Are the jumpers set right for a 5x86 @ 4 x 33? I can't read the small print on the photo to know.

It has jumpers for setting up the FSB and CPU type, no jumpers for multiplier.

aaronkatrini wrote on 2023-04-16, 21:09:

Hi all, […]
Show full quote

Hi all,

I'm looking to find a newer BIOS for my 486 motherboard so that it can properly detect AMD 5x86 133MHz CPUs (currently recognizes them as DX4-100).
From what I could find on the internet it's a Epox one. This was recognized by the ID string : 08/15/95-SIS-85C471B/E/G-2C4I9PA0-00 (AWARD BIOS).
Also this board is very similar : https://theretroweb.com/motherboards/s/epox-pronix-gxa486sg.
However that board has 2x 72p + 4x 30p Ram slots, but mine has 4x 72p. (both boards share the same PCB, just populated differently). This might just be a different revision.
It's a very nice board and I'm quite pleased with it. It has all the bells and whistles that I was looking for on a 486 Mainboard such as VLB slots, no leaky battery, support for faster 5x86 CPUs (according to the labels on the PCB). For reference the board is working fine with a Cyrix 5x86-100GP. Also I'm looking forward for a new blank BIOS chip to arrive so I could also try the BIOS on the TRW. So please, if you have a more recent BIOS, let me know 😀 Thanks!

Hello,

I agree with the other answer.
Without a jumper setting affecting clkmul, you neither can have the multiplier 2 on a DX4 or the multiplier 4 on a Am5x86.
There is nothing a BIOS can do, it's a hardware setting. OK, when the clkmul pin is set correctly, THEN the Bios is needed to correctly display Am5x86 for example.

Re: Automatic multiplier settings on 486

CoffeeOne wrote on 2023-04-16, 22:34:

Hello, […]
Show full quote

aaronkatrini wrote on 2023-04-16, 21:09:

Hi all, […]
Show full quote

Hi all,

I'm looking to find a newer BIOS for my 486 motherboard so that it can properly detect AMD 5x86 133MHz CPUs (currently recognizes them as DX4-100).
From what I could find on the internet it's a Epox one. This was recognized by the ID string : 08/15/95-SIS-85C471B/E/G-2C4I9PA0-00 (AWARD BIOS).
Also this board is very similar : https://theretroweb.com/motherboards/s/epox-pronix-gxa486sg.
However that board has 2x 72p + 4x 30p Ram slots, but mine has 4x 72p. (both boards share the same PCB, just populated differently). This might just be a different revision.
It's a very nice board and I'm quite pleased with it. It has all the bells and whistles that I was looking for on a 486 Mainboard such as VLB slots, no leaky battery, support for faster 5x86 CPUs (according to the labels on the PCB). For reference the board is working fine with a Cyrix 5x86-100GP. Also I'm looking forward for a new blank BIOS chip to arrive so I could also try the BIOS on the TRW. So please, if you have a more recent BIOS, let me know 😀 Thanks!

Hello,

I agree with the other answer.
Without a jumper setting affecting clkmul, you neither can have the multiplier 2 on a DX4 or the multiplier 4 on a Am5x86.
There is nothing a BIOS can do, it's a hardware setting. OK, when the clkmul pin is set correctly, THEN the Bios is needed to correctly display Am5x86 for example.

Re: Automatic multiplier settings on 486

Thanks for the idea 😀

There are some jumpers on the board I couldn't properly identify / make out what they really controll. Will look into it carefully and see if one of them has continuity for that line. Will update with findings.

Yeah most all BIOS found with SIS-85C471B/E/G string are about same date stamp or earlier. A few was dated about same or within weeks of but side notes said release was 1/12/1996.
https://theretroweb.com/bios/?postString=SIS-85C471B/E/G
So most likely none of those would report any difference.....

As alluded above have you identified whether the board misreporting your CPU as a DX4-100 is merely a cosmetic issue or actually affects anything?
So long as these boards don't stop you from booting, the issue is if they autoconfigure your memory & cache wait states, they need to guess your system's bus speed, but the good news is that even if it thinks you have a DX4-100, it still would identify the correct bus speed of 33 MHz.

