US2438908A - Pulse code modulation communication system - Google Patents

Description

April 6, 1948.

w. M. GOODALL PULSE CODE MODULATION COMMUNICATION SYSTEM Filed May 10, 1945 11 Sheets-Sheet 1 //4 //3 FIG] '5 I I CODE CONTROL EXPONEN TIAL CONTROL ELEMENT PULSE CIRCUIT mmva car GENERATOR DELAY MASTER NETWORK 03C.

TERMINAL SAMPLING H GOMMR/NG TRANS. 4 EQUIPMENT CIRCUIT CIRCUIT C I26 1 MON/TOR CIRCUIT /30 v /s/ 132 ms 134 TERMINAL SAMPLING COMPARING k RADIO 1 EQUIPMENT 'CIRCUIT CIRCUIT TRANS I C MON/TOR CIRCUIT i v I /4o l l M2 I l I43 144 9 Z: TERMINAL DECODING RECEIVING PHASE EQUIPMENT CIRCUIT CIRCUIT CONTROL Q: MON/TOR CIRCUIT /50/ /5/ I52 /s4 15a /4 I? R U/T CONTROL R MONITOR CIRCUIT lNl/EN r01? F F/G. 5 FIG 5 W M. GOODALL BY HG HG FIG. .2 FIG. 6 F/G. 9 WWW 2 FIG. 4 FIG. 7 F762 /0 A T 7' ORNE Y P 1943- w. M. GOODALL 2,438,908

PULSE CODE MODULATION COMMUNICATION SYSTEM Filed May 10, 1945 ll Sheets-Sheet-' 2 FIG. 2 2/0) 2) 2/3 2/4) 5 1 Z co/vr/aoL c005 gg V PULSE ELEMENT 52557;

GENERATOR TIMING ccr.

i V DELAY NETWORK PHASE RECEIVING DECODING TERMINAL CONTROL CIRCUIT CIRCUIT EQUIPMENT MONITOR. :j 8 CIRCUIT- -\v 240/0 PHASE '1 DECODING TERMINAL RECEIVER CONTROL CIRCUIT EQUIPMENT MON/TOR CIRCUIT Z TRANSMISSION COMPARING SAMPLING TERMINAL cmcu/r CIRCUIT CIRCUIT EQUIPMENT 226 MON/TOR 3 I 5 cmcu/r 225 2.40/0 4 COMPARING sAuPL ING TERMINAL 6:

TRANSMITTER cmcu/r cmcu/r EQUIPMENT p-p v I 236 H an/70R :3 CIRCUIT INVEN 7' 0/? W MGOODALL BY 40M 2W.

rrow/5v April 1943- M w. M. GOODALL 2,438,908

PULSE CODE MODULATION COMMUNICATION SYSTEM Filed May 10, 1945 ll Sheets-Sheet 3 INVENTOR W M GOODALL BYM.

ATTORNEY April 6, 1948.

PULSE CODE MODULATION COMMUNICATION SYSTEM W. M. GOODALL Filed May 10, 1945 11 Sheets-Sheet 4 [ma-w 11M M.

A r TORNEV April 6, 1948.

FIG. 5

W. M. GOO'DALL PULSE CODE MODULATION COMMUNICATION SYSTEM Filed May 10, 1945 11 Sheets-Sheet 5 Byw h b ATTORNEY April 6, 1948.

Y W. M. GOODALL PULSE CODE MODULATION COMMUNICATION SYSTEM Filed May 10, 1945.

ll Sheets-Sheet 6 lNl ENTOR W M. GOODALL By 14M iww.

A NORA/Ev April 6, 1948. w. M. GOODALL PULSE CODE MODULATION COMMUNICATION SYSTEM Filed May 10, 1945 11 Sheets-Sheet 7 FIG. 7

a) K g //VVE/VTOP W. M. GOUDALL mm 2%.

A TTOIQA/EV l1 Sheets-Sheet 8 L I *H WV V W. MI GOODALL Filed May 10, 1945 April 6, .1948.

v PULSE CODE MODULATION COMMUNICATION SYSTEM Alli]? m. 6t 4 ma 3% INVENTOR W M GOODALL By W z! J." l a ATTORNEY April 6, 1948. w. M. GOODALL PULSE CODE MODULATION COMMUNICATION SYSTEM Fil ed May 10, 1945 11 Sheets-Sheet 9 INVENTIOR W M. GOODALL 1 Qmk m at April 6, 1948. w. M. GOODALL 2,438,908

PULSE CODE MODULATION COMMUNICATION SYSTEM Filed May 10, 1945 ll Sheets-Sheet 10 o 0 Q i g O5 Q1 Q as INVENTOR n. M. GOODALL A TTORNEY April 6, 1948.

W. M. GOODALL PULSE CODE MODULATION COMMUNICATION SYSTEM 1 Filed May 10, 1945 I302 was FIG. I3 Q .A/YAA VV I I l i ll Sheets-Sheet ll 433 I I434 I436 INVENTOR W M GOOD/ILL zwumzw.

ATTORNEY Patented Apr. 6, 1948 PULSE CODE MODULATION COMMUNICA- TION SYSTEM William M. GoodalLOakhurst, N. .1, assignor to Bell Telephone Laboratories, Incorporated, New York, N. Y., a corporation of New York Application May 10, 1945, Serial No. 592,958

13 Claims. (01. 179-15) This invention relates to communication systems for the transmission of complex wave forms of the type encountered in speech, music, sound, mechanical vibrations and picture transmission by means of code groups of a uniform number of signaling impulses, each of which comprises any one of a plurality of different signaling conditions or types of signaling conditions transmitted in succession.

This invention is an improvement upon a similar system described in my patent copending application Serial No. 554,495, filed September 16, 1944.

The object of the present invention is to provide an improved and simplified apparatus and method of generating a plurality of series of permutation code groups of pulses representing a complex wave for transmission over an electrical transmission path in order to secure a greater signal-to-noise ratio for the reconstructed signal wave at the receiving station. In other words, the information transmitted by permutation code pulses may be transmitted over a very noisy transmission system without adding appreciable amounts of noise to the received signal. Another object of this invention is to provide methods and systems for the rapid representation of the instantaneous amplitudes of the complex waves by permutation code groups of pulses in an improved and simplified manner. A feature of this invention relates to coding and decoding equipment which employs a continuously varying exponential wave both for coding pulses at the transmitting end and decoding pulses at the receiving end, instead of a step Wave form as employed in the above-identified application.

