Hi:

UDMA depend on Chipset and HDD features, you can't implement something on the BIOS if the HW don't have the capacities.

Regards!

No, 430TX chipset supports UDMA-2 33MB/s but seems it's the Award BIOS bug when using newer HDDs 66MB/s or higher:

http://www.ryston.cz/petr/bios/award.html

13. UDMA mode limitation 
2In the older Award BIOSes was the HDD set to highest supported mode, it was not problem when only UDMA33 HDDs were produced, but later problem arised: HDD was programmed to e.g. 66 MHz and chipset maximum was 33 MHz, the result was unreliable operation. There are some solutions, switch off UDMA in BIOS setup at all, or use special program supplied by HDD manufacturer and switch off ATA/66 support. Better way is to implement this directly into the BIOS code. FYI: 
3UDMA mode 2 = ATA/33 
4UDMA mode 4 = ATA/66 
5UDMA mode 5 = ATA/100 
6UDMA mode 6 = ATA/133

RiP wrote:

No, 430TX chipset supports UDMA-2 33MB/s but seems it's the Award BIOS bug when using newer HDDs 66MB/s or higher: […]
Show full quote

No, 430TX chipset supports UDMA-2 33MB/s but seems it's the Award BIOS bug when using newer HDDs 66MB/s or higher:

http://www.ryston.cz/petr/bios/award.html

Copy code to clipboard

13. UDMA mode limitation 
2In the older Award BIOSes was the HDD set to highest supported mode, it was not problem when only UDMA33 HDDs were produced, but later problem arised: HDD was programmed to e.g. 66 MHz and chipset maximum was 33 MHz, the result was unreliable operation. There are some solutions, switch off UDMA in BIOS setup at all, or use special program supplied by HDD manufacturer and switch off ATA/66 support. Better way is to implement this directly into the BIOS code. FYI: 
3UDMA mode 2 = ATA/33 
4UDMA mode 4 = ATA/66 
5UDMA mode 5 = ATA/100 
6UDMA mode 6 = ATA/133

Interesting info, but modifiying the BIOS to support this UDMA modes require a lot of effort and knowledge, it's not "trivial" like inject a microcode for a CPU recognition
or change value/strings on the code (like the 34GB bug).

Are you tried a 80 pin ide cable?
an IDE/ISA board?

Regards!

Actually, the BIOS should be modified to detect anything higher than UDMA2 33MB/s correctly.
No, I only use 40-pin cable for old motherboards without any problem. Many users tried 80-pin and got issue.
It's TX430's onboard busmaster IDE controller.

This utility can modify too:
http://www.rom.by/articles/BP/index_english.htm

1=> Patcher can make support of big HDD (up to 137 Gb) with fixing "UDMA"-problem (HDD UDMA66/100/133 on only_UDMA33 MB)

Well they present a utility, but I don't see any download links 🙁

That could be very interesting to retrieve !

Deksor wrote:

Well they present a utility, but I don't see any download links 🙁

That could be very interesting to retrieve !

http://www.rom.by/Award/patcher/bp-4_23.rar
http://www.rom.by/Award/patcher/BP-4_51_beta.rar
http://www.rom.by/Award/patcher/BP-6a9.rar

RiP wrote:

fixed only the 128GB HDD size and not the UDMA 😢
When it's set as Auto, DMA doesn't work at all in Win98 and when it's set as Disabled, DMA works (low CPU usage) but only at half of the speed (16MB/s)

This is just placebo effect. Win98 inbox Standard PCI IDE driver never depends on BIOS to support MDMA/UDMA modes. BIOS IDE configuration only affects INT13h which is mostly used by DOS. The HDD geometry fix is all you need.

@manuelink64 awesome ! Thanks a lot !

Yes, yes, necromancy, I know. First post. Tut tut.
Whatever.

This thread just solved all my socket7 problems, and it'd be rude not to share and to thank... More rude than a necropost even. 😜

For completeness, here's the tale:
My ECS P5TX-B (free, but apparently a little uncommon, and ECS really sucks for support) had both the 32GB limit and UDMA detection problems - it erroneously set UDMA5 mode on my disks while the chipset can only do UDMA2.

