Have you isolated whether this is a problem for both the optical and the hard drive (by booting up with the optical drive disconnected)? Make and model of the hard drive?

Repo Man11 wrote on 2025-01-19, 23:44:

Have you isolated whether this is a problem for both the optical and the hard drive (by booting up with the optical drive disconnected)? Make and model of the hard drive?

Yup, changed IDE cables, disconnected the CD drive, tried several master/slave combinations, nothing ever worked. When enabling DMA I can only reboot and disable it from BIOS, boot in WIN, and re enable it.Then, Windows seems to be forgetting I've ever flagged the checkbox. Interestingly enough, the BIOS allows me DMA and PIO enabling on each combination of PRIMARY/SECONDARY MASTER/SLAVE. I can attach a photo of the menu.

The HDD is a Quantum Fireball 30GB UDMA5, but in the past I've used at least 2 other HDD and the results were the same.

p.s. not sure if this is worth mentioning, but I can't put 80 pin cables on this mobo. It should support UDMA33 only, and the central pin is populated, so a 80 pin cable just won't fit.

p.p.s. actually, I've got 2 x 64 MB DIMM in the RAM area. But, as I was saying, they are compatible and tested OK with memtest86.

Strange. Sounds like the on-board IDE controller loses the drives when the IDE signaling jumps up to UDMA2.

Edit: Perhaps you have a buggy BIOS. Maybe you can follow these steps to patch it:
Re: DMA Mode check does not stick in Windows 98SE?

If that's not it, can you clean the IDE pins on the board? 30 years of grime on the IDE header could be the problem.

Have you tried using the secondary IDE header?

An 80 conductor cable could help if it is a signal issue. You can find 80 connector cables that don't have the key pin. And often you can pull the key plug from block with a needle when they are keyed. It's just a plastic friction peg, no glue. (Or you can remove or bend the pin on the board if you are that sort. Just make sure you pull the correct pin!)

If you have a PCI IDE controller, it would be interesting to see if that works with DMA. That could help rule out hardware vs software.

Other random Checks:
* Make sure you have the latest motherboard BIOS.
* Make sure you don't have power cables wrapped on the ribbon cable.
* Make sure you actually have a Intel 82371AB (PIIX4) on your board.

douglar wrote on 2025-01-21, 17:16:

Strange. Sounds like the on-board IDE controller loses the drives when the IDE signaling jumps up to UDMA2. […]
Show full quote

Strange. Sounds like the on-board IDE controller loses the drives when the IDE signaling jumps up to UDMA2.

Edit: Perhaps you have a buggy BIOS. Maybe you can follow these steps to patch it:
Re: DMA Mode check does not stick in Windows 98SE?

If that's not it, can you clean the IDE pins on the board? 30 years of grime on the IDE header could be the problem.

Have you tried using the secondary IDE header?

An 80 conductor cable could help if it is a signal issue. You can find 80 connector cables that don't have the key pin. And often you can pull the key plug from block with a needle when they are keyed. It's just a plastic friction peg, no glue. (Or you can remove or bend the pin on the board if you are that sort. Just make sure you pull the correct pin!)

If you have a PCI IDE controller, it would be interesting to see if that works with DMA. That could help rule out hardware vs software.

Other random Checks:
* Make sure you have the latest motherboard BIOS.
* Make sure you don't have power cables wrapped on the ribbon cable.
* Make sure you actually have a Intel 82371AB (PIIX4) on your board.

Well, it seems I've got already the latest BIOS available to date (on the net), which is version 1.4 with signature => 03/03/98-i430TX-2A59ISH9C-00. I've checked and hardware is 82371AB indeed, cleaned pins as well. I've tried installing the Intel chipset driver 3.4.0 from theretroweb.com, and that as well doesn't work. Nothing seems to be working.

The post you've linked is interesting, but it leads to dead links and the content is not available even with wayback machine.

Not sure what else I could try here.

Is there any damage to the motherboard?

fix_metal wrote on 2025-01-22, 16:45:

The post you've linked is interesting, but it leads to dead links and the content is not available even with wayback machine.

Not sure what else I could try here.

Sounds like you got the 430tx bios with the DMA issue. The bios reports a higher DMA mode than it is capable of doing. If you try to enable DMA in windows98, windows detects that something is weird and drops back to PIO. If you force dma in the bios, windows locks up. The solution is to get a patched bios or make a patched bios.

Maybe this thread helps: "Bios Patcher" questions

Or this Video: https://www.youtube.com/watch?v=0KBYj1vx3zQ

We have 3 versions of BP Bios Patcher here: https://www.vogonsdrivers.com/index.php?catid … &menustate=77,0

fix_metal wrote on 2025-01-19, 23:41:

It works pretty much in anything, but I never managed to get DMA working. Even though all the options are set properly in the BIOS, WIN98SE will just fail in booting when I enable and reboot. Basically it gets stuck forever at screen loading when the logo shows up. Interestingly enough, the low animated bar keeps on going nevertheless.

I had this exact issue with my ASUS P5A-B. After installing W98SE and enabling DMA Windows then refuses to load.
The solution is to install the IDE controller drivers from the motherboard's install CD before enabling DMA since the default windows drivers do not properly support DMA on some motherboards. Sadly you will need to reinstall Windows since (in my experience) IDE controller drivers will not install properly once you enable DMA with the default Windows drivers.

douglar wrote on 2025-01-22, 17:45:

Sounds like you got the 430tx bios with the DMA issue. The bios reports a higher DMA mode than it is capable of doing. If you […]
Show full quote

fix_metal wrote on 2025-01-22, 16:45:

The post you've linked is interesting, but it leads to dead links and the content is not available even with wayback machine.

