That's the same error and freeze that I got on a Micronics 440LX board when I tried to run it with a Mendocino Celeron which the BIOS didn't support. I've never tried doing a DIY patch to support later CPUs though, so I can't comment on that procedure.
It looks like you're 1 generation ahead of the problem I had but still the same problem. All 3 of your CPUs are based on the Coppermine core, and apparently neither of your boards has BIOS support for those.

You wrote that one of your P3s is 833MHz but I assume you meant 866MHz. Those run on a 133MHz bus.
The 1GHz might be either 100MHz frontside bus or 133FSB. The latter is more common.

If the Biostar M6TBD is an OEM board from a brand name PC, then it probably won't have any option to attempt running at 133FSB. 100FSB is the official limit for the 440BX chipset. It might be willing to boot those CPUs at 100FSB though, with the corresponding loss of CPU speed (they'll run at 3/4 speed). There's an outside chance it might refuse to boot with them at all, but I've only seen that level of obstinance with Intel boards.
At minimum, the Biostar M6TBD ought to at least boot willingly with the Celeron 566MHz once it has support in the BIOS for Coppermines. Most likely it would also boot the P3s, just not at full speed.
The 440LX board is less certain but it ought to be able to run the Celeron as long as it supports the CPU voltage (which apparently it does) and it has BIOS support (which you're trying to add). There were some 440LX boards that ran those CPUs, though I don't think it was common.

Have you found any official BIOS updates for the M6TBD? I'm a little surprised if it never added Coppermine support. Most 440LX boards probably never had support for Coppermines but most 440BX boards had it added eventually. If you find a newer official BIOS then give it a try. Sometimes when OEM boards add support for newer CPUs they don't mention it publicly in the changelog.

Hi shamino, thank you very much for your input.

shamino wrote:

That's the same error and freeze that I got on a Micronics 440LX board when I tried to run it with a Mendocino Celeron which the BIOS didn't support. I've never tried doing a DIY patch to support later CPUs though, so I can't comment on that procedure.
It looks like you're 1 generation ahead of the problem I had but still the same problem. All 3 of your CPUs are based on the Coppermine core, and apparently neither of your boards has BIOS support for those.

I managed to score 2 Deschutes P2s ( a 350 and 400) locally and was finally able to boot both boards (M6TBD with the 440BX and the 440LX based HOT637 v3.1).

I did find BIOS updates for both boards
Neither board was flashable with whatever version of awdflash a tried, including the one bundled with the updates. Flashrom did not want to work either (Linux or DOS) on these boards (it works well on newer boards, like the the CUV4X). Uniflash worked perfectly with the HOT637 and let the Celeron @ 566MHz . I would not want to run it there very long with that slocket (no regulator) as the integrated VRM only goes down to 1.8V .

On the M6TBD, the board was bricked after the update. I was able to bring it back with a hotflash using my spare M6TBD. After unsuccessfully trying multiple BIOSes for various Biostar M6TB family boards, I decided to double-check the copy of the new BIOS I was using and, sure enough, it was corrupt (must have happened accidentally while I was checking the included microcodes in it)! I re-extracted the file from the archive, reflashed with Uniflash and was greeted with the POST screen. This allowed the Celeron 566 to boot successfully albeit detected as a P3 @ 556MHz. As a final step (after having googled for an easy to add CPU microcode to a BIOS), I ran the new BIOS through Bios Patcher version 4.23 and had the Celeron detected properly as Coppermine Celeron @ 566 MHz!

It was my intention to run the M6TBD with a P3 1GHZ @ 750 MHz at its rated 1.75V (the VRM on the M6TBD looks like it can go down to 1.5V according to its datasheet) for the time being and eventually replace it with 100MHz FSB P3 @1000MHz or a modded Tualatin or Tualeron (if/when I can get one cheap locally). Now, I looks like it will probably work (bedtime soon, so will try it tomorrow).

shamino wrote:

You wrote that one of your P3s is 833MHz but I assume you meant 866MHz. Those run on a 133MHz bus.
The 1GHz might be either 100MHz frontside bus or 133FSB. The latter is more common.

