US2365450A - Radio telegraph multiplex system - Google Patents

Description

Dec. 19, 1944. w. H. Buss RADIO TELEGRAPH MULTIPLEX SYSTEM 5 Sheets-Sheet l I Filed April 29, 1942 WW I QQQQQQQQJ' N INVENTOR W 1' BY ..M ATTORNEY N M SN m m m i N QQQ QW Dec. 19, 1944. w. H. BLISS RADIO TELEGRAPH MULTIPLEX SYSTEM 5 Sheets-Sheet 2 Filed April 29, 1942 praan;

nun:

INVENTOR W 1' .5 BY

ATTORNEY Dec. 19, 1944.. w. H. BLISS 2,365,450

RADIO TELEGRAPHMULTIPLEX SYSTEM I Filed April 29, 1942 5 Sheets-Sheet 3 8 78 INVENTOR Z1315 154 BY MV-M/ ATTORNEY Dec, 19, 1944. w. H. BLISS RADIO TELEGRAPH MULTIPLEX SYSTEM 5 Sheet-Sheet 4 Filed April 29, 1942 INVENTOR. fififes/v 54/55.

.4 Wm NE Y Des. 19, 1944. V w. H. BLiSS RADIO TELEGRAPH MULTIPLEX SYSTEM Filed April 29, 1942 5 Sheets-Sheet 5 dots and dashes.

Patented Dec. 19, 1944 RADIO TELEGRAPH MULTIPLEX SYSTEM Warren H. Bliss, Orono, Maine, assignor to Radio Corporation of America, a corporation of Delaware Application April 29, 1942, Serial No. 440,920

28 Claims.

This invention relates to radio-telegraphy and more particularly to multiplex systems wherein a plurality of channels are used.

In present multiplex telegraph communication systems the radiated energy consists of the usual dots and dashes and the sequential selection of the channels is accomplished by means of motor driven synchronous distributors.

In the present invention, however, the use of the usual dots and dashes is dispensed with and in lieu thereof the dots and dashes are converted into impulses of substantially uniform duration and of a frequency determined by the frequency of operation of the multiplex system. Furthermore, the usual motor driven synchronous distributors are replaced by an electronic switching system in which no mechanically moving parts are employed, the electronic switching system functioning to efiect the desired sequential switching between the plurality of channels.

The impulse method of transmission, as will be disclosed in this specification, offers certain particular advantages over the conventional transmission of dots and dashes. When dots and dashes are transmitted, the receiver must not only respond to the actual reception of these transmitted signals but must also be responsive to their duration in order to distinguish between Frequently, due to atmospheric conditions, fading or reflections of waves, the receiving apparatus does not properly distinguish between dots and dashes since the dashes are not infrequently broken up into two or more dots. When an impulse method is used, however, eliminating the use of dots and dashes and thereby eliminating the necessity of distinguishing between dots and dashes at the receiver, the receiver need only respond to the actual reception of the impulses or, in fact to the beginning of each impulse so that the efiects of fading and multiple path reception may be avoided. Furthermore, when the receiver need respond only to the arrival of the impulses, it is immaterial what wave form the impulses may have or what their duration may be.

Furthermore, as will be later described, the present invention provides a circuit arrangement whereby the receiver is actually blocked, or rendered inoperative, between impulses so that false operation because of noise or intervening atmospheric disturbances may be substantially completely eliminated.

Also, through the use of an electronic switching circuit, troubles and difliculties which may inherently arise because of the use of motor driven synchronous distributors are eliminated with the result that reception is improved.

There are several dinerent manners in which multiplexing may be en'ected, one or these being the time division method which is predicated on the principle or assigning the sole use 01' the transmission circuit to each of the plurality of channels successively for predetermined time periods. 'l'his is accomplished by sequentially switching the trunk or transmission circuit to the various channels in a predetermined rixed order. Thus, each channel iunctions as a normal telegraph system during its predetermined time period OI assignment to the circuit.

The particular time division prmclple of operation of this invention 15 one in which a plurality of assignments are made tor each baud. A baud (named for Baudot, the inventor of the live-unit equal-length printer code that carries his name) is an interval of time corresponding to the time necessary to transmit one dot. example, a dash would cover a time interval corresponding to three bands. A baud, therefore, may be taken as the interval of time or the shortest duration of a man; or space element in a given telegraphic code. All other marks or spaces comprising the code are integral multiples thereof. In the present system, assignment of the channels to the transmission system occurs at a rate such that each channel is assigned at least once during each baud.

Since the present system operates on the principle of impulses, the multiplex receiver may be synchronized with the multiplex transmitter by the received intelligence impulses and no additional tone signal need be transmitted to eflect such synchromzation. Although there may be certain short intervals during which no impulses are in fact transmitted, the receiver is so designed as to continue operating at the proper rate of speed until impulses are received to assure the necessary synchronous operation.

For

telligence to be transmitted may be converted into impulses of a substantially constant length, and of a predetermined frequency, the intelligence being represented by the presence or absence of the impulses alone and not by their presence together with their duration and/or intensity.

Another purpose of the present invention resides in the provision of electronic switching means in a radio telegraph multiplex system for assigning the transmission system to one or another of a plurality of channels in a predetermined sequential order.

Still another purpose of the present invention resides in the provision of means for receiving the transmitted impulses and for converting the impulses into signals representing dots. and dashes in order that proper recorders or printers may be operated thereby. v

A still further purpose of the present invention resides in the provision of means for rendering the multiplex receiver ineffective or blocked during the intervals between the received impulses so that spurious signals will not affect the operation of the receiver.

Another purpose of the present invention re-' sides in the provision of an oscillator at the multiplex receiver for controlling the operation of the electron switching means, the frequency of operation of the oscillator being controlled and being maintained in synchronism with the operation of a corresponding oscillator at the transmitter through the use of the received impulses alone.

Still other purposes and advantages of the present invention will become more apparent to those skilled in the art from the following specification particularly when considered in conjunction with the drawings wherein:

Figure 1 shows schematically one form of a radio telegraph multiplex system Figures 2a and 2b show schematically amultiplex receiver for receiving and utilizing impulses sent in. accordance with the present invention and Figures 3a and 311 show a series of curves representing voltage variations at various designated points in the circuits of both the transmitter and receiver.

The complete radio multiplex telegraph system includes both a transmitter and receiver. The transmitter, which is shown schematically in Figure 1, includes three separate sources of signals. For the purpose of the present invention it would be assumed that three channels are employed and that the transmission system is sequentially connected to the three channels in order that the signals of each of the three channels may be, transmitted. One of the sources of signals represented by the rectangle i may be a standard auto head which supplies signals represented by dots and dashes. These signals, when derived from a standard auto head, are originally represented in the form of a perforated tape, the tape being punched on a standard Creed ,or Kleinschmidt perforator prior to being run through the standard auto head. The sources of signals for the other two channels may also be derived from standard auto heads, shown schematically at 2 and 3, or, if desired, from separate five-unit type transmitter heads. The tape which is fed through the five-=unit tape transmitter head normally is punched by a perforator in response to the operation of a teletype machine. It is to be understood, however, that it is immaterial what particular sources may be used for providing the dot and dash signals which are to be transmitted in each of the three separate channels. The signals derived from the source I are available along conductor 4 whereas the signals from the two sources 2 and 3 are available alongconductors 5 and 6, respectively.

Energy for operating the standard auto heads or the tape transmitters I, 2 and 3 is derived from the auto base amplifiers I one of which is associated with each source of signals. In order that the auto heads or the tape transmitters may be operated at proper speed, the auto base amplifiers are supplied with energy of a predetermined frequency along conductor 3. This energy is derived from an oscillator including tubes 9 and i0 and an amplifier tube H. The stabilized control oscillator is similar to the oscillator shown in the U. S. patent to J. N. Whitaker, No. 2,162,520. The frequency of operation of the oscillator is determined by the values of the inductance l2 and the capacitance l3. The two tubes 9 and ill operate in a manner somewhat similar to a multi-vibrator to supply voltage variations of the desired frequency to conductor 54. As described in the patent referred to above, oscillator tube 9 functions as a class A amplifier and operates to drive tube id to saturation on one half of each cycle and to cut-oil on the other half of each cycle. The output from tube It is nearly a square wave and is fed back to the tuned circuit fl2--l3 through a high resistance. The generated voltage variations are then applied to the control electrode of amplifier tube H and after proper amplification appear at the anode of the amplifier tube. The anodes of tubes 9, Hi and M are maintained positive with respect to their cathodes by a source of positive potential which is connected to the terminal 55.

In order that the functions and operation of the system as a whole may be completely understood, the wave forms of the voltage variations appearing in various points of the system have been shown in Figures 3a and 3b. The curves shown in Figures .3a and 3b are designated by letters and these same letters are applied to the schematic diagram of the transmitter and the receiver at the appropriate points.

The voltage variations which are present in conductor 3 and which'represent the dot and dash signals to be transmitted from this channel are represented at curve A in Figure 3a. Similarly, the voltage variations which are present at conductors 5 and 5 are represented by curves D and G, respectively. These voltage variations naturally depend upon the subject matter being transmitted and are continuously changing as the standard auto head or the tape transmitter heads are operated.

The voltage variations that are derived from the amplifier tube H associated with the stabilized control oscillator are represented by curve K in Figure 3a.

