Great newsYou'll be pleased to know that I'm working on emulating the AY-3-4592 keyboard encoder...
Have you read Chris's web pages here and the linked to pages and documents?
One of the unmarked / blank keys is not part of the matrix and is just wired to the connector for user applications. As this requires users specific wiring, it's not likely to be needed in emulaton.
The other unmarked / blank key is also not part of the matrix, but is the system reset key (or break key).
The repeat key again, is not part of the matrix. It just causes the strobe signal to keep changing state, hence the CPU card keeps reading the data (that is still present from the keyboard encoder chip) as 'new' key presses.
The caps key is separate as well.
The rest of the keys are part of the matrix. Apart from the shift, shift lock and control keys, the remainder produce an ASCII code. The shift, shift lock and control keys modify this code, but the final data is still either an ASCII character code or an ASCI control code.
There is a table of the characters and codes in one of the documents. Although I'm not sure how closely this matches up to the actual key caps.
Also the 40 x 25 VDU card used the teletext character set, but the 80 column card may be different (I have looked at this, I but have forgotten the details).
On the subject of the strobe, Acorn included an option for it to be positive. The chip output is the opposite to the normal keyboard card configuration. That may explain why there is so much confusion with this signal.
Mark
Statistics: Posted by 1024MAK — Thu Mar 27, 2025 3:02 pm