All right then, since everybody is asking me to push the limits... I'll try 3x83 and check if the CPU can do 3.5x75 as a first step 😀 😀 We'll see how it goes.

Chkcpu wrote on 2023-12-18, 17:23:

Hi Sphere, No, the Tyan S1564 is not affected by this Win98 UDMA bug because the i430HX chipset uses the 82371SB (PIIX3) southbr […]
Show full quote

Sphere478 wrote on 2023-12-17, 23:46:

@chkcpu, is that pio bug present on tyan s1564?

Hi Sphere,
No, the Tyan S1564 is not affected by this Win98 UDMA bug because the i430HX chipset uses the 82371SB (PIIX3) southbridge that doesn’t do UDMA. The same goes for boards with the i430FX and i430VX chipset.
So your S1564 will happily run Win98 in PIO 4 or MW DMA 2 modes via the on-board IDE channels.

Only socket 7 boards running a 02/1999 or earlier Award BIOS with UDMA supporting chipsets are affected by this bug. So boards with i430TX, ALi154x, SiS-5582/5591/5598, and VIA VP2/VP3/MVP3 chipsets to name a few. But you should be fine when running a 03/1999 or later BIOS on these boards.
It is possible that early Award BIOSes on Slot 1 boards are affected as well, but I have no data on that.

Cheers, Jan

A patch for the Chaintech 5SIM with SiS 5582 (and of course all the others) would be very much appreciated 😉
Already using your patched BIOS (huge THANKS for that!), but an additional UDMA bugfix would be really great.
Having to rely on HDDs with artificially limited UDMA-modes can be quite annoying (the way I got UDMA to work on this board at all).

Wasn't there widespread data loss back in the day when people ran 83 MHz fsb, giving 41.5 on PCI, thus putting pressure on the harddisk controllers and finally the harddisks? IIRC, at least some of the IBM Deathstars weren't very fond of 41.5 MHz PCI speeds?

Never heard about that but I remember that my ISA Sound Blaster Pro2 was giving me some trouble after overclocking the system...

ChrisK wrote on 2023-12-19, 09:22:

A patch for the Chaintech 5SIM with SiS 5582 (and of course all the others) would be very much appreciated 😉
Already using your patched BIOS (huge THANKS for that!), but an additional UDMA bugfix would be really great.
Having to rely on HDDs with artificially limited UDMA-modes can be quite annoying (the way I got UDMA to work on this board at all).

Hi ChrisK, I've put the CH-5SIM BIOS on the list for the UDMA bugfix. Now that I'm redoing the patched BIOS, I will put in the improved 32GB HDD limit bugfix as well. Give me a few weeks to get it ready.

Happy holidays, Jan

mtest001 wrote on 2023-12-17, 20:54:

Thanks for the feedback @Chkcpu and for proposing a fix, that would be very nice. I'll be happy to be the guinea pig for you 😀

It's a pity the BIOS settings for UDMA only propose "AUTO" or "Disabled", and there is no way to manually force it.

Hi mtest001,
The new HOT-569 patch J.3 BIOS version is ready and I sent you a download link via PM.

Please let us know how this new BIOS works.

Regards, Jan

Thanks a lot for this Christmast gift !
I'll give it a try asap and give some feedback here.

Chkcpu wrote on 2023-12-27, 12:38:

The new HOT-569 patch J.3 BIOS version is ready and I sent you a download link via PM.

Please let us know how this new BIOS works.

Regards, Jan

Hello Jan,
Happy to report that the patch "J3" woks well. Now I am able to check the DMA box in Windows98 and it stays checked.

Still a bit disappointed by the performance of the CF card, I was expecting better speed:

...and cherry on the cake, now my 128 GB SSD is detected by the BIOS:

128 GB is a bit overkill for Win98 ! I will never use that space...

Thanks again for the patch !

Chkcpu wrote on 2023-12-27, 12:38:

Hi mtest001, The new HOT-569 patch J.3 BIOS version is ready and I sent you a download link via PM. […]
Show full quote

mtest001 wrote on 2023-12-17, 20:54:

Thanks for the feedback @Chkcpu and for proposing a fix, that would be very nice. I'll be happy to be the guinea pig for you 😀

It's a pity the BIOS settings for UDMA only propose "AUTO" or "Disabled", and there is no way to manually force it.

Hi mtest001,
The new HOT-569 patch J.3 BIOS version is ready and I sent you a download link via PM.