Another feature of this invention relates to methods of operation and equipment for controlling the character of the pulses of each of the permutation code groups of pulses in accordance with the instantaneous amplitude of a complex wave by means of a single electron conduction device which in the exemplary. embodiment set forth herein is exemplified by a multielement vacuum tube frequently called a pentode tube.

Another object of the present invention is to provide receiving equipment which employs a continuously varying exponential wave for decoding permutation code groups of pulses and reconstructing a complex wave therefrom.

Another feature of this invention relates to equipment for controlling the length of pulses,

ble of transmitting the highest radio frequencies. Another object of the present invention is to provide apparatus for modifying the input of complex waves in such a way that they will be more suitable for controlling pulse generating equipment in accordance with the present invention so that a larger portion of the available amplitude range of the complex wave form is employed and at the same time permitting a larger amplitude range of complex wave forms to be employed without overloading the coding equipment. T

Another feature of this invention relates to the use of an instantaneous amplitude compressing device for compressing the instantaneous amplitudes of the complex wave forms applied to the coding equipment prior to the transmission of these complex waves. Another feature of the invention relates to receiving equipment for reconstructing complex wave forms represented by permutation code pulses in which the permutation code pulses represent a compressed wave form at the transmitting station without the use of any expanding equipment at the receiving station to compensate for the distortions of the compressing equipment at the transmitting station. Applicant has discovered that many of the advantages of compression can be obtained by compressing the wave before it is applied to the sampling and coding equipment and then a complex wave regenerated in suitable form at the receiving station from the received code signals without the use of corresponding expansion equipmentand without objectionablev distortion which is usually manifest by undesirable added frequency compoments when signaling waves are instantaneously compressed.

Briefly, in accordance with the present invention equipment is provided for generating synchronizing or control pulses from which a plurality or group of timing pulses are generated. The synchronizing or control pulses are employed to control the generation of a continuously varying exponential wave and also to control the sampling of the instantaneous amplitude of the complex wave form which it is desired to trans-'- mit to a distant station. Prior to this sampling the complex wave is compressed by means of one or more non-linear circuit elements. pression of the wave prior to its being sampled provides a two-fold advantage. First on weak signals it causes the range of instantaneous amplitudes applied to the coding to vary through The com greater amplitude and second when strong signals are applied it tends to prevent the strong signals from overloading the coding equipment.

The magnitude of the samples of the compressed wave together with the continuously varying exponential wave and the output signal-s are applied toa coding tube which" tubei combines the values .enumerated above and applies themto a comparing circuit which compares the output of the above-mentioned coding tube with a continuously varying exponential wave for .determining the character of each of the signaling pulses transmitted over this system. At the receiving station the received pu-lses, -together with a similar exponential wave is applied to-other tubes for causing the reconstruction or the complex wave.

Other means is provided for sampling or deriving an electrical quantity which is af-unction of the amplitude of the complex wave after said complex wavehas:beencompressed. The magnitude ofthe electrical quantity. or samplels then compared with the magnitude of the continuously varying exponential wave and if the'sum of these two waves exceeds a predetermined value pulses of a predeterminedkind orcharacter arextransmitted. Duringor after the transmission of the'pulses in question the magnitude of the sample isreducedby an amount equal .to the-portion magnitude of the sample represented by the pulse transmitted. Thereafter the remaining .portionof the sample is again compared with the continuously varying exponential-wave andcthe above action repeated until all of the pulses in a given permutation code .groupor seriesxhas been transmitted, at whichtime another sample is :obtained, which-in turn is similarly compared with another continuously varying exponential wave similar to the first exponential wavedescribed above,-so that a succeeding series'ofppulses is likewise generated and=transmitted. The amount subtracted'from the electrical quantity .due to thetransmission of .the various pulses .is the total maximum :possible value multiplied by where A is a constant greater than 1 and usually 2 and n isthe ordinal or digital numberof the pulse transmitted.

.At the receiving station a .control pulse ,generator, and .a continuously varyingexponential wavegenerator are provided which are similar to .the respective circuits at the transmitting station. In addition, decodingequipment isprovided in which the received pulses are employed .tosynthesize and reconstruct a complex wave having a wave form similar to the waveform of the .complex wave sampled at the receiving station.

Novel features of this invention which arebclievedtobe characteristic thereof are set forth with particularity in the claims appended thereto.

The invention itself, however,.both as to ,its organization and method of operation, together with. other objects and features thereof, may be .best understood from the following description of an exemplary system embodying the inventionwhen saiddescription is read with reference .to.-the accompanying drawings, in which:

-Figs. 1 and 2 show in diagrammatic form the various elements of an exemplary system embodying .thepresent invention and the mann .common to all the channels at a receiving terminal; in case ,of two-way communication these figures maybe common both to all of the transmitting channels and to all of the receiving .:channels. at the respective stations;

1Fig..11 shows the manner in which Figs. 1 and 2 are positioned adjacent one another;

Fig.12 shows-the manner in which Figs. 3 to 10, inclusive, are positioned adjacent one anotheryand Figs. 13 and 14 show graphs of the various current or potential wave forms at various places in the system.

Figs. land 2 when arranged as shown in Fig. 11 show in diagrammatic form the various circuit elements and apparatus and the manner in which they cooperate one with another to form an exemplary system embodying the present invention. Fig. 1 shows the apparatus and circuits at one station while Fig. 2 shows similar apparatus and circuits at a second station. As shown in Figs. 1 and 2, two transmission paths or channels are provided between the two stations in each direction. The transmission paths as well as the other interconnections between the various elements which cooperate to form the exemplary system described herein are shown as .single lines in Figs. 1 and 2. Persons skilled in the .art will readily understand that the single lines interconnecting the different elements in Figs. 1 and 2 represent the desired or necessary number of conductors which extend between the respective elements of the system. It will be obvious to persons skilled in the art that additional channels in either or both directions may be provided between each of these sta tions. It will also be obvious to persons skilled in the art that additional stations or terminals may .be provided and similar transmission paths orchannels extended between any or all of the various stations or terminals.

