Do you get the beeps when you try to power on the system without RAM and AGP card?
Do you have a POST analyzer card? If you do, does the system go out of the reset state and are there any POST codes sent over port 80h?

Another thing to try is forcing the BIOS into recovery mode from the bootblock code by shorting two highest address lines on the flash chip.

Thanks @kmeaw!

Good question - I get no beeps whatsoever. Fans spin up, and I can hear hard drives spin up, but no other indications of normal operation.

I forgot to mention - I also burned a replacement BIOS rom and swapped that in. The stock BIOS was udpated to version 3.1d, with hard drive support (https://www.wimsbios.com/biosupdates/la ... #gsc.tab=0). This is a 128kb BIOS. I burned official BIOS 3.3, which is a 256kB BIOS (https://www.ecs.com.tw/en/Product/Motherboard … A_Plus/download). No difference with either BIOS installed.

I don't have a POST analyzer.

To force BIOS into recovery mode, would that be shorting pins 2 and 3, for A16 and A15 respectively?

:edit: Well shoot, for less than $15 with the fastest amazon shipping and arrival by 8 a.m. tomorrow, I also have a POST analyzer on the way. 😁 Seems like a good tool for the toolkit.

to clarify, the BIOS eprom is either a 27c010 or 27c020, so if I understand it correctly, the highest address pins would be A15 adn A16, pins 3 and 2 respectively.

The processor is probably fine. I think the motherboard couldn't take the stress. Hopefully it's not the chipset. Maybe a transistor. You could try recapping the board.

tauro wrote on 2024-11-07, 19:00:

The processor is probably fine. I think the motherboard couldn't take the stress. Hopefully it's not the chipset. Maybe a transistor. You could try recapping the board.

Damn. Why would they put that option in the bios if the motherboard can't handle it. I guess it was a misconception on my part. I knew the risks with the cpu, didn't think there was risk for the motherboard because of that.

75MHz never killed anything. If anything it might be caps finally dying of old age. You would need an oscilloscope to peek at Vcore and see if there is huge ripple on it.

YesAffinity wrote on 2024-11-07, 17:37:

To force BIOS into recovery mode, would that be shorting pins 2 and 3, for A16 and A15 respectively?

That is correct, on a PDIP-32 2Mbit flash chip it would be pins 2 and 3. Short them before connecting the power, short them with some metal tool, power it on and remove the tool in a second or a two. The bootblock code would fail the checksum test and go into recovery, initializing only the bare essentials needed to boot DOS from a floppy disk. On some boards even a PCI video won't work but you should see and hear floppy activity if it worked.

I'm happy to report that this PC has resumed normal operation. I'm not certain if it was forcing BIOS into recovery mode that worked or not. I tried it twice, power cycled a couple times with no improvement and no life from the floppies. I was simultaneously testing a Hercules 4500 AGP card, specifically the fan which was seized when I got it, but back to functional with a drop of 3-in-1 oil. Unknowingly, the video card was only partially seated in the AGP slot due to the retention clip of one of the IDE connectors interfering with the card fully inserting. Anyhoo, there were some error beeps, I think 1 long and 2 short, and the system booted.

I subsequently sorted everything out, the system is fully re-assembled with the 512MB of PC-1oo and Hercules 4500 card, and back at 75mhz CPU bus and 300mhz CPU speed. Well, I've got a 350mhz on the way, so that'll be getting upgraded soon. 😀

Thanks all for the help. Truly appreciated!

kmeaw wrote on 2024-11-07, 20:11:

YesAffinity wrote on 2024-11-07, 17:37:

To force BIOS into recovery mode, would that be shorting pins 2 and 3, for A16 and A15 respectively?

That is correct, on a PDIP-32 2Mbit flash chip it would be pins 2 and 3. Short them before connecting the power, short them with some metal tool, power it on and remove the tool in a second or a two. The bootblock code would fail the checksum test and go into recovery, initializing only the bare essentials needed to boot DOS from a floppy disk. On some boards even a PCI video won't work but you should see and hear floppy activity if it worked.

Is there a thread/post with instructions to do this on any type of BIOS (award, phoenix, etc)?

How are you setting all these different multipliers, is your Pentium II unlocked?

Linoleum wrote on 2024-11-08, 13:37:

Is there a thread/post with instructions to do this on any type of BIOS (award, phoenix, etc)?

I am not able to locate the instructions I have seen back then but here is another page that looks good to me: https://mastermindloltricks.blogspot.com/2012 … e-you-dont.html

It won't work with any type of BIOS - most uncompressed BIOSes (the older ones, usually 64kB or less) don't have a recovery procedure, they just beep and halt the machine if the checksum is bad.
A normal PC maps the rom chip so the last byte of ROM is available at F000:FFFF in 16-bit real mode (the mode the CPU is in when it is reset). The first instruction fetched from the ROM is F000:FFF0 (the reset vector) - 16 bytes before the last byte.
This instruction usually jumps into the middle (around F000:C000) of the F000 segment. If the BIOS is compressed, it would be the boot block - a piece of code which enables DRAM, ROM shadowing with write access and runs the decompressor, then checks the integrity of the decompressed code. If the integrity check fails, it is capable of running a very limited set of services to boot DOS from floppy.

All that code is located at the end of the ROM, so by pulling any high (A15 and above) address line would prevent the lower half of the ROM to be read causing a checksum failure and forcing the bootblock code to go into recovery. On an 8-bit data wide 2Mbit chip pulling A15 high would cause reads of 0x0000..0x7FFF,0x10000..0x1FFFF to be misdirected to 0x8000..0xFFFF,0x18000..0x1FFFF. Pulling A16 high would cause 0x0000.0xFFFF to go to 0x10000..0x1FFFF. Shorting A15 and A16 together would have a similar effect because A16 would be initially high as the reset vector points to the higher half of the ROM.

konc wrote on 2024-11-08, 14:07:

How are you setting all these different multipliers, is your Pentium II unlocked?

I assume so, 🤣. Setting them in the BIOS. It shows as 300mhz in Win98.