In order to sequentially connect the transmission system to the three channels in the desired manner, an electronic switching device is provided which includes tubes it, if and i8. These tubes are gas filled triodes and each includes a cathode, a control electrode and an anode. The cathode of each of the tubes is connected to ground by a. cathode resistance which is bypassed by a condenser. The anodes of the tubes are connected, by Way of separate anode load resistances, to terminal which provides a source of positive potential. The control electrodes of tubes Hi, ill and it are each supplied with energy from the anode of tube I I by way of condensers I9. The control electrode of tube It is connected to terminal 29, which provides a source of negative potential, by way of resistances 2| and 22, the resistances being connected in series. The junction of the two resistances is connected to the cathode of tube l8 by resistance 22. Likewise, the control electrode of tube [1 is connected to a source of negative potential by series resistances 23 and 24,-the junction of these resistances being connected to the cathode of tube I6 by resistance 24. A similar arrangement is provided at the control electrode of tube l8 and the junction of these control electrode resistances is connected to the cathode of tube H by a further resistance. Furthermore, the anodes of the gas triodes l6, l1 and I8 are interconnected by means of the commutating condensers 25. In order to provide an adjustable source of negative potential for the control electrodes of the gas filled tubes l6, l1 and 18, a potentiometer 26 is included in the common control electrode circuit in order that an adjustment of the bias potential applied to tubes l6, I1 and I8 may be provided.

Through the operation of tubes l6, l1 and I8, electron switching impulses may be produced which affect the connection of the transmission system with the three channels. When the tubes l6, l1 and I8 are connected as shown in Figure 1 of the drawings, they become conductive in sequence and in succession, the transfer of conduction from one tube to the next being at a rate determined by the frequency of the voltage variations appearing at the anode of tube ll. Assuming, for example, that tube I6 is conductive, the positive potential available at the cathode of that tube will then reduce the normal negative grid-cathode bias of tube I1 (by an amount determined by the values of the resistances 24 and 24) so that when the next succeeding cycle of control voltage variation is available from tube ll, tube will become conductive and tube I6 will be rendered non-conductive due to the interconnection of the anodes of the two tubes by condenser 25. While tube I1 is conducting, the bias potential on the control electrode of tube I8 is rendered less negative to prepare this tube for operation when the next succeeding impulse is received. Accordingly, tubes l6, l1 and I8 are rendered sequentially conductive by the stabilized oscillator wave K and, as a result, control potentials may be derived from the cathodes of these three tubes. These control or selection impulses are represented by curves B, E and H, respectively in Figure 3a.

Curves B, E and H illustrate the actual output wave forms at the cathodes of tubes [6, l1 and i8 andby means of the differentiating action of smallcondensers 21, 29 and 29 wave forms B, E." and H are converted into wave forms 13, E and H, respectively. Although these selection impulses have both positive and negative peaks, only the positive peaks are used.

By an inspection of Figure 3a it may be seen that the frequency of operation of the stabilized control oscillator (tubes 9 and i) is chosen such that the voltage variations will undergo three complete oscillations during an interval corresponding to one baud. Furthermore, it may be seen that the commutator output. impulses represented as B, E and H each occupy a length oi time corresponding to one cycle of the stabilized oscillator wave K and are rectangular in wave form, although the duration of the peak third the duration of one dot (or one-third of a baud). The positive selection impulses derived from the commutator outputs are displaced electrical degrees with respect to each other.

These commutator output impulses represented at B, E and H are then used to affect the electronic switching so that the transmission system may be selectively connected to the three channels. Furthermore, the stabilized oscillation wave that is available from the amplifier I l is applied by way of conductor 8 to the three auto base amplifiers 1 in order to synchronously control the speed of operation to the standard auto head I and the transmitter heads 2 and 3.

As stated above, signals from the three separate sources are available on conductors 4, 5 and 6 and may for the purpose of example, have a wave form such as shown at A, D and G in Fig ure 3a. Three separate tubes 21, 28 and 29 are employed, each of which is associated with a different channel. These tubes are preferably of the screen grid type and include two control electrodes. The signals which are available on conductor 4 are applied to one of the control electrodes or grid number 3 of tube 21, whereas the signals available on conductors 5 and 6 are applied to the corresponding electrode in tubes 28 and 29 respectively. The other control electrode or grid number 1 of tube 21 is supplied with energy from conductor 30 which is connected to the cathode of tube I6. Likewise, the first control electrodes of tubes 28 and 29 are connected to the cathodes of tubes l1 and I8 by way of conductors 3| and 32, respectively. There are present on conductors 30, 3| and 32 voltage variations of substantially rectangular wave form and these voltage variations are represented by the curves B, E and H in Figure 3a. However, the actual selection pulses to be applied to the number one control electrodes of tubes 21, 28 and 29 are as shown by curves B, E and H of Figure 3a.

Tubes 21, 28 and 29 are. normally provided with a potential bias between the first control electrode and cathode such that the tubes are rendered non-conductive in the absence of a controlling selection pulse from the associated gas discharge tube in the electron switching portion of the circuit. In order to provide the desired negative bias potential between the first control electrode and the cathode of each of tubes 21, 28 and 29, the control electrodes are each connected respectively by way of potentiometers 33, 34, and 35 to terminal 36, terminal 36 being supplied with a source of negative potential. By'

adjusting the potentiometers 33, 34 and. 35, the degree of the normal negative bias between the first control electrode and the cathode of tubes 21, 28 and 29 may be regulated. Similarly, a predetermined potential relationship is maintained between the other control electrode or grid number 3 of each of tubes 21, 28 and 29 and for this purpose potentiometers 31, 38 and 39 are provided. These potentiometers are connected between grid number 3 of each of tubes 21, 28 and 29 and the negative terminal 36.

Since tubes 21, 28 and 29 are normally nonconducting, they are individually rendered conductive only when both of the control electrodes of any single tube are supplied with positive imcharge tube 41.

pulses. The first control electrodes of tubes 21, 28 and 29 are supplied with regularly recurring positive impulses derived from the conductors 30, 3| and 32 which are connected, respectively, to the gas discharge tubes l6, l1 and I8 in the electron switching portion of the circuit. The tubes 21, 28 and 29 are then sequentially supplied with positive impulses but the tube still remains nonconducting unless a positive potential simultaneously occurs on the other control electrode or 10 grid number 3. In tube 21, for example, grid number 3 is supplied with energy from conductor 4 and the wave form of the voltage variations is represented by the curve A in Figure 3a. The impulses applied to the first control electrode of tube 21 are indicated by the curve B in Figure 3a. When these positive potentials coincide'in point of time, then current is permitted to flow through tube 21 since the anode of this tube is connected to the terminal to which is applied a positive potential. When current flows through tube 21, a momentary potential drop is produced across its cathode resistor 40/ so that the cathode of tube 21 changes its potential in a positive direction and develops a positive impulse. Curve C of Figure 3a illustrates the oathode output impulses developed by tube 21. negative impulses occurring on lead B (wave form B of Figure 3a) have no efiect on the action of tube 21 since both control electrodes (number 1 grid and number three grid) must be driven positive to produce anode-to-cathode conduction.

A similar operation occurs in tube 28 since the first control electrode is supplied with voltage variations such as represented by the curve E in Figure 3a. while the other control'electrode or grid number 3 has applied thereto voltage variations such as represented by curve D in Figure 3a. When the positive impulses, of these curves coincide, tube 28 is rendered conductive and a voltage drop is produced across its cathode resistor 41 so that the cathode of tube 28 is driven momentarily in a positive direction. Curve F of Figure 3a shows the output impulses of tube 28. Similarly, tube 29 is rendered conductive when both of its control electrodes are simultaneously supplied with positive impulses to produce a voltage drop in its cathode resistor 42. Curve I of Fig. 3a shows the output impulses of tube 29.

The cathodes of tubes 21, 28 and 29 are connected together and are brought to a common junction 43. Accordingly, the voltage variations that are present at each of the cathodes are combined into a single series which is available at the common junction point 43. The voltage variations which are available at this point are then supplied to the control electrode of gas discharge tube 44, the common junction point 43 being connected to the control electrode by means of condenser 45 and resistance 46. The voltage variations appearing at the control electrode of tube 44 are similar to those indicated by curves C, F and I of Figure 3, combined.

Cooperating with tube 44 is another gas dis- These tubes together constitute the impulse shaping section of the system and the tubes are alternately rendered conducting and non-conducting. The anodes of tubes 44 and 41 are connected together by means of a commutating condenser 48 while each of the anodes has a separate load resistance 49 and 50. The load resistances 49 and 50 are connected between the anodes of tubes 44 and 41, respec- The tively, and a point of positive potential, terminal 15. The cathodes of tubes 44 and 41 are connected to ground by cathode resistances 5i and 52 while the control electrodes of the tubes are normally supplied with a negative potential by means of potentiometers 53 and 54. One end of each of the resistance elements of the potentiometers is connected to terminal 55, to which is applied a negative potential, and the other end of each is connected to ground. The movable contact of potentiometer 53 determines the negative biasing potential of the control electrode of tube 44 with respect to its cathode while the position of the movable contact of potentiometer 54 determines the negative biasing potential of the control electrode of tube 41 with respect to its cathode.

