More pictures of markings and damage

Been fixing several 486/386 boards lately. One did have an issue with the -5v, which was required by the video card to get a picture.

That said, there do look to be several vias that are corroded, and may have lost conductivity. You’ll need to trace all of them back to their respective ICs (in all directions) to make sure you’re getting good continuity. If it makes you feel better I have one board that is completely fine that refuses to boot. Nothing is wrong with it electrically that I can tell. It just refuses to give a picture.

A possibility is that the board keeps the processor in reset. Often the reset logic is close to the CMOS battery. You might want to check the voltage at the 486 reset pin, or a pin or via connected to the 486 reset pin that could be accessible without needing to flip the board over. Another indirect method to detect whether the processor is trying to execute code is checking the /OE and /CE pins of the BIOS ROM chip. Both of these pins need to show some activity, that is their voltage must not be exactly equal to the +5V rail.

mkarcher wrote on 2021-01-11, 17:36:

A possibility is that the board keeps the processor in reset.

C10 Reset pin is 4.75V, so it is high. I lost 2 hours tracing all the logic on the board leading to pin C10 instead of just first measuring the voltage on the darn thing. Thanks for the suggestion

jdgabard wrote on 2021-01-11, 17:11:

You’ll need to trace all of them back to their respective ICs (in all directions)

At the moment this is not possible. Unless someone can provide a picture of the traces under and around battery for this exact board, I don't have anything to go on, the corrosion was really bad when I got this board and someone tried cleaning it before, erasing all clues of where copper was before it was washed off.

This board could be HOT-409 Ver: /currently unknown/ board, just found this info by punching in strings in random configuration (opti 82c495sx AND 409 led me to elhvb page directly, weird, never had that luck before!)

Will try to disable cache and check if the CPU setup is good next.

Addition: I found a good picture of the battery area. I have almost none of those traces, they were eaten by the corrosion. I have my hands full tomorrow!

On the other glance, this board is different from mine. Well... alas...

usssentinel wrote on 2021-01-11, 19:58:

mkarcher wrote on 2021-01-11, 17:36:

A possibility is that the board keeps the processor in reset.

C10 Reset pin is 4.75V, so it is high. I lost 2 hours tracing all the logic on the board leading to pin C10 instead of just first measuring the voltage on the darn thing. Thanks for the suggestion

The reset input is active high, so the board keeps the processor in reset, indeed. There are two likely causes for the board keeping the processor in reset: Either the trace that carries the "power good" signal from the power supply has been eaten by the leaking battery, so the reset circuit never gets the signal that voltages have stabilized, or the chipset is trying to synchronize the RESET signal with the processor clock, and due to some missing clock (maybe a broken trace from the oscillator), the chipset waits all the time for a clock edge to lower reset. I've already seen both of these causes happen on actual hardware.

To troubleshoot why the processor is held in reset, the tracing of the logic might turn out useful, so don't count the full two hours as lost.

Well, with days behind me not being able to determine why C10 would be held high, I just removed the whole pin from the socket and it did not help. Luckily I resisted the urge to just snap the pin from the CPU itself, boy would my face be red if I went that route.

Considering I removed all the logic gates from the board that would possibly keep the board in perpetual reset, I managed to bring it down to low at a certain moment, but It did not change anything. I even went ahead and placed the BIOS chip in another board with AMI bios (that was weird experience, it hung at 0E at every post but it was to be expected, different chipset, so I presume the BIOS chip is not blank, something at least tried to happen... or some of the logic gates was shared with something else, I did not bother to put them back yet), also that BIOS chip in this board, same blank stare from POST analyzer, so it is not BIOS chip. At this moment other than pulling out cache chips and everything else that is pullable from this board and testing it on another identical or similar, I'm out of ideas

Rest in Pieces, old faithful Shuttle HOT-409 🙁