circuit.

6. A translating device according to claim 5, wherein the comparing means includes devices responsive to incoming signals to pass over one of said circuits only signals having a duration in excess of the reference interval of the same stage, to produce a code pulse responsive to said signals, and to supply the following stage with a signal having a duration decreased by the reference interval of the passing stage and delayed by this last mentioned duration.

7. A translating device according to claim 5, wherein the comparing means includes devices adapted to respond to incoming signals, to delay signals having a duration smaller than the reference interval of the instant comparing stage, and to transmit them over a selected one of said circuits in their full duration.

8. A translating device according to claim 5, wherein a first one of said circuits in each stage having two circuits comprises an electronic device having two positions of stability, means by which said device is actuated by a signal derived from the signal applied to that stage, when said applied signal has a duration greater than the reference interval of that stage, connections whereby the said electronic device, when actuated, controls the emission of a code pulse and otherwise produces a signalblocking the second circuit, said blocking signal having a duration substantially double that of the reference interval of said stage.

9. A translating device according to claim 4, wherein the last stage of thecascade has a single signal transmitting circuit and the comparing means has devices for passing thereover only signals having a duration greater than the shortest reference interval, and for thereby producing a code pulse.

10. A translating device according to claim 4, wherein, with T representing the maximum modulation duration, and n the number of code elements, the reference intervals are related successively as T/2, T/2 T/2 11. A device for translating recurrent electric pulses modulated in duration by time displacement of their trailing edges into coded pulses consisting of coding units with input and output terminals in cascade connection and in number equal to that of the elements of the chosen code, a first coding unit comprising means for applying primary duration modulated pulses to its input terminals, means for producing from said primary pulses secondary signals delayed by a given reference time interval with respect to said primary pulses, means for superimposing said secondary signals on said primary pulses so as to produce composite signals whose amplitude exceeds a predetermined value only in case the duration of said primary pulses exceeds said reference time interval, an electronic tube, means for applying said composite signals to a control electrode of said tube in such a way that said tube only transmits to its output circuit signals whose amplitude exceeds a predetermined value, means for transmitting said signals of excess amplitude to the output terminals of said first coding unit, means for deriving from said signals of excess amplitude a short duration pulse occurring at a fixed instant of the recurrence period of said primary pulses, an auxiliary circuit for transmitting the said secondary signals to the output terminals of said first coding unit, a second electronic tube, said auxiliary circuit including means for applying said primary pulses to a first control electrode of said second tube so as to render said auxiliary circuit inoperative for the duration of said primary pulses and for applying the said composite signals to a second control electrode of said second tube so as to render said auxiliary circuit inoperative for the duration of the said secondary signals, means for deriving from said short duration pulse a longer pulse having a duration approximately equal to twice said time reference interval, means for applying said longer pulse to said auxiliary circuit so as to render said auxiliary circuit inoperative for the duration of said longer pulse, means for transmitting signals received at the output terminals of said first coding unit to the input terminals of a second coding unit operating in a similar way but having 3, reference time interval equal to one-half of the reference time interval of said first coding unit, means for transmitting signals received at the output of said second coding unit to a third and in turn to successive coding units each having a reference time interval equal to one-half of that of the preceding coding unit, means for transmitting signals received at the output terminals of the last but one coding unit to the in- 11 put terminals of a last coding unit having a reference time interval equal to one-half of that of said last but one coding unit, said last coding unit operating in a way similar to preceding coding units but including only the circuitfor transmitting to its output terminals signals whose duration exceeds the reference interval of its stage, means for applying short duration pulses generated by each of said coding units to a common mixer circuit, said mixer circuit comprising a. delay circuit staggering said short duration pulses in time in a predetermined order and with predetermined time intervals, a circuit for shaping short-duration pulses received at output terminals of said mixer circuit into code pulses of predetermined wave shape and duration, and means for impressing said code pulses upon a transmission circuit,

PAUL FRANGOIS MARIE GLOESS.

LOUIS JOSEPH LIBOIS.

No references cited.