0xDEADBEEF wrote on 2025-04-12, 22:06:

Sweet! I have seen “CPU is unworkable or has been changed.” message too, happens when I switch to different CPU. I think there […]
Show full quote

Sweet!
I have seen “CPU is unworkable or has been changed.” message too, happens when I switch to different CPU. I think there is some "check if user is doing something too stupid" logic in BIOS that prints it 😉
I am excited to hear you are working on it.
This board is pretty cool, 4 pci + 4 isa, built in usb and ATX form factor that makes it compatible with modern ATX cases.

Hi 0xDEADBEEF,

I made some progress with the patched Abit AX5/PX5 BIOS.
I’ve send you a PM with details about the changes I’ve made sofar, and a copy of a patched AB-AX5/PX5 interim BIOS for you to test.

When you are able to test this BIOS, please send a reply about your findings via PM or here in the thread.

Happy testing,
Jan

I recently ended up with an unusual Socket 7 motherboard that would benefit from a modified BIOS. I did give it my best shot with the BIOS Patcher, but that wasn't able to take care of the issues. It's a Chaintech 5HTM0 M10x. Intel HX chipset, 200 MMX is the maximum supported CPU. No bus speeds beyond the official 66 MHz, but the core voltage does go down as low as 2.5 volts; I just happen to have a 2.4 volt K6-3 that could work nicely were BIOS support added. The last official BIOS dates from 1997, and its age shows. No matter what I tried, I haven't been able to get it to shut down properly with an ATX power supply. My hope was to use the BIOS Patcher to cure the shut down issue and to add support for the K6-3, but it wasn't able to solve either issue. APM still has the exclamation point, and it wouldn't POST with the K6-3. Any help would be greatly appreciated.

Repo Man11 wrote on 2025-06-04, 20:44:

I recently ended up with an unusual Socket 7 motherboard that would benefit from a modified BIOS. I did give it my best shot with the BIOS Patcher, but that wasn't able to take care of the issues. It's a Chaintech 5HTM0 M10x. Intel HX chipset, 200 MMX is the maximum supported CPU. No bus speeds beyond the official 66 MHz, but the core voltage does go down as low as 2.5 volts; I just happen to have a 2.4 volt K6-3 that could work nicely were BIOS support added. The last official BIOS dates from 1997, and its age shows. No matter what I tried, I haven't been able to get it to shut down properly with an ATX power supply. My hope was to use the BIOS Patcher to cure the shut down issue and to add support for the K6-3, but it wasn't able to solve either issue. APM still has the exclamation point, and it wouldn't POST with the K6-3. Any help would be greatly appreciated.

Hello, looking at the photo, it seems to me that it has linear regulators, so I would try to start the K6-3 at reduced frequency, for example 200 MHz, but this is only possible if it has jumpers for manual selection, if instead it has automatic recognition of the CPU, it should start at reduced frequency, but I do not know if the failure to start depends on the failure of the CPU.

PC@LIVE wrote on 2025-06-04, 20:52:

Repo Man11 wrote on 2025-06-04, 20:44:

I recently ended up with an unusual Socket 7 motherboard that would benefit from a modified BIOS. I did give it my best shot with the BIOS Patcher, but that wasn't able to take care of the issues. It's a Chaintech 5HTM0 M10x. Intel HX chipset, 200 MMX is the maximum supported CPU. No bus speeds beyond the official 66 MHz, but the core voltage does go down as low as 2.5 volts; I just happen to have a 2.4 volt K6-3 that could work nicely were BIOS support added. The last official BIOS dates from 1997, and its age shows. No matter what I tried, I haven't been able to get it to shut down properly with an ATX power supply. My hope was to use the BIOS Patcher to cure the shut down issue and to add support for the K6-3, but it wasn't able to solve either issue. APM still has the exclamation point, and it wouldn't POST with the K6-3. Any help would be greatly appreciated.

Hello, looking at the photo, it seems to me that it has linear regulators, so I would try to start the K6-3 at reduced frequency, for example 200 MHz, but this is only possible if it has jumpers for manual selection, if instead it has automatic recognition of the CPU, it should start at reduced frequency, but I do not know if the failure to start depends on the failure of the CPU.

