DE4028927A1 - Coding and transmission of originals and images - combines code elements of several lines or channels into code word - Google Patents

Abstract

The information signal coding and transmission combines one or several code elements of several lines, or channels, into a code word, independently of presence of point to point process, or a transit length composition, or a two-dimensional combination etc., and/or of a single-, two-, or multi-position digital code. If MHL process is used, the white and black, and possibly grey, code words, and EOL of two or more lines, are in parallel, and combined into a new code word. If required, filling bits are provided for generating identical code element number for the combined lines. USE/ADVANTAGE - For telefax and TV transmission, recording, and storage, with reduced transmit time and/or bandwidth, and increased reliability.

Description

The present invention is concerned with a method for coding and transmission of information, in particular of templates and pictures and for television.

For coding and transmission of information 2 or So far, several channels via one path are frequency and time-multiplexed methods, such as B. the carrier frequency technology and pulse code modulation has been used. Even with faxes apart from the "point to point" are the one-dimensional running lengths gene and two-dimensional coding known. All these Methods either have too much bandwidth or transmission time or the effort is too great.

The present invention now has the aim, e.g. B. for faxes and for television the information for transmission, Coding recording and / or storage so that against the transmission times and / or over the conventional codings the bandwidth is reduced and / or the transmission si security is increased. This is by the in claim 1 disclosed teaching achieved.

According to the invention, for. B. the transmission time in the Shortened by 2 or more lines, also run length-coded, or color television channels simultaneously codemul tiplex coded and transmitted. This happens in the Way, by the synchronously, or quasi-synchronously arranged Code elements of 2 or more lines or channels in parallel ordered and combined into one code word and transmitted be. When decoding on the receiving side the code elements for the individual lines or channels again assigned.

The invention can e.g. B. be applied to Telex, Teletex, Fax, with digital telephone and data channels, with Ge Community connections and selector switches.

Also in image transmission and in television technology use the invention to advantage. When transmission over cables it may also be necessary that an equal If there is no current, then you become the period provide as a code element.  

Furthermore, the invention discloses an advantageous phase coding of the color tone when watching TV. It will not Phase shift that is coded in the alternating current and which is a measure of the hue, but the Phase shift of the samples, which subsequently in the Period of the coding alternating current is transmitted where at the amplitude encodes the saturation vector.

The invention also shows applications for the duplexver return with an alternating current of a frequency. This is based on the principle of adding two phase-shifted by 90 degrees bener alternating currents, in which the amplitudes of the half-waves represent the information and then the oncoming traffic don't pick up. They are also applications for the double Quadrature amplitude modulation shown, in which the 4 Co alternating currents are summed up twice and one phase position of 0, 90, 90, 180 degrees and at the 2nd summation take a phase position of 45 and 135 degrees.

My patents and Offenle are also considered state of the art Publication: Patents US 47 94 621, 46 75 721, 47 31 798, Canada 12 14 277, European patent application 01 10 427, 01 97 529, 02 39 959, 02 84 019, German disclosures DE 36 29 706.2, 35 14 664.8, 37 19 670.7, 38 02 088.2, 38 05 263.6.

In the following, the invention will be approx ago explained. These represent:

Fig. 1 principle of a code multiplex arrangement,

Fig. 2 previous generation of phase jumps z. B. at the 4 PSK,

FIGS. 3 to 8 and 83 produce phase jumps,

Fig. 9 produce amplitude levels,

Fig. 10, 11 and 13, of a double QAM and vector diagram of a higher-quality encoding,

Fig. 14 vector diagram of a dual QAM,

Fig. 16, 91 arrangement of the encoder marks at a polyhydric Co consolidation by means of amplitude levels and phase position,

Fig. 15 Summary for the generation of phase and amplitude step,

Fig. 17 generating phase jumps,

Fig. 18, 19, 20, 21, 24, 28, 79, 88, 89, 90 code multiplexes examples

Fig. 22, 23 and overview of a television receiver,

Fig. 25, 26, 27 duplex traffic via radio lines, and with only an alternating current with phase adjustment,

Fig. 29 compensate for overlap,

Fig. 30, 31, 32 generating and implementing PWM pulses in half-cycle pulses,

Fig. 33 to 38 generation and implementation of PDM pulses into an alternating current,

Fig see. 39 to 44 according to the invention codes for the remote,

Fig. 45, 46, 62, 63 and doppelbinäre doppelduobinäre array of code elements,

Fig. 47, 48, 49 overviews circuit for television,

Fig. 50 to 55 codes of color television signals,

Fig. 56, 57, 58 multiple use of transmission paths PDM coded signals,

Fig. 59, 60 evaluation of phase-modulated signals,

Fig. Chart 64 via a function of the frequency-modulated oscillation of the amplitude and frequency of vibration Modula tion,

Fig. 65 coding scheme for color television,

Fig. 66, 84, 85, 86 narrow-band code,

Fig. 67 scheme for a television phase encoding,

Fig. 68, 69, 87, a phase encoding for color television signals,

Fig. 70, 71 serial arrangement of television signals,

Fig. 72 principle arrangement for the transmission of television signals, phase encoding,

Fig. 73, 74 Pulse width modulation circuit,

Fig. 75, 76 principle of multiple transmission pulsdauermodulier ter signals via a current path,

Fig. 77, 78 Digital coding of color television signals and shift for the transmission,

Fig. 80, 81 half-period durations of phase encodings,

Fig. 82 diagram for the accommodation of an information channel between 2 television channels,

Fig. 92-110 Coding of several lines or channels.

A simple way of realizing phase jumps is described in FIGS. 3, 4, 5, 6, 7. This is first explained in more detail with reference to FIG. 3. Rectangular pulses with a frequency of 1 MHz are switched on on the transmitting side S. If, as shown in FIG. 3c, a low-pass filter TP 5.5 MHz is switched on in the transmission path, the receiver E almost receives a rectangular pulse. If, as shown in FIG. 3b, a low-pass filter TP of 3.5 MHz is switched on, the vertical slope is no longer present; if, on the other hand, the low-pass filter is reduced to 1.5 MHz, as shown in FIG. 3a, then a sine-like alternating current with the period duration of the rectangular period is obtained at the receiver E. Since the period does not change compared to the rectangular pulse, the phase or frequency of the sinusoidal alternating current shown in FIG. 3a can be changed by changing the periods of the rectangular pulses. Since such a change always occurs at the zero crossing, there is a continuous change and hardly any harmonics are generated, ie the transmission is more narrow-band than in the case of the conventional phase keying. In the receiving point, the change in the period can then be provided as a measure of the phase shift. Such an evaluation circuit will be described later.

In FIG. 4, rectangular pulses are last with different periods T = f T = f 1 and T = f2 shown. According to an analogous arrangement according to FIG. 3a, one would obtain a sinusoidal alternating current with the period durations T = 1 / f, T = 1 / f1, T = 1 / f2 on the receiving side. Since the frequency of the alternating current decreases or increases in phase jumps, the frequency change corresponds to a phase jump. This is clearly shown in FIG. 2, which represents phase scanning of a conventional type. It can be seen in this that a frequency change occurs with each phase change, but not in a continuous manner. It is therefore also difficult to determine the size of the phase jump from the period on the receiving side. In order to keep the frequency changes and thus also the frequency band small, each phase jump can be broken down into steps. This is schematically recorded in FIG. 5. In this T / 2 is the half period of a pulse and corresponds to 180 °. This angle is divided into 36 steps of 5 degrees each. If a phase jump of 40 degrees should occur, the half period T / 2 is shortened 4 times by 5 degrees and of course the other half period as well. The half-period compared to the reference pulse is then T1 / 2. After the phase jump you can either leave this frequency, or you can switch back to the frequency T / 2 by providing a phase jump of 5 degrees in the opposite direction. Compared to the reference phase, there would still be a phase shift of 30 degrees. In FIG. 6, the periods of the reference phase and 4 times the periods of the shortened by 2 × 5 degrees periods are temporally 4 times a drawn. When comparing after the 4th period, the difference of 40 degrees compared to the reference phase can be seen.

In Fig. 7 is a circuit of an embodiment is shown He invention. It is assumed that the period is subdivided into 72 steps, namely with phase jump steps of 5 degrees. 10 measuring pulses are to be assigned to each stage, so that 72 × 10 = 720 measuring pulses are required for the period and 360 measuring pulses are required for the half-period. Only the half periods need to be coded on the transmission side. The second half-period is then controlled by the encoder Cod. If phase-change stages of 5 degrees are provided, 350 measuring pulses are required for the half-period, if the change is to be leading, and 370 for a lagging phase change. The counter Z in FIG. 7 must therefore have at least 370 outputs. The measuring pulse frequency therefore depends on the coding frequency. In the example of FIG. 7, the control alternating current for the measuring pulses is generated in the oscillator Osc. One can thus control the counting element directly via the gate G1, or else generate pulses by means of a Schmitt trigger or another circuit and then switch the counting element Z with these pulses. You can also change the pulse duration by changing the oscillator frequency. Assume that the output Z2 on the counter Z marks 370 measuring pulses, i.e. the lagging phase shift, then such a potential is applied by the encoder Cod via g2 to one input of the gate G2 that when the counter reaches output Z2, via which z. B. the same potential is applied to the other input of G2 that the potential at the output of G2 changes, z. B. from h to l. In the electronic relay ER this has the consequence that plus potential + is applied to the output J. The encoder Cod is connected to the electronic relay ER via connection A. At the next overflow of the counter Z to Z2 is controlled via the connection A ER so that at the output J minus potential - is applied. At the output of ER, bipolar right corner impulses can be taken. One could generate unipolar square pulses just as well. This process is repeated as long as the encoder applies code potential to G2. Are z. B. 5 phase steps are provided for a phase shift, the counter Z is switched 10 times to Z2. At the output Z2, the counter is switched back via the gate G4, R. It can thus be set by a different number of outputs on the counter Z and / or by changing the oscillator frequency, the pulse duration, the number of stages and the size of the stages. These variants are controlled via the encoder Cod. Via fA, the oscillator frequency can be switched, via the connections g2, g3,. . . the number of stages and, if necessary, the phase angle change and the stage size and, via A, the amplitudes of the rectangular pulses J. In the example, 2 sizes + / (A) +, - / (A) - are provided. The rectangular pulses J are then switched to a low-pass filter analogous to FIG. 3 and via a transformer Ü z. B. on the transmission path, possibly with the interposition of a filter Fi.

Starting potential must still be applied to gate G1 via B, so that the oscillator pulses come into effect. With this arrangement The following codings are possible: a leading, a lagging, no phase shift. They can do this also take place gradually. The phase difference or the Be pull phase can be used. In addition, an amplitude dencoding may be provided in stages. A wide one re possibility is the coding at the positive or negative pulse or half wave. Also the The number of rectangular pulses is another code means. Man can also sift out a harmonic of the rectangular pulses. He does this follow e.g. B. at the 3rd harmonic, there are 3 periods  contained in a plus / minus pulse. In these 3 periods are also when the pulse duration is changed, that contain phase shifts.

In various circuits, such as. B. in Quadra turamplitudenmodulation (QAM) 90 ° phase-shifted alternating currents are required. In FIG. 8 is a circuit principle for the generation of such is illustrated phasenverschobe ner alternating currents of the same frequency. Analogously to FIG. 7, the counter Z is controlled by an alternating current which is generated in the oscillator Osz and is guided via the gate G, at the other input of which there is a starting potential B. In the example, 4 rectangular pulses are to be generated, which are 90 degrees out of phase with each other. If the counter Z has 100 outputs, then electronic relays ER1 to ER4 are to be connected to the 25th, 50th, 75th and 100th output analogously to the ER relay in FIG. 7. Rectangular pulses are then generated with these electronic relays, as already described in FIG. 7. Here are still in the ER relay means that always reverse the potential with bipolar rectangular pulses and take away the potential with unipolar rectangular pulses during a run. The rectangular pulses, designated J in FIG. 7, are then sent via the filters Fi1 to Fi4. The resulting alternating current has a 90 degree phase shift compared to that generated by the next output. Instead of phase-shifted alternating currents can be phase-shifted decreases of z. B. Control PAM samples. On the electronic relay ER1, a filter Fi0 is arranged, which, for. B. only lets the 3rd harmonic of the square-wave pulse pass, so that here one obtains 3 times the frequency of the square-wave pulses. The phase shift is then transferred to the 3rd harmonic.

With Fig. 7 you can also generate different amplitude denstufen. Only 2 are marked in the circuit. In Fig. 9 is another way to generate different amplitude denstufen. The z. B. Alternating current generated in Fig. 7 is supplied to a limiter in which the control pulses are generated. The characteristic states are supplied via the connection code, which perform a switchover to the amplitude size determined by the code, specifically in the encoder code. The switchover to a different amplitude size always takes place at the zero crossing. The size of the amplitudes is determined by the resistances R1 to R4, which are arranged in AC circuits. Electronic relays I to IV, which are controlled by the encoder Cod, switch on the various resistances in the AC circuits. At output A you get 4 different amplitudes.

It is also known to encode information using the half-waves or periods of an alternating current; in the case of a binary code, the characteristic states are large and small amplitude values. If 2 such coding alternating currents of the same frequency are phase-shifted and added by 90 degrees, they can be transmitted with an alternating current of the same frequency. In Fig. 10a, b are the channels K1 and K2, which are encoded by the periods as large code elements with the characteristic states Am plitudenwert = 1 and small amplitude value = 0. If one phase is shifted by 90 degrees against the other, they can be added. In Fig. 11, their vector diagram is shown. Channel K1 has vector K1 (u) and channel K2 has vector k2 (v). The two characteristic states of the two alternating currents are designated u1 / uo and v1 / vo. If both are added, the 4 sum vectors I, IV and II, III are obtained. You can see that vectors II and III are no longer on the 45-degree line. The evaluation is therefore somewhat more difficult. For the evaluation of the binary signals, 4 options, which can all be placed on the 45-degree line, are designated in FIG. 11 by (II) and (III). FIG. 13 shows the 4 possibilities, 00, 11, 10, 01. If all 4 possibilities are on the 45-degree vector, as shown in FIG. 11, these can be coded by 4 amplitudes of different sizes, ie with a sinusoidal alternating current. In Fig. 9 a sol che possibility is shown. To transmit binary signals from 2 channels, a multi-value quaternary code, such as B. the 4 PSK or 4 QAM. These codes are spread over a period. In Fig. 9, the positive and negative half-wave are the same size, there is then a DC current freedom in the transmission. The positive and negative half-wave can be used as an additional criterion. One can then distribute the 4 amplitude characteristics, 2 on the positive and 2 on the negative half-wave. These can be the same size, e.g. B. in Fig. 11, I + IV for the positive and negative half-wave. So that this coding alternating current is always above the interference level, the Co dier alternating current must always have a certain size, for. B. as in Fig. 11 (III). The amplitude size IV will then be increased somewhat.