Please let us know how this new BIOS works.

Regards, Jan

Hi, Jan! Is this freshly modified HOT-569(non-A) BIOS version better than the one for the latest version of this mobo HOT-569A, which is long time available for downloading on your site?

Do you know what's the tweak in MODBIN (my mistake as I always use these two tools when patching old Award BIOSes) BIOS Patcher 4.xx/6.00 for Ultra ATA-66/100/133 if applied to intel 430TX mobo. I have one 430TX board with patched BIOS by this software, but still I haven't bother to check if it really supports faster speeds than ATA-33 under Win98 after the patching.

mtest001 wrote on 2023-12-29, 23:25:

Hello Jan,
Happy to report that the patch "J3" woks well. Now I am able to check the DMA box in Windows98 and it stays checked.

Still a bit disappointed by the performance of the CF card, I was expecting better speed:

Thanks for your elaborate test report on the HOT-569 patch J.3 BIOS, looks good! 😀
Both the UDMA and 32GB limit bugfixes appear to work as expected.

About the disappointing read speed you see in Speedsys, this may be because you run this check from DOS. I’m unsure if and how UDMA works in DOS but I believe you need a device driver for that.

This is what I see under a fresh Win98 install on my 430TX board (Abit AB-PX5), without and with the same UDMA bugfix:

This was measured using ATTO Disk Benchmark v2.41 (2007 version) on an 8GB Transend CF 133 card. The lower Write speeds you see here are probably the maximum my CF card can handle.

Here is the new HOT-569 patch J.3 BIOS for anyone who needs it.
Note that this BIOS is not for the HOT-569A. The 569A uses a different BIOS, but when needed I can fix the 569A BIOS as well.

Cheers, Jan

analog_programmer wrote on 2023-12-30, 00:38:

Hi, Jan! Is this freshly modified HOT-569(non-A) BIOS version better than the one for the latest version of this mobo HOT-569A, which is long time available for downloading on your site?

Do you know what's the tweak in MODBIN 4.xx/6.00 for Ultra ATA-66/100/133 if applied to intel 430TX mobo. I have one 430TX board with patched BIOS by this software, but still I haven't bother to check if it really supports faster speeds than ATA-33 under Win98 after the patching.

Hi,

Yes, the new patch J.3 BIOS version for the HOT-569 is now more advanced than the patch J.2 BIOS for the HOT-569A on my website. I will be reworking the 569A BIOS as well, but this will take some time.
Do you have this board and could you test an improved 569A BIOS?

The i430TX chipset was one of the first to support UDMA 33 via the on-board IDE channels. As this ATA-33 support is limited by hardware, only overclocking the PCI-bus could run it a bit faster but I never heard of the tweak you mentioned.
Do you have a reference?

Regards, Jan

Chkcpu wrote on 2023-12-30, 17:20:

Yes, the new patch J.3 BIOS version for the HOT-569 is now more advanced than the patch J.2 BIOS for the HOT-569A on my website. I will be reworking the 569A BIOS as well, but this will take some time.
Do you have this board and could you test an improved 569A BIOS?

Very well! Yes, I have Shuttle HOT-569A variant and also I have XGecu programmer, so I'll gladly test the updated BIOS-mod version for this newest revision of the mobo when it's done.

Chkcpu wrote on 2023-12-30, 17:20:

The i430TX chipset was one of the first to support UDMA 33 via the on-board IDE channels. As this ATA-33 support is limited by hardware, only overclocking the PCI-bus could run it a bit faster but I never heard of the tweak you mentioned.
Do you have a reference?

This UltraATA tweak is from old BIOS Patcher auto-BIOS-modding tool (sorry for the stupid mistake in my previous post as I always use MODBIN before BIOS Patcher with Award BIOSes), but its site is long time ago abandoned and the tool itself is also unsupported and unupdated. I just thought, that as a pro-BIOS-modder you may heard of it and maybe you may take a look into its implementation for the tweak, that it applies (or claims to apply) for the faster IDE transfer modes (sorry, I can't help here, I have the tools to disassemble the exe-binary, but I'm assembly n00b and still barely read this kind of code).