Not sure what else I could try here.

Sounds like you got the 430tx bios with the DMA issue. The bios reports a higher DMA mode than it is capable of doing. If you try to enable DMA in windows98, windows detects that something is weird and drops back to PIO. If you force dma in the bios, windows locks up. The solution is to get a patched bios or make a patched bios.

Maybe this thread helps: "Bios Patcher" questions

Or this Video: https://www.youtube.com/watch?v=0KBYj1vx3zQ

We have 3 versions of BP Bios Patcher here: https://www.vogonsdrivers.com/index.php?catid … &menustate=77,0

In and by itself it looked really promising, and biospatcher 2.43 found several fixed in microcode, including the DMA 33 thing.
Then, I've tried flashing the patched bios, and now the patient is mild (not cold though, which is great anyway): the pc boots with a long beep, then 2 short beeps, then something like a "you're dead" 8 bit speaker video game sound. Video does not work, keyboard as well (not even leds would light). Basing myself on the video it should have been auto recoverying by unplugging the DIN keyboard, which doesn't work.
What it does is immediately try booting from the floppy drive though. Which is a difficoult path to follow through without a screen, but most importantly without a keyboard working...

So well, now I've got even more trouble to solve, as I either have to:

1. Find someone with a programmer to flash this Macronix MX28F1000P EEPROM and get it back to its original BIOS level
2. Find someone from the bay who sells and programs these Macronix MX28F1000P with a provided ROM file

Either ways I wouldn't be able to test anymore. It'd be too risky hot swapping the EEPROM chip live to try tweaks out (moving back to original when it gets bad, then swap and patch a new version on the sacrifical one).

I'm open to ideas here 😀

Nevermind, I've figured out that the boot block minimal data was trying to call out for the floppy drive to boot something. I've come with the idea of using Award Flash or Uniflash with an autoexec.bat, and it worked. Actually, I've later found uniflash readme has even a chapter about it.
At least now I got the PC back on its feet.

fix_metal wrote on 2025-01-23, 00:04:

Nevermind, I've figured out that the boot block minimal data was trying to call out for the floppy drive to boot something. I've come with the idea of using Award Flash or Uniflash with an autoexec.bat, and it worked. Actually, I've later found uniflash readme has even a chapter about it.
At least now I got the PC back on its feet.

Glad to hear it. Flashing roms is always a little scary.

Does the new BIOS get DMA working?

ElectroSoldier wrote on 2025-01-23, 02:27:

Does the new BIOS get DMA working?

It can't even boot, I had to revert to the original one. I've tried tweaking microcode with minimal I/O stuff, but that doesn't do the trick either. I will try with just the UDMA "patching" and hope for the best.
Stay tuned!

It seems I might be out of luck on this one. I've tested passing the bios through biospatcher by just skipping any fix at all, and the pc won't boot anyway even like this.
I've tested the 1.3 bios and for whatever reason it won't recognize the boot sector/MBR on the HDD. Bios correctly scans it, but then it can't boot the OS after POST.
Unless anyone has any better ideas I guess I'll just leave it be no DMA as it's always been. So sad.

With my TX chipset motherboard, I use a PCI SATA card with an SSD. I have no idea if it has this bug (I don't think it does) but that works well either way.

Are you sure the latest BIOS version you have installed is a good version of the BIOS?

ElectroSoldier wrote on 2025-01-23, 21:45:

Are you sure the latest BIOS version you have installed is a good version of the BIOS?

Pretty much sure, yes. I didn't ever flash the mobo before the other day, and I highly doubt whoever got the PC before me did. Consider this PC has been mine since...well, I guess 2003 or 2004? The BIOS string version I have now is identical to the one I had prior to flashing anyway, and it's the same provided by theretroweb.com site => 03/03/98-i430TX-2A59ISH9C-00

I looked on retroweb, there are two different BIOS versions.
Have you only ever tried the one version of the BIOS?

fix_metal wrote on 2025-01-23, 20:36:

Unless anyone has any better ideas I guess I'll just leave it be no DMA as it's always been. So sad.

Hi fix_metal,

Yes I have a better idea, patch the offending BIOS code by hand to fix this annoying bug. 😉

The link douglar posted earlier in this thread was spot-on and the issue you are facing is indeed a BIOS bug, but often called the “Win98 UDMA bug”. This bug is present in all Award BIOSes from before 03/1999 for chipsets that support UDMA. In this timeframe that would mainly affect socket 7 boards with the i430TX chipset, but I’ve seen affected boards with Ali, SiS, and VIA chipsets as well.

As Award fixed this bug in March 1999, most board manufacturers from that era published a BIOS update to fix this bug. But with the latest 03/03/1998 BIOS for your SYE-P55TX board you seem to be out of luck here.

Strange that the BIOS Patcher tool didn’t work on your BIOS, but I can patch this bug by hand and fix the Rev 1.4 BIOS for you. I expect to have time for that this weekend.

I will report back when the BIOS is ready.
Cheers, Jan

Chkcpu wrote on 2025-01-24, 16:14:

I will report back when the BIOS is ready.

Awesome. I updated the storage wiki with your info. https://www.vogonswiki.com/index.php/Storage# … age_Limitations

I can put the batched BIOS on theretroweb.com if you like.