Normally, the tube 41 may be considered as conducting and tube 44 non-conducting. When a positive impulse is applied to the control electrode of tube 44 that tube is rendered conductive and by reason of the interconnection of the tubes and their commutating action, tube 41 is simultaneously rendered non-conductive. Inasmuch as the control electrode of tube 44 may draw grid current, the resistance 46 is also provided to limit the amount of grid current that is permitted to flow.

Once conduction is established by a positive impulse, tube 44 will remain conducting for a predetermined length of time determined by the parameters of the circuit, after which time tube M will begin to conduct and tube 44 will immediately be rendered non-conductive. It will thus be seen that tubes 44 and 41 together constitute a trigger circuit having one degree of electrical. stability. Accordingly, positive impulses of the desired wave form and shape are available from the cathode of tube 44 and these impulses are represented by the curve J in Figure 3a. These re-shaped impulses which are available from the cathode of tube 44 and which are present on conductor 56 are then supplied to a radio telegraph transmitter where they are used to key or modulate a radio frequency carrier. Keyed radio frequency energy is then transmitted from the transmitter 51 to the transmitting antenna 58 for radiation. Any type of radio transmitter 51 may be used and various forms of keyed transmitters are well known in the art.

Through the use of the above described transmitter it is therefore possible to transmit a keyed carrier at a predetermined maximum rate and in a manner corresponding to the intelligence derived from the three separate sources. When such signals are received and properly differentiated three recording apparatus or printers may be controlled to reproduce the transmitted subject matter.

Figures 2a and 2-1) show a receiver designed to respond to the impulses of radio frequency energy which are transmitted by the transmitter shown in Figure 1 and described above. The receiver broadly includes means for detecting and demodulating the received signals together with electron switching means for assigning the received impulses to the proper channel. After the impulses are assigned to the proper channel, wave shaping means are then used to reconvert the impulses into the proper dot and dash symbols which are ultimately used to control a recorder or a printing device.

The incoming received impulses are received Any tected by the radio receiving system 62. known type of radio receiving system may be used. In order to eliminate the possibilities of fading, however, it is recommended that dual unit diversity receivers or even triple unit diversity receivers be used in the well known manner. After the signals have been detected they are then available on conductor 63 at the output of the receiving system 62. The wave forms of the impulses available at this point may be such as those shown at L in Figure 3b. The amplitude of these impulses may not be uniform nor may they be of uniform length. A lack of uniformity in amplitude and/or length is not material since the intelligence is represented only by the presence or the beginning of each impulse. The impulses that are available on conductor 63 are not directly usable, however, since it is desired to reshape and reform the impulses into a more directly usable form. For this purpose a reshaping circuit is provided.

The re-shaping circuit includes tubes 64 and 65 as well as the gas triodes 66 and 61. The detected impulses available along conductor 63 are applied to the control electrode of tube 64 (which is of the. screen grid type and which operates as an amplifier tube). The cathode of tube 64 is connected to ground and the anode is connected to positive terminal 66 by way of anode load resistance 69. Accordingly, voltage variations are produced at the anode of tube 64 as a result of an application of voltage variations to the control electrode of that tube.

The blocking control vacuum tube 65 includes a cathode, a control electrode and an anode. The cathode is connected to ground by way of the cathode load resistance while the anode is connected directly to the positive terminal 66. The cathode of tube 65 is also connected to the grid of amplifier tube 66 by way of resistance 1|.

The blocking circuit also includes a gas diode 12 as well as a conventional vacuum diode rectifier 13. The anode of the gas diode 12 is connected to the cathode of the diode 13 and these two elements are in turn connected to the control electrode of tube 65. A condenser 14 is interposed between the control electrode of tube 66 and ground. The anode of the diode 13 is connected directly to the output of transformer 15 while the cathode of the gas diode 12 is connected to the output of transformer 16. The cathode of the gas diode 12 is also supplied with an adjustable negative potential through potentiometer 11. One end of the potentiometer is connected to ground while the other end of the potentiometer is connected to terminal 18 to which is applied a negative potential. The position of the movable contact along the potentiometer 11 determines the degree of negative bias applied to the cathode of the gas diode 12.

Amplified energy from the anode of tube 66 is applied to the control electrode of tube 66 through series connected coupling condenser 76 and current limiting resistance 86. The tube 66. includes also a cathode and an anode. The cathode is connected to ground by a cathode resistance 6i while the anode is connected to positive terminal 66 by anode resistance 82. The bias of the control electrode of tube 66 with respect to its cathode is determined by the setting of the potentiometer 66. One end of the resistance element of the potentiometer is connected to ground whereas the other end of the resistance element of the potentiometer is connected to the negative terminal 18. A commutating condenser 64 is connected between the anode of gas triode 66 and-the anode of gas triode 61. The triode 61 includes a cathode, a control electrode and an anode. The control electrode of tube 61 is connected to the anode of diode 13 and tothe output of transformer 15 by way of series connected coupling condenser 85 and a current limiting resistance 66. The cathode of tube 61 is connected to ground by way of cathode resistance 81 while the anode of tube 61 is connected to positive terminal 68 by way of anode load resistance 88. The bias of the control electrode of tube 61 with respect to its cathode is controlled by means of potentiometer 89 so that the setting of the movable contact of the potentiometer determines the normal voltage of the control electrode of tube 61 with respect to its cathode. The tubes 66 and 61 are rendered alternately conducting and, accordingly, positive impulses may be derived from each of the cathodes of these tubes. The voltage variations which are available from the cathodes of tubes 66 and 61 are similar but are opposite in polarity with respect to each other. These impulses are shown at M and N in Figure 3b and are utilized in the system as will be explained later. It will thus be seen that tubes 66 and 61 taken together constitute a trigger circuit. A synchronized oscillator is provided for properly controlling the receiver and for properly distributing the impulses to the proper channels. The oscillator in many respects to the oscillator used in the transmitter but differs from the transmitter oscillator in that it is controlled or synchronized by means of the received impulses. Although the received impulses are not regularly recurring but occur in accordance with the transmitted intelligence, the flywheel action and stability of the oscillator are sufficient to maintain proper operating speed even though certain of the synchronizing impulses may be missing. The synchronizing impulses for stabilizing the oscillator are applied to the control electrode of tube 6| 'by way of conductor 95.

The impulses that are available at conductor 63 at the output of the receiver, in addition to being applied to the amplifier. tube 64, are also passed through a low pass filter including series inductance 96 and parallel condensers 91. This low pass filter removes the unnecessary high frequency component of the impulses yet permits the relatively low frequency impulses to be available at the output of the low pass filter. The output from the low pass filter is then applied to the control electrode of amplifier tube 98 by way of conductor 96. The amplifier tube includes a cathode, a control electrode and an anode and the amplified impulses that are available at the anode of tube 98 are applied to coupling condenser I66 and, accordingly, to the control electrode of tube 9! by way of conductor 95. These impulses which are sent out at a predetermined frequency from the transmitter are then used to synchronize the oscillator at the rereceiver in order to insure that its frequency will correspond to the frequency of operation of the oscillator in the transmitter. The output from the receiver oscillator isderived from the movable contact of potentiometer till which is connected across the ends of the tank coil inductance 92, By varying the position of the movable contact along the potentiometer. the intensity of the signals available at the movable contact may be varied. These oscillations are then applied to the control electrode of an amplifier tube I02 which includes a cathode, a control electrode and an anode. The cathode is connected to ground by way of the usual condenser by-passed cathode resistance and the oscillations are applied to the control electrode of tube I02 from the movable contact of potentiometer IDI. The anode of tube I02 is connected to positive terminal 58 by way of the primary of transformer I63. At the secondary of the transformer I 03 there are therefore available amplified oscillations of a predetermined frequency corresponding to the frequency of operation *f the oscillator in the transmitter. The wave form of this voltage variation or oscillation may be such as that represented by the curve 0 in Figure 3b.

The voltage variations generated by the local oscillator and available at the secondary of transformer I03 are applied to the primaries of transformers IM and IE5. The secondary of the transformer I04 supplies energy to the control electrode of tube I08 through a phase shifting network. The phase shifting network is of the double resistance-capacitance type and includes condensers Itl, I08 and variable resistances I09 and H0. The variable resistances are simultaneously adjustable from zero value to their maximum values to alter the phase relationship of the voltage variations applied to the control electrode of tube I06 with respect to the voltage variations applied to the primary of transformer I04. The secondary of transformer I05 likewise supplies voltage variations to the control electrode of tube III through a similar phase shifting network. The particular type of phase shifting network used is immaterial in so far as this invention is concerned. The particular phase shifting network shown by way of example in Figure 2a of the drawings is disclosed and described in U. S. patent application, Serial No. 329,990, filed April 16, 1940, by R. E. Mathes and W. H. Bliss.

Tubes I06 and III each include a cathode, a control electrode and an anode. The cathode of each of the tubes is connected to ground through a usual cathode resistance, the resistance being by-passed by a condenser. The anode oftube I06 is connected to the positive terminal 68 by means of the primary of transformer I5 in order that voltage variations may be applied to the anode of diode I3 and to the control electrode of gas discharge tube 61, whereas the anode of tube I II is connected to the positive terminal 68 by the primary of transformer I6 in order to supply voltage variations to the cathode of the gas diode 12. By adjusting the phase shifting circuits associated with tubes I06 and I II it is possible to produce, at the transformers I5 and I6 respectively, voltage variations having a frequency corresponding to the frequency of operation of the local synchronized generator, the two wave forms being phase displaced by a predetermined amount. The voltage variations which are present at the secondary of transformer may be such as those shown by the curve P in Figure-3b whereas the voltage variations appearing at the secondary of transformer I6 may have a wave form such as that shown at Q in Figure 3b. It will be observed that the phase relationship of the curves 0, P, and Q in Figure 3b is not identical but that the curves are displaced from one another by a particular amount. The reason for phase displacement and the relationship of the displacement with respect to the received signals will be explained later.