The mobo with 430TX chipset which BIOS that I've patched with BIOS Patcher is Zida Tomato 5STX (works perfectly stable with the auto-patched BIOS). Here's the output from the DOS prompt (I always make log-files):

1BIOS Patcher ver. 6.00.alpha_15. |for Award/Phoenix & AMI bioses|
2Attention! Advanced qualification is required!
3
4
5Found 1Mbit Award BIOS!
6Remove "otkat" (backup key)
7
81. New CPU Support           : -> fixed.
92. P3-detect error           : is not needed to be fixed.
103. New Koeffs Support        : -> fixed.
114. 32Gb-problem              : -> fixed.
125. Some HDD detect-problem   : -> fixed.
136. "MB"/"GB" string search   : -> fixed.
147. 65Gb-problem (1-st step)  : -> fixed.
158. 65Gb-problem (2-nd step)  : -> fixed.
169. Error display Freq>999MHz : is not needed to be fixed.
1710.Error display Koefs>9.5x  : not found.
1811.New Stepping Support      : is not needed to be fixed.
1912.Tualatin L2-init error    : not found.
2013.New Freq in Setup open    : not found.
21
2214.Set "Y" as default on exit: -> fixed.
23
24
25Write Allocate addinng: -> fixed.
26UDMA66/100/133 on UDMA33_only_MB patch: -> fixed.
27
28Tweak options addinng:
29
30CBROM V2.08 (C)Award Software 2000 All Rights Reserved.
31[Other] ROM is releaseCBROM V2.08 (C)Award Software 2000 All Rights Reserved.
32Adding modul.tmp  43.6%
33CBROM V2.08 (C)Award Software 2000 All Rights Reserved.
34Adding start.tmp  87.5%
35CBROM V2.08 (C)Award Software 2000 All Rights Reserved.
36Adding tweak.tmp  36.0%
37
38if you can`t see all messages - choose 80x50 mode or run with ">report.txt".
39(c)2002-2004 apple_rom, www.ROM.by
40(c)2006 Angel07, www.cgi-scripts.info

You see the UDMA66/100/133 on UDMA33_only_MB patch: -> fixed. line - its some default tweak/fix for the tool.

It's not a problem for me to reflash the last original Zida BIOS (I think it's ver. 1.30 and I have it) and measure the ATA speeds if I know the suitable tool for DOS or Win9x/Me that will do the job (I'm open for suggestions as I don't remember/know any).

Please, PM me, if I've to explain something more in details, so we don't use this thread as a "chat".

Chkcpu wrote on 2023-12-30, 16:54:

About the disappointing read speed you see in Speedsys, this may be because you run this check from DOS. I’m unsure if and how UDMA works in DOS but I believe you need a device driver for that.

Thanks, I'm sure you are correct. I'll do a benchmark under Windows with the same tool as you.

analog_programmer wrote on 2023-12-30, 21:58:

This UltraATA tweak is from old BIOS Patcher auto-BIOS-modding tool (sorry for the stupid mistake in my previous post as I alway […]
Show full quote

This UltraATA tweak is from old BIOS Patcher auto-BIOS-modding tool (sorry for the stupid mistake in my previous post as I always use MODBIN before BIOS Patcher with Award BIOSes), but its site is long time ago abandoned and the tool itself is also unsupported and unupdated. I just thought, that as a pro-BIOS-modder you may heard of it and maybe you may take a look into its implementation for the tweak, that it applies (or claims to apply) for the faster IDE transfer modes (sorry, I can't help here, I have the tools to disassemble the exe-binary, but I'm assembly n00b and still barely read this kind of code).

The mobo with 430TX chipset which BIOS that I've patched with BIOS Patcher is Zida Tomato 5STX (works perfectly stable with the auto-patched BIOS). Here's the output from the DOS prompt (I always make log-files):