I'm fairly sure it is a BIOS recognition problem - some Socket 7 boards that can work with K6-2 K6-3 and their Plus variants will not POST with these CPUs with the original BIOS. An Evergreen Spectra BIOS would likely solve the issue, but if they made on for this board I wasn't able to find it. I tried reducing the bus speed and the multiplier, but it refused to POST.

Repo Man11 wrote on 2025-06-04, 21:07:

PC@LIVE wrote on 2025-06-04, 20:52:

Repo Man11 wrote on 2025-06-04, 20:44:

I recently ended up with an unusual Socket 7 motherboard that would benefit from a modified BIOS. I did give it my best shot with the BIOS Patcher, but that wasn't able to take care of the issues. It's a Chaintech 5HTM0 M10x. Intel HX chipset, 200 MMX is the maximum supported CPU. No bus speeds beyond the official 66 MHz, but the core voltage does go down as low as 2.5 volts; I just happen to have a 2.4 volt K6-3 that could work nicely were BIOS support added. The last official BIOS dates from 1997, and its age shows. No matter what I tried, I haven't been able to get it to shut down properly with an ATX power supply. My hope was to use the BIOS Patcher to cure the shut down issue and to add support for the K6-3, but it wasn't able to solve either issue. APM still has the exclamation point, and it wouldn't POST with the K6-3. Any help would be greatly appreciated.

Hello, looking at the photo, it seems to me that it has linear regulators, so I would try to start the K6-3 at reduced frequency, for example 200 MHz, but this is only possible if it has jumpers for manual selection, if instead it has automatic recognition of the CPU, it should start at reduced frequency, but I do not know if the failure to start depends on the failure of the CPU.

I'm fairly sure it is a BIOS recognition problem - some Socket 7 boards that can work with K6-2 K6-3 and their Plus variants will not POST with these CPUs with the original BIOS. An Evergreen Spectra BIOS would likely solve the issue, but if they made on for this board I wasn't able to find it. I tried reducing the bus speed and the multiplier, but it refused to POST.

Hi Repo Man11,

Here is the Evergreen Spectra upgrade BIOS for the Chaintech 5HTM you were looking for. 😀

2A59FC3A.zip (removed due to bug)

This 03/2000 upgrade Award BIOS by Unicore Software has all the usual goodies like K6-2(+)-III(+) and 128GiB HDD support, and should also fix your APM issue.

Let us know how this BIOS works on your board.
Cheers, Jan

Chkcpu wrote on 2025-06-05, 10:57:

Hi Repo Man11, […]
Show full quote

Repo Man11 wrote on 2025-06-04, 21:07:

PC@LIVE wrote on 2025-06-04, 20:52:

Hello, looking at the photo, it seems to me that it has linear regulators, so I would try to start the K6-3 at reduced frequency, for example 200 MHz, but this is only possible if it has jumpers for manual selection, if instead it has automatic recognition of the CPU, it should start at reduced frequency, but I do not know if the failure to start depends on the failure of the CPU.

I'm fairly sure it is a BIOS recognition problem - some Socket 7 boards that can work with K6-2 K6-3 and their Plus variants will not POST with these CPUs with the original BIOS. An Evergreen Spectra BIOS would likely solve the issue, but if they made on for this board I wasn't able to find it. I tried reducing the bus speed and the multiplier, but it refused to POST.

Hi Repo Man11,

Here is the Evergreen Spectra upgrade BIOS for the Chaintech 5HTM you were looking for. 😀

The attachment 2A59FC3A.zip is no longer available

This 03/2000 upgrade Award BIOS by Unicore Software has all the usual goodies like K6-2(+)-III(+) and 128GiB HDD support, and should also fix your APM issue.

Let us know how this BIOS works on your board.
Cheers, Jan

Thank you for finding that for me, it was a good way to start my day! Sadly, this BIOS doesn't seem to be their best work. I'm using a sixty gig IDE drive, and I had it attached to a Promise ATA 100 controller card to get around the size limitation, and to pick up some speed. After flashing the new BIOS, the controller card suddenly couldn't detect the drive attached to it? I tried another PCI slot, then it didn't even display the card's BIOS screen. On the POST screen, for both the VGA and the IDE card the IRQ was listed as "NA."

