forgot to mention... ain't a pro in capacitors...

I've never seen bad caps actually damage other components (other then corrosion from the electrolyte).
Usually you just get instability at higher frequency's or worst case no POST.
Also obviously 133MHz FSB is a decent overclock and pretty far out of spec for PCI and AGP (and even the IDE controller IIRC).
So even if the MB is happy at 133 (some were not, even if they had settings for 133), those components need to tolerate 33% out of spec.

So if it was me doing your setup I'd start the system at 66MHz FSB and do some fast stability tests.
Then go 100MHz and repeat.
After that its just a matter of losening all mem and PCI/ISA timings and going for the OC, usually via whatever steps the MB has between 100 and 133.

bofh.fromhell wrote on 2025-12-13, 19:22:

I've never seen bad caps actually damage other components (other then corrosion from the electrolyte). Usually you just get inst […]
Show full quote

I've never seen bad caps actually damage other components (other then corrosion from the electrolyte).
Usually you just get instability at higher frequency's or worst case no POST.
Also obviously 133MHz FSB is a decent overclock and pretty far out of spec for PCI and AGP (and even the IDE controller IIRC).
So even if the MB is happy at 133 (some were not, even if they had settings for 133), those components need to tolerate 33% out of spec.

So if it was me doing your setup I'd start the system at 66MHz FSB and do some fast stability tests.
Then go 100MHz and repeat.
After that its just a matter of losening all mem and PCI/ISA timings and going for the OC, usually via whatever steps the MB has between 100 and 133.

oh dear I never knew IDE controllers get overclocked too 🤣, wouldn't be surprised if hard drives would get burned up. I'll notify if my tests gonna successful but since I ain't got parts, then not anytime soon

Takino-42 wrote on 2025-12-13, 13:15:

I've seen some old boards having their old electrolytes replaced with SMD polymers, is it really possible or it's just better to replace it with equal electrolytes of good quality? Anyway, I would wanna know how stable is 133FSB on this board for other people.

For example, the same Pentium MMX on an MVP3 motherboard with electrolytic capacitors glitched at 100 MHz FSB even at a supply voltage of 3.2 volts (100*2.5=250МГц).
Even with non-branded polymer capacitors, it is stable at 2.8V on the same 100MHz FSB.
But it still doesn't take 100*3=300, alas.
So they have a positive effect.
But if you didn't win the silicon lottery, polymer capacitors won't be a panacea.
133 MHz for 440BX is basically a working option, but not a guaranteed one.
The IDE controller, incidentally, like the PCI bus, will operate at a native 33 MHz.
Provided that the BIOS correctly supports 133 MHz FSB.

In short, it's a decent idea.

In theory it's a good idea.

In practice? Forget about it. The 6163, across rev 2.0 and 3.0 (I have both) are horribly finnicky with polymers installed.

And 133 FSB on 6163? Also forget about it. None of the boards I've tested EVER did 133FSB properly, and that's caused by the Cypress clockgens MSI loved to use during the P3 era. Best bet if you want 133FSB is getting a retail MS-6199VA, which is basically 6163 but with VIA 693A. Might be less performant but at least it does 133FSB properly.

PcBytes wrote on 2025-12-17, 18:49:

In theory it's a good idea.

In practice? Forget about it. The 6163, across rev 2.0 and 3.0 (I have both) are horribly finnicky with polymers installed.

And 133 FSB on 6163? Also forget about it. None of the boards I've tested EVER did 133FSB properly, and that's caused by the Cypress clockgens MSI loved to use during the P3 era. Best bet if you want 133FSB is getting a retail MS-6199VA, which is basically 6163 but with VIA 693A. Might be less performant but at least it does 133FSB properly.

oh yeah I might just go with VIA chipsets then I guess, I saw some good chaintech boards with them supporting up to 1ghz coppermines and having just as good memory bandwidth so that should be worth it. i820s platform are rare n expensive anyway. Very thankful for the advice

shevalier wrote on 2025-12-17, 16:33:

For example, the same Pentium MMX on an MVP3 motherboard with electrolytic capacitors glitched at 100 MHz FSB even at a supply v […]
Show full quote

Takino-42 wrote on 2025-12-13, 13:15:

I've seen some old boards having their old electrolytes replaced with SMD polymers, is it really possible or it's just better to replace it with equal electrolytes of good quality? Anyway, I would wanna know how stable is 133FSB on this board for other people.

For example, the same Pentium MMX on an MVP3 motherboard with electrolytic capacitors glitched at 100 MHz FSB even at a supply voltage of 3.2 volts (100*2.5=250МГц).
Even with non-branded polymer capacitors, it is stable at 2.8V on the same 100MHz FSB.
But it still doesn't take 100*3=300, alas.
So they have a positive effect.
But if you didn't win the silicon lottery, polymer capacitors won't be a panacea.
133 MHz for 440BX is basically a working option, but not a guaranteed one.
The IDE controller, incidentally, like the PCI bus, will operate at a native 33 MHz.
Provided that the BIOS correctly supports 133 MHz FSB.

In short, it's a decent idea.

alright I'll maybe consider it for other boards too!