Report on vintage "Programmer Electronnic Control" alias "RAF Tornado Computer"

On Thu, 8 Mar 2007, Richard wrote: Thank you, Richard. Well, the CPU is made entirely from 74xxx chips - NO BITSLICE chips in ther. Of course they used the SMD military variant of the chips 54xx and date codes range from 1982 to 1986. I am sure that the powersupply was replaced at least once (not wondering looking at the design). The core PCBs are standard DIL-packages and might have been used in other designs (I discovered a design bug there, too ;-) ) Back to the CPU: It consists of three PCBs of 100mm*160mm in size. Two PCBs are identical and contain 6bits of the accumulator and ALU (shift and logic functions only) each. The third board is labeled "function decode" and this contains the rest of the ALU (12 bit adder, subtracter, zero-detect) and the address generator. Data flow etc. is controlled by 6 PROMS on this PCB. So the content of these PROMS contains what I call microcode, i.e. the bitpatterns and the timing sequence required to make the processor work and the complicated multiply-insttuction lasting for many cycles originates there, too. Since these 32-nibble-PROMS with open collector output are soldered to this multilayer (YES: MULTILAYER) PCB, where the vias are made by some rivets, I did not have the heart to unsolder them for readout. Apart from this I do not have got a reader for them. Of course I considered connecting the logic analyzer to the PROMS and record their contents during operation. But due to the interconnection of the open collector outputs it would be difficult to judge wich of the PROMS is pulling the pin to low. So currently I am still at risk, that there is (a) a defective PROM responsible for the freezing phenomenon and that (b) I will lose the unit if I kill one of the PROMS. But that is life. Thanks and best regards, Erik.