I also disabled the onboard IDE controllers and tried an external ISA IDE card, too see if it would make a difference. Unfortunately it didn't help. Any ideas? This motherboard really has me stumped...

Hmm... your symptoms are related to data integrity, but you've tried multiple I/O controllers, multiple drives, different RAM and a different CPU. That leaves two suspects:

1) L2 cache. Try disabling it and see if problems go away.
2) UMA. The chipset shares memory bandwidth between CPU and VGA. Possibly somethingis buggy there. Try an external PCI VGA card.

Thanks! I tried both PCI and ISA vga cards and did not make any difference.

The bios doesn't have an option for disabling the cache unfortunately. Unless there's some sort of hidden way to do it.

I changed the suspicious looking capacitors (one of them was indeed bad), but unfortunately it did not help. I think the next step would be to flash the bios to a newer version, and hopefully that might allow me to disable the l2 cache? Anyone else has any other ideas?

What makes you think a newer BIOS version will add that functionality? Unless specified in release notes, newer BIOS versions will generally only fix bugs.

There is a slightly newer retail BIOS for the board here: https://web.archive.org/web/20101206223259/ht … /IN5598bios.htm (BCM IN5598 = GVC/PB FR500), but the only feature it adds is K6-2 support. As a retail BIOS instead of an OEM one, it might offer more options, but that's certainly not always the case (my PB BIOS for MSI MS-6168v2 has more options than the retail MSI BIOS).

Ooh this sounds like a really annoying intermittent fault, the absolute worst, and I think you've done a great job trying to find the cause, sorry it's not born much fruit yet.

I've got a similar Dell Optiplex GN+ motherboard, not tested it yet as I've still to get a power supply sorted for it as it was a proprietry Dell interface, however I know that it's got a similarly hamstrung BIOS that doesn't have options for L2 Cache Control. I'd like to be able to control L2 so I can slow it down to ~386 speeds for older DOS games.

If I can get it running, I'm going to see if I can mess about with Debug or write a basic DOS .COM executable in assembly to try and change the Cache Control register at a 52h offset to set the disable cache bits.

A quick look at the SiS SiS5597 SiS5598 datasheet and it's got a similar register at 51h "L2 Cache Controller":

Bit 7 L2 Cache exists Selection 0: No L2 Cache 1: L2 Cache Exists When no L2 exists, this bit should be programmed to 0. Bit 6 L […]
Show full quote

Bit 7 L2 Cache exists Selection
0: No L2 Cache
1: L2 Cache Exists
When no L2 exists, this bit should be programmed to 0.
Bit 6 L2 Cache Enable
0: Disable
1: Enable

I don't know how to access these registers yet, the Intel 430TX chipset I have I think uses CONFADD 0CF8h and CONFDATA 0CFCh registers to access these, but I haven't really figured it out yet - there might be some good information on setting these chipset registers online somewhere?

Anyway maybe this would allow you to try disabling the L2 cache for testing?

I've asked here Anyone know of a DOS Utility to disable Level 2 cache on 430TX chipset? to see if anyone knows of existing utilities to disable L2 cache from DOS instead of relying on the BIOS.

Also, maybe disabling L2 is not possible after boot at DOS, I'm no expert on how caches work, I just know L1 can be toggled, so figured L2 would be ok to toggle too?

Good luck!

Ah-ha, so mornings reading, I have figured out:

1. These CONFADD and CONFDATA registers are just standard Mechanism #1 way of PCs to access the PCI Configuration space which is where the chipset registers are: https://wiki.osdev.org/PCI#Configuration_Spac … _Mechanism_.231
2. Googling for DOS tools to access these registers plonked me back to - VOGONS! And a promising sounding tool called TweakPCI: TweakPCI - A DOS utility to view/modify PCI configuration registers

I'll be checking this out next to see if I can use it to examine and tinker with the Cache Control registers - well if I get the board runining that is...

My 2 seconds of reading TweakPCI and the SiS datasheet gives me the following command you could try and run and see if it disabled your L2 cache from DOS:

1tweakpci 5597 1039 b51 6=0

Don't know if I have the Vendor & Device IDs correct (taken from the Datasheet), or just setting a single bit will do the job or if I even have TweakPCI syntax right, but you never know?!

fenderstrat wrote on 2023-02-03, 04:35:

Hi Everyone,

.....

Thanks!

So the system is completely unstable. Normally a Windows (98 for example) installation is a good test, but in this case you won't even come so far, that you can start it.

Are there some timing parameters in the BIOS which you can slow down?
Maybe reset settings to factory default or bios default?

Ok, update: I finally got around to working on this board again. I flashed a newer updated bios and did not change anything, except that this new bios allows me to disable the L2 cache. After disabling the cache, the machine works just fine, no issues whatsoever! So yay, I found the culprit! I'm just going to see if there's something wrong with the cache chip(s?) now. Thanks everyone for your help!