https://theretroweb.com/motherboards/s/biosta … m5alc#downloads

What are you trying to point me to on the page?

I believe you're being pointed to a BIOS upgrade. Also, NT 5.x based systems are listed on the page as supported.

Otherwise, ACPI is probably the culprit. If you can disable APCI in the BIOS, try that. If there is no option or you're still having difficulties during setup, run Windows 2000 setup once more, but Press "F5" as soon as the "Setup is inspecting your computer" phase begins. I usually press F5 a number of times, but it's not necessary. You will be given the option to load a specific HAL instead of letting Windows detect it. Choose "Standard PC" for the HAL. As an aside, I've read pressing F7 instead of F5 selects the Standard HAL automatically, but I've never tried it out.

Ok, I'll try the F5 thing. The board already has a version of the newest BIOS on the page modded by Jan to support drives larger than 32GB

Sukhoikip wrote on 2023-05-26, 22:29:

What are you trying to point me to on the page?

I added the link to help others avoiding looking it up.
Yes, BIOS upgrade.

Maybe unsupported chipset that requires specific drivers? Win 2000 Service Pack 4? or no SP?

The 2000 disk is SP4, the XP is SP3. There is no ACPI by name, but disabling power management made 2000 get past "Starting Windows 2000" once though it immediately bluescreened. Selecting the Standard PC HAL in both 2k and XP changed the second parameter of the recurring BSOD to 0x0000001C.

Do you have caching enabled for Video and/or System BIOS shadowing? If so, try disabling those.

What about COM/Parallel ports? Try disabling those.

Has this BIOS you installed with the modifications been tested with 2000/XP before? I wonder about the results if you revert to the original.

I've had issues with XP refusing to work on computers that I'd thought were running fine with Windows 98; I found that the tight memory timings I had set would not work with XP. They actually didn't work with Win98 either, but I had blamed the instability on the OS. I would try running Memtest86 for a few loops and see if it passes.

I once had a S370 VIA Apollo motherboard, on which Windows with an NT kernel was installed only as an update from Windows 98.
You install Windows 98 from a CD, run an update on Windows XP from it, convert FAT32 to NTFS 🙁

I've tried disabling what you suggested Meatball, now it's back to the original error. I can't revert to another BIOS because the one one there is the only one that supports >32GB disks and fits on a 1Mbit EPROM.

Repo Man11, I don't think the timings are suspect. Even if it's pulling straight from the SPDs, the sticks in there are PC133 running at PC100.

shevalier, I've already tried launching the 2000 installer from 98 and just telling it to set up on the SSD. Same result. I haven't tried installing 98 on the SSD and upgrading it yet as my 98 install disk is missing.

Try installing the OS on another system (with the SSD transferred). After the copy phase is over, bring the disk back to the SS7. This is a reach, but I'm running out of ideas!

Maybe you could post here all of the BIOS settings as configured for us to review?

Do you have a spare non-ssd to try?
If so install it on that disk.

If not I guess you need to look for a guide SSD W2K install.
Some trick might be involved.

Doing the first half of the setup on another PC will have to wait, BSODed before entering the interactive part on 2 separate systems (LGA 775 and 2011-3) with an INACCESSIBLE_BOOT_DEVICE error. I'll have to try my S370 system later.

Nexxen, I do have a CF-IDE adapter and 16GB CF card, though I'm pretty sure NT5.x needs additional fiddling to install on removable media.

Some things to test:

"Enable" Power Management. There should be a "User" setting.

Set Power Management by APM to "No." For Windows 9x, I set it to Yes to disable ACPI, but not for Windows 2000/XP.

Optionally, disable Power Management PM events for COM Ports and Keyboard to "No."

PnP System should be set to "Yes" except for Windows NT 4.0, 3.x, and earlier; Windows 95 can be less fiddly from being off, though, depending on hardware installed.