Do you know whether write-back internal cache is working? Check what Speedsys says for your CPUID.

jakethompson1 wrote on 2023-04-17, 03:27:

As alluded above have you identified whether the board misreporting your CPU as a DX4-100 is merely a cosmetic issue or actually affects anything?

Excellent point ! I looked at the BIOS and also BIOS of EFA 4MHL3G - 12/07/95-SIS-85C471B/E/G (release REV 1.05 01/12/'96) thinking it may have something different but exact same cpu speed post strings.
from hex: 25MHZ, 33MHZ, 40MHZ, 50MHZ, 50DX2, 66DX2, 80DX2, 75DX3, 100DX3, 120DX3. No 133DX
also cpu id strings: 80486DX. 80486SX. 80487SX. 80486SX2. 80486DX2. P24T. DX4. P24D. TI486DX2. CxDX4. Cx486S. Cx486S2. Cx486DX. Cx486DX2. Cx5x86. U486SX. U486SX2. U486DX. U486DX2. Am486DX. Am486DX2. Am486DX4. Enhanced Am486DX2. Enhanced Am486DX4. So even if it does identifies as AMDDX4 it cannot report 133 AFAIK

Hi!
Here some SIS471+404 Mr.BIOSes - https://docs.google.com/spreadsheets/u/0/d/1v … eyuFf8/htmlview#
Here the Mr.BIOSes archive - http://vogonsdrivers.com/index.php?catid=77&menustate=32,29

BIOS versions

@Babasha, did you check the contents of the links, you are sharing, yourself? There are no MR BIOSes available for almost any 486 class motherboards, including SIS471.

@aaronkatrini
The manual lists both options so it is that same board. I cannot find any additions to the model name to tell if it's 2x72/4x30 or 4x72 memory.

aaronkatrini wrote on 2023-04-16, 21:09:

Hi all,

I'm looking to find a newer BIOS for my 486 motherboard so that it can properly detect AMD 5x86 133MHz CPUs (currently recognizes them as DX4-100).
From what I could find on the internet it's a Epox one. This was recognized by the ID string : 08/15/95-SIS-85C471B/E/G-2C4I9PA0-00 (AWARD BIOS).

Hi aaronkatrini,

I have been looking into your 08/15/95-SIS-85C471B/E/G-2C4I9PA0-00 Award BIOS and it has indeed no Am5x86-133 support. It also has the infamous 2GB HDD display limit bug. 😉

As Horun showed us, all 1995 SiS471 BIOSes on TheRetroWeb are limited to Enhanced Am486DX4 support and won’t set the WB protocol enable bit in the chipset when running an Am5x86 in L1 cache WB mode.

However, I have an up-to-date 11/28/95-SIS-85C471B/E/G-2C4I9PA0-00 BIOS for you.
I got this BIOS from a Gemlight board and adapted it for your Epox board, using your 08/15/95 BIOS as a template.

This BIOS is free from the Year 2094 and 2GB Harddisk display limit bugs, and fully supports the Am5x86 and Cx5x86 CPUs, and Harddisks up to 8GB.
You can Enable/Disable the L1 and L2 cache separately from the BIOS, but the L1 WB/WT option is automatic.
This BIOS shows the “L1 cache: WB/WT” option in the BIOS Setup only for the P24T and for Cyrix CPUs.
For the P24D (486DX2WB), Am486DX4WB and Am5x86, this 11/28/95 BIOS can detect if these CPUs are in WB mode and programs the chipset registers accordingly. It then hides the “L1 cache: WB/WT” option, because user interaction for this automatic function is not required.

Please let us know how this BIOS works so we can put it on The Retro Web.
Cheers, Jan

Thank you all for your support!! Wow, words can't express my gratitude to you 😀

I'm still waiting for the new Eprom to arrive, I don't feel risking flashing the current BIOS chip. Will update when I get my hands on it.