Furthermore, the signals transmitted from either .of the stations of Fig. 1 or 2 may be receivedby other receiving devices and systems, as for example, by the system disclosed in my copending application Serial No. 554,495. Likewise the receiving apparatus of Fig. 1 or 2 may receive signals from other types of transmitting systems, as for example, from the transmitting stations shown in my above-identified copending application which is incorporated herein as if fullyset forth herein.

As shown in Figs. 1 and 2, lines 6 and 9 represent twocommunication paths between the two stations. Signals are transmitted over line 6 from the station shown in Fig. 1 to the station shown in Fig.2. Signals are transmitted over line 9 from the station shown in Fig. 2 to the station shown in Fig. 1. These lines 6 and 9 represent communication paths between the two stations which may beof any suitable type for the transmission or conveyance of electrical signaling impulses of proper frequency range. These paths may be usedindependently of each other or in pairs to provide two-way communication circuits. These paths are illustrated in Figs. 1 and 2 by a single tween these stations.

a line and persons skilled in the art will readily understand that these lines represent any and all types of communication paths and combinations thereof including open wire lines, cable conductors, testing pairs, channels of carrier current systems, coaxial lines and cables, wave guides, radio channels and may also include time division multiplex channels. Furthermore, these transmission paths may include suitable amplifiers, filters, predistortion networks, equalizing networks, gain control networks and circuits, phase control circuits and networks, repeaters, repeater stations, as Well as signal operated switching apparatus as well as other signal operated devices which control the direction of transmission. In addition, when the transmission path is over or includes a coaxial cable, the transmis sion system may include any or all of the features disclosed in or referred to in United States Patents 2,343,568, granted to L. W. Morrison, Jr., on March 7, 1944; 2,212,240, granted August 20, 1940, to Lalande et al.; and 2,095,360, granted October 12, 1937, to Green. These features may also be employed for other types of conductors when desired. Any or all of the transmission channels between any or all of the stations may also include regenerative repeaters as well as other equipment associated with each of the respective types of transmission paths suitable for terminating and interconnecting the various types of transmission lines so that they will each cooperate with the adjacent sections to form a continuous transmission channel between the two stations. In addition to lines 6 and 9, two radio channels are shown in Figs. 1 and 2. Transmitting antenna I and receiving antenna 8 represent one of the radio channels while wave guide I3 and horn or radiator I2 and horn II and wave guide I4 illustrate a second radio channel be- A typical example of a radio channel including radio relay repeater stations suitable for the transmission of pulses of the type employed in the exemplary system described herein is described in detail in a series of papers published in the Proceedings of the Institute of Radio Engineers for November 1934, entitled An experimental television system, part 1 of which is entitled Introduction, which is by E. W. Engstrom, part 2 of which is entitled The transmitters, and is. by Kell, Bedford and Trainer, part 3 of which is entitled The receivers, and is by Holmes, Carlson and Tolson, and part 4 of which is entitled "The radio relay link for television signals, and is by Young. Another suitable type of radio amplifier, repeater and transmitter is described in detail in an article entitled High gain amplifier for 150 megacycles by Rodwin and Klenk, published in the Proceedings of the Institute of Radio Engineers for June 1940. Typical wave guide and horn or radiator structures are described in United States Patent 2,206,923, granted to Southworth July 9, 1940. The foregoing publications and patents are hereby made a part of the present application to the same extent as though set forth herein in full.

In addition, a synchronizing path or channel 5 is shown extending between the two stations shown in Figs. 1 and 2. This control path or channel may be similar to other transmission paths between the stations. Furthermore, if it is so desired, the synchronizing signals or the control frequency may be transmitted over one or more of the other transmission paths extending between the stations. Inasmuch as there are numerous types of synchronizing apparatus in the prior art which will operate over the same transmission paths as employed for the transmission of communication systems and since the operation of this type of equipment is well known and understood by persons skilled in the art, it is considered unnecessary to further expand the present disclosure to show details of a typical system of this type. It is understood, of course, that such equipment as will cooperate with the various circuits of the present invention may be provided when it is so desired.

Each of the stations is provided with certain control equipment which may be common to all the circuits terminating at that station or which may be common to a plurality of the circuits terminating thereat. Of course, this common equip-- ment may be provided for each of the individual circuits, if so desired, as is well understood by persons skilled in the art. However, in the system shown in Figs. 1 and 2, control circuits and equipment common to all the channels shown terminating at each of the respective stations are shown at the top of each of these figures.

The common equipment at station I comprises a control oscillator I In which may be of any suitable type as, for example, the type described in detail in one or more of the following patents: 1,476,721, Martin, December 11, 1923; 1,660,389, Matte, February 28, 1928; 1,684,455, Nyquist, September 18, 1928; and 1,740,491, Aifel, December 24, 1929, the disclosures of which are. hereby made a part of the present application as if fully included herein.

The output of the control oscillator is coupled to control a control pulse generator I I I. The output of this generator extends to receiving and monitoring equipment as shown in the drawing and also to delay network H2. The delay network H2 may be of any suitable type of delay network as, for example, one or more sections of one or more of the types disclosed in United States Patent 1,770,422, granted July 15, 1930, to Nyquist. The output of the delay network H2 extends to sampling circuits I22 and I32 and also to a code element timing circuit I I3. The output of the code element timing circuit extends to the coding and comparing circuits I23 and I33. The output of the delay network I I2 also extends to the exponential circuit II4. Similar common equipment comprising a control oscillator 2I9, control pulse generator 2| I delay network 2I2, code element timing circuit 2I3 and exponential circuit H4 is provided at the station shown in Fig. 2.

In addition to the control oscillators H0 and 2H) at each of the control stations, a master oscillator I0 is shown in Fig. 1. This master oscillator may be located at either of the stations and may replace the control oscillator 210 or I II) at either of these stations. However, the master oscillator is frequently located at some central point and provides a control frequency for an entire nation-wide system or for some smaller portion or sections of a large system. Typical oscillators for standard frequency systems suitable for use as the master oscillator and also for the control oscillators are disclosed in United States Patents 1,788,533, Marrison, January 13, 1931; 1,931,873; Marrison, October 24, 1933; 2,087,326, Marrison, July 10, 1933; 2,163,403, Meacham, June 10, 1939; and 2,275,452, Meacham. March 10, 1942. All of the patents referred to above are hereby made a part of the present application as if fully included herein.