Resource Control should be set to "Auto" unless you have a specific IRQ/DMA you need to set for an ISA card. You're not using DOS/Windows 9x, so no need for manual. (unless you're going to dual boot?)

Enable PCI Delayed transaction. This hurts PCI performance as disabled. Only disable with non-2.1 compliant hardware.

Set VGA Frame Buffer to "Disabled," your newer cards will not benefit.

You probably don't need to assign an IRQ to VGA because of Windows 2000/XP - only useful for Windows 9x (on occasion) depending on the Video Card requirements. Optional, but since nothing is working...

From my own experience - installing Windows with the NT kernel reveals unstable PCs.
I would test the memory and CPU at maximum load.
It's trite, but maybe it's time to change the capacitors.

Sukhoikip wrote on 2023-05-27, 13:46:

I've tried disabling what you suggested Meatball, now it's back to the original error. I can't revert to another BIOS because the one one there is the only one that supports >32GB disks and fits on a 1Mbit EPROM.

Repo Man11, I don't think the timings are suspect. Even if it's pulling straight from the SPDs, the sticks in there are PC133 running at PC100.

shevalier, I've already tried launching the 2000 installer from 98 and just telling it to set up on the SSD. Same result. I haven't tried installing 98 on the SSD and upgrading it yet as my 98 install disk is missing.

Overly tight memory timing just happened to be the cause of the instability that (seemingly) worked with Win98 but not with WinXP; there are many others. If it cannot pass memtest, you then know the problem is in the hardware rather than the software. If it can pass memtest, that's also good information to know.

Meatball, I tried the above settings, it generates a different error, attached. After a few POSTs the system stopped POSTing and instead made a couple long beeps followed by endless short beeps. Clearing the CMOS seems like a good point to stop for the night.

Repo Man, I set the timings to "slow" which is the best the BIOS would allow even if the RAM was the problem. The RAM is probably fine as it's the same kit I used for hwbot X265 on my S370, during which it ran for almost 30 hours at PC140. Is there a way to just get the memtest program so I can run it from the DOS prompt rather than burn an entire CD to test one system?

Sukhoikip wrote on 2023-05-28, 02:38:

Meatball, I tried the above settings, it generates a different error, attached. After a few POSTs the system stopped POSTing and instead made a couple long beeps followed by endless short beeps. Clearing the CMOS seems like a good point to stop for the night.

Repo Man, I set the timings to "slow" which is the best the BIOS would allow even if the RAM was the problem. The RAM is probably fine as it's the same kit I used for hwbot X265 on my S370, during which it ran for almost 30 hours at PC140. Is there a way to just get the memtest program so I can run it from the DOS prompt rather than burn an entire CD to test one system?

Based on your description of the most recent POST codes, I wonder if there is a CPU problem. Have you tried a different K6-2? Are you able to swap in a Pentium MMX?

You didn't use the F5 and Standard HAL setting in this scenario, right? You shouldn't, but just checking.

Meatball wrote on 2023-05-28, 03:06:

Sukhoikip wrote on 2023-05-28, 02:38:

Meatball, I tried the above settings, it generates a different error, attached. After a few POSTs the system stopped POSTing and instead made a couple long beeps followed by endless short beeps. Clearing the CMOS seems like a good point to stop for the night.

Repo Man, I set the timings to "slow" which is the best the BIOS would allow even if the RAM was the problem. The RAM is probably fine as it's the same kit I used for hwbot X265 on my S370, during which it ran for almost 30 hours at PC140. Is there a way to just get the memtest program so I can run it from the DOS prompt rather than burn an entire CD to test one system?

Based on your description of the most recent POST codes, I wonder if there is a CPU problem. Have you tried a different K6-2? Are you able to swap in a Pentium MMX?

You didn't use the F5 and Standard HAL setting in this scenario, right? You shouldn't, but just checking.

I did try the Standard PC HAL with the BIOS settings you suggested, didn't realize I shouldn't. I do not have another socket 7 CPU, but I'm not inclined to blame it unless the system doesn't recover.