Also the CLKMUL pin wasn't properly located. According to what pinout diagram I could find online, that pin on the motherboard isn't connected to anything (at least visually). But on the link provided from CoffeeOne, the pin modded by the user on that thread does indeed point to a jumper nearby. I did some experiments messing with that jumper and where it leads to (its jumper J25), according to the manual for selecting UMC CPUs, but that didn't make any change. Maybe it will when I have the new BIOS 😀

GigAHerZ wrote on 2023-04-21, 06:32:

@Babasha, did you check the contents of the links, you are sharing, yourself? There are no MR BIOSes available for almost any 486 class motherboards, including SIS471.

Im flash 5-10 motherboards to/from Mr.BIOS (386 and P1). Sorry I dont check every file/BIOS in Mr.BIOS archive.

Sorry for the long delay on replying, the Eprom I had ordered from China has yet to arrive and a couple of Eproms I bought from a local electronics store weren't new, and I don't have an UV light to cancel them at the moment. Long story short, a friend helped me and send me an Eprom already programmed with the BIOS from @Chkcpu.

The BIOS works fine, it boots right up and shows the string "11/28/95-SIS-85C471B/E/G-2C4I9PA0-00", however I can't get it to recognize the AMD 5x86 133 properly. Same issues as before. Maybe it's me that I'm not properly configuring the Jumpers correctly, but I believe that I've tried all possible combinations 😒

So for the moment I'll leave it as is, perhaps one day I'll get back into it and see if I missed something. For the moment I'll continue to use a Cyrix 100GP which works fine. I would like to thank everybody who participated on this thread. Your help was very much appreciated. Cheers!!

aaronkatrini wrote on 2023-05-05, 20:31:

Sorry for the long delay on replying, the Eprom I had ordered from China has yet to arrive and a couple of Eproms I bought from a local electronics store weren't new, and I don't have an UV light to cancel them at the moment. Long story short, a friend helped me and send me an Eprom already programmed with the BIOS from @Chkcpu.

The BIOS works fine, it boots right up and shows the string "11/28/95-SIS-85C471B/E/G-2C4I9PA0-00", however I can't get it to recognize the AMD 5x86 133 properly. Same issues as before. Maybe it's me that I'm not properly configuring the Jumpers correctly, but I believe that I've tried all possible combinations 😒

So for the moment I'll leave it as is, perhaps one day I'll get back into it and see if I missed something. For the moment I'll continue to use a Cyrix 100GP which works fine. I would like to thank everybody who participated on this thread. Your help was very much appreciated. Cheers!!

Looks like you've given up for now, but in case you want to resume this later, those EPROMs you ordered from China are probably not going to be new either.

As a substitute for those often quite expensive UV eraser boxes, you can use a germicidal UV-C lamp. You want the specified wavelength to be as close to 254nm (2537 Å) as possible. Regular "blacklight" lamps (350-400 nm) are not going to work at all. I picked up one of these for less than €10. This particular brand seems to only be available in the Swedish market, but would guess it's just rebranded China stuff, it's probably available elsewhere under other brands. Or just get something else emitting the correct wavelength.
I've modded my lamp to circumvent the annoying 2 min timer (you need about 15-20 min total exposure time to erase an EPROM) and also raise the power to the LED's spec (measured voltage was quite a bit lower than they can handle), but if you're only gonna erase a few EPROMs that might not be a problem, and you can use it as is.

Otherwise the BIOS EPROM in your picture looks very much like a 27 series one. The 28 series EEPROMs are pin compatible with those - would guess they were intended as drop in replacements - and would let you do away with UV erasing completely. A very handy alternative 😀

If you don't have a programmer already, the Arduino based programmer TommyPROM is very simple to build and use. Put one together myself when I had to program a couple of 450 ns EEPROMs , that simply were too slow for the commercial programmers I've got access to. Since the TommyPROM software is open source, I could adjust the timings myself and get those chips programmed.

Hi,

Eprom was not the issue, I already tested the board with the new BIOS. A correctly flashed Eprom was provided by a friend of mine.
He also provided me with already erased other Eproms for future use 😀

Also i have a programmer (Willem), it's not the most intuitive to use but gets the work done.

Maybe one day I'll find a solution to the UV light, but most of the time I'm not messing with Eproms but rather Eeproms (Socket 7 and up boards), hence at the moment I don't feel the need for an UV light.