In order to segregate the multiplex signals and to assign them to the proper system, an electronic distributor is used at the receiver. The electronic distributor or electronic switching device is quite similar in most respects to the corresponding electronic switching device which included tubes It, II and I8 in the transmitter shown in Figure 1. In the receiver the electronic switching device includes tubes H2, H3 and lit. These tubes are preferably gas triodes and each of the tubes includes a cathode. a control electrode and an anode. The cathodes of the tubes H2, H3 and IM are connected to ground by way of cathode resistances H5, Iii;

and III, respectively, the, cathode resistances being by-passed by appropriate small condensers. The anodes of the tubes are bonnected to positive terminal 88 by way of anode load resistance H8, H9 and I28, respectively.

The control electrodes of tubes H2, H3 and IM are maintained at their proper negative potential with respect to their associated cathodes by means of a potentiometer IZI, the resistance element of the potentiometer being connected between ground and negative terminal It. The movable contact of the potentiometer I2I is connected to each of the control electrodes by means of two series-connected resistances such as I22 and I23 shown associated with tube I I2. The junction of the resistances I22 and I23 are connectedv to the cathode of tube Ill by resistance I22. Similar control electrode-cathode connections exist between the control electrode resistances of tube H3 and the cathode of tube H2 as well as between the control electrode resistances of tube Ill and the cathode of tube H3. By means'of these interconnections together with the commutating condensers I23 *(that' interconnect the anodes of tubes II2, I I 3 and N4), the tubes are rendered conductive in sequence in a predetermined order. The control electrodes of the tubes H2, H3 and Ill are supplied with voltage variations from the secondary of transformer I 03 by way 01 individual coupling condensers I25. The operation of this electron switching device or commutating device is similar to the transmitter version of the corresponding device shown in Figure 1 and as a result of voltage variations being applied to the control electrodes of the electron switching device, voltage variations may be sequentially obtained from the cathodes of the three tubes. Accordingly, a voltage variation such as that represented by the curve S in Figure 3b may be obtained from the cathode of tube H2 whereas voltage variations such as those represented by the curves W and A in Figure 3b may be obtained from the cathodes of tubes H3 and IIrl respectively. The voltage variations which are available at the cathodes of tubes H2, H3 and H4 are applied respectively to conductors I26. I21 and I28.

The output from one' section of the electron switching device is taken from the cathode of tube H2 and is supplied, by way of conductor I26, to the channel selecting device for channel number I. This selecting device or circuit consists of two multi-electrode tubes I29 and I33 and their associated elements. Tube I29 is the marking control tube for channel number 1, while tube I30 is the spacing control tube for that channel. These tubes include at least a cathode, a first and second control electrode and an anode. The anodes of tubes I29 and I30 are connected directly to the positive terminal 68 whereas their cathodes are connected to ground by way of their cathode load resistances I3I and I32, respectively. The first control electrode of each of tubes I29 and I30 are interconnected and are maintained at a proper negative potential with respect to their associated cathodes by means of a potentiometer I33, the resistance element of the potentiometer being connected between ground and terminal I3 2 to which is connected a source of negative potential with respect to ground. The movable contact of the potentiometer I33 is then connected to the first control electrodes of tubes I29 and I30 by way of a conventional control electrode resistance. The first control electrodes of tubes I29 and I30 are supplied with voltage variations from the oathode of electron switching tube II2 by way of conductor I26 and large coupling condenser I35. This wave form, as stated above, is substantially rectangular and may be of the form represented by the curve S in Figure 3b.

Corresponding marking control and spacing control tubes are also used for channels 2 and 3. The marking control tube I36 and the spacing control tube I31 for channel number 2 are also multi-control electrode tubes such as are used in channel number 1 and these tubes are connected in a manner similar to the connection of the tubes I29 and I30 in channel number 1. A potentiometer I39 is also provided for controlling the potential bias between the first control electrodes of these tubes and their associated cathodes. control tubes I39 and nel number 3.

The second control electrode of the marking control tubes I29, I 36 and I39 are all connected together and are supplied with impulses from the cathode of the triode 66. These impulses, as stated above, may be such as represented by the curve M in Figure 3b and are, in fact, the received impulses after re-shaping and re-form-.

ing. Likewise, the second control electrode of the spacing control tubes I30, I31 and I40 for channels 1, 2 and 3, respectively, are connected together and are supplied with impulses from the cathode of gas triode .61. These impulses may have a wave form such as that represented by the curve N in Figure 3b which bears an outof-phase relationship with respect to the impulses available from the cathode of tube 66. The second control electrodes of the marking control tubes I29, I36 and I39, as well as the second control electrodes of the spacing control tubes I30, I37 and I40, are maintained at their proper negative potential with respect to their associated cathodes by means of potentiometers MI, a potentiometer being associated with each pair of marking and spacing control tubes. The resistance elements of the potentiometer are connected between ground and the negative terminal I39. The movable contacts of all of the potentiometers MI are connected to the second control 'electrodes'of the marking control tubes I29, I36 and I39 by way of a grid resistances I42 whereas the movable contacts of all of the potentiometers I M are also connected to the second control electrodes 'of the spacing control tubes I30, I317 and I90 by way of an appropriate control electrode resistances M9.

Similar marking control and spacing I40 are also used in chan- The control electrodes oi.v both of tubes I29 and I30 are so biased with respect to their cathodes that the tubes are normally non-conductive. Neither tube is permitted to conduct unless both of its control electrodes are supplied with a positive impulse. The first control electrodes of tubes I29 and I30 are both controlled by the selection impulses which are derived from the cathode of tube H2 or the electronic switching device by way of large condenser I so that these control electrodes may be supplied with positive control impulses at predetermined intervals as indicated by curve S in Figure 3b. Tube I29 has applied to its second control electrode, impulses from the cathode of tube 66 so that this tube becomes conducting only when a positive impulse derived from tube 66 coincides with the positive impulse derived from tube H2. The resulting conductive conditions of tube I 29 then produce a voltage drop across the cathode resistance of tube I29 to produce marking impulses in channel number 1. These impulses are derived from the cathode of tube I29 and in differentiated form may appear as a. wave form such as that indicated by curve T in Figure 3b. Tube I29 is therefore conductive only when the positive impulses of curve M in Figure 3b coincide with the positive impulses of the selected impulses represented by the curve S in Figure 3b.

Likewise, the spacing control tube I30 is normally biased to a non-conductive condition and is rendered conductive only when both of its control electrodes are supplied with a positive impulse. This can happen only when a positive impulse from the cathode of tube 51, coincides with the positive impulses which are derived from the cathode of the electron switching tube H2. The voltage variations from the cathode of tube 6'! may be such as represented by the curve N in Figure 3b and when these positive impulses coincide with the positive impulses of curve S in Figure 3b then the tube I30 is rendered con-- mutating tube Ht.

ductive to [produce spacing control impulses at the cathode of tube I30 and these impulses, after differentiation, may have a wave form such as that represented by the curve U in Figure 3b.

Similar operations occur with respect to the marking control and spacing control tubes I36 and I31 of channel number 2 to produce marking and spacing control impulses such as represented by curves X and Y respectively in Figure 3b, these impulses being produced as a-result of a combination of the impulses derived from the cathodes of tubes 60 and 61 in conjunction with the channel number 2 selection impulses which are derived from the cathode of the electron switching tube I I3.

Similarly, the marking control and spacing control tubes I39 and I40 that are associated with channel number 3 are normally rendered non-conductive but are permitted to pass current to generate marking control and spacing control impulses such as represented by curves B and C in Figure 3b. These curves are a result of the combination of impulses derived from the cathodes of tubes 66 and 6'? working in combination with the selection impulses that are derived from the cathode of the electron switching or com- Further descriptions of the operation of the circuit will be later presented,

The three channels at the receiver also include locking circuits, the locking circuit for channel number 1 including tubes I44 and I 45 as well as their associated circuits. Tubes I44 and I45 are preferably gas triodes, each including a cathode,

a control electrode and an anode. The cathodes of tubes I44 and I45 are connected to ground by cathode resistors I46 and I41 respectively. The anodes of tubes I44 and I45 are interconnected by commutating condenser I48 and the -anodes are connected to the positive terminal 68 by way of load resistances I46 and I50, respectively. The control electrode of tube I44 is maintained at its proper operating potential with respect to its associated cathode by means of potentiometer I i. Likewise, the control electrode of tube M5 is maintained at its proper operating potential by potentiometer I52. The resistance elements of potentiometers IM and I52 are connected between ground and negative terminal 16. Appropriate control electrode resistances are included between the control electrodes and the movable contacts of the potentiometers. The control electrode of tube M4 is supplied with impulses from the cathode of tube I26 by way of the differentiating or peaking condenser I53. As stated above, these voltage variations (having both positive and negative peaks) may correspond to the curve shown at T in Figure 3?). Likewise, the control electrode of tube I45 is supplied with controlling impulses from the cathode of spacing control tube I30 by way of the small differentiating condenser I56. Due to the interconnection of the anodes of tubes I44 and I45 they are rendered alternately conductive by reason of the positive impulses applied to their control electrodes. When a positive impulse is applied to the control electrode of tube I44 this tube will continue to conduct until a positive impulse is applied to the control electrode of tube I45 at which time tube I45 is rendered conductive and, by reason of the commutating condenser I40, tube I 44 is simultaneously rendered non-conductive. This condition of conduction will continue until the control electrode of tube I44 is again supplied with a positive initiating impulse from the cathode of tube I29. Each time the tube I44-is rendered conductive a potential drop is produced across its cathode resistance I46 so that the voltage variations may be derived from the cathode of tube I44. The negative pulses from condensers I53 and I54 have no influence on the operation of tubes I44 and I45 since the application of a negative pulse to the control electrode of a gas diode has no efiect on the operation of this type of electron tube.

