Notes on the Transmission of Data at 750 Bauds Over Practical Circuits

This paper describes experience over some 12 months of using a 750-baud phase-modulated data-transmission equipment over British Post Office trunk-line circuits, with a brief description of the equipment and test gear and of the lines used. It is concluded that line attenuation and general circuit noise are not limiting transmission characteristics in the magnitudes to be expected. Delay distortion and impulsive noise can be troublesome, but the most serious disability is short interruptions of the transmission path lasting from a few milliseconds to a few seconds. Because of this, the view is expressed that the assessment of the suitability of circuits for data transmission by quoting average digit error rates is unrealistic, and a figure for the number of minute periods per day when error-free operation can be expected is to be preferred; also, that with the duration of the interruptions encountered, automatic error correction would be complicated, and have to be supported by a system of repetition of messages in error. In the interests of simplicity, this could be the preferred solution even for single errors, since they appear to occur infrequently. The view is expressed also that users who cannot rely on a network provided by fixed plant in a civil area, e.g., a military user in the field, may well have a worse interruption problem due to the inclusion of numerous cable couplers in relatively short line circuits, and could better be served by a system of telegraph channels, each working at a slow digit speed, in order to provide a long digit length to combat the interruptions.