Dead keyboard (not membrane)

[Albuzz]

Hi, I’m a newbie here so please forgive any stupid questions.

I’ve just fitted a new keyboard membrane and one of the silly led wires popped out and shorted somewhere (the screen disappeared and locked up). After rebooting the keyboard is totally dead. At least before most of the keys worked. Can anyone point me in the right direction to start looking. I’m pretty sure it’s not ic24 as I’ve subbed it without any difference.

Machine sits on the boot screen waiting for F2.

Also posted on Facebook page so apologies for the duplication.

Thanks
Alan

[ones' complement]

I would immediately suspect IC23 (ZX8302), it is the go-between CPU and IPC (IC24). Also nearest to flying LED lead connector

[Albuzz]

Ok Thanks for that..

I’ve swapped on my Minerva rom and it’s not just the keyboard because the machine hangs after the memory test displaying the Minerva splash screen but where it normally defaults to F2 after a few seconds, it now doesn’t.

It’s over 35 years since I dabbled with electronics so I’m a bit rusty but I do have an oscilloscope to help me. So any pointers to check the Zx8302 would be helpful.

Thanks again
Alan

[ones' complement]

I suspect if the ZX8302 fails in its communication with the IPC it will hang the machine. The ZX8302 communicates with IPC through two lines: COMCTL (pin 29) and COMDATA (pin 35). I would check for activity on those lines (though be very careful not to short either to other pins when testing - the ULAs have proven to be quite fragile). The display proves CPU and ZX8301 are OK.

[Albuzz]

I have a replacement ZX8302 en-route.. I’ll swap it and see what happens before tinkering any further and I’ll update with the results.

I must do something though as a permanent solution for the poor LED connection.

Cheers

[Albuzz]

I’ve now replaced the ZX8302 and it seems that it’s not the culprit..

I’m in the process of getting a readable copy of the circuit diagram so that should help and hopefully I’ll be able to identify the PIN numbers better.

Is it possible for the CPU to be partly faulty in that it’s able to access memory but isn’t able to communicate in terms of I/o requests? As I said the Minerva would normally try a boot from mdv1 then drop to the flashing cursor so I’m guessing it can’t perform any I/O at all?

Thanks
Alan

[ones' complement]

I'd still try looking for activity with IPC (ZX8302 COMDAT/COMDATA, but also check IPC keyboard scanning lines). It's possible it could be PCENL (pin 10), the ZX8302 chip select. More readable schematics here: Issue 5 and Issue 6.