The 32GB limit didn't really bother me, but Windows '98SE dropped to PIO mode (atto bench: ~9.5 MB/S) as soon as any drive supporting >UDMA2 was on the bus. Disabling UDMA in BIOS allowed Windows to use DMA, but with severely limited performance (atto bench: ~15 MB/s).
Both of these modes led to severe loading stutter in many games, and generally miserable performance in any application needing to do much disk I/O. It was certainly no placebo either, the benchmarks were conclusive... This system had disk performance not much better than a fast 486.
Thinking to eliminate hardware problems (e.g. 40 conductor cables, mine being from 1993 IIRC), I fired up an old Debian GNU/Linux (Etch) livecd... Because GNU/Linux pretty much just assumes the BIOS is broken, and probes hardware directly. I got a nice message from the driver that it had selected UDMA2, and benchmarks confirm this. A sweet ~30MB/s with minimal CPU overhead.

Right. That means it's a) just Windows being Windows b) my Windows drivers being broken, or c) the BIOS lying about supported UDMA modes and Windows falling for it.

This led me, after many hours stumbling around ancient mailing lists and archived support FAQs, to the right search keywords and this thread.
All the symptoms match, the chipset, BIOS vendor and dates all line up. Bingo. It's the BIOS. Colour me not very surprised, and not very optimistic about fixing it.

But wait, there's hope yet... Thanks to @RiP and @manuelink64, I had a slightly shady, potentially risky (I don't own a 'prom burner or know anyone who does) way to patch the BIOS...

And it worked. It worked so well I'm pretty floored TBH. 😁
Bios now reporting correct UDMA2 mode on all disks. DMA automagically enabled in Windows. ~25MB/s in atto bench with negligible CPU load. 😀
I'll try a 120GB disk in it sometime soon, but I expect that's fixed as well.

For anyone in the same boat, you'll need:
http://rom.by/Award/patcher/bp-4_23.rar
http://rom.by/Award/patcher/cbrom.rar
http://rom.by/Award/patcher/lha.rar
http://rom.by/Award/patcher/real_microcodes.rar
An image of your broken, pre October 1998 Award BIOS.
The appropriate tool to flash it.
Some courage.

Extract everything to a directory somewhere.
Run bp-4_23.exe <your_rom_file>
Flash.
Reboot.
Profit.

I've stashed the needed patcher files (and a fixed BIOS for my board, based on v2.1d) here in case that site goes away (personal server, there until it isn't), and I humbly suggest someone with a more permanent host might want to mirror them as well (vogonsdrivers perhaps?).
ECS BIOS images and manuals are also a pain to find, the ECS website is hopeless but a mirror can be found at https://www.infania.net/misc/moboarchive/. That might be worth mirroring too IMO. I'm crawling it now (slowly and politely), we'll see how that turns out.

Thanks again to @RiP, @manuelink64, and the genius who created this wonderful patching tool, whoever you are. You made my day. 😀

Ed. Yup, both 40GB and 120GB disks detected and working fine. 😁

https://www.infania.net/misc/moboarchive/ is an illegal mirror of https://www.elhvb.com/mobokive/Archive/index.html

RiP wrote on 2019-03-27, 19:18:

I could find modded BIOS for some Atrend motherboards that fixed the 128GB HDD size and UDMA but sadly the modded BIOS for Taken […]
Show full quote

I could find modded BIOS for some Atrend motherboards that fixed the 128GB HDD size and UDMA but sadly the modded BIOS for Taken TX3 (LGS Prime3C) fixed only the 128GB HDD size and not the UDMA 😢
When it's set as Auto, DMA doesn't work at all in Win98 and when it's set as Disabled, DMA works (low CPU usage) but only at half of the speed (16MB/s)

Can anyone fix the modded BIOS? 😕
http://wims.rainbow-software.org/edwin/

Oh and there's a page for your board too now, didn't realize it was a Taken at first:
https://www.ultimateretro.net/en/motherboards/10464

steve_v wrote on 2021-12-24, 06:13:

Yes, yes, necromancy, I know. First post. Tut tut. Whatever. […]
Show full quote

Yes, yes, necromancy, I know. First post. Tut tut.
Whatever.