The sources of signals I20 and I 30 shown in Fig.

l and also 22!] and 230 shown in Fig. 2 are illustrated in the drawing by microphones. It will be apparent to persons skilled in the art that other suitable sources of signals or complex waves may be employed such as telegraph systems, picture transmission systems, television systems, mechanical vibration pick-ups, photoelectric devices, piezoelectric devices, etc. Each of these sources is connected to the respective terminal equipments I2I, I3I, HI and 23!. This terminal equipment represents all of the equipment connected between the sources or microphones and the transmission circuits and equipment for coding the signals to be transmitted as will be described hereinafter. This terminal equipment may include one or more of the following typical types of apparatus including amplifiers, transmission lines, switching equipment of any suitable type such as manually controlled switching equipment at a manual central oilice, automatic or machine or dial switching equipment such as employed in automatic central oflices as well as various types of transmission apparatus including phase control apparatus, equalizing apparatus, voice controlled switching apparatus, etc. This terminal equipment may also include toll line facilities including either or both two and four-wire circuits, i. e., a single two-way or to and fro communication path or two one-way communication paths operated in opposite directions and toll line switching equipment. When the terminal equipment IIZI, I3I, 22! or 22H includes switching equipment, the transmission circuits described hereinafter in detail embodying the present invention function as trunk circuits.

The complex wave form of signals from the devices IZii, E38, 22?! and 23B are transmitted to the sampling circuits :22, I32, 222 and 232, respectively. Associated with the sampling circuit is a comparing circuit which together with the sampling circuit causes the magnitudes of the samples obtained by the sampling circuit to be represented by permutation code groups of signaling pulses in which each group of pulses comprises the same number of pulses and represents the magnitude of a different sample and in which each pulse may comprise one of a plurality of different signaling conditions.

From the comparing and coding circuits the signals are transmitted through the associated high frequency transmitters to the other station. After the signals have been received at these stations they are tranmitted through a phase or time delay circuit I44, Ifi l, 264 or 254 which may be constructed of a suitable number of network sections of a suitable frequency range of one or more of the types disclosed in the above-identified Patent 1,770,422, granted to Nyquist.

As will be readily understood by .persons skilled in the art, it is necessary that the incoming signals be maintained in a predetermined phase relation with the common equipment at the receiving stations so as to be properly decoded and translated and a complex wave reconstructed therefrom corresponding to the complex wave at the transmitting station. It is, of course, possible to provide phase or delay equipment at each of these stations for each of the transmitting and receiving channels terminating thereat.

However, as shown in Figs. 1 and 2 the phase control equipment has been provided at the receiving terminals or channels terminating at the respective stations. This arrangement has the advantage that the phase control equipment may be adjusted to compensate for the time of transmission over each of the respective channels so that it is possible to employ common control equipment at each of these stations for the operation of each of the channels terminating thereat.

After passing through the receiving and phase control equipment at the respective stations the signals are decoded by the respective decoding circuits I42, I52, 242 and 252 and conveyed to the respective terminal equipment IM, I5I, 2M and 25I and then to the receiving or recording de vices Mil, I58, Mil and 25h, respectively. The terminal equipment I iI, IEI, 2M and 25I may include any of the types of apparatus or equipment described above with reference to the ter minal equipment I2I, ISI, HI and ESL Inasmuch as the circuits individual to each of the channels all operate in substantially the same manner, the circuits of a single channel only have been shown and described herein in detail. It will be understood, however, that similar circuits are duplicated for each of the additional channels extending between any of the stations of the system.

A control pulse generator I I I is provided at the transmitting station and connected to the output of the control oscillator III! or the master oscillator I0. The control pulse generator III is arranged to generate control pulses of both positive and negative polarity in response to the alternating current or voltage applied thereto from the control oscillator I II] or the master oscillater It. In the exemplary embodiment of the invention set forth herein the control pulse generator III simultaneously generates positive and negative control pulses for each cycle of the alternating current applied thereto. The negative control pulse generated by the control pulse generator III is transmitted through a delay network Il2 which in the exemplary embodiment set forth herein consists of a time or phase delay network comprising impedance elements such as inductances, capacitances and perhaps resistances as pointed out above. The output of the delay network II 2 is connected to a code element timing circuit II3, and to an exponential wave generator H 1 and to the sampling circuits such as I22 and I32.

In response to each of the control pulses applied to the code element timing circuit II3 a series of code element timing pulses is generated. Each series of code element timing pulses comprises a group of positive pulses interspersed with an equal number of negative pulses. A positive pulse and a negative code pulse is generated by the code element timing circuit for each pulse of each code group of the signaling pulses as will appear hereinafter. The output of the code element timing circuit is connected to the comparing and coding circuits IZ'S and I33. The exponential generator II is employed to generate an output wave in the form of an exponential function, the magnitude of which changes a predetermined fraction during the time interval assigned to each one of the code element timing pulses, that is, during the interval between the code element timing pulses. In the exemplary embodiment of the present invention the magnitude of the exponential wave is assumed to change by one-half of its immediately preceding value during each of the code element timing intervals. The exponential wave generator generates an exponential wave in response to each of the control pulses applied thereto from the control pulse generator Ill through the delay network H2. The output of the exponential wave generator H4 is connected to the sampling circuits I22 and I32; to the coding and comparing circuits I23 and I33; to the decoding circuits I42 and I52; and to monitoring equipment I25, I35, I45 and I55. The outputs of the comparing and coding circuits I23 and I33 are connected to the respective transmitting amplifiers I24 or the radio frequency transmitter I 34 as well asto the monitoring equipment I25 and I35 and the sampling circuits I22 and I32.

Briefly, the operation of the system is as follows: Assume, for example, that the input speech, music, or picture signal from the microphone or other device I20 has a wave form such as illustrated by the graph or curve I30I shown in Fig. 13. Assume further that the frequency of the master oscillator I or control oscillator III) is such that the control pulse is generated by the control pulse generator at the times corresponding to each of the large dots I302 on curve I30I. At each of these times a control pulse is transmitted through the delay network II2 to the sampling circuit I22 where some electrical quantity, such as, for example, the charge on a condenser, is made proportional to the instantaneous amplitude of the complex wave I30I at that time. In addition, the code element timing circuit and the exponential wave generator are set into operation.

