Sounds like something is shorting out, sounded like when one of my molex connectors earthed to the case by accident it shut down the system to protect the motherboard.

The PC actually boots up once it has cooled down for a while with the case interiors untouched. Even in such a scenario, the power button needs to be pressed for a few seconds with the popping sound coming and the it starts. The non booting issue happens immediately after it is used quite extensively for a short while.

Check to see if any components feel hot

If you haven't done this already, swap the CR2032 battery for a new one.

I doubt it's the cause of your issues, but some systems behave oddly when the CMOS battery is depleted.

At this point it's mandatory to take apart and continue with a thorough troubleshooting.

Was just about to suggest that strip to barebones and see if it posts

The PC actually posts "for now". Tried switching off and on multiple times and it worked. However, the behaviour is very erratic. At times the board posts with the slight touch of the button. Other times the button needs to be held down for a few seconds with the popping sound. Then moments later never posts. Earlier, I did try different steps such as connecting only the PSU and monitor and later bare minimum peripherals such as HDD, FDD, etc. However the issue was occurring randomly and it became impossible to narrow down to the possibly faulty component.

I also had replaced the CMOS battery but in vain.

I had a visual inspection of the caps and none were bulged at the top. However, one cap near the ATX power connector was a little bent as in not straight vertically. I suspect it is swollen at the base as the board is still in the case. Might have to remove the board to view that.

Could the problem actually be the button itself

My bets are on a failed cap. Erratic behavior is an indicator of a component semi-failing. Caps are the number one suspect in this case. My suggestion is to buy a replacement cap set, with focus on the caps on the power rails. Caps might not show visual indication yet the still fail.

Good luck, and don't feel bad about it. It's inevitable on P4 mobos.

Greywolf1 wrote on 2025-01-05, 14:26:

Could the problem actually be the button itself

I suspected that and even tried with a different power cable (with a button) and it was the same scenario. Infact the board was actually posting after removing the 3.5mm speakers wire from the sound card. After plugging it in wouldn't do so. Then I removed just the USB connector of the speaker and it posted.

wbahnassi wrote on 2025-01-05, 14:26:

My bets are on a failed cap. Erratic behavior is an indicator of a component semi-failing. Caps are the number one suspect in this case. My suggestion is to buy a replacement cap set, with focus on the caps on the power rails. Caps might not show visual indication yet the still fail.

Good luck, and don't feel bad about it. It's inevitable on P4 mobos.

Absolutely agree! Sounds like capacitors shorting out, but are able to receive just a tiny bit of charge until they are at a level where input current is within current limiter's operating range

Hello all,
Yesterday, I took out the motherboard to check the caps. Noted quite a significant dust buildup and used a brush to clean it off. Then connected the board using the graphics card. However, it wouldn't post. So, I used the onboard video and it posted. After a while, I connected a different video card and then this time also there was no vga output. Hence, I removed the card and again connected the onboard video. This time the board didn't post no matter what I tried. Had removed, cleaned and reseated the RAM on both the slots, took out the BIOS battery, used a different PSU. Had also tried connecting a PCI video card, but there was no output.

The fans spin and the LED light on the board is ON. Thats it. No output on the monitor. So, could it be assumed that the motherboard is officially dead or is there any hope left?

stealthjoe wrote on 2025-05-10, 06:09:

Hello all,
Yesterday, I took out the motherboard to check the caps. Noted quite a significant dust buildup and used a brush to clean it off. Then connected the board using the graphics card. However, it wouldn't post. So, I used the onboard video and it posted. After a while, I connected a different video card and then this time also there was no vga output. Hence, I removed the card and again connected the onboard video. This time the board didn't post no matter what I tried. Had removed, cleaned and reseated the RAM on both the slots, took out the BIOS battery, used a different PSU. Had also tried connecting a PCI video card, but there was no output.

The fans spin and the LED light on the board is ON. Thats it. No output on the monitor. So, could it be assumed that the motherboard is officially dead or is there any hope left?

It's troubleshooting time with pics and a multimeter.
Probably a cap as suggested above.

Post pics in quantity and quality. Thanks 😀

Hello. Below is the image of the motherboard. The caps seem to be fine visually.

At times I do notice a very faint beeping sound something like the voltage fluctuating. The cpu fan tends to slow down and speed up slightly.

https://theretroweb.com/motherboards/s/intel- … v2-brownville-2
You must check voltages on all mosfets.

Do you have a post card?

Nexxen wrote on 2025-05-11, 08:26:

Do you have a post card?

If you are referring to the PCI diagnostics card, then yes. It was showing 0000 on the led and the numbers never changed unlike on a proper working motherboard where it cycles through many numerals.

Whilst a dead CPU wouldn't be my first guess, that delidded Pentium 4 looks like a potential concern. How have you been managing to ensure adequate mounting pressure without a heat spreader? Do you have another Socket 478 CPU to test with?

Nexxen wrote on 2025-05-11, 08:26:

You must check voltages on all mosfets.

I did test the 2 MOSFETs below the 4 pin power connector on the above image. It measured 0.58V from Source(+ve probe) to Drain (-ve probe) which I believe is expected. However as per my understanding, it needs to be OL from Drain (+ve probe) to Source (-ve probe). However it measured ~ 0.8v for me. Any ideas?

sfryers wrote on 2025-05-11, 08:42:

Whilst a dead CPU wouldn't be my first guess, that delidded Pentium 4 looks like a potential concern. How have you been managing to ensure adequate mounting pressure without a heat spreader? Do you have another Socket 478 CPU to test with?

Actually the lid (heat spreader) is stuck to the heatsink and it was this way ever since I got this board nearly 5 years ago. It was working fine since then.

I tried to remove the lid from underneath the heatsink but in vain. So in this case it prevents me from using another cpu with the same heatsink.