A reduction of e.g. B. binary-coded alternating currents with the half-waves or periods as code elements is already known. The prerequisites for this are phase shifts in the sampling. The present invention shows a further possibility of reducing the frequency, in particular binary-coded information. In Fig. 1 is a channel K with a binary code 1, 0, 1, 1,. . . recorded. If the frequency of the channel is to be reduced in 2 channels at half the frequency, then 2 serially arranged binary values of channel K must be distributed in parallel to channels Kv1 and Kv2, e.g. B. the 4 values 1, 0, 1, 1 of the channel K, the value 1 on Kv1, the value 0 on Kv2, the value 1 again on Kv1 and the further value 1 on Kv2. You can always save a value, or you can transfer the values at different times. This must be taken into account in the evaluation. Simultaneous transmission of 2 channels has already been shown in FIGS. 11 and 13. As can be seen from FIG. 13, 4 combinations are possible.

In Fig. 10 4 coding alternating currents K1-K4 are shown with the code elements period and the characteristic states of large and small amplitudes worth the same frequency. If you want to transmit all 4 on the basis of the QAM, they must have the following phases, K1 = 0 degrees, K2 = 90 degrees, K3 = 90 degrees and K4 = 180 degrees. K1 / K2 and K3 / K4 are combined to form a coding alternating current in accordance with FIG. 9 and added. The vector diagram for this is shown in FIG. 14. It can be seen that 16 combinations are possible. Furthermore, it can be seen from this that only 4 values lie on the 45-degree vector. During the evaluation, the leading or lagging phase shift must also be taken into account for the other values. The phase-shifted alternating currents are generated in an arrangement as shown in FIG. 8 and supplied to 2 arrangements according to FIG. 9, these alternating currents being mutually phase-shifted by 90 degrees.

One can also use a total alternating current and simple coding add alternating current, prerequisite is a 90 degree phase shift against each other. This creates 8 combinations opportunities.

Also 4 channels can be coded multiplexed, as shown in FIG. 1, transmitted (Kv1, Kv2, Kv3, Kv4). Then 16 combinations are necessary. You can also provide well-known codings, such as. B. the 16 PSK, the 16 QAM, the 8 PSK. A coding is required here in each case if phase shifts are provided according to the present invention. Instead of the closely related characteristic states in the double QAM according to FIG. 14, any coding can also be carried out. In Fig. 16, the encoding by 30 degree phase differences and amplitude levels 3 and 4 is carried out. If you want even greater security, you can still divide the 4 amplitude levels BPh. Steps can still be accommodated on the zero line. So you can provide each half-wave for such coding. However, if you want to carry out a transmission via wired transmission paths, it is advisable to transmit the negative half-wave with the same coding, so that one is free of direct current. The same method can be used to downsize. In Fig. 1, the channel should only be transmitted at the quarter frequency. In each case 4 binary elements 1 and 0 arranged in series are arranged in parallel as provided in FIGS . 1a, b. The values 1, 0, 1, 1 of channel K are then divided in parallel into channel Kv1 "1", channel Kv2 "0", channel Kv3 "1" and channel Kv4 "1". In the encoder, the predetermined coding point is then determined for the respective combination and transmitted to the phase and amplitude of the coding alternating current. The phase is determined in FIG. 7, if necessary one can also use this to encode the amplitude, and in FIG. 9 one can then code the required amplitudes. In Fig. 15 the overview is shown for this. In the encoder Cod, the coding point is determined on the basis of the combination of four. The phase encoder generates the half-waves or periods with a corresponding phase and the amplitude encoder generates the associated amplitudes. A phase encoder can look analogous to FIG. 7 and an amplitude encoder analogous to FIG. 9. A phase jump always means a change in the period duration. This change, i.e. frequency change, cannot be maintained with any further phase change, or one can switch back to the original frequency in the next period or half-period. Since in the latter case the alternating current has a different phase, a reference phase is required for the evaluation. As can be seen from FIG. 4, with the aid of the circuit of FIG. 7 any phase can be maintained, that is to say the frequency that was generated during the phase change can be maintained. The phase changes are always made in the present case at the zero crossing. In the Fig. 16 can be a reference phase BPh provide a Phasenver shift is made from the lag and lead from 2 × 30 degrees.

FIG. 17 shows the generation of the phase jumps of FIG. 16 according to the principle of FIG. 7. The 360 degree angle is characterized by 3600 pulses. If there is only a change in amplitude with the reference phase, the counter is always switched through from 0 to 360 degrees. The control takes place via the encoder Cod, which has already been described in FIG. 7. The change in amplitude takes place as shown in FIG. 7 or in FIG. 9. If the phase jump Ph1 in FIG. 16 is to take place, then if a direct current freedom is required, every half cycle must be switched to the output 195 . A reference phase is not necessary for the evaluation because, as long as no further phase change takes place, the clear phase is determined by the period duration. If the coding is on the vector Ph3, the period is 330 degrees, ie there is always a switchover at the output 165 . The phase shift is always related to the period. Would z. B. bezo conditions in the latter case, the phase shift to the half period, a downshift would have to take place at the output 150 . Other methods of generating phase jumps can also be used.

The phase jumps are evaluated in a known manner by dimensioning the periods by means of an excessive Control speed of counters, e.g. B. in the European ischen patent application 8 61 04 693.6.

In the evaluation of Fig. 14, a reference phase is erforder Lich. The amplitude points 1 to 4 are directly on the reference phase position, while the other 12 coding points are arranged leading and lagging the reference phase. The signals are assumed to be those of a television system. The reference phase is then determined during the blanking time and control signals are simultaneously transmitted. Only the amplitude values on the reference phase are used. The signals are fed to the input set EST from the transmission path ÜW ( FIG. 12). Once they go to a delimiter B and once to a code evaluation CA. In the limiter, the positive and negative half-waves are converted to Jp and Jn pulses. In the comparison device VE, the phase of the pulses coming from the transmission path is compared with a reference phase pulse JBn. In Fig. 12a which are upstream, and the lagging the reference phase pulse Jv, Jn, JB shown, which are compared with the determined reference phase from a coding pulse JBN. The 3 possible phase values leading, lagging or reference phase are given for code evaluation. In this, the amplitude values are determined and, in conjunction with the leading, lagging or reference phase, who then determines the coding points and transmits them via S for further use. The coding of the reference phase in the blanking time can, for. B. look so that one sends 4 times the point 2 and 4 times the point 4 on the reference phase. The evaluation of these takes place in the reference phase evaluation BA. From this a reference phase pulse JBn is then given to the comparison device.

In FIG. 18, a further embodiment of the OF INVENTION shown dung. The 5 channels K1 to K5 are to be transmitted code-multiplexed only via one channel or path. The z. B. binary-coded information of these 5 channels is first stored in the memory Sp. In FIG. 20, z. B. the steps of the binary characters are shown and already synchronized. 5 steps or pulses S1, 2, 3 arranged in parallel are therefore to be coded. . . The steps of S1 are 1 - 1 - 0 - 1 - 0. 5 bits are required for coding these 64 combinations. In the example, these are encoded with the amplitudes of the half-waves of an alternating current with the characteristic states of large and small amplitude values and with a leading and a trailing phase jump of 36 degrees, as shown in FIG. 19. The binary values are supplied from the memory Sp in FIG. 18 to the encoder Cod and are converted into a corresponding code in the latter. In the decoder on the receiving side, the corresponding steps are reassigned to the 5 channels according to the code.

In Fig. 21 a further application of the invention for the coding and transmission of the signals in color television is presented. The luminance signal is tapped at 6 MHz. This principle has already been disclosed in published patent application P 32 23 312. The colors red and blue should each be tapped at 1.2 MHz, ie one red and blue tap hits 5 luminance taps. The luminance taps are labeled I, II, III, IV, V. These samples are coded with 8 bits, binary coded in the example. With tap III, the taps for red and blue must also be made. The red and blue samples are binary coded in the example with 6 bits. During the transmission of the 5 luminance samples, the code for the color samples red and blue is also sent at the same time. With the tapping of red and blue one could start with the transfer of the color and with the sampling I of the luminance signal. You can also save all 5 luminance samples and only start transmitting all television signals after the 5th sample has been taken. In FIG. 21, the binary codes are all net up signing signals to be transmitted. The 8 bits 1-8 of the luminance samples are arranged in parallel. Serial are then arranged under 9, 10 digital sound and other signals T + So, the 6 bits of the red signal and again the sound and other signals and under 11, 12 again the sound and other signals and the 6 bits of the blue signal. It is useful if one still stores the luminance samples I to V at the transmitter and sends the color codes for red and blue with the previous luminance samples, so that the receiver does not need to store the 5 luminance samples. Then only the red and blue samples have to be saved. The sound and other signals must also be saved and then sent to the loudspeaker at the same time as the picture. These signals can of course also be placed in the blanking time. In the example, 12 bits are required for the transmission of a luminance sample for the sound and other signal samples and for the color sample taking. An example of the coding of these 12 bits is shown in FIG. 21b. 5 half periods of an alternating current are provided for this. The binary code consists of code elements of the half-waves with the characteristic states of large and small amplitude values. In addition, a leading and lagging phase shift of 36 degrees is provided, so that you get 12 bits.

An overview of such a television transmitter is shown in FIG . The control element StO controls the TV camera FK, also supplies the other control signals such as blanking and synchro signals A + S. The red, green and blue signals are fed once to the Y matrix YM and red and blue to the color processing FA. At the same time, a capacitor K is seen which taps the luminance signal Y, the color signals r + bl and the sound and other signals. At tap 3, a criterion for color processing is given via connection 3a. In this a tap from the red and blue signal is made before and both values are stored in the capacitors C1 and C2. The FA is supplied by the Y matrix with a Y value which is present at the third tap, so that the color difference signals ry and by are held by the handles 6a and 6b. - You can only tap the color separation signals. - The sound and other signals are fed via 6c and 6d to the concentrator via the TSo module. All values are fed from the concentrator to a memory Sp. From the memory, the signals are timely z. B. as described in Fig. 21a, supplied to an analog / digital converter. This is encoded in accordance with FIG. 21b. During the blanking period, a switchover to the concentrator K1 via U takes place. B. send the code word a few times with only zeros. Other signals So can also be sent during the blanking time. You can also mark the beginning of a line with a zero code. During the line, the sequence and the number of half-waves dictate synchronization. In the present code, a nominal frequency of 15 MHz is required. If you only want to use an amplitude code, 2 alternating currents with 18 MHz each are required, which could then be shifted by 90 degrees and transmitted together. Which method is used here is only a question of economy and safety. In the example, the phase jump before or after hurrying is determined by the period. No reference phase is then required. Of course, multistage amplitude codes and / or phase codes can be used to reduce the frequency. At the input tone T you can e.g. B. apply the PAM signal, which is then tapped more often within the 8 kHz time. There are numerous ways to use tap 6c / 6d. In the Fig. 23 is a partial overview of a television receiver is set to represent. The signals are fed to the demodulator DM via the HF oscillator and mixer and the amplifier V. In this z. B. the signals as shown in FIG. 21b are recovered and fed to the decoder DC. The color signals are then passed on to the matrix Ma. The Y signal is also connected to this. At the exit of the matrix one obtains e.g. B. the color difference signals RY, GY, and BY, which are led to the television tube like UY. The decoder DC then delivers the blanking and synchronizing signals AS, the sound and other signals.

An example is shown in FIG. 24, in which the code for the code multiplex is obtained from a plurality of alternating currents. It represents a binary code in which the half-waves of the alternating currents serve as code elements and in which a large and a small amplitude value form the characteristic states. The characteristics to be transmitted consist of rectangular pulses with a frequency of 1000 Hz, as shown in FIG. 24b. 20 channels are to be transmitted in a code-multiplexed manner. The half-waves of the alternating currents 1000, 1500, 2000, 2500 and 3000 Hz are provided for this. Each channel can of course be fed multiple channels with a lower bit frequency in a time-multiplexed manner. The same number of bits could also be achieved with 2 alternating currents at 2000 Hz and another 2 alternating currents at 3000 Hz, each of which would have to be 90 degrees out of phase with one another, so that they could be added during transmission. It is already known how best the synchronization between the individual channels is established (instruction sheets of the DBP issue 4/6 year 79), and it will therefore not be discussed any further. The digitized speech or several speech channels can be transmitted simultaneously in the same way.

With an amplitude coding one can perform duplex operation with the same alternating current. For this it is necessary that the counter-coding alternating current is 90 degrees out of phase. This principle is shown in FIG. 25. The code can be digital, a binary code, according to the patent DE 30 10 938 or also analog according to the Canadian patent 12 14 277. With half-waves as code elements, the frequency is 32 kHz with digital coding and 4 kHz with analog coding. In FIG. 25, S1 and E2, the microphone of the handset of a subscriber and S2 and E1 of the other party. In S1 there is also a coder in which the coding produces alternating current from the language. From S1 the coding change current goes through a fork G, the connecting or connecting line RL to the fork G of the opposite party and to the receiver E1. In this is additionally a decoder, which selstrom again produces the speech from the coding. The coding alternating current from S1 is the synchronizing alternating current. This is branched off from E1 via a phase shifter 90 degrees to S2, in which it may be amplified. Now speaks S2, a 90 ° out-of-phase coding alternating current is sent via G, RL, G to E2, decoded there and transmitted to the listener as speech. If e.g. B. is spoken briefly at the same time, an alternating current arises on the transmission path RL. An extinction is not caused. This principle can be used for duplex traffic in data transmission. Further examples in this regard are disclosed in laid-open document 38 02 088.

This method can of course also with radio z. B. used in radio relay. In Fig. 26, a related overview is recorded. The transmission alternating current is also provided here as an encoding alternating current. Pre-stage modulation is advantageously used. The transmission alternating current is generated in the oscillator Osz1. In the analog / digital converter A1 / D1, the base signal is converted into an AC digital code. - It is even easier to provide an arrangement according to FIG. 7 as an oscillator and encoder. From the encoder, the electronic relay is then controlled so that J large and small rectangular pulses are present at the beginning, which are then formed in the low-pass filter TP to a sinusoidal alternating current. - The coding alternating current then reaches the output stage E and the transmitting antenna via amplifiers (not shown). In the final stage you can also provide a branch circuit in which the harmonics are phase-shifted by 180 degrees, which are then fed back to the main circuit for compensation. On the receiving side, the useful signals are fed to an amplifier V via a fixed tuning circuit and then forwarded to the digital-to-analog converter D2 / A2. The analog signal is then z. B. forwarded via an operator. Via the amplifier V, the alternating transmission current is also branched off to a phase shifter of 90 degrees Ph and then switched on to the Oszilla gate Osz2. With this the oscillator is synchronized. Via the converter A3 / D3, not shown, the amplifier and the amplifier E are then operated in the opposite direction. The receiver E1 is connected exactly like the receiver E2, only the phase shifter is not necessary.

