I just found this: http://forums.storagereview.com/index.php?/to … on-an-sis-5571/

Maybe it has to do with DMA or other IDE mode enabled in BIOS? I don't know if it is enabled or not, but I can check the BIOS in a few hours.

There was no DMA option in the BIOS.

OK, I managed to find an old 20GB IDE drive at the bottom of a bin (no, not outdoors 🤣).

I plugged it in and. . .
It gets farther!
It detects the drive as the right size. It allows me to partition and format it. It copies all the windows 2000 files. It reboots and loads the OS. This is farther than before. I gets all the way to the last file copy screen and reboots in the middle of the copy 😜 Tried this twice and it rebooted at slightly different points during the install.

I think the problem has been resolved, but I am still uncertain. I think this hard drive has problems of its own and that's why it's rebooting. I do think the mobo is OK, just not 100% sure.

Assuming the problem has been fixed by using an old hard drive like this, can someone recommend an IDE to SD card adapter that is known to work with older systems. It's weird because I thought I read reviews of people using mine for older systems. I am not sure what it is about the newer controllers that don't work with the board. I know when I booted with the new old 20GB drive, it said LBA Mode 4. I don't know if that is something to consider.

Maybe I should use compact flash instead. Would that be more likely compatible?

Have you tried using something like OnTrack Disk Manager like they did in the old days?

Nprod wrote:

Have you tried using something like OnTrack Disk Manager like they did in the old days?

I have not. Since I loaded the 128GB BIOS patch and the BIOS is reading the size correctly, the detection is not the problem. There is something wrong with the IO. Would using a drive overlay also help with these problems or only with the size limit?

I just bought a compact flash adapter in hope that it will work better than the SD card. Apparently CF is PATA so it doesn't need a converter chip. Only problem is a lot of the CF cards I look at say UDMA and I know that this machine doesn't support it, so hopefully it will default back to PIO.

Synaps3 wrote:

Nprod wrote:

Have you tried using something like OnTrack Disk Manager like they did in the old days?

I have not. Since I loaded the 128GB BIOS patch and the BIOS is reading the size correctly, the detection is not the problem. There is something wrong with the IO. Would using a drive overlay also help with these problems or only with the size limit?

Actually, the BIOS might still be the problem. It is very well possible that the parameters are properly read with the update, but the BIOS still can't handle a drive that large...

Quick Google on disk limits https://www.tldp.org/HOWTO/Large-Disk-HOWTO-4.html

Also, check for other problems that may cause IO instability (failing power supply? faulty memory?).

yawetaG wrote:

Actually, the BIOS might still be the problem. It is very well possible that the parameters are properly read with the update, b […]
Show full quote

Synaps3 wrote:

Nprod wrote:

Have you tried using something like OnTrack Disk Manager like they did in the old days?

I have not. Since I loaded the 128GB BIOS patch and the BIOS is reading the size correctly, the detection is not the problem. There is something wrong with the IO. Would using a drive overlay also help with these problems or only with the size limit?

Actually, the BIOS might still be the problem. It is very well possible that the parameters are properly read with the update, but the BIOS still can't handle a drive that large...

Quick Google on disk limits https://www.tldp.org/HOWTO/Large-Disk-HOWTO-4.html

Also, check for other problems that may cause IO instability (failing power supply? faulty memory?).

Good to know. I'll see if I can find an 8GB SD card and try that. Everything I've tried so far has been larger (except the DOS 4GB install - which might be why that one worked) because I thought the BIOS patch would have fixed the problem. What's the point of patching if it only shows the right number.

I'll try the 8GB and report back if it works.

In addition, I ran DOS benchmarks and I ran MemTest86 for 6 hours with no problems. The problem is related to the IDE controller or BIOS.

Synaps3 wrote:

yawetaG wrote:

Actually, the BIOS might still be the problem. It is very well possible that the parameters are properly read with the update, but the BIOS still can't handle a drive that large...

Quick Google on disk limits https://www.tldp.org/HOWTO/Large-Disk-HOWTO-4.html

Also, check for other problems that may cause IO instability (failing power supply? faulty memory?).

Good to know. I'll see if I can find an 8GB SD card and try that. Everything I've tried so far has been larger (except the DOS 4GB install - which might be why that one worked) because I thought the BIOS patch would have fixed the problem. What's the point of patching if it only shows the right number.