One thing I want to ask, about how many hours under the sun does it take to get erased? I live in Southern Europe (Rome, Italy) and so there is plenty of sunny days.

aaronkatrini wrote on 2023-05-06, 13:17:

Hi, […]
Show full quote

Hi,

Eprom was not the issue, I already tested the board with the new BIOS. A correctly flashed Eprom was provided by a friend of mine.
He also provided me with already erased other Eproms for future use 😀

Also i have a programmer (Willem), it's not the most intuitive to use but gets the work done.

Maybe one day I'll find a solution to the UV light, but most of the time I'm not messing with Eproms but rather Eeproms (Socket 7 and up boards), hence at the moment I don't feel the need for an UV light.

One thing I want to ask, about how many hours under the sun does it take to get erased? I live in Southern Europe (Rome, Italy) and so there is plenty of sunny days.

Heya,

Understood that, but thought you might want to experiment further in the future. Having the means to erase EPROMs yourself, or using 28 series EEPROMs, will make that a lot easier 😀

That would probably take weeks, if not months, even with Mediterranean summer UV indexes. An UV-C LED of a few milliwatts will produce magnitudes higher concentrations of light in the required relatively narrow band than what will reach the earth from the sun.

It's easy enough to test though. Just expose a couple of those chips to direct sunlight, dump them every other day or so and compare the images with previous dumps to see how/if they change. The last bits can take a [very] long time to erase since there's randomness involved, photons with the right wavelength has to hit the chip in the right place and so on.

Also keep in mind that window glass will absorb UV-C light almost completely - there's a reason for using quartz for the chip window - so just putting them in the windowsill won't do. The sunlight has to be direct. Also, the same goes for skin grease etc, make sure to clean the chip window thoroughly before 😀

Chkcpu wrote on 2023-04-21, 09:21:

Hi aaronkatrini, […]
Show full quote

aaronkatrini wrote on 2023-04-16, 21:09:

Hi all,

I'm looking to find a newer BIOS for my 486 motherboard so that it can properly detect AMD 5x86 133MHz CPUs (currently recognizes them as DX4-100).
From what I could find on the internet it's a Epox one. This was recognized by the ID string : 08/15/95-SIS-85C471B/E/G-2C4I9PA0-00 (AWARD BIOS).

Hi aaronkatrini,

I have been looking into your 08/15/95-SIS-85C471B/E/G-2C4I9PA0-00 Award BIOS and it has indeed no Am5x86-133 support. It also has the infamous 2GB HDD display limit bug. 😉

As Horun showed us, all 1995 SiS471 BIOSes on TheRetroWeb are limited to Enhanced Am486DX4 support and won’t set the WB protocol enable bit in the chipset when running an Am5x86 in L1 cache WB mode.

However, I have an up-to-date 11/28/95-SIS-85C471B/E/G-2C4I9PA0-00 BIOS for you.
I got this BIOS from a Gemlight board and adapted it for your Epox board, using your 08/15/95 BIOS as a template.

EPox-GXA486SG-Rev-J.1-BIOS.zip

This BIOS is free from the Year 2094 and 2GB Harddisk display limit bugs, and fully supports the Am5x86 and Cx5x86 CPUs, and Harddisks up to 8GB.
You can Enable/Disable the L1 and L2 cache separately from the BIOS, but the L1 WB/WT option is automatic.
This BIOS shows the “L1 cache: WB/WT” option in the BIOS Setup only for the P24T and for Cyrix CPUs.
For the P24D (486DX2WB), Am486DX4WB and Am5x86, this 11/28/95 BIOS can detect if these CPUs are in WB mode and programs the chipset registers accordingly. It then hides the “L1 cache: WB/WT” option, because user interaction for this automatic function is not required.

Please let us know how this BIOS works so we can put it on The Retro Web.
Cheers, Jan

Hi Jan, may I ask an off-topic question.
You patched a lot of BIOSes, so I would like to know the following:

Many "modern" 486 BIOS have LBA support, but they do have a 7.8GB Bios limit (something a bit below 8GB I believe).
I am talking of the Asus VL/I-486SV2GX4 and the Asus PVI-486SP3, I have those 2.
Would it be possible to overcome the 8GB limit with a modified Bios?