The tube I44, as stated above, is supplied with marking control impulses from the cathode of tube I29 and tube I45 is supplied with spacing control impulsesfrom the cathode of tube I30. After tube I44 has been rendered conductive by a marking control impulse this tube continues to conduct until the spacing control impulse is eflective to render tube I 45 conductive and tube I44 non-conductive. Accordingly, when marking and spacing control impulses derived from tubes I20 and I30, respectively, are applied to the locking circuit including tubes I44 and I45, voltage variations such as those shown by the curve V of Figure 3b may be derived from the cathode of the gas triode I44. Since the marking and the spacing control impulses 'are generated in accordance with the received impulses the voltage variations available at the cathode of tube I44 and at conductor I55 will correspond to the signals transmitted from channel number 1 of the transmitter and may be used for controlling an ink recorder I69 or a printer at the received apparatus.

A similar locking circuit including tubes I56 and I51 are associated with channel number 2 of the receiver and the circuit elements asso- [response to the marking and spacing impulses and by reason of their actions a voltage variation may be derived from conductor I 62 which is connected to the cathode of tube I56. These Voltage variations may be similar to those shown by the curve Z in Figure 3b. When an appropriate recorder or printer I63 is provided, these voltage variations may then be used to reproduce the subject matter transmitted on channel number 2 from the transmitter.

Similarly, a locking circuit is provided for channel number 3 at the receiver and this circuit includes tubes I64 and I65 as well as their associated circuits. These tubes are connected in a manner similar to the connections associated with the locking circuit for channels 1 and 2 and tubes I64 and I65 are supplied with control impulses from tubes I36 and I40 respectively. The marking control impulses which are derived from the cathode of tube I39 are applied to the control electrode of tube I64 by way of coupling condenser I66 whereas the spacing control impulses that are derived from the cathode of tube I 60 are applied to the control electrode of tube I65 by way of coupling condenser I 61. As stated above, the marking control impulses and the spacing control impulses for channel number 3 may correspond to the wave form shown by curves B and C in Figure 3b and these two wave forms together with the tubes I64 and I65 are effective to produce a voltage at the cathode of tube I64 and at conductor I68 corresponding to the wave form shown at D in Figure 3b. These voltage variations are then applied to an appropriate recorder or printer.

When a circuit such as that shown in Figures 2a and 2b is used, therefore, triple channel multiplex signals may be properly segregated into their proper channels and appropriate voltage variations maybe produced in order that ink recorders I69, I63 or. I10 associated with the three channels may becaused to record the transmission from the corresponding units I, 2 and 3 in Figure 1.

The operation of the system as a whole will now be described. In order to facilitate a complete understanding of the operation of the system it will be assumed that the smallest unit of time division in the multiplex system is one-third of a baud which, in many causes, is standard practice. Furthermore, for the purpose of explanation of the system an operation speed of 40 bauds per second will be assumed. It is to be understood, however, that in actual practice the maximum usable speed would be determined largel by radio transmission conditions.

The basic control for the entire system is the stabilized control oscillator at the transmitter shown in Figure 1. This oscillator includes tubes 9 and I0 and stabilization of the oscillator may be maintained by the inductance I2 and the capacitance I3 alone or a crystal may also be included for better frequency stabilization if necessary. It the system is to operate at 40 bands per second with an assignment time of one-third of a baud per circuit then naturally the frequency oi operation of the oscillator will be 120 cycles per second. The output from the oscillator is supplied to a class A amplifier H which, in turn, supplies energy to the three auto base amplifiers 1. These amplifiers have sufllcient output energy to supply 120 cycle power for driving the auto heads or tape transmitters by means of small synchronous motors. The signals delivered from three transmitting units i, 2 and 3 must be properly phased and synchronized with the control frequency supplied by the master oscillator. The oscillations supplied by the master oscillator are shown by the curve K in Figure 3a while curves A, D and G represent typical signals derived from the auto heads or transmitter heads I, 2 and 3, respectively. Each baud for each channel starts at the center of an oscillator cycle with number 1 channel bauds starting on cycles i--4-1-l0, etc., number 2 channel bauds starting on cycles 25-8| I etc., and number three channel bauds starting on cycles 36-9l2, etc.

Since the electron distributor or switching device of Figure 1 provides the control for successive channel switching, its operation will be taken up next. The electron distributor or switching device includes gas triodes l8, I1 and i8 and by reason of the circuit arrangement above described only one of three tubes is in a state of conduction at any particular instant. As

each oscillation cycle from the stabilized oscillator is applied to the control electrodes of the gas triodes l6, l1 and ill, the conductive condition of the tubes is passed along and rotated among the three triodes so that they are successfully rendered conductive. As each successive tube is placed in a conductive condition the last tube that was conducting is rendered nonconducting. Assuming that at some particular instant tube I1 is conducting then tube 18 is the next in order of firing. All three tubes have a common grid bias voltage from potentiometer 2B but at the instant under consideration the total bias on tube 18 is more positive than that of tube IS. The conducting tube I! has a highly positive grid because of plate to grid conduction.

The reason that tube 18 has a more positive bias than tube 16 is because the cathode of tube I1 is several volts positive above ground potential due to current flow through its cathode resistance and this raises the grid potential of tube it by reason of the connection of the cathode of tube ii to a point along the control electrode resistance for tube i8. Consequently, when the next positive half cycle is applied to the control electrodes of tubes it, i1 and it from the stabilized oscillator only tube it is rendered conductive while tube i1 is rendered non-conductive. ibe it is not affected in any way at this time. When tube it is rendered conductive then tube ii is rendered non-conductive as stated above by reason of the commutating action of condenser 25. Tube it will continue to conduct for one cycle of operation of the stabilized oscillator and when the next succeeding positive impulse is applied to the control electrodes tube It is rendered conductive and tube it rendered non-conductive. Thus the electron distributor will make one complete cycle of operation for three cycles from the stabilized oscillator with the result that each tube of the electron switching circuit is rendered conductive once for each baud.

Wave forms shown at B, E and H in Figure 3a show the outputs from the distributor tubes of the electron switching device and these outputs are derived from the cathodes of tubes l6, l1 and I8. The condensers which by-pass the cathode resistances for the tubes are helpful to improve the wave form and to assist in the commutating action. With the auto head or transmitter head drive motor properly phased as previously indicated. the electron distributor tube for each channel will fire or become conductive about one-halfcycle of the oscillator after the beginning of each baud for that particular channel. This time relationship is clearly indicated in Figure So by comparing for example curves A and B or curves D and E. Furthermore, the time relationship of these waves may be readily seen by comparison with the wave form represented at K which is the output from the stabilized oscillator.

The signals A, derived from number one channel, for example, are applied to the second control electrode or grid number three of tube 21 while the selection impulses B, from distributor tube iii are applied to the first control electrode or grid number one. Both of these control electrodes are biased beyond tube cut-oil? so that no plate current can flow in the tube unless positive signals are applied simultaneously to both control electrodes. For tube 21, therefore, there will be pulses of plate current only at the instance when the positive impulses of curves A and B coincide, and the cathode potential swing caused by these plate current pulses appear as indicated by the wave form shown at C in Figure 3a. One of these pulses is produced for each dot, three for each dash" and none for the spaces between the dots and dashes. The curves shown at C, F and I in Figure 3a represent voltage variations at the control electrode of tube 44 which are the same as the voltage-variations at the cathodes of tubes 21, 28 and 29.

The wave forms shown at F and I in Figure 3a are the outputs or driving impulses from channels number 2 and number 3, respectively. Tubes 28 and 28 are controlled in a manner similar to the control of tube 21 and these tubes respond respectively to signals from the transmitter heads 2 and 3 and control impulses from electron switching tubes l1 and I8.

Since the anodes of tubes 21, 28 and 29 are all connected to positive terminal l5 the outputs from these tubes are derived from their cathodes. All oi the outputs are combined at point 43 and are applied to the control electrode of tube 44 in combined form through condenser 45. The complete voltage variation. therefore, at the control electrode of tube 44 would actually be a curve resulting from the combination of curves 0, F and I.