Indeed, it is an 800MHz 133Mhz FSB CPU

I have been busier than expected, so tomorrow turned into next week.
That being said, I am happy to report that my socket 370 SL52R Pentium 3 @ 1GHz seems to boot just fine in my M6TBD @ 750Mhz (100MHz FSB) . I have the Asus slocket jumpered for 1.75v and I dare hope that is indeed what the board's SC1164CS VRM is feeding it (Vcore is not displayed in the BIOS).

Is there any DOS or Linux application that would allow me to confirm that as I would rather avoid the risk of overvolting the CPU for the duration of a Windows (98SE) installation ?

Alternatively, does anyone know whether Asus slockets have Vcore testpoint where I could measure directly ?

Glad to hear it runs, that's excellent news.
For long term monitoring of sensors I normally use Speedfan, but my experience with it has been on newer systems. I don't know how likely it is to support the sensors on a board from the Slot-1 era. If it does work then it can be set to sound an alarm or run arbitrary commands under whatever conditions you specify, although the setup can be confusing. I haven't used it on Win98 but I'm pretty sure at least some versions support it.

In your case though I wouldn't trust software for the initial confirmation, especially for voltage. Instead check the voltage with a multimeter to really be sure. The voltage sensors on motherboards are sometimes very inaccurate so you can't really trust them without at least verifying them once.

The easiest way to check Vcore voltage is usually on the MOSFETs in the CPU VRM area. If you put the negative probe against the PSU and the positive on the large tab on the back of a MOSFET in that area, on half of those you'll see the input voltage (+5v) and on the other half of them you'll see the regulated Vcore output. I think it's expected to sag very slightly beyond that point so it will read on the slightly high side.
Alternatively you can measure on an inductor (they come immediately after the MOSFETs in the circuit) or on the positive leg of a Vcore capacitor (which is even closer to the CPU in circuit), but that's usually not as convenient and there's more risk of accidentally shorting something. If your slocket has it's own capacitors then they might be more accessible though. If you're worried about shorting anything then wrap the probe with tape.

With a slocket that has it's own VID jumpers you can probably even test how a board reacts to a setting without a CPU actually being installed. The board should still see the VID from the slocket and set the Vcore.

Since the CPU is significantly underclocked you could even get away with undervolting it if you want. That would reduce the heat / power usage, which grows in proportion to the square of voltage.

With a slocket that has it's own VID jumpers you can probably even test how a board reacts to a setting without a CPU actually being installed. The board should still see the VID from the slocket and set the Vcore.

That is great idea, shamino ! It reminded me that I actually that I also have an older 370SP rev 1.0 slocket without Coppermine support that has VID jumpers and that could be used to run tests without having a CPU and cooler installed (per your suggestion). As a bonus, that slocket has clearly a identified and easily accessible Vcore test point (the Asus one does not) ! This ended up being even easier than probing the MOSFETs, inductors or the Vcore cap .

The result ? According to my multimeter, the board responds properly to VIDs below 1.8V and can safely power a Coppermine CPU as least as low as 1.75 volts .

Thank you very much one again for your input and suggestions, shamino,

So If anybody wants a slot 1 440BX board with two ISA slots, onboard fast ethernet and unofficial Coppermine compatibility which includes proper voltage support(BIOS update and patch required), the Biostar OEM M6TBD is a possibility. They are available rather cheap as new old stock from at least one seller (probably much cheaper than a new SE440BX-2 if you combine shipping on multiple units) on a certain online auction site (as mentioned in another thread). The only caveat so fear is the exclusive use of Choyo brand electrolytic caps on the board, so an eventual recapping will likely be required (maybe not for a long time if it is kept cool and low ripple power is applied).

As a late follow-up, I can know confirm that my M6TBD with Bios Patcher 4.23-ed BIOS combined with a PowerLeap PL-iP3/T ver. 2.0 (the one with voltage jumpers, which I set to 1.5V as per CPU rating) and Celeron 1300 (Tualatin) seems to work fine .