A phase shifter based on the principle of FIG. 7 is shown in FIG. 27. This also provides for compensation for small frequency fluctuations. For this purpose, a counter Z is provided with 1000 outputs. During a half-wave of the alternating transmission current, the counter runs through these 1000 outputs. The control pulses Js are generated in an oscillator, not shown. With a 90 degree phase shift, a half-wave encounters a phase shift of 45 degrees, which corresponds to 250 outputs. The coming alternating current half-waves from amplifier V are fed to a limiter, so that square wave pulses Jp and Jn are produced at the output of the same. These impulses are connected to the control element St. The control impulses Js and the start indicator Be are also applied to this. The control element is switched in such a way that only whole Jp or Jn pulses are effective at the counter element. If the counter has reached output 1000 during a pulse Jp, gate G11 comes into the working position. A Jn pulse is connected to the gate G12 and, after the end of the Jp pulse, the potential is briefly switched on by the delay of the monostable element mG4. G12 takes effect and applies potential to one output of G13, 1 potential has already been applied from G11 to the other input of G13. A potential change now occurs at the output of G13, which reverses polarity at the output. As a result, G17 generates a downshift potential for the counter. Such a potential is also applied to the gates G8, G9 and G10 that they control one of the monostable elements mG1, mG2 or mG3 in cooperation with the assigned outputs 1000 , 999 , 1001. Since the Jp pulse has controlled the counter to 1000 , the gates G9 and mG2 now took effect. Now with the next Jn pulse the counter is controlled on the output 250 , the gate G6 is effective, which controls the electronic relay ER, which generates a rectangular pulse according to FIG. 7, which is formed into a half-wave in the low pass . Gates G15, G14 and monostable element mG5 are arranged for the output marking for the Jn pulse. The monostable link mG2 holds e.g. B. up to exit 260 . G6 then returns to the starting position. The electronic relay remains in this position until the next marking of output 250 . If only frequency 999 is reached due to a frequency fluctuation, then gate G8 is marked in place of G9 and mG1 and G5 are brought into effect when output 249 is reached . If the output 1001 is reached, G10 and mG3 are brought into effect and the gate G7 when the output 251 'is reached. Such frequency fluctuations are also passed on to the 90 degree phase-shifted alternating current. The control element is shown in detail in FIG. 27a. The pulses Jn and also the start signal are connected to the gate G3. If both are present, G3 takes effect and brings the bistable member bG into the working position, which now applies working potential to the gate ter G1. Only now can the Jp pulse take effect. The control pulses Js now reach the counter via the gate G2, which is only a potential reversal gate. The other operations on the counter are already described.

In Fig. 27, the negative half-wave can either be generated by the Jn pulse, or the passage of the positive half-wave is repeated, the respectively marked outputs being stored.

The code used in the invention may preferably be an amplitude and / or phase code, such as e.g. B. such is shown in Fig. 16. In the case of a pure amplitude code, it is also possible to provide 2 code alternating currents of the same frequency, one of which is then shifted by 90 degrees during transmission and is subsequently added to the other.

The principle of the invention also serves for the transmission of digitized speech. In Fig. 28 are 5 code alternating with a binary code, wherein the characteristic states of a large and a small amplitude value of the respective half-shaft, is provided. The frequencies are 8, 12, 16, 20 and 24 kHz. You get 20 bits, 2 alternating currents of the same frequencies, but 90 degrees out of phase, are provided, so you get 40 bits, ie with 8 bit code words, as shown in Fig. 28a, you can use it to transmit 5 digitized voice channels .

In Figs. 21 and 22 meet each line at a Abgriffsfre per line frequency of approximately 30 kHz (PAM) 2 Tonabgriffe which z. B. at the beginning of the respective image line and in the middle of the image line, the distance is then 32 µs. Each tap is then converted into an 8-bit code in the analog / digital converter A / D and is then, as shown in FIG. 21a, sent with the following 5 luminance code words. In Fig. 21a z. B. with I / 9, 10, 11, 12 and V / 9, 10, 11, 12. The taps during the picture change time must z. B. can be determined by a time measurement. The coding then also takes place during the picture change time.

Any code, such as the AMI or HDB3 code, can of course be used for the code multiplex. In the examples, an amplitude code is often used in which the code elements consist of the half-waves or periods of a sinusoidal alternating current with the characteristic states of small and large amplitude values. A code element corresponds to one bit. Are z. B. 12 bit for the CVBS and audio signal, 12 half-waves are required. The coding can be carried out synchronously with the taps since the length of the code words does not change. If, on the other hand, a phase code or a phase code is provided, the period duration changes with each phase change, so that the signal taps are no longer synchronized with the code in the case of a periodic tap and in the same direction with phase changes. To compensate, there are 2 options - except for buffer storage - once with each phase change until the next phase change, restore the nominal frequency, e.g. B. in Fig. 4, the nominal frequency f2 and there is a phase change T = f1 and the following codings have the same phase changes, the following codings are coded with the nominal frequency f2. Only when phase f1 changes again does a phase change take place with respect to the reference phase, ie the reference phase must be saved at the receiver. This can e.g. B. in the blanking time from the transmitter. Another way to avoid overlaps between two taps is that the transmitter uses a measurement solution with each code word between the end of the code word and the previous and the following tap. If there is a risk of an overlap in the leading or lagging direction, then whoever interposes the code words with the smallest or largest period durations. In Figs. 29a and 29b such Darge are up. This can be avoided by storing lines. In Fig. 19 a code element has 6 different levels and 2 digits the code word, consequently 6 to the power 2 combinations are possible, that is 36 combinations. With 32 combinations you get 5 bits. In FIG. 21b, a code element can also assume 6 levels, so that 6 to 5 = 5184 combinations are possible with 5 digits, that is to say at least 12 bits. With 12 bit you get 4096 combinations.

In Fig. 22, the PAM for the sound in the TSO element is generated and z. B. placed half-line at 6c. The connections 6c and 6d are not required if the sound and the other signals are placed in the blanking time, so that the concentrator K1 then takes over these tasks.

With the help of Fig. 21, 22 and 23 should be shown how to z. B. can also use the code division while watching TV. The transmission frequency can of course be significantly reduced if you see more amplitudes and / or phase steps before. You can also with different carriers, such as. B. in the patent application P 32 29 139.6 Fig. 9 see, or combine with different current paths. So you can z. B. in Fig. 28 with 8 kHz transmit a 64 Kbit voice channel, with a binary code. 2 digits are each marked by the two half-waves of an 8 kHz alternating current, 2 further digits by the 2 half-waves of an alternating current that is 90 degrees out of phase. These two alternating currents are summed and transmitted as an alternating current via the one current path. The same is done via a second current path, so that the code word has 8 digits and 2 levels, so that 256 combinations are obtained. Decoding is carried out on the receiving side after the evaluation of the half-waves and, of course, intermediate storage. The coding can also be duobinary.

Another method, in particular analog signals such as speech, tones, the luminance signal in television, the color signals in television, telecontrol values, transmitted with frequency modulation and with less bandwidth, consists in using the pulse duration modulation PDM the size of the PAM pulses in PDM - convert pulse lengths. These PDM pulses can then be converted into alternating current pulses e.g. B. be converted according to the method of FIG. 7 delt. The pulses are then formed by the half-waves or periods of an alternating current, the periods lasting or half-periods of the half-waves or periods becoming equal to the length of the PDM pulses.

The spectrum of the frequency-modulated Schwin used so far supply contains a large one above and below the support Number of side vibrations, so that a very wide band is required for the transfer. The bandwidth required is larger than twice the frequency swing. In the Circuitry according to the invention can predominantly be digital Switching means are used, so that an inexpensive Her position is possible.

Below is the method using drawings explained in more detail. First, known circuits are repeated explains the u. a. are necessary for production (European cal patent application 02 84 019). 2 embodiments of the Invention are described below. First they are Principles of the two versions summarized. The In formation is pulse modulated once and in the Follow in pulse durations using the equidance procedure changed, or the information is immediately with Coded in pulse durations using the sawtooth method. These Pulse durations are then linked to the pauses between the pulse durations to rectangular pulses and subsequently with Hil fe from filters to sinusoidal coding alternating currents changes. The pulse durations and pauses are transformed with Help of counter elements in connection with electronic Switches. The pulse duration then corresponds to the duration of one Half period or period of the coding alternating current. Is the Pulse duration small, is the frequency of the half-wave or period with coding alternating current high, the pulse duration is long, so  is the frequency of the half-wave or period during coding alternating current small. The off occurs on the reception side evaluation, for example, by measuring the half or perio last. So here is a frequency and Phase modulation.

In the second embodiment, the pulse duration pulse, in FIG. 32 PD1, PD2 and the pause between the pulse durations ( FIG. 32, P) - the pulse duration and the pause correspond to e.g. B. each the distance between 2 taps, in Fig. 30a designated tp - supplied to an electronic relay in which bipolar rectangular pulses are then generated. The frequency-modulated coding alternating current is then generated with the aid of filters.

In Fig. 7 it is shown how with the aid of a counter Z in conjunction with the frequency of the advance or measuring pulses generated in the oscillator Osc, the time of a pulse is determined. The respective output of the counter then marks the time. This is then provided in connection with gates for the control of an electronic relay ER. This then generates bipolar rectangular pulses.

The function is as follows. In the oscillator Osc are the indexing or measuring pulses for the counter Z creates. These arrive at the counter via gate G1 member Z as long as the start character at B is present. in the Examples are only the outputs Z1 and Z2 of the counter needed. These outputs are at gates G2 and G3. Half the period of the rectangular pulse J is the size of the sum of measuring pulses up to Z1 is from the encoder Cod on g3 h potential, so that when reaching the output Z1 at the output of G3 there is a potential change that the electronic relay ER causes the square pulse to break up. If this was a positive impulse, the next will be Impulse negative. The counter is then in this position switched back again. The gate for this is the output Z2 G4 provided. From the encoder, the Oszil lator frequency be increased or decreased, so that one e.g. B. markie with the respective outputs different times could ren. A connection A also comes from the encoder Cod ER, with which you can control different pulse sizes J.  

The square-wave pulses are given to the line as a sinusoidal coding alternating current via a low-pass filter TP, the transformer U and filter Fi. The half or period of the coding alternating current is the same as that of the rectangular pulse. The principle of converting the square-wave pulses into a sinusoidal alternating current is shown in FIG. 3. Are z. B. rectangular pulses with the frequency 1 MHz with a low pass 5.5 MHz band-limited, you get, as shown in Fig. 3c, still fairly steep edges. A low-pass filter of 3.5 MHz was used in FIG. 3b . It can be seen that the edge steepness has noticeably decreased here. In Fig. 3a, a low pass of 1.5 MHz is turned on, the receiver here has a sinusoidal alternating current. The period durations are the same as those of the rectangular pulses, ie the period durations can be used as a measure of the frequencies or phases. In FIG. 7, this principle has been det angewen in the conversion of the rectangular pulses J in a code alternating with the aid of low-pass filter TP.

In FIG. 4, rectangular pulses are recorded various periods, expressed by the frequencies f, f1 and f2. These rectangular pulses have different phase shifts or different frequencies. It can be seen from this that one can cause phase jumps or frequency jumps by changing the period durations, so that this also gives frequency modulation. In FIG. 5, such a phase or frequency hopping is carried out stepwise. This is enough for the bandwidth to be small. As can be seen from FIG. 6, a total phase shift of 40 degrees is obtained with phase jumps of 5 degrees each 180 degrees with 4 phase jump steps.

In Fig. 30a PAM-encoded pulses from a signal Inf represents are provided. These are converted into pulse duration pulses using an equidistance method, as shown in FIG. 30b. The distance between the PAM pulses ( FIG. 30a tp) corresponds to a pulse duration PD and a pause P, as shown in FIG. 30b. Pulse duration modulation can also be carried out using the sawtooth method. This method is shown in FIGS. 31 and 32. The pulse durations are rectangular pulses PD1, PD2,. . . Furthermore, the symmetrical PDM and the bipolar PDM are known (see also book "Modulation method" by Stadler 1983).

In Fig. 35 an embodiment is illustrated in accordance with the He invention. In the pulse duration modulator PDM the pulses are e.g. Generated as shown in FIG. 30 or 32, and guided to the gate G1 through G5. At the other input of the gate G1, the measuring pulses Jm, z. B. 100 kHz frequency. As long as there is a PD pulse at G1, the measuring pulses Jm are effective at the output. Via the potential reversing gate G2, the measuring pulses reach the counter Z, which is controlled with these pulses. The number of outputs on the counter corresponds to z. B. the distance between 2 PAM pulses, in Fig. 30a tp. The tap frequency is 10 kHz, then the counter would have 100,000 outputs. The frequency hub is determined by the largest and smallest amplitude value of the information Inf, denoted by gw and kw in FIG. 30a. The outputs A of the counter Z lead to gates G3 and the outputs of the gates to gates G4. The respective PD pulse that blocks gate G4 is located at the other input of gate G4. Only when the PD pulse is no longer there can the output potential become effective via G3 at G4. ER now receives a potential change indicator for the next rectangular pulse via G4. The beginning of the rectangular pulse is marked by the respective PD pulse. The next right corner pulse is determined by the pause P ( Fig. 30b, P). A potential is applied from ER to gate 5 via P, so that the measuring pulses Jm at gate G1 become permeable again. The counter Z is now switched to the gate G6 output. When the next PD pulse comes again, G6 takes effect and the counter is switched back to the starting position via R. At the output of ER are then rectangular pulses RJ the size of the half-periods such as that of the PD pulses and the pauses P. In the filter Fi who the rectangular pulses to sinusoidal half-waves fmo, so the information is frequency modulated. The half-periods of the useful signal modulation frequencies then move between the half-periods on the counter labeled kw and gw. In Fig. 33, e.g. B. kw = 15 kHz, the center frequency 10 kHz and in Fig. 34 gw = 7.5 kHz. In the example, the pulse durations can change by half. This is a matter of dimensioning the pulse duration modulation circuits. The half-waves of the pauses have a lowest frequency of 7.5 kHz in FIG. 33 and a greatest frequency of 15 kHz in FIG. 34. The amplitudes of the half-waves always remain the same. The evaluation on the receiving side is carried out by measuring the half-periods. A synchronization is not necessary, since the zero crossings of a period when encoding with the help of a PAM encode the taps at the same time, so it is only necessary to convert the positive half-waves into PAM pulses. The PAM pulses are then lagging one perio de on the receiving side.

The redundancy of the pauses in Fig. 35 can be avoided if, for. B. stores the PAM pulses and retrieves the next PAM pulse after each PD coding. However, synchronization is then required at the receiver. When using the PAM on the transmission side, the tap frequency would have to be synchronized from time to time. In Fig. 36, the basic circuit diagram of such a circuit is shown on the transmission side. The PAM pulses are stored in the memory Sp. The next impulse is called from ER via AR. The next pulse was already stored as a PDM pulse in memory Sp1. The counter Z is now controlled via the control element St and set to a corresponding output. ER also brought the counter back to the starting position via R. The control impulses Jm are also on the control unit. With the retrieval of the PDM pulse, a PAM pulse is also given to the pulse duration modulator from the memory Sp and stored therein as a PDM pulse until the Sp1 memory is free again. It is advisable to provide 2 Sp1 memories, which are then alternately connected to the control unit after each call from ER. At the end of the PDM pulse, a pulse end criterion is given to ER via the counter Z, G1, G2. The rectangular pulse PD generated by ER is reversed to the next, the counter is switched back via R and the call of the next PDM pulse is initiated via AR. The rectangular pulses RJ are passed on via a filter. Half waves with the half-period durations of the PDM pulses then arise at the output of the filter, as are shown in FIG. 37. In FIG. 38, the PD pulses and possibly pauses in FIGS. 30b and 32 directly control the electronic relay ER. A polarity reversal takes place after each rectangular pulse. With the uninterrupted sequence of PD pulses, which has been obtained by storing it, as shown in FIG. 36, the ER relay can also be controlled in FIG. 38. A polarity reversal is only required after each pulse. In Fig. 38, the beginnings of the PD pulses are only marked via PDS if a continuous transmission of PD pulses is to take place. In the case of a pulse / pause transmission, the start and end of a pulse are marked anyway. If you want to be free of direct current during transmission, the respective pulse must be coded equally by a positive and negative half-wave. This can e.g. B. done by storage in a shift register, then a halving of the respective overflowed outputs or a halving takes place by means of a computer during the evaluation. A division into 2 half pulses can also be achieved using the symmetrical PDM.