I pulled the card and attached the cable to the board's IDE port, but it would only load Windows in Safe Mode. So I decided to do a fresh install, and that went okay, but for two things: the exclamation point for the APM (and the shutdown issue) came back, and when I tried to load the driver for the NVS280 video card (45.23 force installed as an FX5200) it wouldn't work because of the IRQ issue (To be clear, this is with just the video card installed, no other PCI or ISA slots are being used). This is all despite my disabling all unnecessary ports in the CMOS settings, trying different PCI slots, and experimenting with the APM settings.

The good news is that it will POST and display the correct speed with the K6-3.

I wonder if the Evergreen BIOS is for the 5HTM1 rather than the 5HTM0? They are very similar, but Chaintech offered separate BIOS files.

Repo Man11 wrote on 2025-06-05, 20:09:

Thank you for finding that for me, it was a good way to start my day! Sadly, this BIOS doesn't seem to be their best work. I'm u […]
Show full quote

Chkcpu wrote on 2025-06-05, 10:57:

Hi Repo Man11, […]
Show full quote

Repo Man11 wrote on 2025-06-04, 21:07:

I'm fairly sure it is a BIOS recognition problem - some Socket 7 boards that can work with K6-2 K6-3 and their Plus variants will not POST with these CPUs with the original BIOS. An Evergreen Spectra BIOS would likely solve the issue, but if they made on for this board I wasn't able to find it. I tried reducing the bus speed and the multiplier, but it refused to POST.

Hi Repo Man11,

Here is the Evergreen Spectra upgrade BIOS for the Chaintech 5HTM you were looking for. 😀

2A59FC3A.zip (removed due to bug)

This 03/2000 upgrade Award BIOS by Unicore Software has all the usual goodies like K6-2(+)-III(+) and 128GiB HDD support, and should also fix your APM issue.

Let us know how this BIOS works on your board.
Cheers, Jan

Thank you for finding that for me, it was a good way to start my day! Sadly, this BIOS doesn't seem to be their best work. I'm using a sixty gig IDE drive, and I had it attached to a Promise ATA 100 controller card to get around the size limitation, and to pick up some speed. After flashing the new BIOS, the controller card suddenly couldn't detect the drive attached to it? I tried another PCI slot, then it didn't even display the card's BIOS screen. On the POST screen, for both the VGA and the IDE card the IRQ was listed as "NA."

I pulled the card and attached the cable to the board's IDE port, but it would only load Windows in Safe Mode. So I decided to do a fresh install, and that went okay, but for two things: the exclamation point for the APM (and the shutdown issue) came back, and when I tried to load the driver for the NVS280 video card (45.23 force installed as an FX5200) it wouldn't work because of the IRQ issue (To be clear, this is with just the video card installed, no other PCI or ISA slots are being used). This is all despite my disabling all unnecessary ports in the CMOS settings, trying different PCI slots, and experimenting with the APM settings.

The good news is that it will POST and display the correct speed with the K6-3.

I wonder if the Evergreen BIOS is for the 5HTM1 rather than the 5HTM0? They are very similar, but Chaintech offered separate BIOS files.

Although this Unicore BIOS is clearly indicated for the Chaintech 5HTM Rev M10x, which is the 5HTM0 (the Rev M20x is the 5HTM1), you are right about this one being less than stellar.

I found that they forgot to adapt the PCI Routing table to the board’s layout, effectively breaking the PnP logic! This will prevent IRQ assignments to PCI devices installed in any of the 4 PCI slots.

I’ve fixed the Unicore BIOS by making the PCI Routing table identical to the table in the original Ver 0A14 BIOS. I’m confident that this will fix your issues with the 2A59FC3A Unicore BIOS.

Here is the fixed 5HTM0_J1 BIOS.

Cheers, Jan

Wow, thank you very much for fixing that Jan! With your fixed Evergreen BIOS, everything is working fine! The only exception is the that the APM in System Devices still ends up with an exclamation point, and no matter what I try, it still shuts down as though is has an AT power supply. I've never encountered an issue like this before, and at this point I think it just how it is. I guess I should find an AT case that this board will work with.

Repo Man11 wrote on 2025-06-07, 03:01:

Wow, thank you very much for fixing that Jan! With your fixed Evergreen BIOS, everything is working fine! The only exception is the that the APM in System Devices still ends up with an exclamation point, and no matter what I try, it still shuts down as though is has an AT power supply. I've never encountered an issue like this before, and at this point I think it just how it is. I guess I should find an AT case that this board will work with.

