What I tried:
- Disabled all power-saving and power-up options in BIOS
- Enabled and disabled Keyboard Power Up jumper (KBPWR)
- Tried different PSUs, including old ATX1.0 with -5V and -12V rails
- Tried different processors - Katmai, Mendoccino, Coppermine
- Tried different SDRAM modules
- Installed new CMOS battery
- Reflashed the BIOS including Bootblock and ESCD - Tested all official BIOS versions: 1011A, 1012, 1013B (without hwmonitor).
Flashed using Aflash v2.21 with the aflash /boot command.

None of this helped.

Yes it is normal .. for Asus boards from around 2014, but not pre 2000 😀
Afaik ASUS P2-99 is 100% same design (PCB and components) as full P2B of same revision with only two changes - ZX chipset and only 2 DIMM slots populated.
Asus never published nice schematics, only some scribbled over Xeros of their general designs they called BX98, but they did publish boardviews.

1.04 only available SCSI server version https://www.chinafix.tech/asus-motherboard-p2 … 1-04-boardview/
https://www.chinafix.tech/asus-motherboard-p2 … 1-10-boardview/
https://www.chinafix.tech/asus-motherboard-p2 … 1-12-boardview/
you can view those using https://github.com/OpenBoardView/OpenBoardVie … ases/tag/9.95.2
Its like a diagram, click on a pin and it shows you all the connections

Power on happens when ATX PS_ON is pulled low to ground. One thing that could happen is damaged resistor R134a responsible for pullup of that pin to 5V, without it random things might happen. This resistor hides between ATX connector and KB_PWR jumper (with USB connectors to the left).
What sometimes happens someone has trouble with removing ATX cable and decides to wedge a screwdriver to unlatch it - R134a is directly under ATX connector latch and could have gotten crushed.

rasz_pl wrote on 2026-04-08, 19:32:

Yes it is normal .. for Asus boards from around 2014, but not pre 2000 :) Afaik ASUS P2-99 is 100% same design (PCB and componen […]
Show full quote

Yes it is normal .. for Asus boards from around 2014, but not pre 2000 😀
Afaik ASUS P2-99 is 100% same design (PCB and components) as full P2B of same revision with only two changes - ZX chipset and only 2 DIMM slots populated.
Asus never published nice schematics, only some scribbled over Xeros of their general designs they called BX98, but they did publish boardviews.

1.04 only available SCSI server version https://www.chinafix.tech/asus-motherboard-p2 … 1-04-boardview/
https://www.chinafix.tech/asus-motherboard-p2 … 1-10-boardview/
https://www.chinafix.tech/asus-motherboard-p2 … 1-12-boardview/
you can view those using https://github.com/OpenBoardView/OpenBoardVie … ases/tag/9.95.2
Its like a diagram, click on a pin and it shows you all the connections

Power on happens when ATX PS_ON is pulled low to ground. One thing that could happen is damaged resistor R134a responsible for pullup of that pin to 5V, without it random things might happen. This resistor hides between ATX connector and KB_PWR jumper (with USB connectors to the left).
What sometimes happens someone has trouble with removing ATX cable and decides to wedge a screwdriver to unlatch it - R134a is directly under ATX connector latch and could have gotten crushed.

Thanks for the help!
Boardview will be useful, but I'm new to board repair.

If I correctly located R134A using boardview, it's not soldered.

I also have fully functional ASUS P3V133 (Apollo Pro 133) board, and it's also missing in that location.
P2-99 and P3V133 are incredibly similar - the only differences are the chipset and the presence/absence of some components. It looks like ASUS used universal PCB based on the P2B-series.

I also noticed one thing:
- If I remove the CMOS battery, the board stops turning on and off automatically when AC power is applied.
However, pressing the PWR_SW button causes the classic cold boot issue: the first time I press it, the board turns on for half second and then turns off. But the second time I press it, the board turns on normally.
It looks like the problem is somewhere on the battery side.

same on my p2b

hmm, maybe my memory is failing me already and P2B did do the power on/of trick on AC power?
or maybe C61 is damaged?
You could always try experimenting with soldering a 1Kohm resistor into R134a spot (or more generally between +5V and ATX pin 14 PSon).

rasz_pl wrote on 2026-04-09, 19:15:

hmm, maybe my memory is failing me already and P2B did do the power on/of trick on AC power?
or maybe C61 is damaged?
You could always try experimenting with soldering a 1Kohm resistor into R134a spot (or more generally between +5V and ATX pin 14 PSon).

Thanks, I'll try that!

I also noticed some oxidation on the Multi I/O chip pins. I'll try cleaning them and washing the motherboard first.