This thread just solved all my socket7 problems, and it'd be rude not to share and to thank... More rude than a necropost even. 😜

For completeness, here's the tale:
My ECS P5TX-B (free, but apparently a little uncommon, and ECS really sucks for support) had both the 32GB limit and UDMA detection problems - it erroneously set UDMA5 mode on my disks while the chipset can only do UDMA2.

The 32GB limit didn't really bother me, but Windows '98SE dropped to PIO mode (atto bench: ~9.5 MB/S) as soon as any drive supporting >UDMA2 was on the bus. Disabling UDMA in BIOS allowed Windows to use DMA, but with severely limited performance (atto bench: ~15 MB/s).
Both of these modes led to severe loading stutter in many games, and generally miserable performance in any application needing to do much disk I/O. It was certainly no placebo either, the benchmarks were conclusive... This system had disk performance not much better than a fast 486.
Thinking to eliminate hardware problems (e.g. 40 conductor cables, mine being from 1993 IIRC), I fired up an old Debian GNU/Linux (Etch) livecd... Because GNU/Linux pretty much just assumes the BIOS is broken, and probes hardware directly. I got a nice message from the driver that it had selected UDMA2, and benchmarks confirm this. A sweet ~30MB/s with minimal CPU overhead.

Right. That means it's a) just Windows being Windows b) my Windows drivers being broken, or c) the BIOS lying about supported UDMA modes and Windows falling for it.

This led me, after many hours stumbling around ancient mailing lists and archived support FAQs, to the right search keywords and this thread.
All the symptoms match, the chipset, BIOS vendor and dates all line up. Bingo. It's the BIOS. Colour me not very surprised, and not very optimistic about fixing it.

But wait, there's hope yet... Thanks to @RiP and @manuelink64, I had a slightly shady, potentially risky (I don't own a 'prom burner or know anyone who does) way to patch the BIOS...

And it worked. It worked so well I'm pretty floored TBH. 😁
Bios now reporting correct UDMA2 mode on all disks. DMA automagically enabled in Windows. ~25MB/s in atto bench with negligible CPU load. 😀
I'll try a 120GB disk in it sometime soon, but I expect that's fixed as well.

For anyone in the same boat, you'll need:
http://rom.by/Award/patcher/bp-4_23.rar
http://rom.by/Award/patcher/cbrom.rar
http://rom.by/Award/patcher/lha.rar
http://rom.by/Award/patcher/real_microcodes.rar
An image of your broken, pre October 1998 Award BIOS.
The appropriate tool to flash it.
Some courage.

Extract everything to a directory somewhere.
Run bp-4_23.exe <your_rom_file>
Flash.
Reboot.
Profit.

I've stashed the needed patcher files (and a fixed BIOS for my board, based on v2.1d) here in case that site goes away (personal server, there until it isn't), and I humbly suggest someone with a more permanent host might want to mirror them as well (vogonsdrivers perhaps?).
ECS BIOS images and manuals are also a pain to find, the ECS website is hopeless but a mirror can be found at https://www.infania.net/misc/moboarchive/. That might be worth mirroring too IMO. I'm crawling it now (slowly and politely), we'll see how that turns out.

Thanks again to @RiP, @manuelink64, and the genius who created this wonderful patching tool, whoever you are. You made my day. 😀

Ed. Yup, both 40GB and 120GB disks detected and working fine. 😁

God bless you @ steve_v !!!!!!

I've been banging my head on the PCchips M577 for a week now, trying everything under the sun; and could not get win98SE to boot with "DMA" checked (in windows), and also UDMA mode in BIOS.
It was always one or the other; not both.

Although, in DOS the speed was very good with UDMA on (instaled Win98 amazingly fast) , I started suspecting something was still wrong with the bios (07/19/2022 J3 patched BIOS by Joszef Csongradi and Jan Steunebrink), when at boot, the HDD was detected in it's maximum state (UDMA 5), on an ATA33 chipset (VIA MVP3) 😳

Needless to say, in win98, HDD sequential speeds were crap.
Could not go above 10MB/s no matter what driver, setting, HDD, ide connector... i tried

For anyone having the same problem, i'll attach the modded BIOS below.
It's the J3 Patch BIOS from the "990309J3.ZIP" on this website "http://www.steunebrink.info/k6plus.htm", but I also ran it through "Bios Patcher", and applied ONLY 2 fixes that addressed proper HDD & UDMA detection.