It might be that it also failed to detect the correct settings with much smaller hard disks.

I'll try the 8GB and report back if it works.

In addition, I ran DOS benchmarks and I ran MemTest86 for 6 hours with no problems. The problem is related to the IDE controller or BIOS.

What happens if you run a benchmark that does read/write cycles on the hard disk?

Furthermore, check if the problem only occurs some time after booting, or is it present right at boot? If the former is the case, you're looking at something that fails after the system has heated up, e.g. a bad solder joint. Also check the capacitors on the mobo for bulging or other damage, and maybe pop open the power supply (when disconnected from the mains!) to have a look there.

yawetaG wrote:

What happens if you run a benchmark that does read/write cycles on the hard disk?

Furthermore, check if the problem only occurs some time after booting, or is it present right at boot? If the former is the case, you're looking at something that fails after the system has heated up, e.g. a bad solder joint. Also check the capacitors on the mobo for bulging or other damage, and maybe pop open the power supply (when disconnected from the mains!) to have a look there.

It is present right at boot and it seems pretty predictable where the boot process fails.

The 8GB SD card arrived and it has the same problem. I am still not sure I have the drive parameters correct. I've been letting the BIOS do the detection on its own and it gets the correct size, so I did not pay too much attention to the cyls, heads, etc.

Here is what the BIOS says (this is the auto detection of the 8GB SD card):

1OPTIONS    SIZE   CYLS     HEAD     PRECOMP     LANDZ     SECTOR     MODE
22(Y)       7946    966       255     0         15399       63        LBA
31          7948    15400     16      65535     15399       63        NORMAL
43          7948    1925      128     65535     15399       63        LARGE

I have tested both LBA and NORMAL modes. When booting under LBA mode the BIOS reports "LBA Mode 4" and under NORMAL mode it says, "CHS Mode 4". Is this mode 4 too old to be compatible with CF or SD cards?
What I find strange is that I've heard of people using CF or SD cards on much older systems like 486s and report them working. I don't understand this.
There is also one other strange thing about this comp. It can not reboot itself. It just goes to a blank screen and then I have to turn the power off and on manually. I think this is just because I am using an ATX to AT power supply convertor.

I'm almost ready to either send this mobo back to the seller I got it from or try to sell it myself (with the above caveat). It's been consuming a lot of my time and the board I even got to replace it didn't work either - for other reasons.
It still very well might work with a normal hard drive under 8GB, but it is not worth it to me even if it does.

I have moved on to a new motherboard, but I just wanted to update this thread so that people can gain some insight if they are having a similar problem.

I am not 100% sure, but I believe my problem with this mobo's IDE controller was that it was not UDMA compatible and thus will probably only work with a real ancient hard drive. I have moved to the GA-5AX and the CF card works just fine. So if you're looking at using socket 7 system with a SiS chipset make sure you either use a real old HDD (less than 8GB) or MAYBE an ancient CF card that doesn't use UDMA.

Synaps3 wrote:

I have moved on to a new motherboard, but I just wanted to update this thread so that people can gain some insight if they are having a similar problem.

I am not 100% sure, but I believe my problem with this mobo's IDE controller was that it was not UDMA compatible and thus will probably only work with a real ancient hard drive. I have moved to the GA-5AX and the CF card works just fine. So if you're looking at using socket 7 system with a SiS chipset make sure you either use a real old HDD (less than 8GB) or MAYBE an ancient CF card that doesn't use UDMA.

I'm not so sure about that diagnosis...

It's true that the SIS 5571 only supports PIO4/DMA2, but IDE is fundamentally backwards compatible when it comes to standards. I've frequently run HDDs supporting UDMA on PIO4 chipset boards and they work - albeit at slower transfer speeds of course. So lack of UDMA support itself isn't the issue. I rather suspect that whatever beta BIOS patch you have installed that shows >128GB HDDs correctly is just plain buggy. Either that or as Yawetag suggests there might be other issues upstream of the I/O subsystem. With the GA-5AX you're using different RAM for starters, so that's still a possible cause.

dionb wrote:

I rather suspect that whatever beta BIOS patch you have installed that shows >128GB HDDs correctly is just plain buggy.

This is also my feeling.
For my DFI K6 board there were also offered BIOS patches, but they did not work reliably.
For this reason on my retro systems I do not bother dealing with IDE drives at all. An IDE or SATA optical drive for booting, and all HDDs on SCSI.