After washing this board the issue still persists, even with the 5vsb and PS_ON resistor replaced with an 100k ohm (original was 0ohm)...
I get 1.84v on the fan power connector and 5v on the tachometer signal pin.
Any ideas?

I don't know what would be a typical cause for the board to be stuck in a Reset state. But I believe in always looking at voltages as a first step.

Use a multimeter (BIOS readings are unreliable) to check for voltages on the board, see if you find anything that looks weak or unsteady. Primary voltages from the PSU, battery, Vcore, and see if you can find DDR (spec is 2.5-2.7V for DDR400, 2.3-2.7V for older DDR), and anything else you think of.
I don't know what voltage the nForce2 chip uses, it might be the RAM voltage.

If all the voltages look good, then there's still the possibility of ripple caused by bad caps. Ripple is harder to measure and requires an oscilloscope.

If you have a good 12V on the board, but it's not getting to the fan headers, then you might have a blown transistor on the fan outputs. That tends to happen when excess current is drawn by an overpowered fan, or by a fan that has seized up.
Some boards have a separate transistor near each fan connector, and some have them shared.

The Asus A7N8X-series were insanely specific in their RAM requirements, failing to work correctly with many DIMMs that worked with no issues on other nForce2 boards. Not saying that is definitely the case here, but test with multiple DIMMs before assuming RAM isn't (one of the) cause(s)

After doing more testing, I found that the board gets hung up on memory error. I tested with some RAM sticks that worked with other KT400 chipsets, but none worked.

As for this weird behavior in the first post, once it is stuck in reset, once it isn't. I have confirmed that PS_ON and PWR button aren't shorted to GND prior to applying power. Already reflowed the weird SSOP16 Asus chip, but no go.

You've got a good CR2032 battery installed I assume?

It's possible you're still dealing with bad caps, even if they aren't visibly damaged or swollen.

They can create some really crazy problems, and the nature of caps charging\discharging means that the problems can be intermittent or seem to change when you do unrelated things.

Just a thought.

Oh... have you tried different power supplies? Preferably totally different models.

Armed with courage from previous repair and knowledge, I tackle this trainwreck after (almost) 5 months.

Corrected the mainboard model in the opening post (it is actually A7V8X-X, not A7N8X-X).

I bought this board with a Athlon CPU which was supported. Sadly it was completely dead (no POST execution on any supported mainboards nor cracks in the die). I think the cause for this could be a bad PSU the previous owner used. It may had a bad power rail which started to damage components, resulting in whatever this issue is.

The "memory error" POST code is 9F, which there's no info about it for AWARD bios'es. On PC speaker I get long beeps which go on forever (indicating memory error). What's odd is that the board goes directly to 9F regardless if there's memory inserted.
Already tried to update the BIOS from 1006 to 1013 but to no avail.

grjr wrote on 2025-06-08, 01:27:

You've got a good CR2032 battery installed I assume?

Yes (even tried to use my benchtop PSU set to 3.2V), but it doesn't change anything.

Ozzuneoj wrote on 2025-06-08, 02:00:

It's possible you're still dealing with bad caps, even if they aren't visibly damaged or swollen. […]
Show full quote

It's possible you're still dealing with bad caps, even if they aren't visibly damaged or swollen.

They can create some really crazy problems, and the nature of caps charging\discharging means that the problems can be intermittent or seem to change when you do unrelated things.

Just a thought.

Oh... have you tried different power supplies? Preferably totally different models.

Maybe bad caps near the RAM slots, those could only cause the 9F POST code issue. Or it's the chipset being half-dead.
I already tried 3 different PSUs, all known good so those shouldn't be an issue.

My primary suspect for the whole power shenanigans would be the SIO since is responsible for power on/off and fan stuff. My MS-7448 had a similar issue. It was turning on automatically on after applying power and powering on/off rapidly. The issue was a bad SIO which I've since swapped.

As for the RAM errors, I have three suspects.
First is the SIO somehow making all this mess.
Second is the chipset being half-dead as it is responsible for RAM connectivity. Already tried to heat it up with a hot-air gun, but that didn't change anything.
Third, as Ozzuneoj said, bad caps. The ones near the DIMM slots aren't of the highest quality, so these would be the primary suspect.

Attaching photos of the current state of the mainboard.

Any ideas?

kotel wrote on 2025-10-26, 14:17:

Any ideas?

Measure the memory supply voltage. (>2.5V. For Asus - I think, 2.65V it must be good. They always make voltage higher then standart)
Measure the reference voltage of the memory interface — the five-pin regulator between the memory and the ATX20 connector.
The output should be "Ram Supply/2."

https://theretroweb.com/motherboards/s/asus-a7v8x-x
It looks like yours has some damage near the power connector - is R454 missing?