The impulse shaping circuitincludes gas triodes 44 and 41. These tubes are alternately rendered conductive and commutation between the two tubes is accomplished by the presence of the commutating condenser 48. Normally, tube 41 is in a state of conduction but whenever a driving impulse from any one of the three channels (1. e., from any one of the three tubes 21, 28 and 29) is applied to tube 44, this tube becomes conductive since the applied impulses extend in a positive direction. As soon as tube M is rendered con-.

ductive tube 41 is rendered non-conductive. Due to the action of the commutatlng condenser 48, however, tube 41 again assumes a conductive conas will be explained later.

dition and tube 44 is rendered non-conductive after about one-third of a cycle of control since after the expiration of such a time the anode of tube 41 will have assumed a sufficient positive potential to permit conduction of tube 41. In other words, when tube 44 is subjected to a control impulse such as represented by curves C, F and I, it is rendered conductive but the conductive condition persists for only about one-third of an assignment cycle (one-ninth of a baud). This results in the development of a short positive rectangular pulse at the cathode of tube 99 for each driving impulse applied to the control electrode of that tube. The wave form of the voltage also reveal that the impulses have a time duration corresponding to about one-third of a cycle of the stabilized oscillator. This wave form is then used to key a transmitter such as represented at 51 in the usual manner'well known in the art.

Through the operation of the transmitter it may be seen, therefore, that relatively short rectangular impulses of radio frequency are transmitted, the impulses having a uniform time duration and being of a uniform intensity. The intelligence to be transmitted is represented solely by the presence or absence of the impulses. Furthermore, the impulses are representative of intelligence from three separate sources.

The operation of the receiver will now be described and, broadly, the receiver operates to segregate the received impulses and properly assign them to one or another of the three channels and to thereafter transform the received impulses into impulses which may be directly used. 7 The radio frequency impulses which are sent out by the transmitter at a distant point are picked up by the receiving antenna 6| and are amplified and detected by a receiving system 62. In actual practice it may be desirable to use more than one receiver and it isto be understood that two-unit or three-unit diversity reception may be employed.

Due to fading. noise, multi-path or other phenomenon of radio transmission. rectified re-' ceived impulses shown by curve L in Figure 3b vary considerably in form and amplitude from those radiated by the transmitter and represented by curve J in Figure 3a. In order that these pulses may .be more readily used, a reshaping circuit is provided which includes tubes 64, 65, 66, 67, I2 and 13. This part of the circuit also operates to block the circuit to any extraneous noise impulses which might occur between received signal impulses.

The incoming amplified and rectified impulses are passed through a signal blocking and amplifier tube 64 to a gas triode 66. The blocking action of the amplifier tube 64 is accomplished by plate current saturation. Normally, the grid of tube '64 is only slightly positive with respect to its cathode since tube 65, which controls this grid voltage, has its plate current nearly cut-oft The impulses from .the receiver 62 extend in a negative direction as indicated by curve L and when they are applied .to the control electrode of tube 64 the anode po- 'tential of the anode of that tube will rise or change in a positive direction due to a drop or cessation of anode current.

The gas triode 69 is normally not conducting so that when its grid potential is changed in a positive direction by the impulse from the anode of tube 64 it will be rendered conductive. When tube 66 is rendered conductive tube 61, whicn was previously conducting, will be rendered nonconductingby reason of the commutating condenser 84. Tube 61, however, will again assume its conducting condition approximately onehalf cycle from transformer 15 (see curve P in Figure 31)). It will be noticed that the gas triode 66 by reason of its circuit arrangement and control electrode potential will respond and become conductive on a wide range of amplitudes and applied signals. The bias on the control electrade of tube 66 is adjusted by means of potentiometer 83 so that any signal above the normal noise level will be effective to render the tube conductive. Variations in amplitude of the applied signals, therefore, will have'no efiect on the re-shaping impulse as developed across cathode resistor 8! unless, of course, there is a complete fading or absence'of the particular impulse. Variations in the arrival time will, however, produce variations in the length of the re- "formed impulses but this variation in length will be taken care of by means to be explained later. The curve represented at M in Figure 3 is the wave form of the re-shaped impulses as derived from the cathode of tube 66 whereas the curve shown at N represents the re-shaped impulses as derived from the cathode of tube 61. Since the tubes 66 and 61 are alternately conductive the two wave forms M and N are complementary, the impulses of the former extending in a positive direction and the impulses of the latter extending in a negative direction. Both of these wave forms are ultimately used in the channel selector tubes.

The received impulses are also used to assure synchronous operation of the receiver with respect to the transmitter. The receiver includes an oscillator circuit similar to the oscillator shown in Figure 1 at the transmitter, the oscillator including tubes 99 and 9|. Tank-circuit including inductance 92 and condenser 92' in general determines the frequency of operation ofthe oscillator but the inductance 92 has connected in parallel therewith a series arrange- (ment of a condenser 93 and an adjustable resistance 94. The adjustable resistance 94 is provided in order that the oscillator frequency may be adjusted within predetermined limits to matchthe operating frequency of the oscillator at *the transmitter.

The received impulses which are available at the output of the receiving system 62 are applied to a low pass filter comprising inductance 96 and condensers 91. Thi low pass filter has a cutoff frequency of about 200 cycles per second for an assumed fundamental control frequencyof cycles per second. A low pass filter is used in order to remove all undesired and unnecessary higher harmonics from the received signals and to permit only the passage of the fundamental component which is applied to the control electrode of tube '98 where it is reversed in polarity and subjected to a certain degree of amplification.

The output from tube 98 is then applied to the control electrode of tube 9| and functions as the synchronizing medium. Although these received impulses are irregular as to their actual existence, the absence of one or even several impulses will not affect the operation of the oscillator at the receiver since initially the oscillator is tuned to substantially the frequency of operation of the oscillator at the transmitter and, furthermore, since a certain flywheel or inertia action is inherent in the'oscillator. It the synchronizing impulses that are applied to both tube 9i and lead from the amplifier tube 98 are not in proper phase with the voltage variations from the anode of tube 90 then the composite output from tube at will cause the oscillator to shift its phase in order to bring it into exact synchronous operation with the oscillator at the transmitter. This synchronous operation will be maintained if the oscillator is reasonably stable and if its natural frequency of operation has originally been adjusted to approximately that of the transmitter oscillator. Thewave form supplied by the receiver oscillator is reprethrough transformers I04 and I05. The phase shifting circuits are of the double resistancecapacitance type which are capable of shifting the phase relationship by nearly 180 electrical degrees. The outputs from the phase shifting circuits are then applied to the control electrodes of tubes I06 and III and these tubes, in turn, supply energy to transformers I5 and 10. The secondary or transformer I6 then supplies energy to the gas diode 12 while the secondary of transformer I5 supplies energy to the gas triode 61 and to the anode of diode I3. Curve Q in Figure 3b shows the energy delivered by the secondary of transformer I6 whereas curve P shows a wave form of the energy supplied by the secondary of transformer 15. The phase relationship of these wave forms may be readily seen by inspecting Figure 3b.

As stated above, tube 66 is rendered conductive when the received impulses are applied thereto. Conduction of the tube 66 renders tube 61 nonconductive and tube 61 re-assumes its conductive condition by reason of the voltage variations represented by curve P beingapplied to the control electrode of the tube. When these voltage variations approach their maximum positive value to initiate operation of tube 61, condenser I4, which is connected to the control electrode of tube 65,

is charged through diode 13 from the same volt-' age variations. The positive charge on this condenser causes the plate current of tube 65 to increase because its grid potential is changed in a positive direction. The higher value of the plate current flowing through the cathode resistance increases the positive bias on tube 64 which produces plate current saturation and renders the amplifying tube 60 ineffective to signals from the radio receiving system 62. This is the blocking action previously referred to which removes any possibility of false operation of the gas triode 66 by reason of noise peaks or disturbances occurring between transmitted intelligence impulses. The tube at is un-blocked or rendered responsive to received impulses when condenser M is discharged through the gas diode E2, the gas diode being rendered conductive due to a fall in the negative half cycle from transformer it (see wave form Q). The instant of unblocking or the instant that tube 04 is rendered responsive to received impulses can be adjusted by means of potentiometer 'll which controls the negative bias normallyapplied to the cathode of gas diode I2 from terminal 18. The blocking action of tube 84 is represented by the curve R. in Figure 3b which in reality shows the wave forms of the voltage variations which are applied to the control electrode of tube 54 from the cathode of tube 00. In reality, therefore, the control electrode of tube 84 has applied thereto a wave form equivalent to curves L and R, combined.

From this portion of the circuit, therefore, it is possible to re-form or re-shape the received impulses and to also render the circuit unresponsive except during intervals when an impulse y be received. The system is completely blocked during the intervals between signal impulse reception so that false operation by reason of electrical disturbances or noise is imp ssible.

The multiplex receiving system also includes an electron distributor or switching device similar to the corresponding elements in the trans mitter. In the receiver the tubes that are included in the electronic switching device are II2, I I3 and H4. These three tubes operate in a manner substantially identical to the operation of the corresponding tubes I6, I1 and I8 in the transmitter and, accordingly, any further description of that operation is believed to be unnecessary. Controlling impulses are derived from the cathodes of tubes H2, H3 and H4 and the wave form of these controlling impulses is represented by curves S, W and A of Figure 3, respectively.

