GB582177A - Multiplex high frequency electrical pulse signalling systems - Google Patents

Abstract

582,177. Pulse modulation; multiplex systems. STANDARD TELEPHONES & CABLES. Ltd., CHATTERJEA, P. K., and HOUGHTON, L. W. April 7, 1943, No. 5587. [Class 40 (v)] In a multi-channel pulse signalling system of the kind in which the channels a ... d, Fig. 1, are interlaced sequencies (e) of phase-modulated pulses, the channels are separated at the receiver by deriving from consecutive pairs a, b; c, d, &c. further pulses (f) of durations equal to the time interval between the respective pairs, differentiating the pulses (f) to obtain the pulse form (g) and then separating the positive and negative pulses from (g) to obtain a sequence (h) in which only the channels a, c are reprosented, and a sequence (i) in which only b, d are represented. A repetition of the process results in complete separation of all the channels. The system is shown in Fig. 1 with four channels, but the number may be any power of two. Fig. 3 shows the general arrangement, the received pulses (e) being applied to a circuit 73 which converts them to the form (f); they are then differentiated and separated into the trains (h), (i) by limiters 74, 75. The process is repeated for each of these trains to give the pulses related to each channel at the terminals 82 ... 85. In order to secure correct chanelling, despite the presence of additional interference pulses, or the absence of some of the signal pulses, several methods may be employed. (a) The pulses of one channel (say No. 1) may carry a characteristic modulation, to which a control unit 88 responds. When this characteristic is present at the unit 88, a correcting unit 89 is maintained inoperative. Should the characteristic modulation not be received in channel (1), the unit 89 comes into action to supply or eliminate one or more pulses until correct working is again attained. A similar control unit 86, 87 is arranged at the output of the preceding separation stage 74 traversed by the pulses of channel 1. (b) Alternatively, a distinctive modulation may be provided for each channel, and at the receiver, the output from channel (1) may be applied to a selector 104, Fig. 11, which may be a frequency filter in the case when the distinctive modulations are of different frequency. The output from the filter is rectified and applied as bias to a valve 110 which serves to eliminate a pulse and to a valve. 129 which serves to insert an extra pulse. When the system is working normally, the valve 110 merely allows the received pulses to pass unchanged to the next stage BB, but should an interference pulse put the channels out of synchronism, the filter output changes, in the negative direction, reducing the saturation bias on the valve 110 and causing the pulses passed to the stage BB to fall off in amplitude until a pulse fails to operate that stage. If the want of synchronism is in the opposite sense due to missing pulses, the output from the filter 104 changes in the positive direction, so as to remove a cut-off bias from the amplifier valve 129, which thereupon transmits one or more pulses from the valve 110<1> through a delay network DN to the stage BB. The missing pulses are thus restored and the channels again brought into synchronism. In Fig. 11, the stages BB join up consecutive pairs of pulses to form pulses of variable duration as at f, Fig. 1, and the differentiation process necessary to produce the form g is effected by capacities C1 and resistances R. The valves 98, 99 effect the separation into the forms h, i and the valves 100 ... 103 the further separation required at the next stage. (c) The channel synchronization may also be effected by using at the stages BB multi-vibrator circuits having a free frequency corresponding to maximum permissible time modulation of the pulses. (d) By averaging or integrating the series of received pulses, any loss of signal pulses or any additional interfering pulses may be made to provide a control voltage for effecting the desired correction.