The outputs of the sampling circuit and the exponential wave generator are compared in comparing and coding circuits I23. If the sample obtained from the complex wave, as'described above, is greater than the amplitude of the exponential wave at the time of the positive pulse from the code element timing circuit H3, a signaling pulse is transmitted from the comparing and coding circuit I23. If the amplitude of the wave from the exponential wave generator is greater than the amplitude of the sample at this time, that is, is greater than the charge of the condenser referred to above, no signaling pulse will be transmitted at this time. Assuming, for example, that the charge or potential on the condenser in the sampling circuit will be greater than the amplitude of the exponential wave when some one or more of" the positive pulses are received from the code element timing circuit. At these times pulses will be transmitted from the comparing and coding circuit I23. These pulses are transmitted through the transmitting amplifier I24 over line 6 to the receiving station. These pulses. are also transmitted to the monitoring equipment I25. As an incident to the transmission of the pulses in question a potential condition is developed in the comparing circuit which is then amplified and limited and then applied to the sampling circuit. The potential applied to the sampling circuit due to the operation of the comparing circuit causes the charge on the condenser in the sampling circuitto be reduced when the corresponding pulses are transmitted. The charge is reduced by an amount represented by the pulse transmitted. Thereafter the remaining potential of the condenser is again compared with the magnitude of the exponential wave and the above process or cycle of operations repeated thus causing the transmission of a series of the pulses representing the magnitude of the sample and, hence,-the

instantaneous amplitude of the complex wave-- form. A similar series of pulses is transmitted representing the magnitude of each of the samples or instantaneous amplitudes. Each of the series of pulses is sometimes calleda permuta.

10 tion code combination of pulses. Pulses of this type are sometimes also called binary pulses since they may be considered to represent the digits of a binary number and occupy positions corresponding to the ordinal positions of such a number.

In the exemplary embodiment described hereinafterthe magnitude of each sample is represented by five code pulses, thus permitting the representation of .32 different amplitudes by each of the permutation code group of pulses. If it is desired to more accurately determine and represent the instantaneous amplitudes of the complexwave form more pulses in each code group may'be employed. For example, if seven pulses are employed, 128 diflerent amplitudes may be represented. In other words, 2 difierent amplitudes may be represented by the permutation of the 11. pulses of two different types, i. e., current and no current, or positive and negative current. It will ,be readily apparent to persons skilled in the art that in order to change the number of code elements representing each amplitude it is only necessary to change the number of pulses generated by the code element timing circuit 3 in response to each of the control pulses transmitted through the delay network II2 and to change the constants of the exponential circuit and perhaps other circuit constants.

After all of the pulses of a given permutation or code group of pulses have been transmitted, another control pulse from the control pulse generator III will cause the equipment to sample a complex wave format a succeeding interval of time and the above process is repeated. The interval of time between the samples of the complex wave together with the number of different amplitudes transmitted representing each of the instantaneous amplitudes determine the accuracy and fidelity of the reproduction of the complex wave form at the receiving station as will be readily apparent to persons skilled in the art.

Furthermore, it is necessary to sample the complex waves at a rate at least twice the highest frequency desired to be transmitted and represented in the reconstructed wave form at the receiving end of the system.

Monitoring equipment I25 is provided at the transmitting station and is connected to the control pulse generator III, to the exponential wave generator H4 and to the output of the comparing and coding circuit I23 and is arranged to decode each group of pulses and generate a pulse having an amplitude proportional to the instantaneous amplitude represented by the code groups of pulses transmitted from the transmitting station. Each of the pulses of varying amplitude generating in the monitoring circuit is applied to a low-pass filter and then to some receiving device or indicator I26 so that the operation of the system may be monitored at the transmitting station. A monitoring circuit is shown in Fig. 1 for each channel. It will be understood by persons skilled in the art that a monitoring circuit may and usually will be common to a number of the channels and connected to any desired channel by a key, a plug and jack, or other suitable equipment.

The station shown in Fig. 2 is similarly provided with a control pulse generator 2 which generates a pulse inresponse to each oscillation received over channel 5 and oscillator 2 I0. However, in the event that channel 5 becomes in- 11 operative the control pulse generator 2|| wil still be controlled by oscillator 2H1. Consequently it may be possible to operate the system for a considerable period of time without receiving a synchronizing or controlling frequency over channel 5.

The receiving station is likewise provided with a delay network 2|2, code element timing circuit 2|3, and exponential wave generator 2|4 which operate in substantially the same manner as described above with reference to the corresponding equipment at the transmitting station. The output of the exponential wave generator is connected to the decoding circuit 242 to which is also applied the received signals after being amplified by amplifier 243. The decoding circuit 242 response to the pulse groups received to produce pulses of differing amplitude depending upon the amplitude of the exponential wave at the time the individual pulses are received. The pulses of differing amplitude are then applied to a low-pass filter where the complex wave form is reconstructed and regenerated. The output of the terminal equipment 24| is transmitted to the receiving device 240. Equipment 2M may include any or all of the equipment mentioned above with reference to terminal equipment 22 I. It will be readily understood by persons skilled in the art that equipment 22| and 2 may each include any of the different devices or equipment referred to above but that both of these devices do not necessarily have to include the same types of equipment but may do so.

Monitoring equipment 245 is provided at the receiving station and enables the attendants to check the operation of the receiving apparatus. This equipment, as shown in Fig. 2, is individual to the transmission path 6. It also may be common to all the paths or any portion thereof terminating at the station of Fig. 2 in which case it will be switched to the transmitting or receiving end of any path over which it is desired to observe the transmission.

Reference will now be made to Figs. 3 to 10 inclusive, when arranged as shown in Fig. 12, together with Figs. 13 and 14, which set forth in detail the various circuits and apparatus and the methods in which they cooperate to form a typical exemplary system embodying the present invention.

Figs. 3 to 7 inclusive, illustrate in detail common equipment employed at one station at which it is assumed the complex wave form is applied to the system. This station will frequently be called a transmitting station hereinafter.

Figs. 8, 9 and 10 show the equipment as a second station at which complex wave form is reconstructed and delivered to utilization circuits and equipment. This station will frequently be called hereinafter the receiving station. In addition, Figs. 3 to 10 inclusive, show only a single one-way communication path and the equipment thereof in addition to the equipment common to a plurality of paths.

Figs. 3 and show details of the circuit at the transmitting station which are common to a number of transmission paths, while Fig. 8 and the upper portion of Fig. 9 show similar circuits at the receiving station. The other figures show equipment which is individual to each transmission path. It will be obvious to persons skilled in the art that to provide additional communication paths in either direction between the two stations it is only necessary to provide additional equipment similar to the equipment already 12 shown in the drawings for each additional transmission path.

