I cannot see C6 & C24 but they may be in failure mode following overheating. What are their expected values? Could you change them?

wbahnassi wrote on 2024-07-30, 04:33:

* Had a leaky battery. Traces restored by previous owner

In cases like this I tend to suspect he might have missed one or more traces.

Obiwan wrote on 2024-07-30, 06:39:

I cannot see C6 & C24 but they may be in failure mode following overheating. What are their expected values? Could you change them?

I put photos showing the two caps. Let me quote them again here:

C24 is just above the hex-inverter Toshiba chip:

C6 is to the right of the SPKR connector:

I don't know their values. They have these two lines printed on them:
" 5'3 "
" Z "
which I don't know how to interpret. I could try removing them first without replacement, but wanted to avoid that in case this behavior is expected (low resistance on one, conductive in one way only on the other).

majestyk wrote on 2024-07-30, 07:50:

In cases like this I tend to suspect he might have missed one or more traces.

The corroded traces are three thick lines coming directly from the power connector and go all the way to the ISA pins. I checked for continuity on those lines and they are ok. If a trace is still missing, I would expect that the board never POSTs, but it actually did POST 2 or 3 times before totally giving up, hence I'm feeling there's a component that was on its last breath and it now it fully died after a few attempts.

I've had my fair share of joy and pain with battery leakage damaged boards.

I do agree with majestyk here. I think there are probably traces in other layers that have been compromised.

This is a shoot in the dark but, what about trying to fill all ISA slots with random cards to see if that does something?

I remember one M396F that worked but only if you placed cards in certain slots, or if you bent it slightly. It was on its way out and it's been a donor for some time.

I wish you luck!

C24 & C6 are capacitors. Their capacitance is numbered in (micro)Farads on them and mesured thanks to a multimeter.
Most of them are polarized so be careful when you replace them. https://en.wikipedia.org/wiki/Capacitor#Capacitor_markings

I'm not sure they are the source of your problem.

wbahnassi wrote on 2024-07-30, 04:33:

* Had a leaky battery. Traces restored by previous owner * CPU gets hot after power on * ISA address lines show activity on osci […]
Show full quote

* Had a leaky battery. Traces restored by previous owner
* CPU gets hot after power on
* ISA address lines show activity on oscilloscope
* Reset line is not held high
* Detected only 45Ω resistance across C6
* Detected 0Ω short across C24 only in one direction, flipping the probes shows OL

Like others, I suggest looking very close at the board around where the battery leaked.
Make sure there are no visible traces with damage (if you can access points where the traces
have visible endpoints, it wouldn't hurt to verify they are connected). Also look to see if there
are any points where the acid may have penetrated through-hole traces in the board and
possibly damaged internal layers.

Odd CPU getting hot - I'm assuming this is with heat-sink/fan removed but it should still
take a bit of time before you observe it getting overly hot.

Address lines showing activity suggests CPU is running (or crashed, trashing)
If you can do a one-shot scope capture, check what the address lines on the ROM
do for the firsts few fetches right after RESET and see if it is different with the ROM
removed. Again, assuming you've already looked at chip select to the ROM and
confirmed that it's getting accessed - now try to see if it's being read correctly!
(With a bit of work and a dual trace scope, you can figure what address and data
are being read in the first few accesses)

RESET depends on mainboard - CPU reset itself is probably pull-low to reset,
and who knows what the mainboard may do to condition the reset... I'd verify
cold power-on reset at the CPU socket. (Also try with button on reset pins so you
can see what it normally does).

I've run into boards where RESET didn't happen with certain normally present
hardware missing - in particular I recall one board which didn't RESET when the
hard-drive wasn't connected .. but there may be other stuff, like someone mentioned,
try Video card, other ISA cards etc.

A cap doesn't conduct DC at all, and it's behaviour and DC voltage measurements
depend entirely on whats connected to it "in circuit". Schematic would help a LOT,
perhaps you can see/figure out what they are connected to.

A CAP can't be infinitily high resistance (OL) in one direction, and 0 Ohm in another,
again whats in the circuit - by itself it would be OL to DC all the time.
Often if connected to a solid-state device (Diode, transistor or IC)
on one side, the circuit at the point where the CAP is connected might
show those characteristics - but very surprising if you see true 0 Ohm in
one direction, most solid-state devices have a small voltage drop in
the "conducting' state (typically 0.7V for a diode.

The resistance scale you are using can affect this (keep in mind that many
Ohmmeters will provide more incircuit current when used on very l ow
resistance scales).

Best of luck!

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal