Bun's Lab
So, the post codes can mean wildly different things depending not only the BIOS maker but also version. Older AMI BIOSes up to 1991 point to a memory issue. Newer ones up to 1993 to something completely different. This 74F00 hex inverter is of course highly…
Ahh, I couldn't resist. Got a late night urge to fix this thing.
🔥1
This NAND is definitely bad. Three gates don't do what they should. Their inputs are low, but so are their outputs. Only one gate was working. Problem is, the holes are so tight that it was impossible to get it out without cutting it off.
😢2
Bun's Lab
This trace is also bad
Continuing: this alone didn't help. Still broken.
Bun's Lab
Photo
Because the RAM slots were so badly affected, I ended up tracing them out.
Everything checks out, including the 74F244 buffers to the RAM. /CAS however is perpetually high after a short burst of activity during start up, so there is never even any access to the RAM after that.
Everything checks out, including the 74F244 buffers to the RAM. /CAS however is perpetually high after a short burst of activity during start up, so there is never even any access to the RAM after that.
Pin 13 comes from the CPU, pin 11 (OUT) goes to the chipset. The purple IN connection was broken. Now it is always high. I have yet to find where it comes from.
Edit: found it, it comes from the other of the two chip set ICs. The one that seems to handle the ISA bus and the RAM address and control lines.
Edit: found it, it comes from the other of the two chip set ICs. The one that seems to handle the ISA bus and the RAM address and control lines.
I'll probably have to take off the uppermost SIMM socket. I can see corroded traces underneath. And although I have traced it all out and everything checks out, there could be still sufficient unneutralized base under there that provides enough of a current path to throw off the chipset. Will try rinsing it again with vinegar and alcohol. If that doesn't help, I'll have to take off the socket. But that's going to be really ugly, judging from my experience with the NAND gate.