The PDM pulses of FIGS. 32 and 32a can also be connected directly to a filter Fi corresponding to FIG. 38. In order not to let the bandwidth become too large, it is then expedient, as shown in FIG. 32a, to place the information in the sawtooth voltages in such a way that the difference in the length or width of the pulses does not become too large. The PD pulses according to FIG. 30b can also be applied directly to the ER switching means. After each pulse, a polarity reversal or no potential must then be applied to the rectangular pulses. The rectangular pulses would then be unipolar. In order to also reduce the bandwidth in the equidistance method with direct control of the ER switching means, one would have to carry out a larger direct current bias (in the case of unipolar PAM) when generating the PAM pulses, or a corresponding dimensioning of the circuit for generating the PDM.

In Fig. 39 4 channels are shown with a half wave lencoding with the characteristic states of large and small amplitudes worth. The frequency is the same for all 4 channels. These 4 channels are used to encode the color television signals. 8 bits are for the Y signal (luminance signal), 4 bits each for channel a and b. 2 bits each in channels a and b are provided for sound and other signals T + S. Channel c is available for coding the red signal and channel d for coding the blue signal, each with 6 bits. 2 channels are then combined according to FIG. 11 vector I, (k1, k2) with the codes I, (II), IV, (III), so that a total alternating current corresponding to FIG. 9 comes about. The phase position of the total alternating currents is then set to 0 degrees and 90 degrees. These two total alternating currents can now be transmitted on the basis of quadrature amplitude modulation, so that a narrow band is required for the transmission of all color television and other signals. Transmitted as double QAM, ie channel a + b quadrature amplitude modulated and channels c + d quadrature amplitude modulated, the channels being 0 °, 90 °, 90 ° and 180 ° in phase relationship and their total alternating currents being 45 ° and 135 ° phase angle, and that If the two total alternating currents are again modulated in quadrature amplitude, the evaluation is more difficult, as can also be seen from FIG. 11. (In the case of a single QAM, vectors I, II and III are created.)

The 4 channels or their binary values can also be transmitted codemultiplexed. In FIG. 40, the binary values of the 4 Ka are shown again ducts. According to FIG. 41, two rows of FIG. 40 are to be combined into 8 bits. In FIG. 39, the frequency of the alternating currents is 6 MHz; 18 MHz are then required for the coding. If a duobinary coding corresponding to FIG. 62 with the half-waves were used as code elements in FIG. 41, bandwidth would be increased somewhat compared to FIG. 39, but the frequency would be 3 times as high. If you combine rows 1, 2, 3 and 4, 5, 6, i.e. 12 bits each with this duobinary code, a 3-level code word with 8 digits is required for row 1, 2, 3. 8 digits mean 4 periods, so a frequency of 2 × 24 MHz would be required, al too high for this purpose. In FIG. 45, a 4stufiges code element is shown, in 4 digits, this gives 256 possibilities. A coding according to FIG. 41 would result in a frequency reduction to 36 MHz. In FIG. 63, a 6stufiges code element is illustrated. In order to code 3 rows of FIG. 40 serially as 12 bits, 5 digits would be required here. So 30 MHz would still be required. In addition to the 3 amplitude levels, two phase levels or periods are also provided. In the Fig. 46 3 amplitudes and 3 are shown Phasenstu fen. If 2 rows with 12 bits each are formed from the arrangement in FIG. 40, 3 digits are required for each row, ie 6 digits for both rows, ie a frequency of 18 MHz is necessary.

In Fig. 43, the color television signals are arranged differently. 8 bits for a Y tap (luminance, pixel B) are serial 4 bits each, the colors red or blue serial 3 bits each in rows III + IV. The 4th bit in rows 3 and 4 is intended for sound and other purposes. The color red or blue comes with every 2nd Y signal, ie these alternate continuously. If the vertical rows 1/2 and 3/4, as shown in FIG. 44, are combined, they result in more favorable conditions when encoded. With 4 stages, 3 stages are required, a frequency of 18 MHz is then required. If the rows 1/2 and 3/4 are arranged in parallel, ie 16 bits, 4 digits are required for a coding according to FIG. 46, ie 12 MHz frequency.

The double QAM of FIG. 39 can be frequency-modulated in order to have even more security in the transmission. The total alternating current has only small frequency changes, so that, as can be seen from FIG. 64, the frequency-modulated oscillation can nevertheless be transmitted in a narrow band. From this figure it can be seen that the half-period T / 2 becomes very small with a frequency increase, that is, the frequency increases sharply. With a modulation frequency Mf and an amplitude u, the half-period T / 2, with double ampli tude 2u, the half-period is smaller, while with additional double frequency M2f, the half-period is significantly reduced.

In Fig. 47 is an overview of a television station is provided, in which the codes explained in Figs. 40, 41, 43 and 44 th are used. The analog tapped signals come from the multiplexer (not shown) into the analog memory ASp and from there the samples are passed on to one or more analog / digital converters. The digitized signals are then stored in the digital memory DSp and subsequently fed to the folder. In this they are arranged according to FIGS. 40, 41, 43 and 44. In this way they are fed to the encoder. According to the predetermined code z. Encoded as shown in FIG. 45 or 46 or 62 or 63 and supplied to the modulator MO. The transmission alternating current is fed from the oscillator to the modulator and the modulated transmission alternating current is passed to the power amplifier to the antenna via amplifier stages (not shown). An overview of the receiver for the evaluation of the coded signals is shown in Fig. 48 Darge. The alternating transmission current comes via the receiving antenna E into the stages tuning circuit / amplifier, mixing stage / oscillator Mi / Osc, via the intermediate frequency amplifier ZF to the demodulation stage - the input is switched like a superimposed receiver when receiving radio - the code alternating current is at the output of the demodulator available. This is switched into the decoder. The signals tapped in the transmission multiplexer are again received here, such as the Y, ry, by, tone and other signals S and the various circuits.

In Figs. 50 and 51 analog codings of the remote ink are presented visual signals. In Fig. 50 an alternating current of the same frequency is provided as a code AC. The half-wave amplitudes are the code elements. The Abgriffsfolge is y, r, y, bl, y, T + S, etc. The transmission of this analog coding th signals takes place on the basis of the frequency modulation, so that a narrow band -. Only one frequency Figure 64 70763 00070 552 001000280000000200012000285917065200040 0002004028927 00004 70644 ä and also receives transmission security.

An analog code is also provided in FIG. 51. It is a phase encoding. The analog code is given by different half-period durations. The amplitudes of the half waves are always the same size, it is a kind of frequency and phase modulation. The individual signals are again arranged in series, in the example y, r, y, bl, y, T + S. The transmission takes place at a tap frequency of the Y signal at 6 MHz. If all signals are tapped, including the r, bl and T + S signals, a tap frequency of 12 MHz is required.

In FIG. 52 a coding corresponding to FIG. 51 is provided, only the sound and other signals T + S are coded by a superimposed amplitude code. It is a binary code with a large and a small amplitude. The values of the Y and the r + bl signals are determined by the half-periods. In synchronism with the PDM pulse z. B. to the ER relay of FIG. 36 given the respective amplitudes, in which a rectangular pulse with low or high voltage is then generated. The amplitude code elements can e.g. B. be assigned to several channels, such as sound stereo, etc. In FIG. 55, the 4 half-wave code elements are assigned to different channels 4 ver.

An evaluation of the PDM, PPM or PFM pulses encoded with the half-period is shown in FIG. 59. This is done again with the help of sawtooth tension. At the beginning of a half-wave, ie at the zero crossing, the generator of the sawtooth voltage is switched on. After the half-wave at the next zero crossing, z. B. by means of a field effect transistor switches the sawtooth voltage for a short time to a capacitor and stored in this. The half-period T / 2 is then equal to the voltage value T / 2 or analogous to the size of the voltage value. The half-period of 1 corresponds to the voltage value u1, that of 2 corresponds to that of u2, etc. If speech was modulated with 8 kHz pulse amplitude on the transmission side, the voltage u1, u2, u3 must be tapped on the receiving side at the same frequency and converted into voice alternating current will. In the case of a time-multiplex tapping of several channels, the stored values u1, u2, u3,. . . be distributed again with the same frequency of the time-multiplexed tap. The production of the original information can e.g. B. done in such a way by the evaluated code u1, u2,. . . after the channel allocation forms staircase and this staircase signal leads over a low pass. Such transformations are known and will therefore not be discussed in more detail.

In the same way as in Fig. 59 the PDM pulses, PPM pulses can also be decoded. In Fig. 60 this is Darge. The distance T / 2 of the pulses is reshaped into PAM pulses using the sawtooth method and saved. From the stand T / 2 corresponds to the voltage u1 etc.

When transmitting television signals according to the principle of FIGS. 36 and 38, the evaluated signals must be distributed synchronously on the receiving side. During the blanking time, synchronization pulses must be sent so that the sampling frequency can be determined on the receiving side, according to the transmission frequency. The sum of the largest half-periods occurring per line must not exceed 54 µs. This is the time allowed for one line with a 4: 3 aspect ratio. As a result, the half-periods must also be measured in the transmitter. Under certain circumstances, a filler code, e.g. B. includes the smallest or largest periods in a particular sequence. Of course, you can also provide other fill codes. In addition, the blanking time must also be provided as a fill code. In Fig. 61, the smallest and largest half-periods k and g are Darge. Such can e.g. B. alternately sent.

On this basis, several channels can be combined via a transmission path. Such an example is shown in FIG. 56. Channels 1 to n are combined in terms of pulse amplitude with the multiplexer Mu, which is known. These PAM samples are stored in the memory Sp, called up by the PDM and, as already described, fed to the counter via a control unit St, to which the control pulses Jm are connected. The other switching operations are the same as e.g. B. described in FIG. 36. After the pulse duration modulator PDM, the pulses can also be processed directly according to FIG. 38. On the receiving side, of course, must be synchronized and distributed according to the tap frequency of the multiplexer.

In the Fig. 57 is another possibility is the Mehrfachausnut wetting a current path shown. In order to be able to separate the code alternating currents in terms of frequency, such control pulses are used that the frequency ranges of the code alternating currents have such a distance that a correct evaluation is possible, for. B. a filter in the receiving station. In Fig. 57 Z1 is the one converter with the control pulses Jm1 and Z2 the other converter or counter with the control pulses Jm2. In FIG. 58, the frequency position of the two channels is illustrated. T / 2I and T / 2II are the lowest frequencies of the two channels. The angular stroke f2 brings you closer to the frequency range of channel T / 2I. In the example there is still a distance from Ab. This can be chosen so that inexpensive filters can be used.

A few codes are shown below with which data can be coded and transmitted with one frequency, in the example television signals. In FIG. 53, a binary code is shown, are provided in the denwert as code elements, the amplitudes of half-waves with the characteristics conditions of large and small Amplitu. A bit can then be coded with a half wave. There are 8 bits for the Y signal, 6 bits for the red and blue signals, and 2 bits for the sound (digitized) and other signals. Red and blue are alternately, such as. B. shown in Fig. 51, coded. With 6 meg taps for the Y signal, a coding alternating current at 48 MHz would be required here. A binary binary coding is provided for this in FIG. 54. The coding alternating current then has a frequency of 27 MHz. One can transmit these coding currents again frequency-modulated, the frequency band is not too wide, as is shown in FIG. 64. The transmission security is even greater. In Fig. 66 a possibility is recorded how one can digitally transmit a message without modulators in narrowband. Each code element is assigned a plurality of periods of an alternating current of a frequency, which are determined by the time Og, that is to say a predetermined number of periods. The coding is assumed to be binary. With each change of state, that is 1 to 0 or 0 to 1, the transition takes place continuously, designated U in FIG. 66. The amplitudes for zero have the size Ak and those for the characteristic state 1 the size Ag. If the same values come one after the other, the amplitude size is not changed. With 5 identical values one would send a period number of Og with the same amplitude five times. The transition to another at their characteristic state is z. B. ge calculated to the following characteristic state, so z. B. Ü + O = Og.

The coding of information according to FIGS. 53, 54 and 66 result in very narrow frequency bands. This can also be used in television technology. So you could u. U. accommodate additional channels between the individual TV channels. An example of this is shown in FIGS. 42 and 82. In Fig. 42, the carrier BTz is seen before this. The carrier is also used for modulation. He is accordingly z. B. of Fig. 66 co dated. As can be seen from FIG. 42, the carrier or carriers for the sound information are also provided in the image channel gap. At VHF it is e.g. B. he required to provide a series re sonanzkreis at the next higher television channel, in Fig. 42 it is the pass or suction curve. The resonance resistance is only as large as the loss resistance. The Nyquist flank is hardly affected. When the composite signal is modulated at 38.9 MHz, a suction circuit Rr is arranged after the residual sideband filter, as is shown in FIG. 82. Such a series resonance circuit is easy to implement. In Fig. 82 is a block diagram of the placement of an information channel between two television channels is shown. The image signal B (luminance signal), the modulated color carrier oscillation F and the blanking and synchronizing signals AS are added in the adding stage to the composite signal. The FBAS signal is then fed to a modulator Mo with the carrier frequency of 38.9 MHz via an amplifier stage. The amplitude-modulated signal is then fed to the residual band filter, so that the lower sideband is partially suppressed, as is known. After the RFi, the series resonant circuit is arranged in the circuit. The resonance frequency here is 37.9 MHz. The resonance curve is designated RR in FIG. 42. At 6 MHz above the image carrier, the additional channel receives a carrier or coding frequency of 189.25 + 6 = 195.25 MHz in the example. This is then the frequency of the coding according to FIG. 66. In the example, who summarizes the multiple channels time-multiplexed on a PAM basis (K1-X) and supplied to the encoder. In this is still a PAM / PCM converter, which converts the serial incoming PAM pulses of the channels K1-X into a binary, duobinary or other code. An oscillator feeds the encoder with the carrier frequency of 195.25 MHz. This alternating current is then modulated analogously to a code from FIG. 66 with the PCM pulses. The modulated carrier is then fed to a decoupler E, to which the modulated sound carrier is switched on. Both signals are then optionally via an amplifier of a crossover W, to which the carrier of the composite signal is connected, in the example 189.25 MHz. The sound carrier has a frequency of 194.75 MHz. All carriers are thus guided to the common antenna via the switch. The series resonant circuit for the additional channel of 195.25 MHz is thus arranged in the television channel with the transmission frequency of 196.25 MHz. The evaluation of this additional channel can, for. B. according to the sound channels.

At very high transmission frequencies you can use the additional channel also accommodate a television channel.

The digital code of FIG. 66 can also be modified as an analog / digital code. An amplitude change occurs for each value. For illustration purposes, the PCM technology is used. With this technique, e.g. B. 256 quantization stages are provided. 256 periods could therefore be provided for the largest PAM value. At half the size, 128 periods would be necessary. If this were the following value, these periods, which are always of the same size, would assume a different amplitude size to distinguish the coding from the beginning or end of a value. In practice one would of course do this differently, e.g. For example, assign 400 periods to the largest value. The value 1 would then have 143 periods, half the size would have a period number of 143 + 128 = 271. Depending on the frequencies available, even higher period numbers can be used.