It's tested on the PC Chips M577, but it also should be good for the "Amptron PM-9900" .. and maybe these clone boards :

Protac MB 5770
Aristo AM-577
EURONE EM-5577S
EURONE EM-5577
Minstaple PM 577

Hi B24Fox,

Thanks for sharing the modded PCChips M577 BIOS with a fix for the Windows 98 UDMA bug by the “BIOS Patcher” tool.
This stimulated me to improve the modded patch J.3 BIOS as well.

When I was preparing the M577 patch J.3 BIOS last year, with the improved 32GB and 64GB HDD limit bug fixes, my patch for the UDMA bug was not yet ready. So unfortunately the M577 patch J.3 BIOS still has the UDMA bug.
But I have a working and tested fix for the UDMA bug now and I put it in the modded M577 BIOS, resulting in a patch J.4 version. Here is a copy for you to try.

http://www.steunebrink.info/bios/990309J4.zip

Of course, all the other fixes from the J.3 BIOS are present in this patch J.4 version as well. In addition, I’ve changed the Setup Defaults for the UDMA mode from disabled to Auto for all IDE devices.

Please let us know how this patch J.4 BIOS works on your M577.

Regards, Jan

Chkcpu wrote on 2023-10-14, 14:23:

Hi B24Fox, […]
Show full quote

Hi B24Fox,

Thanks for sharing the modded PCChips M577 BIOS with a fix for the Windows 98 UDMA bug by the “BIOS Patcher” tool.
This stimulated me to improve the modded patch J.3 BIOS as well.

When I was preparing the M577 patch J.3 BIOS last year, with the improved 32GB and 64GB HDD limit bug fixes, my patch for the UDMA bug was not yet ready. So unfortunately the M577 patch J.3 BIOS still has the UDMA bug.
But I have a working and tested fix for the UDMA bug now and I put it in the modded M577 BIOS, resulting in a patch J.4 version. Here is a copy for you to try.

http://www.steunebrink.info/bios/990309J4.zip

Of course, all the other fixes from the J.3 BIOS are present in this patch J.4 version as well. In addition, I’ve changed the Setup Defaults for the UDMA mode from disabled to Auto for all IDE devices.

Please let us know how this patch J.4 BIOS works on your M577.

Regards, Jan

WOW! Thank you Jan!!
The added text by "BiosPatcher" was covering the cpu detection line/text in the POST screen, and was driving me crazy!
Tried everything to remove it, but couldn't...
This couldn't have come at a better time!!

I will test the new BIOS this next week; and come back with a response.

Once again, thank you! 🥂

Chkcpu wrote on 2023-10-14, 14:23:

Hi B24Fox, […]
Show full quote

Hi B24Fox,

Thanks for sharing the modded PCChips M577 BIOS with a fix for the Windows 98 UDMA bug by the “BIOS Patcher” tool.
This stimulated me to improve the modded patch J.3 BIOS as well.

When I was preparing the M577 patch J.3 BIOS last year, with the improved 32GB and 64GB HDD limit bug fixes, my patch for the UDMA bug was not yet ready. So unfortunately the M577 patch J.3 BIOS still has the UDMA bug.
But I have a working and tested fix for the UDMA bug now and I put it in the modded M577 BIOS, resulting in a patch J.4 version. Here is a copy for you to try.

http://www.steunebrink.info/bios/990309J4.zip

Of course, all the other fixes from the J.3 BIOS are present in this patch J.4 version as well. In addition, I’ve changed the Setup Defaults for the UDMA mode from disabled to Auto for all IDE devices.

Please let us know how this patch J.4 BIOS works on your M577.

Regards, Jan

I just flashed my PCChips M577 mobo with your latest J.4 BIOS. And the HDD seems to work just as well as the "J.3 + BiosPatcher" BIOS... but without the annoying text that covers the CPU in the POST screen 😁
Thank you so much Jan!

There was a WD400 (40gb hdd) that i tried at one point, and it's size was not properly detected by the bios (detected as only 8GB), but was detected ok in windows (98SE).
That happened on J.3 .. w/ and wo/ the BiosPatcher patch.
I will report back when I get the chance to retest that HDD also.