In order to better understand the operation of the system, the common equipment, shown at the top of Figs. 1 and 2 in diagrammatic form and shown in greater detail in Figs. 3, 5, 8 and the upper portion of Fig. 9, will be described at length.

Figs. 5 through II! have not been complicated by showing the connections of the filaments or heaters of the various tubes to a suitable source or sources of power because persons skilled in the art will understand that the filaments or heaters are supplied with power during the operation of the system.

A master oscillator 5H) and a second oscillator 525 are shown in Fig. 5. If the master oscillator 5| 0 is located at the transmitting station shown in detail in Figs. 3, 4, 5, 6 and 7, the local oscillator 525 may be dispensed with. However, in the case of master oscillator 5H] located at some other station, such as a master frequency standard for a large number of stations, systems, or for the entire country, both oscillator 5H3 and the local oscillator 525 will usually be employed. Master oscillator 5H] may be of any suitable type such as the type disclosed in the above-identified Meacham or Marrison patents. The local oscillator 525 will then include some control apparatus for maintaining its frequency in synchronism with the frequency of the master oscillator 5|0 similar to the equipment described in greater detail in the above-identified patents. Oscillator 525 is connected over a synchronizing line 580, which is shown in Figs. 5 and 8 as a coaxial line. Line 580 extends to the receiving station shown in Figs. 8, 9 and 10. Here the coaxial line 580 terminates in the local oscillator Bill, which is similar to the oscillator 525. While the synchronizing line 580 is shown as a coaxial line, persons skilled in the art will readily understand that any suitable type transmission path may be employed which is capable of transmitting the synchronizing frequency employed.

Control pulse generator The local oscillator 525 or the master oscillator 5H1 is connected to a multivibrator circuit comprising tube 5| The multivibrator circuit 5| l operates to generate square waves which are usually of the same fundamental frequency as is received from oscillator 525 or 5H). Multivibrator circuits are well known in the prior art. Typical multivibrator circuits suitable for use in the present system are described in greater detail in United States Patents 1,744,935, granted to Van der Pol January 28, 1930, and 2,022,969, granted to Meacham on December 3, 1935, and in an article by Hull and Clapp, published in the Proceedings of the Institute of Radio Engineers for February 1929, pages 252 and 271. See also section 4-9 entitled Multivibrator, on page 182 of Ultra-high Frequency Techniques, by Brainerd, Kochler, Reich and Woodrufi'. The output of the multivibrator 5| is coupled through condenser 5|2 and resistance 5|3 to amplifier tube 5|4.

Condenser 5|2 is made variable so that it, together with resistance 5|3, may be employed to control the length of the synchronizing pulse derived from multivibrator circuit 5| If the time constant of condenser 5|2 and resistance 5|3 is large the output pulse will be relatively long, whereas, if thet'ime constant of condenser 5|2 and resistance 5|'3 is small or short, the output noted that resistance I3 is connected between the grid of tube 5M and the positive B battery potential instead of the ground or some negative value as in the usual case. By connecting resistance 5I3 between the grid and positive B battery potential or some other high positive battery, the sides of the pulses are made steeper, the generation of very short pulsesis improved and the more satisfactory operation of the system is obtained.

The ouput of the amplifier tube SM is in turn coupled to tubes M5 and 5I6. Tubes 5I4, 5I5 and 5| 6 are amplifier tubes which are overloaded by the magnitude of the pulses applied to them so that these tubes tend to limit the magnitude of the pulse repeated through them and at the same time tend to make it square in wave shape. Amplifiers of this type are sometimes called limiting and at other times clipping amplifiers because they limit or clip off or suppress the upper portion of the input waves applied to their grids or control elements. A single stage limiter is shown in Figs. 8-6 on page 282 and described on page 283 of Ultra-high Frequency Techniques, by Brainerd, Kochler, Reich and Woodrufi, first published in July 194.2 by D. Van Nostrand Company, Inc. The ouput of tube 5I5 is coupled to a power tube which is employed to supply sufficient power for the output pulses of the circuit so that they may be employed to control other circuits of the system. The output of tube 5 is arranged to supply both positive and negative pulses as shown by the graphs MI and 523 in Fig. 5. The negative pulses are obtained from the anode resistor 526 connected to the plate or anode of tube 5I1, while the positive pulses are obtained from the cathode resistor 5I9 connected to the cathode of tube 5".

It will be obvious to persons skilled in the art that in cases where a large number of circuits are supplied from thepulse generator shown in Fig. 5, additional output stages may be connected in parallel with tube 5I I, having their input circuits connected in parallel with the input circuit of tube 5H, or they may be driven by this tube by having their inputs connected to the output circuit of this tube. Such arrange-- ments are well understood and frequently employed where one tube does not supply sufiicient output energy.

The negative pulses from the plate of tube 5H pass through a delay network 522 where they are delayed slightly in time with respect to the positive pulses 52L The purpose of this delay network willbecome apparent hereinafter. Delay network 522 may be of any suitable type employing reactive elements in any well understood manner as in the art pointed out above. The output of the pulse generator, as shown in Fig. 5, is diagrammatically.indicated by the small'curve 52I for positive pulses and curve 523 for negative pulses, in order to facilitate the reading and understanding of the operation of the circuit.

In order to further facilitate the understanding of the operation of the system, reference will also be made to Figs. 13 and 14 which show graphs of the wave form of the voltage of currents at various places in the system under cerasserts tain assumed conditions. Curve i-30l of Fig. 13 represents a complex wave form of a type suitable for transmission over the system described herein. As will be described hereinafter, the magnitude or amplitude of this wave form is sampled or measured at frequently recurring intervals of time. For purposes of illustration, the operation of the system will be described in detail for one such measurement, which is assumed to be madeat the point I303. As shown in Fig. 13 by dots I302, this wave form is sampled at frequent intervals represented by the dots I302. The frequency of the sampling determines the highest frequency component of the complex wave which may be represented by the successive permutation code signals and thus reproduced at the receiving end of the system.