In Figs. 68 and 69, a method for coding the chrominance signals is shown in red and blue. In this case, the color characteristic values are tapped at a predetermined frequency and modulated on each carrier, which are 90 degrees out of phase with respect to one another. The carriers have at least twice the tap frequency. These are summed up. The total alternating current contains the position of the color vector in the color circle due to the phase shift compared to a comparison alternating current. This phase shift is determined by the period or remaining period compared to the comparison alternating current. Storage of these values is necessary for a double carrier frequency up to half the number of taps in a line, and for three times the carrier frequency 1/3 of the taps in a line. The values of the phase shift are included in the transmission in half or period of an alternating current. In the case of cable transmission, a period in which the positive and negative half-wave have the same values can be provided to maintain the freedom from direct current. In the example, the carriers have a triple tap frequency.

In FIG. 68a, the sampling pulses P1, P2, P3,. . . of the color difference signal BY shown. These are expanded in a step-like manner - shown in broken lines. A step-like expansion is accomplished by a capacitor storage with a certain time constant. In FIG. 68b, the taps P 1, 2, 3 are. . . of the color difference signal RY with the step-like extension. In FIGS. 68c and d, the two carrier alternating currents are illustrated with the step-shaped modulated signals. The two modulated carriers are now added. A sum current is then obtained as shown in FIG. 68e. The amplitude corresponds to the size of the color vector, this is a measure of the saturation of the color and the phase shift compared to a comparison phase then corresponds to the color tone in the color wheel. This is already known for the NTSC and PAL systems and is therefore not dealt with in more detail. The output or comparison phase Vg is shown in FIG. 68f. The phase shift therefore always remains in the example during the 3 periods of the carrier Su. Immediately in the event of a phase change, the half-period cannot be measured, which is why in the example at least 3 periods are provided until the next phase change. As can be seen from FIG. 68g, a coding half period is composed of 2 constant periods KP and the actual coding phase shift Ph, which is almost a period at 359 degree phase shift. The individual processes before the transfer of the phase shift to the period is shown in FIG. 69. For this, 3 color circles are shown with the phase shifts 60, 120 and 240 °. The start of the measurement is designated in FIG. 68g and in FIGS . 69a, b, c with Ph0 °. In the example, the burst would have a phase angle of 0 ° as shown in FIG. 69d. A burst is not necessary in the example because the transmission is determined by the absolute period value. It is advisable to code the beginning of the serial arrangement of the code elements for each line. The half period, i.e. the pulse of z. For example, in Fig. 75 controls the ER relay, begins in Fig. 68f with Be and lasts the two periods and additionally the size of the phase shift Ph. In Fig. 69a, a phase shift of 300 ° is measured with a phase shift of 60 ° . The total pulse is then equal to the two perio + the length of 300 °. This pulse is e.g. B. ER relay amplified and then converted to a sinusoidal coding alternating current via a filter, as already described. The length of the half-period in a television line is thus smaller than the distances between the sum of the stair pensignale the color difference signals. Therefore, the sum of the half-periods must also be measured and, if necessary, a filling half-period must be inserted, to which one appropriately assigns the duration of 3 periods of the comparative alternating current. The alternating currents me shown in FIGS. 69a, b, c are total alternating currents Su. In Fig. 69b the color angle is 120 °, measured 240 ° and in Fig. 69c the color angle is 240 °, measured 120 °. The measured period is added to the two constant periods KP. In the examples, the saturations are 100% and 70%, respectively. As with the PAL system, these go into the amplitudes. In FIG. 69d of the phase comparator AC is Darge provides. The period of a period of the total alternating current includes the phase angle of 360 degrees. To get even greater accuracy, you can provide 180 degrees for a period by making an additional mark. If the phase shift is up to 180 degrees, a phase shift over 180 degrees is measured, as can be seen from FIGS. 69a and b. Here it is only necessary to measure the phase shift of the positive half period. Since only the period is required for the coding of the pixel size in the half-periods, one can code by means of an amplitude code that the angle is more than 180 °. During the transfer, you can then transfer twice the value of the angle. In Fig. 69a the angle size dw can then be twice as large. The amplitude code must then be evaluated in the receiving station and the additional 180 ° angle must be taken into account. The color and pixel coding can be transmitted in parallel in a manner similar to that disclosed in FIG. 58 or in series as shown in FIGS. 70 and 71. In the examples, the pixel pickup is twice as fast as that of the color signals. Since the code alternating current is to take place in an uninterrupted sequence of positive and negative half-periods, there is no synchronism between tapping and coding. For this reason, more or less storage is required on the sending and receiving sides. At the receiver, the pixel and color signal assignment must be made exactly with the tap frequency of the transmitting side. The color vector size, designated VS in Fig. 69a, is encoded by the amplitude size. This is saved like the period. A possible arrangement for the transmission is shown in FIG. 70. The required frequency is determined by the image tap number and the color type tap number. If 832 pixels are to be tapped in one line and a half period is required for each pixel, 416 periods are necessary for the pixels. For the color coding, 1 half-period must be provided for every 2 pixels, i.e. 213 periods for one line. These 629 periods are e.g. B. a time of 52 µs available. The frequency, and in fact the smallest frequency of the coding alternating current, is then predetermined. The blanking interval of 12 µs receives the same frequency. Since the coding half-periods are always smaller than the calculated ones, filling half-periods must be provided which expediently have the greatest period. Another code can of course be provided for these. The pixel half-cycles always have the same amplitude size, while in the color half-cycles the color vector is coded with the amplitude size, that is, the saturation is coded. The largest amplitude size of the color half-periods, which are then also provided for the pixel half-periods, can be provided as coding for the filling half-periods. The storage of the amplitude size of the color or saturation vector can be carried out by means of a capacitor which is switched to the coding alternating current Su via a diode. The pixel half-periods B (Y) can also be overlaid with a binary or duobinary amplitude code, with which the speech and other signals are then digitally encoded, as already described in FIGS. 52, 55. In Fig. 71 a pixel coding is still assigned a binary coding for a phase angle of the hue greater than 180 °. In Fig. 69a z. B. is found that the remaining half-period dw is positive, so that the negative half-wave no longer needs to be measured. These 180 ° are determined by this coding B + 180 °. The value dw becomes twice as large during transmission, so that the accuracy increases. The further pixel half-period B + T / S is overlaid with a binary or duobinary amplitude code, with which the digitized speech and other control signals are then encoded. The half-period F includes the hue angle in the half-period and, analogously in the amplitude size, the color or saturation vector.

In FIG. 72, the principle of an interconnection of the half-periods is shown by the amplitude code. An electronic relay again supplies rectangular pulses RJ. The periodic duration of these rectangular pulses is marked via Ph. In the folder Or, memories are provided in which the pixel taps may have already been implemented in half-periods. A memory for storing the color angle KP + Ph is also provided. The amplitudes of the rectangular pulses are fed to the electronic relay ER in the folder Or in synchronism with the half-periods. The analog amplitude size of the color vector is retrieved from the memory via FA. About T + S, the digitized sound and other signal amplitudes are retrieved from a memory and the electronic relay in the order z. B. Fig. 70 supplied. In Fig. 7 such an electronic relay is shown with several on the plitudenstufen. At the output of ER are then the rectangular pulses with the corresponding period lengths and the amplitude levels. In the filter Fi, these are then converted to a sinusoidal alternating current.

The direct measurement of the phase angle is also possible. The zero crossings from BE Fig. 68g then have to be counted in the sum of men alternating current, so that one then z. B. can determine the zero crossing M in Fig. 69b. From this point the measurement is carried out up to Ph0 °. Ph in Fig. 68g could then e.g. B. distribute to 90 ° of the phase shift. The 3 remaining 90-degree angles would then have to be coded with 180 °, similar to FIG. 71.

The coding alternating currents of Fig. 70, 71 will be the transmission alternating current is modulated and transmitted. In principle, the receiver is then switched as shown in FIG. 23. The input signals are routed via the tuning circuit / amplifier HF mixer / oscillator via the intermediate frequency amplifier V to the demodulator DM. The demodulated coding change current z. As shown in FIG. 70 to the decoder DC is supplied. In the case of a serial arrangement of the coding half-periods, the distribution of the image and color values ( FIGS. 70, 71) must take place in accordance with the image point and color difference signal taps. Therefore, it will be useful to send an alternating current at the sampling frequency in the blanking time, which is then synchronized for the synchronization of the alternating current provided in the receiver as a distributor. The amplitudes of this synchronizing alternating current of the blanking time can still be provided for a binary or duobinary coding. The pixel values can be e.g. B. evaluate accordingly in FIG. 59. The evaluation of the color code elements with the half-period values and the amplitude sizes, which if necessary is converted into a length similar to that shown in FIG. 74a, is best carried out on a computational basis. The sound (stereo) and other signals, which are PCM coded, can be demodulated in a known manner.

In Fig. 67a, the pixel taps BAb and the Farbdif reference signal taps FAb are shown and additional how to z. B. with serial transmission ( Fig. 70) the coding alternating current (Cod).

Fig. 79 shows a particularly convenient encoding. The encoding method shown in Fig. 79 works with binary code elements which have the form of one period or a half period of a sine wave and where the binary values 0 and 1 tude through a small or large amplification of the period or half period of the sine wave are represented as it is known from the above-mentioned US Pat. No. 4,675,721. In Fig. 79 it is assumed that the code elements are composed of alternating positive and negative half-periods of a sine wave, with a relatively small amplitude of the half period (z. B. twice as large) the binary value 0 and a relatively large amplitude of the half-period represents a binary 1 . The code shown in Fig. 79 allows the representation of a nine-digit binary number. It is structured as follows: The bits of the first and second digits of the nine-digit binary number to be coded are represented by the first, positive half-wave 1 and the subsequent second, negative half-wave 2 of a sine wave of the period P, as shown in curve ª in Fig. 79 is shown.

The bits of the third, fourth and fifth digits of the nine-digit binary number to be coded are represented by the successive half-waves of a sine wave of the period 2P / 3, as curve b in FIG. 79 shows. The sixth, seventh, eighth and ninth bits of the nine-digit binary number are represented by the successive half-waves of a sine wave of the period P / 2, as curve c in FIG. 79 shows. The sine waves in curves A, B and C are all shown in the first period P1 to simplify the illustration with the same amplitude, so the encoded binary number would consist of nine ones.

In the following periods P2, P3,. . . then they will following nine-digit binary numbers in the corresponding Way coded. In the curve a is the period P2 Value 00 and in the period P3 the value 10 is shown. The oscillation trains arising during coding correspond According to the curves a, b and c are additively overlaid and can then be transmitted over a single line will. The vibrations are on the receiving side a, b, c separated by filters and can then in known Way to be decoded, e.g. B. by measuring the duration the respective half-period, as described in the US patent 47 94 621 is described.

The coding method described with reference to FIG. 79 can be modified in a wide variety of ways. For the gradation of the period lengths of the oscillations for coding the successive bits, it is only necessary that a whole number of code elements (full or half sine oscillation periods) have space within the period P, which is reserved for coding a character. At the beginning and at the end of each coding period P, the different coding oscillations a, b, c etc. must always have the same phase position. You can also work with duobinary amplitude levels, so that you get 3⁹ combinations. With an additional, superimposed coding alternating current of this type, phase-shifted by 90 °, 3¹⁸ combinations are then available. A duobinary coding is e.g. B. used in the European D-MAC system.

The coding capacity of the method of Fig. 79 can thereby be further increased in that each Codierschwingung a b, c, and so a phase shifted by 90 degrees second Codierschwingung the same period assigns, as with respect to, for example, in the curve a 'in Fig. 79 curve a is shown. The two vibrations, e.g. B. a, a ', the same period but in quadrature phase are then superimposed, resulting in a vibration a'', which has the same period as the superimposed individual vibrations gene and therefore separated on the receiving side by a filter and by synchronous modulation can be broken down into the component vibrations again. If this measure is applied to the method described with reference to FIG. 79, then eighteen binary characters can be transmitted in a coding period P.

Fig. 77 is a circuit arrangement according to the invention for coding a color television signal, which contains a luminance signal L, color signals I and Q, synchronization signals, if necessary, additional signals S and sound signals T. The luminance signal is tapped at a predetermined frequency and fed to an analog / digital converter A / D, which converts the samples into 8-bit code characters. The code elements (bits) of the code characters are fed to a memory Sp. The color signals I and Q are tapped simultaneously at a frequency that is equal to half the tapping frequency of the luminance signal, and buffered in a buffer ZSp. The signals stored between are alternately supplied to an analog / digital converter A / D via a switch U4, which converts the samples into 6-bit code characters. The code characters are then fed to the memory Sp. The synchronization signals and possibly other signals S and the sound signals T, which can be stereo signals, sound of different languages, etc., are tapped alternately or simultaneously at a predetermined frequency, temporarily stored in a buffer ZSp and an analog via a switch U5 - / Digital converter A / D supplied, the z. B. 8- or 16-bit code characters are generated, which are also supplied in the memory Sp. An encoder Cod is coupled to the memory Sp, of which the eight code elements (bits) of the luminance signal, three code elements of the I or Q signal via switches U1 to U3 and a code element of the S or T signals are retrieved, as shown in Fig. 78 in column I. The short or long lines here mean a bit with the value 0 or 1. These twelve bits are converted by the encoder Cod into a common code character, for example as was described with reference to FIGS. 19 and 79. With the successive code characters, see columns I to IV, etc. in Fig. 78, a conventional high-frequency transmitter M is then modulated and the modulated high-frequency oscillation is emitted via an antenna S. On the receiving side, the high-frequency oscillation is received by an antenna E, processed and demodulated in a conventional receiving part ET in a known manner, and the demodulated code oscillation is fed to a decoder Dcod, as was explained with reference to FIG. 23. The code words are demodulated in the decoder Dcod. The eight bits of the respective samples of the luminance signal L are then available simultaneously on a group 1/8 of eight output connections; on a group 9/11 with three outputs the color signal bits, which are buffered in a buffer Sp1 and on an output 12, the S / T bits that are buffered in a memory Sp2. The luminance signal bits are converted into an analog luminance signal L in a digital / analog converter D / A. The color signal bits are temporarily stored in the buffer Sp1 and if six bits of an I and a Q signal sample value are available, the now complete color signal code characters are converted into analog color signals by a digital / analog converter D / A an analog memory Sp3 is fed to a corresponding output I or Q.

The S / T bits are stored in the buffer Sp2 and, if full code characters are available, in a digital-to-analog converter D / A in corresponding Analog signals implemented in an intermediate as needed store Sp4 again temporarily or directly can be processed further. During the blanking period can the transmitter M via a switch U6 others Code signals X are supplied on the receiving side available at an output AT of the decoder DCod stand and an appropriate recycling can be.

Fig. 83 shows simple 4-PSK phase shift keying. The nominal frequency is f corresponding to a nominal period of 360 °. The coding is carried out by phase changes from + 45 ° to a phase of 405 ° = f1, from + 135 ° to a phase from 495 ° = f2, from -45 ° to a phase from 315 ° = f3 and a phase change from -135 ° to a phase of 225 ° = f4. The generation is carried out analogously to the arrangements of FIGS. 7 and 8. This arrangement replaces the previously used coding according to FIG. 2. In FIG. 4, such phase jumps are generally described.

