PCI debug POST card maybe ?
Broken BIOS, coin battery discharched ?

I've got the same board any it runs perfectly. I use it as a testing board as it's support goes from p54 to k6-3 and MII and back, with all possible voltage combinations 😀

It sounds to me like the CPU is a bit unwilling. I had this once with an MMX that had been fried.
Does the MMX work on your board? If yes, I'd go for a misbehaving CPU. Getting up does mean at least something works, but not necessarily everything works.
Cyrix MII is notorious on heat problems, so maybe it's gone like myth MMX was? Trying to overclock with a slightly higher voltage may fry it even with some socket A cooler.
Did you try the CPU on another boards that supports 2,9v?
Did you try a 66x2 setting? That's "safe" well below spec to rule out instability of the CPU.

Did you use a single ram stick op 64MB pc133 so timing problems are definitely no issue? (Although ram problems should give the beep)

An issue I have with my board, is it's not happy to take any PCI graphics card. Only AGP (currently a SiS 6326DVD).

zyga64 - I just placed a brand new Duracell CR2032 so CMOS isn't the issue.

waterbeesje -

Tried any CPU I know that was working (a quite broad range of Pentium-S CPUs, an AMD K5 PR100 and an non-MMX Cyrix 6x86) and even tested them on two other known working boards - Acorp 5VIA3P and 5VIA77.

All POST'd first try on the Acorp boards, nada on the PCChips, and all set up accordingly (except for the 5VIA3P which I had to use the 75MHz FSB setting for the M II).

The stick I used I have used on all working SS7 mobos I have (and S7 too) and all have POST'd successfully with it. Even 430TX boards loved that stick. (it's a Micron/Crucial stick, for the record) Single-sided, btw.

As for GPU, I tried a Rage Pro Turbo AIW, a Miro Crystal S3 Trio 64V+ 2MB PCI and a S3 Trio 3D/2X 4MB AGP.

All CPUs still work in my 5VIA77 just fine, but no POST on the M577.

Those graphics cards should be compatible.
So that rules out broken CPU, broken/incompatible ram and graphics...

Back to the basics.
Did you connect some sort of primary IDE device? Some bioses need that.
Could it be one of the caps is reversed by accident while recaping?
Is there any way to see if there BIOS is still good?
Could you try another BIOS that works on another board?

Tried another BIOS chip and no POST either. That chip contains an earlier version of the BIOS (07/11/98 S, newer one is the latest vanilla 1999 BIOS).

No other devices than basic bare-bones bench setup: PSU, CPU+GPU, RAM.

As for caps, I checked polarity three times for each cap changed during recap.

Now that I've been thinking, would baking the motherboard (someone did this with a 286 mobo IIRC) fix this M577?

Your issue sounds like PCChips "sudden silent death" syndrome, when the board just gives up and dies. It seems to be very common with their boards from the Slot 1/Super 7 era, I have 3 boards that have died in that way, and no matter how much troubleshooting or recapping, I was never able to bring them back to life.

One user here speculated that the low quality of their PCBs may affect BGA joints of the chipset, so baking could help in that case.

Oh wow, really never heard of that. Guess it's worth whipping out my solder paste and automotive heatgun and try to reflow the northbridge.