The re-shaped impulses, as represented by curve M, are still representative of the composite signals from the three sources at the transmitter and electronic means are provided in the receiver for properly allocating the received impulses to their respective channels. Selection and assignment of the proper re-shaping impulses to channel number one, for example, is accomplished by the operation of tubes I29 and I30. .The general operation of these tubes is similar to that of the channel switching tubes 21, 28 and 28 in the transmitter. Each of tubes I29 and I30 include two separate control electrodes and both of the control electrodes of each of the tubes are biased so that no plate current flows in a tube unless control signals or impulses are applied in a positive direction to both of the control electrodes of that tube simultaneously. All of the re-shaped impulses from the gas trl-' ode 56 (wave form M in Figure 3b) are applied to the second control electrode or grid number 3 of tube I29. Grid number 1, or the first control electrode of tube I29 has applied thereto the selection impulses from the cathode of tube II2 (curve S in Figure 3b). All of these impulses extend in a positive direction. I It will be observed, however, that the selection impulses from the cathode of tube IIZ are applied to the first control electrode of tube I29 through condenser B35. Whenever the potential of grid number 3 or the second control electrode of tube I29 is driven in a positive direction due to the presence or an incoming signal impulse and whenever the first control electrode is driven positive by the wave form from the electron distributor, a sharp pulse of plate current is permitted to flow through the tube to cause a voltage drop in the cathode resistance I3I of the tube. The wave form of this voltage drop, after passing through condenser IE3, is represented by the curve T in Figure 3b. For the purpose of identification, these impulses will be called the marking impulses. From an inspection of curves M and S in Figure 3b it may be seen that a marking impulse shown by curve '1 is produced when are-shaped incoming signal impulse coincides with a channel number one selection impulse. These marking impulses are a reproduction of the original driving impulses of channel number one as may be seen by comparing curves T and C. Variations in the arrival time of the incoming impulses will have no effect on these marking impulses at the receiver provided the application of the selection impulses I to the tube I29 occurs slightly later than the latest signal impulse that might be received by reason of multi-path reception. It will be observed, therefore, that the selection impulses represented by curve S 'are slightly delayed with respect to the reshaped signal impulses represented by curve M. Naturally, the marking impulses are likewise delayed since tube I29 cannot become conductive until both of its control electrodes are drivenin a positive direction. The leading edge of the marking impulses shown in curve T, therefore, coincide with the leadingedge of the selection impulses represented by curve S.

The so-called marking impulses represented by the curve T which are produced in the cathode output circuit of tube I29 cannot be directly utilized since they indicate only the beginning of a dot or a dash or they may represent the continuation of a dash. In order toascertain whether or not the marking impulses represent dots or dashes, tube I30 is provided and this tube, like tube I 29, has two control electrodes. The second control electrode or grid number 3 of tube I30 has applied thereto voltage variations or signal control impulses from the cathode of tube 61 (see curve N) while the first control electrode or grid number one has applied thereto selection impulses from the cathode of tube II2 by way of conductor I26 and condenser I35.

Both of the control electrodes are normally negatively biased to prevent conduction of the tube and the tube is permitted to conduct only when both controlelectrodes are simultaneously driven in a -positive direction. The wave form applied to the second control electrode is the inverse of the re-shaping signal impulses shown at curve inspection of Figure. 31), it may be seen that the impulses derived from the cathode output of tube I30, which for thepurpose of identification will be called spacing impulses, are initiated by the simultaneous occurrence of both the selection impulses represented by curve S and the inverse reshaped impulses represented by curve N.

Tube I29. therefore, produces the marking impulses for channel number one and tube I 30 supplies the spacing impulses for that same channel.

A locking circuit is used to convert these two groups of impulses, namely, the marking and spacing impulses, into dots and dashes corresponding to the dots and dashes represented by curve A. In channel number one, for example, the marking impulses from the cathode resistor I3I of tube I29 are applied by way of condenser I53 to the control electrode of gas triode I44.

asoacco 1 Likewise, the spacing impulses from tlie cathode resistance I32 of tube I30 are supplied by way of condenser I54 to the control electrode of gas triode I45. The anodes of these two gas triodes are interconnected by commutating condenser I48. 'I'hetwo tubes are alternately operated and both the marking and spacing impulses extend in a positive direction. Tube I44 is therefore rendered conductive (and tube I45 simultaneously non-conductive) whenever a marking impulse is applied to its control electrode and likewise tube I44 is rendered non-conductive and tube I55 conductive when a spacing impulse is applied to the control electrode of the latter tube. The marking and spacing impulses. therefore, control the periods of conduction and non-conduction of from the standard auto head I in the transmitter have been reproduced and are available on conductor I55 in that same form. These dot and dash voltage variations may then be directly applied to a standard ink recorder or printer such as represented at I60.

Channel number two is also provided with tubes I36 and I31 for producing marking and spacing control impulses which are represented by the curves X and Y in Figure 3b and these marking and spacing control impulses are applied to the commutating tubes I56 and I 51 so that voltage variations of dot and dash wave form are produced for channel number 2 as represented by the curve Z in Figure 3b. A similar operation occurs with respect to channel number 3 wherein marking and spacing control impulses are produced as represented by curves B and C in Figure 3b and these exercise control over tubes I64 and I65 to produce voltage variations of dot and dash wave form for channel number 3 as represented by curve D in Figure 3b.

From the above description it is believed that a complete conception of the advantages of the present invention will be apparent to those skilled in the art of radio telegraph multiplex operation. The description given above implies that the radio transmitter be relatively closely associated with the source of signals and, further, that the printers or recorders be relativelyclosely associated with the receiving apparatus. In actual commercial practice, however, this would not be the case in many instances. mitter or the receiving apparatus are displaced a relatively great distance from the source of signals or the printers, it is to be understood that wire lines or other appropriate connections may be used. Accordingly, conductor 56 in the transmitter may have some appreciable length and likewise conductors I55, I62 and I68 may also have some considerable length.

Although the complete radio telegraph multiplex system is described herein in more or less detail it is to be understood that various alterations and modifications of the present invention may become apparent to those skilled in the art and it is desirable that any and all such modifications and alterations be consideredwithin the Where the trans-' purview of the present invention except as limited by the hereinafter appended claims. A I claim:

1. A communicating system comprising means for transmitting impulses of uniform intensity and duration, the impulses being separated by predetermined time intervals or multiple integers of the predetermined time intervals and occurring always at a fixed phase relation, a receiver, said receiver having means for reshaping the received impulses, separate means for producing marking and spacing control impulses from the reshaped impulses, and means responsive to the marking and spacing control impulses to produce a series of impulses having substantially uniform intensity and of varying length.

2. A telegraph transmitting system comprising means for transmitting a series of impulses of uniform intensity and duration, the impulses being separated by time intervals corresponding to integer amounts of a predetermined time interval and occurring always at a flxed phase relation, a receiver, said receiver having means for reshaping the received impulses, means for renderlng the receiving means ineffective between received impulses, means for producing marking and spacing control impulses from the reshaped impulses, and means responsive to the marking and spacing control impulses to produce a series of uniform intensity impulses having time durations corresponding to the spacing of the received impulses.

3. A communication receiving system comprising means for transmitting a series of impulses of uniform intensity and time duration, the impulses being separated by predetermined time intervals or multiple integers of the predetermined time intervals in accordance with the intelligence transmitted, a receiver, said receiver having means for reshaping-the received impulses, means for rendering the receiving means ineffective between impulse receiving intervals. means for producing marking and spacing control impulses from the reshaped impulses, and means responsive to the marking and spacing control impulses to produce a series of impulses having substantially uniform intensity and of a time duration determined by the separation of the received impulses.

4. A communicating system comprising means for producing a series of signals of substantially uniform intensity and of varying length in accordance with the information to be transmitted, means for converting the signals into a series of impulses of substantially uniform intensity and duration, and means for spac ng the impulses by an integer number of fixed time intervals in accordance with the lengths of the original series of signals, means for transmitting the converted impulses to a remotely located receiver, means responsive to the received impulses for producing marking and spacing control impulses, and means responsive to the marking and spacing control impulses for reproducing a series of signals of substantially uniform intensity and of varying length corresponding to the original series of signals.

5. A telegraph communicating system comprising a source of dot and dash signals of substantially uniform intensity and of varying length in accordance with the information to be transmitted, means for converting the signals into a series of impulses of substantially uniform intensity and duration, and means for spacing the impulses by an integer number of fixed time intervals in accordance with the lengths of the original dot and dash signals, means for transmitting the converted impulses to a remotely located receiver, means responsive to the received impulses to produce reshaped impulses, means responsive to the reshaped impulses for producing marking and spacing control impulses, and means responsive to the marking and spacing control impulses for reproducing a series of dot and dash signals ofsubstantially uniform intensity and corresponding in length to th original dot and dash signals.

6. A communicating system comprising means for producing a series of signals of substantially uniform intensity and of varying time duration and spacing in accordance with the information to be transmitted, means for converting the signals into a series of impulses of substantially uniform intensity and time duration, and means for spacing the impulses by an integer number of fixed time intervals in accordance with the duration and spacing of the original series of producing marking and spacing control impulses,

and means responsive to the marking and spacing control impulses for reproducing a series of signals of substantially uniform intensity and of varying time duration and spacing corresponding to the original series of signals.