In Fig. 80 different periods are Darge presents. If the evaluation is carried out by dimension, it is expedient to determine clearly measurable period duration differences, as was done at 0 °, a °, b ° and 90 °. The distances between 1, b, a and 2 should also be pretty much the same. In Fig. 81 phase differences of over 90 ° are provided. A disadvantage is that the frequency changes are very large. In the case of cable transmission, in order to be free of direct current, the coding must be carried out with the same positive and negative half-wave.

In Fig. 19b is a particularly advantageous coding for code multiplex information transmission Darge provides. Half-waves of an alternating current are provided as code elements, which are sent in an uninterrupted sequence of positive and negative half-waves. An amplitude code with 3 characteristic states, i.e. a binary binary code, is provided. A period then has two digits. In order to increase the number of digits, a second coding alternating current of the same frequency, which is however phase-shifted by 90 ° is provided. You get 4 digits from each period. From the two coding alternating currents one obtains 3 to 4 combinations from a period, these are 81 combinations. In the figure, the half-wave aP1 has the characteristic state 1, the half-waves aP2 = 1, aP3 = 2,. . ., the half-waves aP11 = 0, aP12 = 0, aP13 = 1,. . . For the transmission, both coding alternating currents will be added and transmitted as only one total alternating current.

The narrowband transmission of information according to FIG. 66 can not only be used for television, but also very generally for the transmission of digital or analog information via radio or cable. An example of the voice or sound transmission is shown in FIG. 84. The 4 voice channels K1-K4 are tapped at 4 × 8 = 32 kHz by the multiplexer Mu. These PAM pulses are fed to a PAM-PCM converter and, in the example, converted into a binary amplitude code. The amplitude of the positive and negative half-wave of an alternating current is provided as a code element (see patent DE 30 10 938). The 256 kHz pulses are converted into 4 periods of corresponding amplitude in the encoder Cod. This is done with an arrangement according to FIG. 9. A generator generates 1024 kHz and supplies this alternating current to the arrangement of FIG. 9. As already described in FIG. 66, the change in amplitude levels is carried out continuously. In Fig. 85 this is for Fig. 84 represents Darge. The characteristic state 1 (1) is followed by a 0 (2) which has half the amplitude. The 4th period of characteristic state 1 (1) and the 1st period P5 of characteristic state 0 (2) form the transition to characteristic state 0. The same is with the 8th and 9th period, with a change from characteristic states 0 to 1 Darge is the case. This alternating current, code alternating current, is raised in 4 multiplication circuits n = 4, 2, 3 and 4 to a frequency of 98304 kHz and switched to the transmission antenna. This multiplication results in 96 periods from each period. The evaluation of the Codewechselstro mes z. B. with the help of a diode rectification or with the help of coherent demodulation. As a result, it is then distributed to the individual channels. The language or sound is then restored in a known manner, e.g. B. by making the pulses step-shaped and then giving them a low-pass filter.

Through resonance circles one can change several parallel coding disconnect currents. The transfer can also be based single sideband modulation. The coding can also be carried out in the final stage, whereby in the Follow a conversion to higher frequencies that can.

Instead of a binary-coded transmission, the PAM pulses can be pressed directly onto the periods of the alternating coding current. Each pulse can then e.g. B. 100 times as a period with the same amplitude. For speech with a tap frequency of 8 kHz would be z. B. a frequency of 800 kHz is required. The principle of transmission with the half-wave amplitudes has already been patented in Canada under No. 12 14 277. You could e.g. B. also combine 20 channels time-multiplexed and transmit in this way narrowband. A channel could be used as a comparison channel e.g. B. are always coded with the largest possible amplitude. In order to keep the code alternating current always above the noise level, the samples are given a corresponding DC voltage specification. The principle of the invention is shown in FIG . FIG. 86a shows the PAM pulses P1, 2, 3,. . . Fig. 86b shows the associated with the pulses with periods of equal half-wave amplitudes. P1 includes the positive and negative half-waves aP1, aP2, etc.

In FIGS. 84, 85, 86, the amplitudes of the half waves were the characteristic states for the encoding of the information. Instead of the amplitudes, the phase can also be provided as a characteristic state. You can then make the phase jump in stages. This is explained in more detail in FIGS. 5 and 6. A gradual, even phase change can simply be brought about by a frequency change. In Fig. 6, the phase shift of a period T4 / 2 compared to a comparison period T / 2 is 10 °. With 4 periods the difference is then 40 °. If the number of phase stages is very large, only a small change in frequency compared to the comparison frequency is required. The comparison frequency is assumed to be 1 MHz. If a phase shift of 0.5 ° should now take place per period, after 720 periods there is a phase shift of 360 ° compared to the comparison alternating current. The frequency for the leading and lagging phase shifts is then 720/719 and 720/721. Circuits according to FIG. 7 are expedient for such continuous frequency changes. Has such a counter z. B. 721 outputs, then the comparison alternating current is controlled up to output 720 and the leading phase shift up to output 719 and the lagging phase shift up to output 721 . The alternating current frequency of the oscillator should therefore be very high at 1 MHz, so that lower frequencies are generated via the counter and then the alternating current tapped in FIG. 7 at Fi is multiplied. Such frequency generator circuits are already disclosed in my U.S. Patent 4,794,621. In such a coding, the number of periods and the size of the amplitudes can also be provided as parameters in such a coding. If the phase angle of 45 ° is to be used as the characteristic state, 90 periods have to be counted and a phase comparison with the period of the comparison alternating current has to be made in the 90th period. Such circuits are known from color television (burst).

In the color coding and color transmission circuitry, greater storage of the coding was required. In Fig. 87, a possibility is now shown how one can also directly transmit the color signals coded in this way at most with a delay of a value. The principle has already been explained in FIGS. 33 to 35. In the example, the carrier and the comparison alternating current are assigned 6 times the frequency of the tapping frequency. In Fig. 87a is the comparison frequency and in Fig. 87b the phase shift ent containing total alternating current. After 2 periods of the comparison alternating current calculated from the last color tap, the phase shift is measured on the total alternating current. The actual phase shift is Ph, as already explained in detail in the description of FIGS. 68, 69. - You could also save these alone. - Together with the 2 following periods KP these z. B. transmitted with a half wave of the period n duration KP + Ph. The duration P is also transmitted with a half-wave with the period duration P. Since P and KP + Ph are the same as the 6 periods of FIG. 87a, no large intermediate storage is required. The saturation can then be placed in the amplitude size. The transmission of the color signal values can also be followed by a plurality of periods according to the principle of FIG. 84 or 86. A conversion of the phase shift duration into a PAM pulse can, for. B. be carried out according to the principle of FIG. 59. A distinction between the phase shift duration and the duration of the break can, for. B. done by the size of the amplitudes. The saturation of the color could also be when the Y signal is encoded by the half-period, such as. B. in Fig. 51, code by the amplitude of this half wave.

In order to be able to transmit even more information in a narrowband manner, two coding alternating currents can be provided, which are phase-shifted with respect to one another by 90 degrees and have the same frequency. The principle is shown in FIG. 88a. The signal S is fed to an encoder Cod. This divides them into 2 coding alternating currents in accordance with FIG. 19b. A binary or duobinary etc. amplitude code can be seen. One encoding alternating current has the frequency f and the other has the frequency f 90 °, that is, it is phase-shifted by 90 degrees. The generation of such phase-shifted alternating currents can, for. B. done according to the principle of FIG. 8. As a result, the frequencies are then multiplied and interconnected in the adder Ad and the transmission path, for. B. the transmit antenna supplied. As can be seen from FIG. 88, when the amplitudes change in the alternating currents shifted by 90 °, there are also phase jumps or frequency jumps. However, this means that a wider frequency band is required. A binary coding is provided in FIG. 88. One vector changes from Uk + U to Uk and the other from VK + V to VK. The sum vector can thus take the phase from Üvo to Üuo. In order to make this phase change possible without a large frequency change, the amplitude changes are carried out in stages, namely with stages 1-n. Each period is assigned a level, which means that a large number of periods are required to change a binary value. 2 Examples of how such changes can be made are shown in FIGS. 89 and 90. In FIG. 89, circuits corresponding to FIG. 9 are provided for this. Corresponding resistance values R1 to Rn are switched into the AC circuits according to the number of stages. The alternating current f is phase-shifted once by 90 ° and once switched directly into the arrangement in FIG. 9. In FIG. 88a the circuit of FIG. 89 is contained in the encoder Cod. These amplitude stages can also be generated according to the principle of FIG. 7 and the phase-shifted alternating currents according to FIG. 8. The electronic relay ER switches corresponding amplitude stages 1 to n according to the code. The rectangular pulses present according to ER are converted into sinusoidal alternating currents via filters and into interconnected in the adder.

From the described possibilities of narrow-band coding, higher-quality coding can also be provided. An example is shown in FIG. 91. It represents a 9-level coding. The levels are 100, 150, 200 periods, these are one phase leading Phv and one lagging phase, so that another 6 levels are obtained. 4, 5 and 6 is the 2nd digit. The change in position can, for. B. are displayed by an amplitude change. A code-multiplexed application is also advantageously possible. The Fig. 78 may, for. B. are so shaped. 8 bit luminance, 4 bit for the other signals. The following sequence then results: 8 + 4 (red), 8 + 2 (r) + 2T / S, 8 + 4 (blue), 8 + 2 (bl) + 2T / S, etc. For sound and other signals 1 bit per luminance signal is sufficient. Only in the case of television are some parallel channels of this type necessary in order not to obtain frequencies that are too high.

The application of the inventions to facsimile machines is explained below. An overview circuit of fax machines is shown in FIG. 92. The reading unit L has the task of converting the original to be transmitted into analog electrical signals. They are then converted into digital signals in the encoder Cod and modulated accordingly for transmission in the modem mod. The AS interface unit adapts to the telephone network. The received signals go to the modem / decoder Decod via the AS and are converted back to their original form in this unit. The paper is then recorded in the recording unit Az. A central control ZSt controls the remote copying system and coordinates the remote copy transmission. Control takes place from control panel B.

First, the application of the invention to group 2 devices will be explained. The scanning methods, such as the CIS method, are not dealt with because they have nothing directly or indirectly to do with the invention. Group 2 scans point by point regardless of the information content. Since there are only white / black differences, there are only 2 types of code elements, i.e. binary code elements. According to the invention, the stored code elements, here binary code elements, of 2 or more lines are arranged in parallel and combined and transmitted to form a code word. Scanning 2 or more lines at a time is not as economical as storing. A simultaneous transmission of 6 lines Z1-Z6 is provided in FIG. 93. In this example, only 1 code element is used for the code word from each line. S1, S2, S3,. . . are the binary code elements that are combined to form a code word. You could e.g. B. Combine S1 + S2 into a code word. A code for code division multiplex coding is e.g. B. shown in Fig. 19b. In order to encode 6 binary values, 64 combinations are required. In the method of FIG. 19b, if the half-waves are only carried out in two stages, one and a half periods per alternating current are necessary. With 3-stage, i.e. duobinary execution, you can accommodate significantly more lines. With this method, a comparison alternating current corresponding to the burst in color television is required for the decoding. One of the two alternating currents can be used for this. This can e.g. B. can then be provided as a starting character. If it is necessary to be free of direct current, the period can be provided as a code element. A phase or combined phase / amplitude code can also be used for the code-multiplexed summary. In Figs. 4, 19a, 45, 46, 63 are, for. B. such. You can also provide multiple AC coding currents, such as. B. explained with FIGS. 28, 79. A phase jump keying can be generated more easily via square-wave pulses according to FIGS. 4, 7, 38. In the Fig. 94 3 phase jumps are shown namely 360 °, 360 ° -90 ° and 360 ° + 90 °. The change in the period represents the phase jump. FIG. 95 shows an example of such a generator of phase jumps. An electronic relay is labeled ER. In the example, it switches plus and minus potential to the output. The duration of the connection is determined by the encoder Cod. The duration of the connection is then equal to the period of the generated rectangular pulse. With this method different phase differences or phase positions can be generated. The rectangular pulses are then converted into a sinusoidal coding alternating current with a filter Fi. In Fig. 94, the rectangular pulse 3 has the same period as the rectangular pulse 1, but this is 90 ° out of phase. Fig. 94 thus has Phase 3 stages; If 2 amplitude levels are added, 5 to 4 combinations can be achieved with 2 periods, so that more than 9 bits can be coded, ie 9 lines could be transmitted at the same time. In FIG. 96, a code having 2 phase levels is leading and lagging v n 2 shown amplitude levels.

In Fig. 97 an overview of a multiplex code encoding is illustrated. The respective values are tapped via a multiplexer and fed to a memory. The values tapped at the same time by the encoder are coded in a code-multiplexed manner and fed to the transmission path via the interface unit. The start and end or only the end character of a line marking or line group can be marked by one or more parallel code words. In Fig. 98 there are 4 × 6. In the process of QAM z. B. according to FIG. 19b, possibly also with a phase coding, one can in the telephone bandwidth 2 or more codes in parallel with different frequencies or frequency positions, such as. B., accommodate, so that the transfer time can be reduced further 24 and 79 shown in Fig..

There are 16 shades of gray in today's halftone transmission. 4 bits are required for coding them. With coding according to the principle of QAM Fig. 19b one could e.g. B. with 2 periods with an amplitude binary code 2 lines transmitted simultaneously. Of course, this is also the case with a phase code. B. possible according to Fig. 19a.

In group 3 of fax machines z. B. also used the one-dimensional run length coding according to the MHC method. With this method as well, according to the invention, a further reduction in the transmission time can be achieved without any loss of transmission security. Such a coding example is shown in FIG. 99. In this 4 lines are provided for simultaneous transmission. - There can also be more or fewer lines. - Line 1 begins with the start character, in the example lines 2-4 begin with the same characters. If there is only one code element from each line, e.g. B. S1 = 4 coded, 4 bits are required for a code word. With 2 code elements per line, i.e. in the example S1 + S2, 8 bits are required. With the help of filler bits, every 4 lines can be brought to the same number of code elements. The fill bits can also only be provided after the end indicator. The code word EOL (End of Line) can be distributed over all lines at the end of the line, the same also applies to the start character, which can be distributed over the first 4 lines. In Fig. 100, the beginning and the end identifier EOL is shown distributed over the 4 lines. It is expedient to provide a memory after tapping, as shown in FIG. 97. With a binary code and QAM coding according to FIG. 19b with a half-wave as a code element, only one period each is necessary for the 4 bits. Any code can of course be used for code-multiplexed coding. Since a code word always has the same number of code elements, an error detection can be made possible by counting. The sequence of positive and negative half-waves can also be used. The same applies to phase coding. For the identification of the last line, a special identifier distributed over all 4 lines can be provided, or the very last line can be used as a final identifier, e.g. B. provide.

Another method of reducing transmission time is below explained using an example. An example after the MhC code is used for this:

Line 1: EOL, 21 ws, 3 sw, 6 ws,. . .
Row 2: 2 ws, 6 sw, 23 ws,. . .
Line 3: 10 ws, 6 sw, 23 ws,. . .
EOL = 11 ws and 1 sw.

