Check if you didn't switch the polarity of one of the caps while recapping..
That could explain the symptoms. Next check if all solder point are really good and connect, a broken powerline because a cap doesn't connect also can cause the same.
No other options that I know..

Did the board atleast boot before the recap?

Nvm1 wrote:

Check if you didn't switch the polarity of one of the caps while recapping.. That could explain the symptoms. Next check if all […]
Show full quote

Check if you didn't switch the polarity of one of the caps while recapping..
That could explain the symptoms. Next check if all solder point are really good and connect, a broken powerline because a cap doesn't connect also can cause the same.
No other options that I know..

Did the board atleast boot before the recap?

No dice - caps are soldered in the correct orientation (silkscreens are huge on this board) and yes, it did ONCE boot but after the recap. I don't get it because the same caps powered a hungry Pentium D just fine.

Solder points are also fine. I triple checked that during recapping.

PcBytes wrote:

CPU sometimes gets warm and other times it stays cold.

First thing you have to do now, is figure out why that happens. CPU VRM is a buck converter and it should not give out randomly, unless some power saving mode kicks in from the system. Try to set CPU core voltage to 3.3V and w/o CPU inserted measure output voltage at the coil/filter caps/CPU socket, look for anything unusual. Also check RAM VRM voltage.

Just measured CPU voltage - 3.23v (with 3.2v setting, board doesn't have exact 3.3v setting) at the cap near the socket.
RAM measures 5.1v on one capacitor (near the corner of the board) and 3.3v on the one next to it.

CPU core and RAM voltages are in specs w/o load, check if CPU I/O voltage is 3.3V. Retest CPU core and I/O voltages with a non MMX pentium.

You should also take a look at ICs in QFP packages for bent pins/debris. Reseat BIOS and inspect socket above and beneath, sticking sharp pins beneath BIOS socket might short or puncture solder mask and short with nearby traces.

SSTV2 wrote:

CPU core and RAM voltages are in specs w/o load, check if CPU I/O voltage is 3.3V. Retest CPU core and I/O voltages with a non MMX pentium.

You should also take a look at ICs in QFP packages for bent pins/debris. Reseat BIOS and inspect socket above and beneath, sticking sharp pins beneath BIOS socket might short or puncture solder mask and short with nearby traces.

CPU I/O voltage measured 3.2v without CPU, 3.3v with Pentium non-MMX 150MHz.

All ICs are OK, no bent pins or debris whatsoever. Reseated BIOS a lot of times (I have another BIOS chip for it that I got in a pack of ICs).

Any other things to try? I was hoping to do an unusual build with it if I get it to work - P200MMX, 256MB RAM, a Geforce 5600 128MB, Sound Blaster Audigy SE, 80GB HDD and maybe Windows XP. 🤣

Hmm, not sure what else could be checked. If it did POST once, then the chipset and BIOS are OK, probably a bad solder joint/corroded via or CPU socket/cut trace/missing SMDs/shorting semiconductors are causing you grief. I see that this mobo uses QFP northbridge, I'd take a really good look at its pins, try pushing every single one gently with a needle. You can also try pushing/bending board in every direction and while doing that, hit reset several times.

SSTV2 wrote:

Hmm, not sure what else could be checked. If it did POST once, then the chipset and BIOS are OK, probably a bad solder joint/corroded via or CPU socket/cut trace/missing SMDs/shorting semiconductors are causing you grief. I see that this mobo uses QFP northbridge, I'd take a really good look at its pins, try pushing every single one gently with a needle. You can also try pushing/bending board in every direction and while doing that, hit reset several times.

I tried slightly bending the board and even pushing the CPU while locked in socket. Nada.

I also inspected the QFP southbridge (the NB is BGA) for bent pins, and all are straight, and make good contact.

Board came in good condition, no corrosion or anything, just a fine layer of dust.

Southbridge* my bad, give it a bath then? 😁

SSTV2 wrote:

Southbridge* my bad, give it a bath then? 😁

Didn't change anything.

I've been thinking, should I try replacing the buck regulator chip near the CPU? I have another discarded P5MVP3 that has a HIP6008CB chip near the CPU (the one that I'm troubleshooting has something else starting with "A")

When CPU is cold it might mean CLK signal is absent and/or RESET signal is active.
Depending on POST card model it may show RESET and CLK signal activity .

Roman555 wrote:

When CPU is cold it might mean CLK signal is absent and/or RESET signal is active.
Depending on POST card model it may show RESET and CLK signal activity .

I don't have a POST card.

CPU seems to heat up very slowly (i.e after half an hour it's barely warm) though.

You can measure with a voltmeter RESET signal level on pin A15 of any PCI slot. If it has a high level (3.3v) - it's ok.
CLK is on B16 pin, but you can't check it so easy.

Roman555 wrote:

You can measure with a voltmeter RESET signal level on pin A15 of any PCI slot. If it has a high level (3.3v) - it's ok.
CLK is on B16 pin, but you can't check it so easy.

So CLK is absent due to RESET being held? I could look if A15 gives 3.3v.

PcBytes wrote:

Roman555 wrote:

You can measure with a voltmeter RESET signal level on pin A15 of any PCI slot. If it has a high level (3.3v) - it's ok.
CLK is on B16 pin, but you can't check it so easy.

So CLK is absent due to RESET being held? I could look if A15 gives 3.3v.

Not at all. I didn't mean that. It's rather a southbridge can't deactivate RESET if PLL doesn't supply it with CLK

POST card is priceless.

Actually, it is very cheap. It's the wait that kills 😁

I suppose the contents of the BIOS chip is OK, since it has booted before.

quicknick wrote:

POST card is priceless.

Actually, it is very cheap. It's the wait that kills 😁

I suppose the contents of the BIOS chip is OK, since it has booted before.

I'm too broke at the moment 🤣

Yes, the BIOS chip contents are OK.

What should I look for that would hold the SB in RESET state?

UPDATE:
Verified pin A15 of PCI slot, 3.36v.