Vertical lines I 304 in the next set of curves represent positive pulses 52I supplied by the pulse generator shown in Fig. 5. Likewise, the vertical lines I305 represent the negative pulses generated by the pulse generator shown in Fig. 5. As shown in Fig. 13, the negative pulses I305 are delayed for a short interval of time and consequently follow pulses I304. This is apparent in Fig. 13 when it is assumed that time increases in the positive direction to the right. It is further assumed in the exemplary system described in detail herein that samples of the complex wave are arranged at a rate of approximately 8,000 times a second. Persons skilled in the art will understand that any other suitable frequency may be employed so long as this frequency is appreciably higher than the highest frequency component of the complex wave necessary or desirable to transmit to the distant station. In other words, the frequency of oscillators 5!!) and 525 is either 8,000 cycles a second or else frequency changing apparatus is connected between them and the multivibrator 5I I in a manner well understood in the art, so that the multivibrator 5H operates at a frequency of approximately 3,000 cycles a second. In other Words, the positive pulses I304 as well as the delayed negative pulses I305 are supplied by the pulse generator shown in Fig. 5 at the rate of 8,000 per second. Consequently, there are microseconds between the pulses I304. Similarly, a period of 125 microseconds elapses between the pulses I305. The negative pulses from the delay network 522 are supplied to the amplifier tube 3I0 in Fig. 3 over lead 524 where they are amplified and changed to positive pulses due to the operation of the amplifier tube 3I0. Tube 3I0 is shown in Fig. 3 as a pentode tube. It will be obvious to persons skilled in the art that this tube may be replaced by any suitable triode tube or other multielement tube having the desired or necessary power or current capabilities.

Code element timing circuit The output of tube 3") is connected to the code element timing circuit comprising tubes 3| I, 3I9, 320 and 32L Tube 3II is normally biased so that substantially no current flows to the anode circuit of this tube. When the amplified and delayed synchronizing pulse is applied to the grid of tube 3 from the output of tube 3I0 it causes the grid of this tube to become positive for the duration of this pulse. During this time condenser 3I2 in the cathode circuit of tube 3 will receive a positive charge. Curve I305 of Fig. 13 shows the potential of the upper terminal of condensers 3I2. The short positive section designated 'I320 "above the delayed synchronizing pulse I305 represents the positive potential applied to the upper terminal of condenser 312 for the duration of this synchronizing pulse. Thus, during the application of the delayed synchronizing pulse 1335 to the grid of tube 311 after it has been changed into a positive pulse by tube 310, energy is stored in the oscillating circuit comprising condenser 312 and inductance 313. At the end of the synchronizing pulse the grid of tube 311 again becomes sufficiently negative so that substantially no current flows in the oathode circuit of this tube. The tube, therefore, does not thereafter materially alter or change the currents or potentials of the oscillating circuit until the next synchronizing pulse is applied to the grid thereof.

At the end of the synchronizing pulse the oscillating circuit comprising condenser 312 and inductance 313 starts to oscillate in a manner shown in curve 1303 of Fig. 13. As shown in Fig. 13, the current in the resonant circuit comprising condenser 312 and inductance 313 makes substantially six complete oscillations between the applications of the synchronizing pulses to the grid of tube 311. Thus, after substantially six cycles the oscillations of the resonant circuit comprising condensers 312 and inductances 313 are started over in proper phase relationship with oscillator 525.

The upper terminal of the oscillating circuit which is connected to the cathode of tube 31 1 is also connected to the grid of tube 319. Tube 319 operates as a cathode follower since the output resistance comprises resistance 313 connected in the cathode circuit of tube 3i3. As is well understood by persons skilled in the art, tubes acting as so-called cathode follower stages have an extremely high impedance in their input circuit. Consequently, the connection to the grid of tube 319 does not materially alter or afiect the oscillations of the resonant circuit as described above. Such properties as well as other properties of vacuum tubes operated as cathode followers are well known to persons skilled in the art (see The Cathode Follower, by C. E. Lockhart, parts 1, 2 and 3, published in Electronic Engineering for December 1942, February 1943 and June 1943, respectively). The foregoing publications are hereby made part of the present application as if fully included herein and apply to the operation of the cathode follower tube 319 as well as to the other cathode follower tubes described hereinafter. Tube 319 is also provided with a so-called decoupling network in its anode circuit comprising resistance 333 and condenser 336.

The output of cathode follower tube 319 is coupled through a coupling network comprising resistance 3E5 and condenser 315 to the grid of the left-hand section of tube 323. The coupling network comprising resistance 315 and-condenser 316 is provided to cause the grid of the left-hand section of tube 320 to follow the potential applied to the cathode of tube 313 with as little distortion as possible. Condenser 31B is employed to compensate for the input capacity of the left-hand section of tube 323. When the ratio of the condenser 316 to the input capacity of tube 323 is the same as the ratio of resistance 315 to resistance 331, the potential of the grid of the left-hand section of tube 323 is substantially a constant fraction of the potential of the cathode of tube 319 independently of the applied frequency. The foregoing relationship holds so long as the constants of the various circuit elements employed as well as the constants of the tubes employed restantially flat-topped or square.

16 main substantially constant and independent of frequency.

The output of the left-hand section of tube 323 is coupled to the right-hand section of tube 320 and this section in turn is similarly coupled to the input of tube 321. The two sections of tube 320 are biased to operate as an overloaded or limiting amplifier and cause the output hignals to be sub- Tube 321 will usually be of a type which is capable of furnishing sufficient power, current or voltage of the desired wave forms. Graph 1301 of Fig. 13 shows the output wave form of the code element timing circuit. As shown in Fig. 13, the tops and bottoms of the waves are shown to be fiat and the corners square. As is well understood by persons skilled in the art, the corners will generally be rounded somewhat in practice and the tops and bottoms of the waves will not be fiat. However, the slight rounding of the wave form is a function of the various parameters of the circuit and may be reduced to any desired degree. Inasmuch as this rounding does not materially interfere with the operation of the circuit as described hereinafter and as an aid in drawing the curves, the graph 13111 has been shown as a square wave with square corners. In addition, the graph 1301 has been shown as comprising waves having equal tops and bottoms. That is, the top and bottom of the square wave is of substantially the same length or time duration as shown in Fig. 13. Such an arrangement is not necessary because the top and bottom of the wave need not be the same length but may vary in length over a considerable range without adversely affecting the operation of the system. However, in an effort to simplify the drawing and aid the understanding of the invention, both the top and bottom of the wave have been shown in Fig. 13 as being of equal length.