According to the invention, only the numbers are encoded. The digits 1 to 0 are to be coded white and the digits 1 to 0 are also black. This requires 20 combinations, i.e. 5 bits. If a line is completely white, then if it is not exactly the first or last line, an end indicator and 1728 white samples must be coded. With a binary code and a coding according to FIG. 19b, 5 half-waves are required. (10 half-waves with no direct current). 4 × 5 half-waves are required for the white scans and 3 × 5 half-waves for EOL. You can also provide a duobinary or phase amplitude code, then you get by with even fewer half-waves. Of course, you can also use previously used codes. Of the 32 possible combinations, 10 are intended for white, 10 for black, a special combination could now also be provided for other license plates, such as EOL etc. So that no more than 3 digits can be coded in one number, one can e.g. B. from 1000 to 1728 encode the first two digits by their own combination, i.e. 10, 11, 12, 13, 14, 15, 16 and 17. Whether the taps are white or black is determined by the code of the following Digit set.

Another variant for coding and transmission is the following. Because the barrel lengths always alternate between white and black you can first find all white lengths and then in then encode and transmit the black lengths. condition for this is that the same number of taps is always used will, e.g. B. 0 to 9 or 00, 01,. . . 99, so always either one or two digits. A separate code word is then used for Black coding provided. As a result, only those Digits sent for black. With 2-digit run lengths a code for 100 combinations is required. Then come the special codes, such as for EOL, switching to black etc. added. Again, you can have 2 or more lines at the same time encode and transmit.

FIG. 19b shows an amplitude code in which the half-waves of two coding alternating currents with 3 amplitude stages are provided as code elements. The half waves are transmitted in an uninterrupted sequence of positive and negative half waves. The coding alternating currents are 90 ° out of phase with one another. - The half-waves can of course also be designed in two stages, i.e. in binary form. - One period of the two alternating currents then results in 4 code elements. Duobinary you get 3 to 4 and binary 2 to 4 combinations. Both are added for the transmission. The code elements of the one alternating current are aP1, aP2, aP3,. . . and that of the other alternating current aP11, aP12, aP13,. . . The separation of the two alternating currents on the receiving side is not dealt with in any more detail because one is already known. If the respective code word has an odd number of code elements, an alternating current will alternately be assigned to the odd code element.

With alternating white / black barrel length coding, for example you can get by with 16 combinations if you only the digits 1 to 8 are used. With these digits you can the numbers 1-8, 11-18, 21-28,. . . 81-88, 111-118,. . . 181- 188, etc. encode. The numbers 9, 10, 19, 20 etc. must then are coded by non-assigned numbers, in which only the Numbers 1-8 occur. With binary coding, 4 bits required. The numbers 1-8 mark white and the numbers  9-16 black. You could z. B. the barrel lengths only 8 digits then you would come up with one digit per barrel length out. For EOL you would need a special code from the digits Set 1-8. You could also use the two-digit number make, so 1-72. The number 9 is then z. B. with 81, 10 with 82, 19 with 83,. . . 40 with 88, 49 with 73, 50 with 74,. . . 70 replaced with 78. You can also z. B. make until 1728. The lengths not marked with the numbers 1-8 z. B. 89, 90, 91-99, 100, 101,. . . must then be encoded with numbers over 1728 will. But this code will be set so that often occurring numbers get as few digits as possible, like it is also the case with the MHC code. Again you can have 2 or Combine more lines and transfer them at the same time. At 2 lines are e.g. B. in place of 2 to the 4, 2 to the 8 combinations required. All types of coding can also be found here use.

All of the aforementioned methods can also be used in the MRC also apply to the MMR code.

A further reduction in the transmission time is then possible if first the lines of the entire template are saved and the lines with the same or almost the same Coding lengths are summarized, e.g. B. if each 4 can be summarized that 4 are the same or almost the same be summarized. This is also beneficial for the MHC code.

Color templates or color images should also be transmitted on the basis of faxes, in particular also via telephone connections, that is to say not only via ISDN connections. The exemplary embodiment of the invention relates to the coding and transmission of color television signals, in particular for fax machines. The coding can of course be done exactly as with color television, e.g. B. according to my US patent 46 75 721 or patent applications DE P 32 23 312, 32 26 382, 37 09 451. Some very advantageous types of coding are explained below. In Fig. 101, the primary color taps of green, red and blue (gr, r, bl) are converted directly into a binary code - a multi-level code can also be used. This creates the channels gr, r, bl. These colors can be coded with 8 bits. All 3 channels are combined in a code-multiplexed manner for the transmission. Instead of a binary amplitude code, you can also use a duobinary code, as in the MAC system. A combined phase / amplitude code, as shown in FIGS. 45, 46, brings considerable time advantages. Are z. For example, in Fig. 101 3 code elements are combined in series and in parallel, 9 bits are required. In the case of a binary amplitude code based on FIG. 19b, half-wave coding requires 2 1/2 and 2 periods of the two alternating currents, which are encoded alternately on the two alternating currents. A duobinary code requires 3 half-waves per alternating current. With a phase code with 3 phase levels and 2 amplitude levels, 4 digits of an alternating current are necessary. In Fig. 101, the color separation signals for green, red and blue were immediately encoded and transmitted. In FIG. 102, the luminance signal Y of 8 bit is now, the red color difference signal with bit 6 and the bit encoding of blue 4. (Of course, one can also provide 8,6,6). The intensity of the colors is taken into account here. The coding is carried out in exactly the same way as in FIG. 101. In Fig. 103 only 2 channels are provided, one for the Y signal and the other for red + blue, to which only half the number of bits is assigned, and in which the code elements are arranged alternately. Here, several code elements of both channels can be encoded and transmitted at the same time. Coding on the principle of double quadrature amplitude modulation DQAM according to FIGS . 9, 10, 11, 39 is also possible. The first QAM will be carried out on the principle of FIGS. 9 and 11 and thus coded red and blue and with a further coding alternating current Y.

One can carry out the fax operation also in duplex traffic over only one telephone line, if one applies the principle of FIGS. 25, 26 and 27.

Another type of coding and transmission of the color signals is shown in FIGS. 104 to 109. - This method can of course also be used for color television, whereby you can add 1 bit for the sound. As already shown in FIG. 68, the PAM taps can be made in a step-like manner. In Fig. 104, Y is tapped at twice the frequency as red and blue. Red and blue are tapped alternately with Y. Red and blue are also continuous stairs like the Y signal alone. All signals can assume the values plus and minus. As already described in FIG. 68, red / blue and Y are each modulated onto an alternating carrier current which has the 2 or more times the tap frequency. Both carrier alternating currents are then added for transmission or also for further use, as described in FIG. 68. The sum vector is thus transmitted, from which the two vectors for red / blue or Y can then be obtained again. Each period of the sum carrier contains the phase shift with respect to the two vectors. After 2, 3 or 4 periods, the transient process is ended even after each change in amplitude of the two vectors, so that no errors are obtained in this regard during the evaluation. Since the staircase signals are arranged bipolar, phase jumps of almost 360 ° are obtained, as is the case with color transmission on television. FIG. 107 shows the plus / minus vectors of Y and r / bl and their sum vectors SU. To avoid such phase jumps, the Y signal is arranged unipolar in FIG. 105. It can be seen from the vector diagram in FIG. 108 that only more phase jumps of up to 180 ° can then occur. The Y signal receives a DC bias for this purpose. If the staircase signals of red and blue are now unipolar ( FIG. 106), only phase jumps of up to 90 ° arise, as can be seen from FIG. 109. Of course, this principle can also be applied to color television. The two alternating currents are separated in a known manner. At the beginning of the scanning of an image, one of the carrier alternating currents can be sent as a phase comparison in the evaluation.

In Figs. 68 and 69 is shown a method, how to encode the color signals using PAM-step signals with the interposition of a respective carrier, which are added to a sum carrier in the form of a phase shift in conjunction with the amplitude. The phase position of the sum vector was transferred to the half period of a half wave. The dimension of the phase shift in the total alternating current could, as can be seen from FIGS. 69a and 69c, begin with the positive or negative half-wave. With a positive start, 6 zero crossings were counted for the dimension and 5 zero crossings with a negative start. An embodiment of such a measurement is shown in principle in FIG. 110. The total alternating current is applied to module G1 by using diodes to determine the positive or negative start. This is reported to the counter Zä, to which the total alternating current SU is also connected. The start of the measurement is signaled at the block Zä with pho. The respective memory Sp1 or Sp2 also receives the start signal. Switch S1 always switches one of the two memories on the counter. The respective memory takes the time measurement, i.e. the half-period, from the start character. The amplitude of this half wave is determined by the amplitude of the total alternating current. The amplitude cannot be determined immediately at the start of the measurement, which is why the memories Sp1 and Sp2 are required. The voltage values of the respective amplitudes are connected to the capacitors C1 and C2 via a FET, FET1 / FET2 and stored. The respective voltage value is then switched to the electronic relay ER via the switch S2. This value then determines the amplitude of the rectangular pulse J. From the memory, the z. B. can be a type of shift register, then the length of the respective pulse is determined at the ER. A filter can be used to obtain a sinusoidal alternating current from the rectangular pulses.

In the case of A4 templates, the vertical has about 1100 lines to do. Are always about the same length when transferring groups Lines combined, the lines must also be coded. It must can be distinguished whether it is the 1, 2,. . .20,. . . or 1100th line. These Line coding can of course also be used as an EOL identifier will.

A simple reduction in transmission time is also done in the way achieved by only using theirs from the lines with only white taps Coded number sequence, e.g. B. 10, 21,. . . Line, or just by looking at the transmission sequence of the lines of the respective white line Code for "white line" there. Especially with business letters z. B. in A4 format is not only between the lines written, but also on the head, d. H. a lot at the beginning and end of writing number of white lines. At z. B. 1728 taps per line is then to transmit only the line code word. The labeling can also be done by the sequence is done together with a code word. This method is not only in the case of code multiplexing, multi-line transmission at the same time applicable, but with all known methods and also with the standardized procedures of groups 1, 2, 3 and 4, i.e. also with one-line transfer. There is always a storage on the transmission side and depending on the case also required on the reception side. With only sw / ws- Pure white lines do not have to be printed out, son it can be the same with a white line as when saving in electro African typewriters can be switched to the next line. If you provide a line coding, you could e.g. B. the EOL label train as line code. A hardware of such memory is from  the Brother AX-45 electronic typewriter B. known. At a code-multiplexed multi-line coding and transmission can e.g. B. with 4-line transmission after the lines of equal value the lines with the next shorter coding lengths z. B. summarize and transfer, you must always also the code number of the respective Transfer the line. The order of transmission can, however done indiscriminately. The determination of the white lines can e.g. B. by The white taps are counted in one sequence. Are z. B. only 1728 white taps available, so there is a white line. There is many options here.

For most of the examples listed here, it is useful to to provide a memory with processor on the receiving side. The paper recording can also be made in several lines, the recording being units do not necessarily have to have the line spacing. One can e.g. B. the recording units a mutual distance of 10, 20th or assign 50 lines, e.g. B. 1, 11, 21, 31 - 2, 12, 22, 32 etc. at 4zeili eng recording and a space of 10 lines between the 4 recording units.

Claims (49)

1. A method for coding and transmitting information, in particular templates and images, and for television, characterized in that one code element or more code elements of 2 or more lines ( Fig. 93, 99) or channels ( Fig. 101, 102) can be combined to form a code word, regardless of whether there is a point-by-point or a run length summary or a two-dimensional summary or similar methods and / or whether a digit code is provided with one, two or more digits. 2. Method for the coding of templates and images with one-dimensional run length coding, preferably according to the MHC method, characterized in that the code words of white and black and possibly gray and the EOL 2 or more lines are arranged in parallel ( Fig. 99, Z1 , 2, 3, 4) and combined to form a new code word, filling bits being provided for the production of an identical number of code elements for the combined lines. 3. Method for coding templates and images preferably for Fax machines, characterized in that the numbered Barrel lengths for white and black with one, two or more digits be encoded and that the other signals such as EOL a predetermined Unassigned number is assigned to every same digit for white and black is assigned a different code combination. 4. Method for coding templates and images, preferably for fax machines, characterized in that the numerically recorded Barrel lengths for white and black, in continuous succession of white Run lengths and black run lengths are sent, with a flag the number of digits is intended for the change to black for the number of run lengths is always the same, in particular the Change in one line is provided. 5. Methods for coding templates and images, preferably for fax machines, characterized in that to reduce the number of code elements the code words and thus the transmission time is only a predetermined one Number of digits (e.g. 1 to 8) and then not encodable Numbers from the predetermined digits are assigned numbers not used will. 6. The method for coding colored templates and images, in particular for faxes and television, characterized in that the color values in particular red and blue or the color difference signals and the luminance or Y signal are pulse amplitude modulated, the tap frequency for the Y signal is an integer multiple of the tap frequency of the color signals, the taps of the color signals take place alternately and synchronously with the taps of the Y signal, the PAM pulses, are formed into staircase pulses so that the color signals are the one staircase signal and the Y Signal form the other staircase signal ( FIGS. 104, 105, 106), each staircase signal is modulated onto a carrier with the 2 or multiple tap frequency of the Y signal, the carriers being phase-shifted by 90 ° with respect to one another for further processing or Transmission is an addition of both carriers. 7. The method according to claim 6, characterized in that a bipolar PAM or stair signals are provided for all signals. 8. The method according to claim 6, characterized in that for the color signals a bipolar PAM or staircase signal and one for the Y signal unipolar PAM or a unipolar staircase signal is provided. 9. The method according to claim 6, characterized in that all PAM or staircase signals are provided unipolar. ( Fig. 106). 10. A method for coding colored templates and images, in particular for faxes and television, characterized in that the binary-coded color separation signals of the 3 primary colors are arranged in series and in parallel, combined in series with one, two or more digits, coded multiplexed. ( Fig. 101). 11. A method for coding colored templates and images, especially for faxes and television, characterized in that the binary coded signals with different bits Y, red and blue (color difference signals) are arranged in parallel and in series, if necessary summarized in two or more digits (e.g. Fig. 102, y = 3 digits, r = 3 digits, bl = 2 digits + 1 × Y), coded multiplexed. 12. A method for coding colored templates and images, especially for faxes and television, characterized in that the same number of bits as the Y signal is assigned to both color signals, the code elements of the colors being switched alternately ( Fig. 103). For the transmission, the same number of code elements of the color signals and the Y signal is combined in a code-multiplexed manner. 13. The method according to claims 6-12, characterized in that 2 or several lines can be combined in a code-multiplexed manner.   14. A method for coding information, characterized in that the digital and / or analog coding and the transmission of information of one, two or a plurality of channels with less bandwidth than the single channel or the sum of the bandwidths two or a plurality of Channels makes up in such a way that the synchronously or quasi-synchronously arranged code elements of the channels to be transmitted are arranged in parallel ( Fig. 20) and thus combined together to form a code word and / or that the digital or analog information to be coded, if necessary with the interposition of intermediate stages (e.g. PAM) are converted into PDM pulses, so that means are further provided which convert the values of the PDM pulses into the half-period or period duration of half-wave or periods of a sinusoidal or sinusoidal alternating current ( FIG. 35 , ER, Fig. 36, ER, Fig. 38 ER). 15. Procedure for transmitting digital signals a variety of different information channels about a single transmission channel in which each of the digital signals from a sequence of code units exists, characterized in that occurring simultaneously Code units from the various information channels be represented by a common code word, so that one of the several sequences of code units only sequence of code words arises; that the consequence of the code words are transmitted over the single transmission channel becomes; that the transmitted sequence of code words is decoded, for each codeword transmitted several code units for the different information channels are generated and that the code units associated receiving information channels are supplied. 16. Method for generating a frequency / phase modulation, characterized in that means are provided which convert information or a signal into pulse durations, where either the unipolar or bipolar pulse duration rectangular pulses directly by generation via a pulse duration modulation circuit or indirectly through generation  using the control of an intermediate electronic Relay or equivalent means or directly be supplied to such a screening agent, that too replaced by the properties of the transmission path if necessary can be that the rectangular pulses in sinusoidal half-periods or periods are transformed. 17. A method for generating a frequency modulation, characterized in that means are provided which convert an information or signal ( Fig. 30a, Inf) into pulse durations ( Fig. 30b, 32), that switching means for measuring the pulse durations, in particular Counter switching means ( Fig. 35, Z) are provided, which at the same time mark the pulse durations (e.g. Fig. 35, Z, A), the marking circuits are thus in connection with pulse duration pulses via gates with an electronic switching means ( Fig. 35, ER) that the beginning and the end of the respective pulse duration pulse encode a periodic signal, in particular a square pulse, furthermore screening means are provided that only sinusoidal or sinusoidal alternating currents and / or harmonics reach the line ( FIG. 35 , fmo). 18. A method for generating a frequency modulation, characterized in that means are provided which convert information or a signal into pulse durations and in that switching means are also provided which the continuous pulses in an uninterrupted sequence (Pd, Pd, Pd,... ) or the pulse duration pulses and the associated pauses ( Fig. 32, PD1, P, PD2). 19. A method for color television, characterized in that all signals are combined in a code-multiplexed manner, the color, sound and other signals in a code-multiplexed manner can be assigned to several Y signals as required, and in that the receiving side is designed as a superimposed receiver (superheterodyne) is, behind the demodulator ( Fig. 23, DM) the decoder is arranged with the timely distribution of the decoded signals. 20. A method for coding the color television signals, characterized in that the y signal, red signal, y signal, blue signal, Y signal, sound + other signals are tapped in a continuous sequence that the PAM values are on the half-period or period duration of half-waves or periods of an alternating current are transmitted, namely with the same amplitude or that only the sequence Y, r, Y, bl is provided and the sound and other signals by a binary or duobinary amplitude code ( Fig. 55) is coded in such a way that an amplitude value corresponding to the code is assigned to each half-wave or period, the 4 amplitude values ( FIG. 52) being able to be assigned to different channels in a code-multiplexed manner. 21. A method for coding the color television signals, characterized in that the television signals are coded only with one frequency ( Fig. 53, 54, 66) in such a way by these serially arranged code elements by the amplitudes of the half-waves or periods with the characteristic values of large or small amplitude value or small, medium and large amplitude value are formed for all signals or that the code is formed from a plurality of periods with 2 or 3 characteristic quantities and a continuous transition between the quantities ( Fig. 66, Ü) , where necessary, this code is intended to accommodate a channel in the gap between the conventional channels. ( Fig. 42). 22. Device for transmitting television signals, which contains a luminance signal (L), a first and a second color signal (I, Q), synchronizing signals (S) and sound signals (T)

