(By Keith Crook – November 2003)
The purpose of the S.I., which was totally achieved, allowed any peripheral to connect to any processor in the 1900 range. This enabled peripherals to be added or upgraded without changing the processor and, equally, for the processor to be changed to another in the range whilst keeping the peripherals. Three areas were standardised.
1. The cabling was fitted with pins and the processor with sockets; a certain cable type and 75-way plug and socket were specified after extensive testing of alternatives.
2. The signals passing between the peripheral and the processor were exactly defined in terms of pin number function and timing. The commands from the processor and the status responses from the peripheral were 6-bit items. Signal lines from the processor include an A-line to select the peripheral, an L-line to indicate the limit of the transfer and so on. Those from the peripheral were signals like Data Request (R) and Interrupt (B).
3. The electrical interface was exactly specified. The specification recommended that the same interface circuit boards were used in all the processors and the same boards in all the peripherals. This greatly reduced development effort and spares holdings.
Together with the Executive Standard Interface, software applications were fully portable across the range regardless of the processor or any specific printer, tape drive or disc system.
The 1964 specification which applied to the early 1900s provided two sets of timing rules. The lower speed catered for data rates up to a maximum in the region of 1.5 Mb/s.
The higher speed became necessary as faster discs became available and was introduced in 1967.