Great to hear the PCI table fix works!
And your K6-III works as well, but keep an eye on the linear regulator temperature. These K6-III’s are quite power hungry and draw a lot of current. 😉

The APM should work as well with this BIOS. To ask the obvious, is Power Management not Disabled in the BIOS?

To let Win98 have full APM control with out the BIOS interfering, I usually use these BIOS settings in the POWER MANAGEMENT SETUP menu:

Power Management: User Define
PM Control by APM: Yes

Doze Mode: Disable
Standby Mode: Disable
Suspend Mode : Disable
HDD Power Down : Disable

Cheers, Jan

Chkcpu wrote on 2025-06-07, 17:58:

Great to hear the PCI table fix works! And your K6-III works as well, but keep an eye on the linear regulator temperature. These […]
Show full quote

Repo Man11 wrote on 2025-06-07, 03:01:

Wow, thank you very much for fixing that Jan! With your fixed Evergreen BIOS, everything is working fine! The only exception is the that the APM in System Devices still ends up with an exclamation point, and no matter what I try, it still shuts down as though is has an AT power supply. I've never encountered an issue like this before, and at this point I think it just how it is. I guess I should find an AT case that this board will work with.

Great to hear the PCI table fix works!
And your K6-III works as well, but keep an eye on the linear regulator temperature. These K6-III’s are quite power hungry and draw a lot of current. 😉

The APM should work as well with this BIOS. To ask the obvious, is Power Management not Disabled in the BIOS?

To let Win98 have full APM control with out the BIOS interfering, I usually use these BIOS settings in the POWER MANAGEMENT SETUP menu:

Power Management: User Define
PM Control by APM: Yes

Doze Mode: Disable
Standby Mode: Disable
Suspend Mode : Disable
HDD Power Down : Disable

Cheers, Jan

I've reinstalled Windows 98 over a dozen times with three different BIOS, different hard drives, and I've tried every variation of Power Management settings in CMOS while also clearing CMOS after BIOS flashes, setting it to defaults, but it still ends up with the exclamation point and powers down as though it's connected to an AT power supply rather than an ATX. I've never had this issue with any motherboard before, and at this point I'm assuming that it is due to some hardware quirk of this particular motherboard or the chipset revision or something.

While I did try out the K6-3, after having it on for a short time the regulator heatsinks were too hot to touch! My thermometer showed them hitting about 57 degrees C, while only the low 40s with the 166 CPU, so I swapped in my Spectra 400 that has a K6-3+ - fitting since it has a Spectra BIOS.

Another interesting quirk - the pins for the USB port are in a nine pin configuration, so I assumed that a standard nine pin USB back plate would work; but it turned out that the ninth pin is on the opposite side on this motherboard! Once I saw that, I measured to make sure that the +5 pins were where they were supposed to be, then switched to a dual five pin USB backplate. The USB ports work fine.

Thanks again for both finding and fixing the Evergreen BIOS!

Hi, Jan!

I tried something similar to this patch

Chkcpu wrote on 2021-11-24, 17:55:

#Part 3 – What to patch in an uncompressed Award BIOS […]
Show full quote

#Part 3 – What to patch in an uncompressed Award BIOS

After two episodes of preliminaries, it is time to start patching! 😀

1) Let’s start with the Year 2094 bug. A well known bug in 1994-1995 Award BIOSes.
Actually this bug and its fix were already addressed here recently by jakethompson1, but I will also show it here for completeness sake.
This bug causes the year to jump forward to 2094 at each boot-up whenever the date was set to 1-1-2000 or later.

At POST 0B, the BIOS checks if the RTC uses valid values by reading the CMOS registers for seconds, minutes, hours, day, month, year, and century. It then uses a short table with maximum and minimum values for each of these 7 registers.
The minimum values table contains the hex numbers: 00, 00, 00, 01, 01, 94, 19, and this is the binary signature you have to look for. If you don’t find this signature, your PC should be free of this bug.

The error is in the routine that walks this table, but the fix is setting the minimum year from 1994 to 2000 by changing the last 2 bytes in this table.
The new table should become: 00, 00, 00, 01, 01, 00, 20 to fix this bug.