• - A first A / D converter (LA / D), the input of which is supplied with an analog luminance signal and digital luminance signal samples occur at the output;
• a first memory (ZSpIQ) which has an input for the first color signal and a second input for the second color signal and stores signal values of these two color signals and makes them available at a first or second output,
• a changeover switch (U4) with two inputs which are connected to the outputs of the first memory and with one output,
  a second A / D converter (A / D, U4) whose input is coupled to the output of the first switch (U4) and to an output,
  a second memory (ZSpST) with an input for the synchronization signal and a second input for the sound signal, for storing signal values of these two signals and two outputs at which the signal values are available,
  a second changeover switch (U5) with two inputs, which are coupled to the outputs of the second memory and with one output,
  a second A / D converter (A / D, U5) this input is coupled to the output of the second switch (U5) and to an output,
  a buffer (Sp) for storing the digital luminance signal samples, the alternating digital color signal samples and the alternating digital synchronization and sound signal samples, the buffer (Sp) having a first group of outputs at which the code elements of the successive digital luminance signal samples are in parallel are available, with a further group of outputs at which code elements of the digital color signal samples are available and a further output at which the bits of the synchronizing sound signal samples are available;
• - An encoder (COD) with a first group of inputs, which are coupled to the first group of outputs of the buffer (Sp) and each receive the bits of a digital luminance signal sample in parallel, with a second group of three inputs, each via one Switches (U1, U2, U3) are coupled to a pair of outputs of the buffer (Sp), at each of which a bit of the first or second color signal is available, and a final output, which is coupled to the synchronization signal tone signal output of the buffer wherein the encoder provides a code word each representing a full digital luminance signal sample, part of a color signal sample and part of a synchronizing and tone signal sample, a pair of successive code words containing two parts of a color signal sample which together make up a complete color signal sample, and four successive code words contain a complete sample of the first color signal and a complete sample of the second color signal and alternately a complete sample of the synchronization signal and the audio signal ( FIG. 77, FIG. 78).

23. Device for receiving transmitted television signals, which contain a sequence of code words, each of which has a plurality of components of a television signal with arrangements in the input stage up to the demodulator, as in the case of a radio heterodyne receiver, a demodulator for evaluating the code words sent to a decoder ( FIG. 23, DC) , to which the demodulated code words are fed, which has a plurality of outputs at which the components of the television signal are available in parallel ( FIG. 23), if necessary with the interposition of storage means. 24. Device for the transmission of additional signals between a first television channel and a second, higher-frequency television channel, with screening means for generating a Nyquist edge of the television signals of the first television channel, characterized by a series resonance circuit for removing the television signals from a narrow frequency band between the channels and an arrangement for introducing a sinusoidal vibration ( Fig. 82, 195, 25) 25. The method according to claims 1 to 24, characterized in that the pulse duration pulses and pauses or when storing pulse duration pulses in an uninterrupted sequence control electronic switching means immediately (ER Fig. 36, 38) that the respective pulse duration or pulse duration pause in one Period duration or half period duration is converted from unipolar or bipolar rectangular pulses and that sieving means are provided which make the sinusoidal half waves or periods from the rectangular pulses in an uninterrupted sequence of positive and negative half waves. 26. Process for evaluating distances z. B. between pulses or of half or period durations, characterized in that at the beginning of the distance marking ( Fig. 60, 1) or at the zero crossing of the half period means for generating a sawtooth voltage are started and that at the end of the distance marking or at the 2nd Zero crossing of the half-period ( FIG. 59) means are connected to the sawtooth voltage which have a dimension thereof or that means are provided (FET) which store this voltage in particular in a capacitor. 27. The method according to claims 1 to 26, characterized in that multiple use of current paths is carried out in such a way that several information channels are combined time-multiplexed ( Fig. 56) or in that the control pulses for the counter elements receive such a frequency ( Fig. 57, Jm1, Jm2) that their encoding alternating currents do not get any overlap when transmitted via a current path. 28. The method according to claim 1-14, characterized in that a multi-level amplitude code (binary, duobinary, etc.) and / or single- or multi-level phase code and / or an analog amplitude and / or phase code is provided for the coding, in particular for the multiple use or reduction of the frequencies in telex ( FIGS. 18, 19, 20) in television ( FIG. 21) in teletex, data transmission ( FIG. 24) in digital voice transmission ( FIG. 28) is provided. 29. A method for the transmission of information based on quadrature amplitude modulation (QAM), the coding being carried out by the amplitudes of the half-waves or periods of an alternating current, characterized in that 4 channels are transmitted in the manner with only one alternating current of the same frequency, wherein the code elements are formed by the amplitudes of the half-waves that the phase-correct ordered coding alternating currents ( Fig. 10, 90 degrees phase difference from a to b and c to d) or sampling, so in the phase position, if necessary, with intermediate storage of the sampling of one or several channels for the transmission are changed so that the channels combined to form 2 addition AC currents ( FIGS. 10, 10b / c, 10d / e) are mutually phase-shifted by 90 degrees and that these two AC addition currents are added again ( FIG. 10, 0 / 90/90/180 degrees). 30. A method for the transmission of information on the basis of quadrature amplitude modulation (QAM), the coding being carried out by the amplitudes of the half-waves or periods of an alternating current, characterized in that the alternating addition current is formed in such a way that the result of the addition of the coding states ( Fig. 11, K1 / K2, K3 / K4) resulting characteristic states ( Fig. 11, I, II, III, IV) are transferred to the amplitudes of an alternating current of the same frequency, which are not based on the predetermined vector ( Fig. 11, I) lying characteristic states ( Fig. 11, II, III) are arranged coded on this ( Fig. 11, (II), (III)) and thus transmitted by the amplitudes of the quasi-addition alternating current, takes place on the receiving side the evaluation by determining the phase position of the quase addition alternating current and by determining measuring points 45 degrees ahead and behind the quasi addition alternating current, further means are provided for Assignment of the coded identification states to the channels. 31. A method for the transmission of information, in particular for telecommunications systems, characterized in that for the multiple use and / or duplex traffic in the transmission of analog and / or digital information, an alternating current of only one frequency is provided by the information to be transmitted in both directions (incoming, outgoing) is encoded by the amplitudes of the half-waves or periods of only one alternating current and transmitted in an uninterrupted sequence of positive and negative half-waves, the encoding alternating currents in both directions are phase-shifted from one another by 90 degrees, and the alternating current circuits in both directions are still the same dimensioned ( Fig. 25) or such separating means provided ( Fig. 25, G) that a reliable evaluation of the coding alternating current assigned to it takes place in the respective receiver. 32. The method according to claims 1 to 31, characterized in that counting switching means are provided as means for generating the square-wave pulses, which have switching pulses (e.g. sine or square-wave pulses) which have an excessive frequency compared to the square-wave pulses ( FIG. 7 , Osz), are controlled, control switching means are also provided ( Fig. 7, Cod), each marking a predetermined number of switching pulses at the outputs of the counter switching means according to the respective code ( Fig. 7, g2, G2, g3, G3) and which influence an electronic relay ( Fig. 7, ER) in such a way that square-wave pulses with the coded pulse duration are sent, alternatively, if necessary, different potentials are assigned to the pulses ( Fig. 7, J). 33. Method for the transmission of analog or digital pulse amplitude modulated information of two channels, characterized in that the pulses of both channels are converted into a staircase signal ( Fig. 68), that 2 carriers, which are 90 degrees out of phase with each other with such a frequency, are also provided that they form a whole multiple compared to the sampling frequency, the staircase signal is modulated onto the assigned carrier for a predetermined number of periods, furthermore the carrier alternating currents are added and the phase shift of the total alternating current compared to a comparative alternating current ( Fig. 68) is measured and one or several periods of the comparison alternating current added to a total period, the amplitude of the total alternating current is impressed on the coding half period with the total period. 34. The method according to claims 1 to 33, characterized in that a duplex traffic is provided by radio ( Fig. 26) by phase shifter arrangements are provided in a receiving station, which phase-shift the received alternating current on QAM basis for synchronizing the transmitter associated with this receiver, whereby tolerances are assigned to the switching means determining the reception position, preferably only one polarization plane being provided. 35. Method for coding information, characterized in that that narrowband encodes the information in such a way is by the digital and / or analog by the amplitudes the half-waves or periods and / or the number of Half waves or periods and / or by the phase relationship of the Half waves or periods encoded information into a multitude same code elements and / or in multiples of Number of the same code elements and / or in a variety of Elements that add up to the sum of the respective code element (Phase steps) is converted. 36. The method according to claims 1 to 35, characterized in that as a code element a variety of half-waves or periods of an alternating current with the same frequency the amplitude characteristics binary, duobinary or tribinary are provided, where appropriate, the change in the amplitudes done continuously. 37. The method according to claims 1 to 36, characterized in that that an analog digital code is provided, by dividing the PAM pulses into a large number of quantization levels and these in an equal or multiple Number of periods of an alternating current of the same frequency converts, with the following code element in each case an amplitude change is marked, a predetermined number Periods are provided as a constant default. 38. The method according to claims 1 to 37, characterized in that a digital or analog phase code in which is provided by the phase or frequency jump, or the analog values into a variety of phase or frequency stages is divided. 39. The method according to claims 1 to 38, characterized in that the evaluation of the half-periods or periods by dimension, the values being immediate by the absolute value or by means of a comparison phase or be determined by the difference phase. 40. The method according to claims 1 to 39, characterized in that by combining the different encodings higher-quality codes are formed.   41. Method for the simultaneous transmission of 2 signals, for example the color signals when watching television, via a channel, characterized in that the two signals are tapped simultaneously ( FIGS. 68a, b), that the PAM taps are formed in a step-like manner A carrier with a multiple frequency of the tap frequency is provided for each signal, both carriers being phase-shifted from one another by 90 °, a comparator carrier with the frequency of the carriers is also provided, which is synchronized with the tap frequency, in the following both carriers are summed, with the second or a subsequent period of the comparison carrier, the phase comparison is carried out with the total alternating current, with the phase jump subsequently being coded. 42. The method according to claim 41, characterized in that the phase shift with the assigned periods ( Fig. 87, Ph + KP) is formed into a half-period and the difference to the period of the tap frequency is formed into a pause half-period ( Fig. 87b, P) and that both are transferred as a period. 43. The method according to claims 41 and 42, characterized in that saturation by the amplitude size of the total alternating current based on the amplitudes of the is transmitted for useful and pause half-waves to be transmitted, is marked. 44. The method according to claims 35 to 41, characterized in that that the phase shift and possibly pauses can be encoded by a variety of periods. 45. A method for coding information based on the QAM, in which in particular the half-waves or periods are coded by binary or duobinary characteristic states, characterized in that the vector change takes place through a multiplicity of stages by means depending on the code are switched on by changing the electrical values, step by step, causing a corresponding step-by-step change in the amplitudes in the sequence ( FIGS. 88, n to 1 and vice versa). 46. Procedures for coding and transmission of information, esp special of templates and pictures, characterized in that for such methods for lines with only equivalent samples (e.g. only white taps) can be coded, whereby these for the transfer only by the code of the same value and the lines follow or by coding the line sequence and the same value be marked. 47. Procedures for coding and transmission of information, ins special of templates and pictures, characterized in that groups of lines of equal value and / or the same or closest lying length for groups for code multiplexing simultaneous Transfer of 2 or more lines can be combined, the Sequence of the tap is numbered and coded and transmitted with storage means are provided at the receiver in which the Information with code for the sequence of line information stored and is provided for decoding and recording that The transmission sequence of the groups is not tied to length. 48. Groups according to claims 46 and 47, characterized in that that counting elements are provided in which lines are counted can only be determined with taps of equal value. 49. Methods for decoding information, especially for Fax, characterized in that on the receiving side for the Recording 2 or more recording units are provided the 1, 2 or a large number of lines are equally spaced from one another can be assigned, whereby the assignment by the central control of the respective line information.