Emponentz'al circuit In addition to the code element timing circuit shown in Fig. 3, an exponential circuit is also provided comprising tubes 323 and 331. As shown in Fig. 3, tube 323 is a so-called twin triode tube. However, both sections of this tube are shown connected in parallel in order to secure the desired characteristics or current carrying capacity. This tube, therefore, is the full equivalent of any suitable three-element tube and may be so considered.

The output of the amplifying tube 310, which amplifies and inverts the delayed synchronizing pulse, is coupled to the grid circuit of tube 323. The cathode circuit of tube 323 comprises a condenser and resistance network comprising condenser 325 and resistance network 326 and 324.

Tube 323 is normally biased so that substantially no current flows inits anode circuit. When the positive delayed synchronizing pulse is applied to the grid of this tube it will cause current to flow in the anode-cathode circuit of this tube, which in turn will cause the upper terminal of condenser 325 to be charged to a relatively high positive potential. When the synchronizing pulse is removed, the upper terminal of condenser 325 will start to discharge through resistance 32B and 324. The potential of the upper terminal of condenser 325 will therefore decay or fall in accordance with the well-known exponential law. In other words, when the value of the resistances 324 and 326 and the value of condenser 325 are both fixed for any given set of operating conditions, the curve of the potential of the upper terminal of condenser 325 against time follows the well-known exponential law. See section 123, beginning on page 453 of Principles of Electrical Engineering, by Timbe and Bush, published by John Wiley and Sons, Incorporated, 1923 (first edition). This publication is hereby made a part of the present application as if fully included herein.

Although many suitable values of the relative magnitudes of condenser 325 and resistances 324 and 326 could be chosen to provide many different time constants suitable for operation in circuits described herein, in the exemplary embodiment employed in the present invention, suitable time constants for this network comprising resistances 324 and 326 and condenser 325 has been chosen so that the potential of the upper terminal of condenser 325 decreases one-half its value during each of the signaling pulse intervals. That is, the time constant of the condenser resistance network comprising condenser 325 and resistances 324 and 326 are so selected that the potential of the upper terminal of condenser 325 at the end of any and all of the pulse intervals will be equal to substantially one-half the value of the potential of the upper terminal of this condenser at the beginning of the respective pulse intervals. I

The grid of the output tube 33! is connected to the junction point of'resistances 326 and 324. Thus, it is possible to select the desired fraction of the voltage upon the upper terminal of condenser 325 for'controlling tube 33!. Tube 33! acts both as a pulse amplifier tube and also as a cathode follower tube. Thus, both positive and negative exponential waves may be obtained from the output circuits of tube 33!. The exponential wave obtained across the output cathode resistor 329 is substantially of the same wave form as the potential wave upon the upper terminal of condenser 325. Such a curve is illustrated by graph !308 in Fig. 13. The output obtained from the anode of tube 3!3 will be inverted and appear as curve I309 of Fig. 13.

Compression and sampling circuit Microphone 420 represents any suitable source of complex waves or signals to be transmitted as described above. The source of complex waves 42!! is connected through terminal equipment 42! to transformer M0. The secondary of the transformer M is connected to potentiometer M2 and also to a non-linear circuit element 4! I. The function of these two elements is to compress the amplitude of the complex wave received from terminal equipment 42!. The nonlinear element 4!! may be of any suitable type and may include any suitable elements such as cuprous-oxide rectifiers, selenium rectifiers, thyrite, etc. As the input potential becomes higher in value, the resistance element 4!! will decrease in value, thus tending to reduce the variation of magnitude of the signaling voltages delivered to potentiometer M2. The movable contact member of potentiometer M2 is connected to the control element or grid of the lefthand section of tube 4 !6. Thus by varying the potentiometer M2 and making a compensating change in the signal level received from the terminal equipment 42! the amount or degree of compression can be easily and readily varied. If the movable contact of potentiometer is set near the bottom and the signal level from the equipment 42! raised, the compression of the signals as applied to the grid of the left-hand section of tube M6 is increased. If, however, the movable member is moved up and the signal 18 level from 42! reduced to compensate for the change of the potentiometer, then the signals applied to the grid of the left-hand section of tube M6 will be less compressed.

'If no compression is desired, key 453 will be operated. With key 450 operated the compression equipment is disconnected from the signal transmission path and the signals are conveyed directlyfrom the terminal equipment 42! to the grid of the left-hand section of tube 4!6 without compression or other change.

The grid or input of the right-hand section of tube M6 is connected to the negative synchronizing lead 524 from the delay circuit 522'. The right-hand section of tube M6 is biased so that normally considerable current flows in the output or anode-cathode circuit of tube 4l6. As a result of this current flowing through the de-' coupling resistance 9 and the anode or'output resistance 4|8 of tube MS, a relatively low 'value of potential is applied to the anodes of the twin triode tube 4l6. Consequently, a relatively low potential is applied to the grid of tube 4!!. As a result, substantially no current flows in the anode-cathode circuit of tube 4!! at this time. However, upon the application of a negative synchronizing pulse to the grid of the right-hand section of tube 4!6,"substantially no output current flows in the anode circuit of the right-hand section of tube 4l6. Consequently, the current flowing through resistance 4!B falls to a value determined by the grid potential of the left-hand section of this tube, which is in turn determined by the instantaneous amplitude of the complex wave applied to the grid of the left-hand section of tube 6 at the instant of time that the de- I complex wave applied to the grid of the left-' hand section of tube M6 at this time. As a result, current will flow in the anode-cathode circuit of tube 4!! and thus charge the upper terminal of condenser 426 to a value determined by said instantaneous amplitude of the applied come plex wave form. Upon the completion of the application of the synchronizing pulse' to the grid of the right-hand section of tube 416, the right-hand section of this tube will again conduct current and lower the potential of grid of tube 4!! to such a low value that substantially no further current flows in the cathode circuit of tube 4!!. The upper terminal of condenser 426 is connected to the grid circuit of tube 43!, which tube acts as a cathode follower and tends to accurately repeat the wave form or potential of the upper terminal of condenser 425. Tube 43! acts as a cathode follower tube since its output is obtained from the cathode resistor 433.

Consequently, the grid of tube 43! presents an extremely high impedance to the condenser 426 and will not therefore appreciably alter or inter-- fere with the potential of the upper terminal of condenser 426.

Comparing circuit and coding The cathode of tube 43! is coupled through

Images