Copy code to clipboard

1BIOS Patcher ver. 6.00.alpha_15. |for Award/Phoenix & AMI bioses|
2Attention! Advanced qualification is required!
3
4
5Found 1Mbit Award BIOS!
6Remove "otkat" (backup key)
7
81. New CPU Support           : -> fixed.
92. P3-detect error           : is not needed to be fixed.
103. New Koeffs Support        : -> fixed.
114. 32Gb-problem              : -> fixed.
125. Some HDD detect-problem   : -> fixed.
136. "MB"/"GB" string search   : -> fixed.
147. 65Gb-problem (1-st step)  : -> fixed.
158. 65Gb-problem (2-nd step)  : -> fixed.
169. Error display Freq>999MHz : is not needed to be fixed.
1710.Error display Koefs>9.5x  : not found.
1811.New Stepping Support      : is not needed to be fixed.
1912.Tualatin L2-init error    : not found.
2013.New Freq in Setup open    : not found.
21
2214.Set "Y" as default on exit: -> fixed.
23
24
25Write Allocate addinng: -> fixed.
26UDMA66/100/133 on UDMA33_only_MB patch: -> fixed.
27
28Tweak options addinng:
29
30CBROM V2.08 (C)Award Software 2000 All Rights Reserved.
31[Other] ROM is releaseCBROM V2.08 (C)Award Software 2000 All Rights Reserved.
32Adding modul.tmp  43.6%
33CBROM V2.08 (C)Award Software 2000 All Rights Reserved.
34Adding start.tmp  87.5%
35CBROM V2.08 (C)Award Software 2000 All Rights Reserved.
36Adding tweak.tmp  36.0%
37
38if you can`t see all messages - choose 80x50 mode or run with ">report.txt".
39(c)2002-2004 apple_rom, www.ROM.by
40(c)2006 Angel07, www.cgi-scripts.info

You see the UDMA66/100/133 on UDMA33_only_MB patch: -> fixed. line - its some default tweak/fix for the tool.

Hi analog_programmer,
Thanks for explaining this “UltraATA tweak”, however I think there is a misunderstanding about this tweak. So I reply here in the thread for all to read.

Yes, I’m familiar with this nice BIOS patcher tool from “apple_rom”. Although my BIOS patching method is quite different, the resulting fixes are largely the same.

The wording of the “UDMA66/100/133 on UDMA33_only_MB patch:” line of the BIOS patcher tool is open for interpretation and can be misread as meaning more than this fix is for.
It means that UDMA66/100/133 DRIVES can now be used on UDMA33_only motherboards, without being reverted to PIO4 mode. This is essentially the same Win98 UDMA bugfix I’ve just put in the HOT-569 BIOS and will be using for your 569A BIOS as well.
Of course, with this fix these UDMA66/100/133 drives will then correctly operate at UDMA33 speeds, not faster… 😉

I hope this clarifies the issue.

Very well! Yes, I have Shuttle HOT-569A variant and also I have XGecu programmer, so I'll gladly test the updated BIOS-mod version for this newest revision of the mobo when it's done.

When ready, I will send you the new patched 569A BIOS via PM for testing.

Happy New Year!
Jan

Hi, Jan! Thank you very much for that clarification. I see, in short, it's just a fix for HDD-compatibility and has nothing to do with UltraATA speeds.

I'll gladly test your newer patched BIOS version for Shuttle HOT-596A board when it's ready.

Happy new 2024th year!

P.S. Some stupid question: If your HDD volume compatibility BIOS patches are almost the same as the ones applied by the BIOS Patcher software, why BIOS Patcher applies its HDD-patches over already manually patched by you BIOSes? I've never used such HDD-volume double patched BIOSes, just tried what more will be autopatched (mostly for adding support for more CPUs) and noticed this strange BIOS Patcher behavior.

Chkcpu wrote on 2023-12-30, 16:54:

This is what I see under a fresh Win98 install on my 430TX board (Abit AB-PX5), without and with the same UDMA bugfix:

I tested both my CF card and my SSD with Atto Benchmark and I still find the results a bit low:

It seems there is a cap at around 25 MB/s. Not that it makes any significant difference when using the system though. The overall experience feels "normal" and of course better than with E-IDE drive.

Sphere478 wrote on 2023-12-17, 23:46:

Btw, some pmmx 200s had 3.5x unlocked. You can try 1.5x setting and see if you get 3.5x

I tried 1.5x and the CPU interpreted that as 2.5x.

Also I could not have a stable setup at 3x83. Windows refused to boot but DOS mode was Ok. I believe I need to work on raising my memory timings.

ChrisK wrote on 2023-12-19, 09:22:

A patch for the Chaintech 5SIM with SiS 5582 (and of course all the others) would be very much appreciated 😉
Already using your patched BIOS (huge THANKS for that!), but an additional UDMA bugfix would be really great.
Having to rely on HDDs with artificially limited UDMA-modes can be quite annoying (the way I got UDMA to work on this board at all).

Hi ChrisK,
The new CH-5SIM patch J.3 BIOS version with the UDMA bugfix is ready and I sent you a download link via PM.

Please let us know how this new BIOS works.

Regards, Jan

Hi Jan,

Great, thank you!
I'll test it as soon as possible and report back the results, though it may take a few days.