Kahenraz wrote on 2022-10-29, 06:28:

I can use Socket 7 CPUs in it, and it clearly states support for up to 233 on the board. It's definitely Socket 7.

Really? Coz I have this board in my main retro PC and max it can take is P200MMX (66x3). Unless there's some undocumented jumper positions that I'm not aware of. Didn't bother me since it's a DOS PC and going over 200mhz makes compatibility problems with some titles anyway.

Try multiplyer "x1.5" - it´s interpreted as 3.5 by the Pentium MMX.

In case it doesn´t work the "BF0" pin at the CPU socket needs to be pulled "high" with a 4K7 resistor.

After that the 233 speed might still be reported wrong by BIOS during POST, but benchmark tools will show the correct speed.

There is a sticker on the board that tells you to set the multipler to 1.5 for 233Mhz.

One of the great things about this board (besides having integrated opl3, wavetable opl4, s3) is support for 128GB drives with latest bios. TBH it's probably the most stable computer I ever had and it's 25+ years old. 😁

Is that with the official BIOS?

Kahenraz wrote on 2022-10-30, 11:50:

There is a sticker on the board that tells you to set the multipler to 1.5 for 233Mhz.

I don't have the sticker nor manual mentiones it, so thanks for the tip, although I doubt I will ever go 233 route, it is nice to know it actually supports it.

The sticker is visible on my board at the bottom right. See the photo in the first post of this thread.

Kahenraz wrote on 2022-10-30, 12:02:

Is that with the official BIOS?

Yes, I think it's latest official, 1.00.07.DH0
Showing wrong HDD values in BIOS but works like a charm.

EDIT: I see now in your pics that you have the same BIOS version already.

Other great thing about this board, during BIOS post you can press key combo for turbo switch with "CTRL ALT -" which takes it to "fast 386 DX40" levels that is good enough for playing Wing Commander (and possibly other speed sensitive games). 😀

I'll have to try that. What CPU would pair well with the turbo?

I've tested it with p166mmx and p200mmx and they both went to 386dx40 speeds, maybe a little faster...
One can also achieve that with SETMUL, but this seems more convenient.

This motherboard is also very robust, provided that you won't pair it with garbage tier PSU.

Kahenraz wrote on 2022-10-29, 01:30:

I managed to fix the overbright issue by flashing another BIOS using shadow ROM with svbl. […]
Show full quote

I managed to fix the overbright issue by flashing another BIOS using shadow ROM with svbl.

Luckily, this tool specifically supports the Triton chipset (Intel 430HX). Here is the command I used:

svbl.com /c:triton.cps /b:86C375_5.bin

Neither s3bright.com nor s3black.com worked with this onboard card. I know that these utilities work with other cards, since I have tried them in the past with my S3 Trio 3D/2X.

I've attached everything needed to get this working, in case anyone else finds themselves with this same problem in the future.

Here is a link to the other relevant thread:

S3 AGP Cards (and possibly others) Too Bright

This is where to obtain s3bright (which doesn't work with this card):

Re: S3 AGP Cards (and possibly others) Too Bright

And s3vbefix (which also doesn't work with this card):

S3VBEFIX - PC\MS-DOS TSR fix for S3 VESA 2.0 video BIOS

I also tried using DEBUG to modify the BLANK pedestal, as described by wbc, which also does not work with this card:

Re: S3 AGP Cards (and possibly others) Too Bright

Copy code to clipboard

1o 3c4 08
2o 3c5 06
3o 3c4 27
4i 3c5
5<at this step subtract 08h from given value>
6o 3c5 <resulting_value>
7q

I found a working BIOS here, and tied 86C375_1.bin, 86C375_2.bin, 86C375_3.bin, 86C375_4.bin, and 86C375_5.bin. 1, 3, and 4 all had the overbright problem. There was some error with 2 that made it fail to load. The only one to produced the correct black levels was 86C375_5.bin.

http://chukaev.ru54.com/video_en.htm

Both S3BLACK.COM and S3BRIGHT.COM won't work with S3 chips prior the ViRGE/GX2 as they have the BLANK pedestal located on a different register (wbc mentioned this when posting that solution). The correct commands for DEBUG should be the following:

1o 3c4 08
2o 3c5 06
3o 3c4 1a
4i 3c5
5<subtract 20h from given value>
6o 3c5 <resulting_value>
7q

Guys, is it possible to use 50ns EDO with this board or do only 60ns and 70ns modules work? Would like to know before ordering, have none 50ns modules to try it myself.

Jura Tastatura wrote on 2022-11-02, 07:41:

Guys, is it possible to use 50ns EDO with this board or only 60ns and 70ns modules work? Would like to know before ordering, have none 50ns modules to try myself.

With the Mr.BIOS there are 55/60/70ns settings as I remember

50 and 40ns tend to be 3.3v be ware

Sphere478 wrote on 2022-11-02, 22:46:

50 and 40ns tend to be 3.3v be ware

Thanks for clarification! Not going to mess with it then, wouldn't want to burn my favourite board. 😀

Even if you would use 40 or 50nS RAM, the BIOS probably wouldn´t be able to set the latencys/wait states etc. accordingly and treat the memory as 60nS.

I have to say this thread is an absolute goldmine of information, thank you very much.
I also have this motherboard - with the S3 chipset as well as the general midi. I bought it as a whole PC, used, with a 133 CPU in it in a really yellowed case for next to nothing and am now rebuilding it.
Seems like a super base for a retro DOS machine. Might swap out the CPU for a 233...