The photos posted the board looks pretty dusty, there is that dirty (damaged?) area near the ATX power connector, a pile of bent pins particularly in the bottom right, and what I am assuming is a bunch of flux residue (?) on the underside.

I would start by giving it a good clean and fixing all bent pins. Then check voltages, shorts, continuity, the usual stuff.

shevalier wrote on 2025-10-26, 18:16:

Measure the memory supply voltage. (>2.5V. For Asus - I think, 2.65V it must be good. They always make voltage higher then stand […]
Show full quote

kotel wrote on 2025-10-26, 14:17:

Any ideas?

Measure the memory supply voltage. (>2.5V. For Asus - I think, 2.65V it must be good. They always make voltage higher then standart)
Measure the reference voltage of the memory interface — the five-pin regulator between the memory and the ATX20 connector.
The output should be "Ram Supply/2."

DDR memory is 2.69V. CPU Vcore is 1.65V.
The RT9173A regulator has the following voltages on pins:
1 - 3.36V
2 - GND
3 - 3.36V
4 - 1.34V
5 - 1.34V
TAB - 3.36V

JidaiGeki wrote on 2025-10-27, 09:58:

https://theretroweb.com/motherboards/s/asus-a7v8x-x
It looks like yours has some damage near the power connector - is R454 missing?

R454 is a 0 ohm resistor pulling PS_ON to 5VSB. I've removed it before for testing purposes (and tried putting in a 1k and 100ohm resistors as pull up but to no avail). Now it's just a dab of solder.

Shponglefan wrote on 2025-10-27, 14:38:

The photos posted the board looks pretty dusty, there is that dirty (damaged?) area near the ATX power connector, a pile of bent pins particularly in the bottom right, and what I am assuming is a bunch of flux residue (?) on the underside.

I would start by giving it a good clean and fixing all bent pins. Then check voltages, shorts, continuity, the usual stuff.

The bent pins aren't shorting anything, but just for good measure I've straightened them.
The brown dirt is the flux residue, also cleaned up now (there wasn't any bridges underneath them).

Also checked the resistor arrays around the DIMM slots but found no legs shorted. There was a boardview for this board, but I can't seem to find it anymore 😒

kotel wrote on 2025-10-27, 15:37:

DDR memory is 2.69V. CPU Vcore is 1.65V. The RT9173A regulator has the following voltages on pins: 1 - 3.36V 2 - GND 3 - 3.36V 4 […]
Show full quote

shevalier wrote on 2025-10-26, 18:16:

Measure the memory supply voltage. (>2.5V. For Asus - I think, 2.65V it must be good. They always make voltage higher then stand […]
Show full quote

kotel wrote on 2025-10-26, 14:17:

Any ideas?

Measure the memory supply voltage. (>2.5V. For Asus - I think, 2.65V it must be good. They always make voltage higher then standart)
Measure the reference voltage of the memory interface — the five-pin regulator between the memory and the ATX20 connector.
The output should be "Ram Supply/2."

DDR memory is 2.69V. CPU Vcore is 1.65V.
The RT9173A regulator has the following voltages on pins:
1 - 3.36V
2 - GND
3 - 3.36V
4 - 1.34V
5 - 1.34V
TAB - 3.36V

Looks good.

Also check the AGP 1.5V power supply - it can also power many other components in the northbridge .

It might also be a good idea to temporarily solder a few new capacitors in parallel with old.
If you solder them to the back of the board, you can simply desolder them later.
Find a schematic for any motherboard with the same chipset and see what other "small LDO regulators" are available to power various auxiliary circuits.

AGP 1.5V seems to be okay (1.66V). It also goes to the NB Vcc.

Looking at the boardview, it seems like the NB is responsible for PWR_ON. The power switch pin goes to some resistors and then directly to the NB (according to the boardview). Datasheet for IT8712F does mention pin 72 as "PWRON#" but depending on how asus programmed it it can also serve as general purpose IO.

Next I'll solder new DDR VRM caps on the backside and see if it was the issue.

Otherwise, I guess I'll continue probing the voltage pins. I don't think it will be the SIO since the chances of it screwing up the bus connection so much are pretty low, but never zero.

Okay, the 2.5V, 3VSB and 5VSB rails are all fine on the mainboard. Couldn't find any other main power rails which are needed fro proper functioning of the mainboard.

Soldering fresh caps in series with the older ones also do nothing. Still, the board hangs at a 9F post code.

Any ideas?

https://www.alldatasheet.com/datasheet-pdf/vi … T/ICS93738.html
But I can't even imagine how to check this other than by replacing it.
Even with an oscilloscope set to the correct frequency.

Q10 is connected properly? from pic it looks maybe not (or it's a shadow?)