Note that this silly bug appeared in mid 1994 when Award tried to make the BIOS millennium compliant. Before that, the BIOS didn’t do any century checking, and when Award fixed this bug by the end of 1995, they simply removed this century check again! 😉

with Award BIOS 4.50 for PCChips M912 motherboard modified by you (M912_J2.BIN) as it has Y2k bug (seen here).

I couldn't find the "00, 00, 00, 01, 01, 94, 19" hex values in this BIOS, but I found these "00, 00, 00, 01, 01, 00, 87" for lowest date values:

I changed "00, 00, 00, 01, 01, 00, 87" hex values to "00, 00, 00, 01, 01, 00, 19" (and corrected checksum bytes at the end of the BIOS-dump file):

I flashed the freshly modified BIOS and it works fine, but unfortunately this did not resolve the Y2k bug:

What can I do next, except to try changing those "00, 00, 00, 01, 01, 00, 87" hex values to "00, 00, 00, 01, 01, 00, 20"?

P.S. This BIOS works fine with HDD bigger than 504 MB (tested with 1.2 GB HDD), but I don't know what's the upper limit for HDD volume.

analog_programmer wrote on 2025-06-08, 09:11:

What can I do next, except to try changing those "00, 00, 00, 01, 01, 00, 87" hex values to "00, 00, 00, 01, 01, 00, 20"?

Ok, just tried this mod with changing the original "00, 00, 00, 01, 01, 00, 87" hex values to "00, 00, 00, 01, 01, 00, 20" and the result from SPEEDSYS report for Y2k-bug is same as changing "00, 00, 00, 01, 01, 00, 87" hex values to "00, 00, 00, 01, 01, 00, 19" - the bug is still there 🙁

P.S. I'm sorry for publishing this new comment, but some time ago the option to edit my old posts was taken away by some admin-script.

Hi analog_programmer,

Welcome to the league of BIOS patchers!

However, this Y2K bug in the M912_J2 BIOS is just a rollover bug from December 31, 1999 to January 1, 1900. We are well into the 21th century now, so this bug is of no concern and can be left alone.
Actually, the code in this 1995 BIOS will correct the date to the 21th century at the next boot-up, when it finds a date before 1980, which means only a date from 1/1/1980 to 12/31/2079 is valid in this BIOS. So we are good to go for another 55 years!! 😉

Note that the "00, 00, 00, 01, 01, 94, 19" hex values from the 1994 Award BIOS are the cause of another bug which is called the Year 2094 bug. This silly bug causes the year to jump forward to 2094 at each boot-up whenever the date was set to 1-1-2000 or later.
These 7 bytes are the minimum values for the RTC’s Seconds, Minutes, Hours, Day, Month, Year, and Century registers respectively.

The M912_J2 BIOS doesn’t suffer from this Y2094 bug and uses the "00, 00, 00, 01, 01, 00" hex values as minimum for the RTC. Yes only 6 values now, the century check was removed from this table and is done by hard-coded instructions in POST_B to fix the Y2094 bug.
So the “87” byte following this table isn’t any RTC value, but the first byte of a 2-byte instruction “87, DB”. This decodes as the xchg BX, BX instruction that interchanges the CPU’s BX register by itself. Not a very useful instruction, often used for short delays, but used here only as padding bytes.
So you didn’t break anything by changing this “87” value. 😉

About the maximum supported HDD drive size in this BIOS, it is 8GB.

Cheers, Jan

As always very detailed explanation. Thanks a lot, Jan!

Surely the Y2k-bug isn't relevant since a quarter of century ago, but it's annoying to see some software "saying" - hey, your BIOS is crippled 😀 I'll not try to fix the unfixable anymore. But it was a good experience to understand more about those calendar-related BIOS bugs.

And since we're still on the M912 Award BIOS topic. Do you have any plans for 3-rd patch attempt for proper Am5x86 CPU BIOS recognition?

analog_programmer wrote on 2025-06-08, 15:17:

And since we're still on the M912 Award BIOS topic. Do you have any plans for 3-rd patch attempt for proper Am5x86 CPU BIOS recognition?

Yes, I’m still planning to release an M912 patch J.3 BIOS, eventually.

Although the Am5x86-133 and Cx5x86-133 now run fine on the J.2 version, the incorrect speed indication on the Am5x86 still needs to be fixed.
Also the lower IDE transfers speed of this Award BIOS troubles me.
I still need to find solutions to these issue before releasing a definitive patched M912 Award BIOS.