7. A multiplex communicating system comprising a plurality of sources of dot and dash signals representative of the subject matter to be transmitted, a plurality of electron discharge tubes corresponding to each of the plurality of signal sources and individually energized thereby, electron switching means for rendering the electron discharge tubes individually and sequentially conducting at a predetermined rate to produce a single series of impulses occurring at said predetermined rate, all of the impulses being appreciably shorter than the time interval of a dot andbeing of substantially constant amplitude and time duration, the presence or absence of the impulses being determined inaccordance with the dots and dashes originating from the plurality oi separate sources, means for transmitting the produced impulses to a remotely located receiver, a wave reshaper for reshaping the received impulses to rectangular wave impulses having a ration less than that of a dot, a plurality of electron discharge paths. means for applying the reshaped impulses to all of the electron discharge paths to condition the paths for conduction, electron switching means operating in synchronism with the electron switching means in the transmitter for individually and sequentiallyrenden ing the electron discharge paths conducting to produce marking and spacing impulses and to segregate the received impulse into channels of said sources a series of impulses occurring at a predetermined rate, all of said produced impulses 'being (if substantially constant amplitude and time duration and the presence or absence of the impulses being determined in accordance with the dots and dashes originating from the plurality of separate sources, means for transmitting the produced impulses to a remotely located vidually and sequentially rendering the electron discharge paths conducting to produce marking and spacing impulses and to segregate the .received impulses into channels corresponding to the original plurality of sources of signals, and

means responsive to the produced marking and spacing impulses for producing a plurality of series of dot and dash signals corresponding to the original series of dot and dash signals.

9. In a multiplex communication system having a plurality 01 channels and a transmission medium, an electronic distributor for sequentialiy coupling said channels to said-transmission medium, said electron distributor comprising a plurality of electron paths, each path including a cathode, a control electrode and an anode, a separate impedance-for connecting each cathode to a point of fixed potential, means including a separate impedance for maintaining each anode positive with respect to its associated-cathode, means for connecting the cathode of each path to the control electrode of another path in a predetermined order, means including an electron storage device for connecting the anode of each path to the anode of another path in the said predetermined order, means fo simultaneously applying a series of control impulses to the control electrodes of all of said paths whereby the plurality of electron paths, each path including a cathode, a control electrode and an anode, a separate cathode resistance for connecting each cathode to a point of fixed potential, means including a separate anode resistance for maintaining each anode positive with respect to its associated cathode, means including an impedance for connecting the cathode of each path to the control electrode of another path in a predetermined order, means including a condenser for connecting the anode of each path to the anode of anothen path in the said predetermined order, means for applying a series of regularly recurring control impulses to the control elec trodes of all of said paths simultaneously whereby the electron paths are rendered individually and sequentially conducting in the said predetermined order in response to the application of control impulses to the control electrodes thereaseaaoo of, and connections from corresponding electrodes oi said paths for rendering the channels of said system sequentially operative.

11. In a multiplex communication system having a plurality of channels and a transmission medium, an electronic distributor for sequential- 1y coupling said channels to said transmission medium, said electron switching arrangement in cluding a plurality of ionizable electron discharge paths each including a cathode, a control electrode and an anode, means including individual resistances for connecting each cathode to a point of fixed potential, means including a plurality of individual resistances for normally maintaining each anode positive with respect to its associated cathode, means including a resistance for connecting the cathode of each path to the control electrode of another path in a predetermined order, means including a condenser for connecting the anode of each path to th anode of another path in the said predetermined order,

'means for normally maintaining the control electrodes negative with respect to their associated cathodes, and means for simultaneously applying control impulses to all of said control electrodes whereby said electron discharge paths may be rendered individually and sequentially conduct= ing in the said predetermined order, and connections from corresponding electrodes of said paths for rendering the channels of said system sequentially operative.

12. In a multiplex communication system having a plurality of channels and a transmission medium, an electronic distributor for sequentially coupling said channels to said transmission medium; said electron switching arrangement ineluding a plurality of ionizable electron discharge paths each including a cathode, a control electrode and an anode, means including individual resistances for connecting each cathode to the negative terminal of a source of potential, means including a plurality of individual resistances for connecting each anode to the positive teral of the source of potential, means including an impedance for connecting the cathode of each path to the control electrode of another path in a predetermined order, means including an electron storage element for connecting the anode of each path to the anode of another path in said predetermined order, means for normally maintaining the control electrodes negative with respect to their associated cathodes, and means for applying positive control impulses to all of said control electrodes simultaneously whereby said electron discharge path may be rendered in-- dividually and sequentially conducting in the predetermined order, and connections from corresponding electrodes of said paths for rendering the channels oi said system sequentially operative.

13. A multiplex transmitting system comprising a plurality of individual sources of signals of substantially uniform intensity and of varying time duration and spacing representative of individual messages to be transmitted, said signals extending in a positive direction, a plurality of electron discharge paths associated with and corresponding to the plurality of sources of signals, each path including an electrode structure having a cathode, a first and second control electrode and an anode, means including a loadimpedance for connecting each cathode to a point of fixed potential, means for maintaining the anodes positive with respect to their associated cathode, means for normally maintaining both the first and second control electrodes 01 each electron discharge path sufflclently negative in normally preclude conduction of said electron discharge paths, means for applying the signals from each or the plurality of sources to one of the control electrodesof its associated electron discharge path, an electron switching device, connections including differentlator circuits between said switchin device and said electron discharge paths for supplying sharp positive control pulses to the other control electrode of each of the electron discharge paths for individually and sequentially preparing said electron discharge paths for conduction, whereby each electron discharge path is rendered conductive upon the presence of signals from its associated source when sharp positive control impulses are simultaneously applied thereto in order to produce voltage variations at its cathode, said sharp positive impulses being appreciably shorter in time duration than a dot, and means for combining the voltage variations produced at the cathodes of the electron discharge paths whereby a series of impulses will be produced having a frequency corresponding to the frequency of operation of the electron switching device, said series of impulses being of substantially constant amplitude and constant time duration but of a length appreciably shorter than a dot, and said last impulses being spaced by an integer number of fixed time intervals in accordance with the duration and spacing of the original sources.

14. A multiplex telegraph transmitting system comprising three individual sources of dot and dash signals representative of three individual messages to be transmitted, said dot and dash signals extending in a positive direction, three electron discharge paths associated with and corresponding to the three sources of signals. each path including a cathode, a pair of control electrodes and an anode, means including an impedance for individually connecting each cathode to a point of fixed potential, means for maintaining the anodes positive with respect to their associated cathodes, means for normally maintaining both of the control electrodes of each electron discharge path suficiently negative to normally preclude conduction of said electron discharge paths, means for individually applying the signals from each of the three sources to one of the control electrodes of its associated electron discharge path, an electron switching device for supplying positive control impulses to the other control electrode of each of the electron discharge paths for individually and sequentially preparing each electron discharge path for conduction, whereby each electron dischargespath. is rendered conductive only when both of its control electrodes are driven in a positive direction to produce voltage variations at its cathode, and means in.- cluding a self restoring trigger circuit having only one degree of electrical stability for combinin the voltage variations prod ced at the cathodes of the electron discharge pa hs whereby a single series of impulses will be produced, said impulses being of substantially constant amplitude and constant time duration but having a length shorter than a dot, and said impulses being spaced at time intervals where n is an integer and t is a predetermined time interval as determined by the dot and dash signals from the three signal sources. I

15. A multiplex signal regenerator receiving system wherein impulses shorter than a dot but of substantially constant amplitude and time duration are received, and wherein the impulses are separated by time intervals at where n is an integer and where t is a predetermined time interval, the spacing of the impulses being an indication or the intelligence received, comprising an oscillator at said receiver, means for controlling the frequency of operation of the oscillator by the received impulses, an electron switching device responsive to the voltage variations supplied by the oscillator, a plurality 01 pairs of electron discharge paths, each electron discharge path including a first and second control electrode, means for connecting the electron switching .device to the first control electrode of each of the pairs of electron discharge paths to individually and sequentially prepare each pair of paths for conduction, means for applying the received impulses to the second control electrode of each of the electron discharge paths, whereby marking and spacing control impulses will be produced by each pair of discharge paths in accordance with the received impulses, and whereby the impulses received by the multiplex system will be properly segregated into a plurality of channels, blocking means coupled to the output of said receiver for preventing noise occurring between received impulses from affecting said paths, and means responsive to the produced marking and spacing control impulses for increasing the lengths of said impulses to thereby produce a plurality of series of dot and dash signals of substantially uniform intensity.

16. A multiplex telegraph receiving system wherein impulses of substantially constant amplitude and time duration are received, and wherein the impulses are separated by predetermined time intervals or multiple integers of the predetermined time interval, the spacing of the impulses being an indication of the intelligence received representing a plurality of messages, comprising an oscillator at said receiver, means for controlling the frequency of operation of the oscillator by the received impulses, means including said oscillator coupled to the output of said receiver for rendering the receiver ou put ineffective during time intervals between received impulses, an electron switching device responsive to the voltage variations supplied by the oscillator, a plurality of pairs of electron discharge paths, each electron discharge path including a first and second control electrode, means for connecting the electron switching device to the first control electrode of each of the pairs of electron discharge paths to individually and sequentially prepare each pair of paths for conduction. means for applying the received impulses to the second control electrode of each of the electron discharge paths, whereby a separate series of marking and spacing control impulses will be produced by each pair of discharge paths in accordance with the received impulses. and whereby the impulses received by the multiplex system will be properly segregated into a plurality of channels, and means responsive to the produced marking and spacing control impulses for producing a plurality of series of dot and dash signals of substantially uniform intensity.

1'7. A communication receiving system includina means for receivin a series of impulses of substant ally constant intensity and time dura-

Images