The board POSTs and correctly shows a Tualatin Celeron as being present, albeit at 1290 MHz (CPU-Z shows 1297 MHz) .

Prime95 has been running for over an hour now in burn-in mode (max heat option chosen) without issue and the Startech Fan370pro ( https://www.startech.com/ca/Computer-Parts/Fa … m-fan~FAN370PRO ) I have been using barely gets warm to the touch during the test .

I finally have a Tualatin with 2 ISA slots . Me happy! 😀

darry wrote:

Hi,

I have been lurking here for years and have learned a lot along the way from all the knowledgeable folks. A great thank you to everyone!

Welcome aboard! 😁

My questions are the following : If I were to patch the BIOS with updated CPU microcode that includes these CPU types, is there […]
Show full quote

My questions are the following :
If I were to patch the BIOS with updated CPU microcode that includes these CPU types, is there a fairly good chance that I might boot one of these boards successfully with said CPUs (at stock clock or underclocked) ?
In other words has someone actually "fixed" a similar issue on an Award BIOS based board with a DIY updated microcode patch ?

I know I am not dealing in certainties here, but I would like to hear about some similar experiences before I spend time/money on older CPUs , an EPROM flasher (or hot-flashing in another board).

P.S. I also have a working CUV4X socket 370 board but it only has 1 ISA slot , whereas the older slot1 boards have 2 or more , which I could definitely use.

Thanks in advance for any comments/suggestions .

Do you happen to know a good resource for adding microcode to BIOS files? Something that would be good for a beginner? It sounds fun to tinker with 😜

Tetrium wrote:

Welcome aboard! 😁

Thank you, happy to be here!

darry wrote:

Do you happen to know a good resource for adding microcode to BIOS files? Something that would be good for a beginner? It sounds fun to tinker with 😜

For legacy hardware, BIOS Patcher makes it almost too easy.
Socket 370 and older work well with 4.23 , more recent hardware is somewhat supported in more recent versions.
http://www.rom.by/articles/BP/index_english.htm
http://www.majorgeeks.com/files/details/bios_patcher.html

For a more hands-on approach, reading through the following gives good basic info .
http://home.arcor.de/frogge/pepper/bmreport1.html

For more modern patching endeavors, like adding Xeon socket 771 microcodes to a socket 775 board (my trusty P5K WIFI Deluxe AP loves that Xeon E5450):
http://www.delidded.com/lga-771-xeon-microcode/

I believe BIOS Patcher works well with Award BIOS versions. Does it work equally well with AMI and Phoenix?

gdjacobs wrote:

I believe BIOS Patcher works well with Award BIOS versions. Does it work equally well with AMI and Phoenix?

Award only, unfortunately.

Ah, okay. Yeah, that's not so great.

darry wrote:

Thank you, happy to be here! […]
Show full quote

Tetrium wrote:

Welcome aboard! 😁

Thank you, happy to be here!

darry wrote:

Do you happen to know a good resource for adding microcode to BIOS files? Something that would be good for a beginner? It sounds fun to tinker with 😜

For legacy hardware, BIOS Patcher makes it almost too easy.
Socket 370 and older work well with 4.23 , more recent hardware is somewhat supported in more recent versions.
http://www.rom.by/articles/BP/index_english.htm
http://www.majorgeeks.com/files/details/bios_patcher.html

For a more hands-on approach, reading through the following gives good basic info .
http://home.arcor.de/frogge/pepper/bmreport1.html

For more modern patching endeavors, like adding Xeon socket 771 microcodes to a socket 775 board (my trusty P5K WIFI Deluxe AP loves that Xeon E5450):
http://www.delidded.com/lga-771-xeon-microcode/

Cheers! 😁

The arcor site seems to be down?

Tetrium wrote:

The arcor site seems to be down?

It literally just disappeared .
http://web.archive.org/web/20161213110830/htt … /bmreport1.html