Unfortunately, we don’t have the UMC498F chipset datasheet. So I recently started to reverse engineer the complicated AMI WinBIOS to see what it does differently from Award.
Thanks to the Golden tip from @TheMobRules earlier in this thread, I’m now able to decompress these 486 WinBIOSes. 😀

But Man, what a spaghetti code these AMI BIOSes use!! 🙁 Must have been programmed in a higher language and with all compiler optimizations enabled!
In contrast, the Award BIOS code is easy to follow and is clearly written in assembler, directly by humans. 😉

Jan

Chkcpu wrote on 2025-06-09, 12:08:

Yes, I’m still planning to release an M912 patch J.3 BIOS, eventually. […]
Show full quote

Yes, I’m still planning to release an M912 patch J.3 BIOS, eventually.

Although the Am5x86-133 and Cx5x86-133 now run fine on the J.2 version, the incorrect speed indication on the Am5x86 still needs to be fixed.
Also the lower IDE transfers speed of this Award BIOS troubles me.
I still need to find solutions to these issue before releasing a definitive patched M912 Award BIOS.

Unfortunately, we don’t have the UMC498F chipset datasheet. So I recently started to reverse engineer the complicated AMI WinBIOS to see what it does differently from Award.
Thanks to the Golden tip from @TheMobRules earlier in this thread, I’m now able to decompress these 486 WinBIOSes. 😀

But Man, what a spaghetti code these AMI BIOSes use!! 🙁 Must have been programmed in a higher language and with all compiler optimizations enabled!
In contrast, the Award BIOS code is easy to follow and is clearly written in assembler, directly by humans. 😉

Jan

Great! I can help with testing for your new M912 Award BIOS patch revision J3 on M912 ver. 1.4 motherboard with real cache chips + Am5x86-P75 (133 MHz). Just send me a PM, if you need testers 😉

I don't like these "fancy" AMI winBIOSes at all. Probably their graphical interface is related to your suggestion for the higher level programming language has been used for them instead of assembly.

Chkcpu wrote on 2021-11-25, 17:44:

Transplanting a BIOS from another board is not always successful, but with SiS471 boards this usually works fine, and I used this 11/28/95 BIOS successfully on several SiS471 boards, including my Chicony CH-471B. But I do adapt the default chipset register programming via Modbin to that of the original BIOS for each board, to maximize success.

Hey Jan

I'd like to adapt the Default Chipset Register using modbin - but I can only edit the Data - not the Indexes. Is there any way to completely clone both Indexes and Data? The bios I want to edit is the MiTAC/Trigon PWA-CAMARO 11/07/2001-VP4-686A-6A5LHM3CC-00 and I'd like to adapt the chipset registers to match those from your Tekram P5M4-M+ Mod from your K6plus page. I did cross flash that Bios to my P5M4-M+ and it does work - but with some limitations.

Anders

adegn wrote on 2025-06-19, 15:40:

Hey Jan […]
Show full quote

Chkcpu wrote on 2021-11-25, 17:44:

Transplanting a BIOS from another board is not always successful, but with SiS471 boards this usually works fine, and I used this 11/28/95 BIOS successfully on several SiS471 boards, including my Chicony CH-471B. But I do adapt the default chipset register programming via Modbin to that of the original BIOS for each board, to maximize success.

Hey Jan

I'd like to adapt the Default Chipset Register using modbin - but I can only edit the Data - not the Indexes. Is there any way to completely clone both Indexes and Data? The bios I want to edit is the MiTAC/Trigon PWA-CAMARO 11/07/2001-VP4-686A-6A5LHM3CC-00 and I'd like to adapt the chipset registers to match those from your Tekram P5M4-M+ Mod from your K6plus page. I did cross flash that Bios to my P5M4-M+ and it does work - but with some limitations.

Anders

Hi Anders,

Indeed, Award’s Modbin tool only allows you to change the contents of chipset registers that are pre-programmed in the BIOS, and not the register index itself. So you can’t add registers you like to change with Modbin.

From your other thread, I understand you are trying to fix a UDMA issue on your Tekram P5M4-M+ BIOS. I will reply there with my view on this.

Cheers, Jan