it does not like to read 1.44mb that works perfect in my others pc .
they are boot disk dos 6.22 that i insert when ontrack ask to push spacebar and insert a boot disk.the pictures are the reads of the disks

Typical causes are cache problems, in this instance maybe enabled on the board while chips absent, memory problems, since it is not changed by modules, something above 64kb (or POST should catch it) and below 4MB. (Otherwise 4MB might work fine) so one of the address lines either dirty in the simm sockets or a line scratched under the board or something.... ORRRRR the 14.318 reference oscillator has drifted off by a lot and floppy timing is screwed.... or timing to ISA bus FDC is disrupted.

Gonna see if I can dig out the board...

BitWrangler wrote on 2026-08-21, 03:17:

Typical causes are cache problems, in this instance maybe enabled on the board while chips absent, memory problems, since it is not changed by modules, something above 64kb (or POST should catch it) and below 4MB. (Otherwise 4MB might work fine) so one of the address lines either dirty in the simm sockets or a line scratched under the board or something.... ORRRRR the 14.318 reference oscillator has drifted off by a lot and floppy timing is screwed.... or timing to ISA bus FDC is disrupted.

Gonna see if I can dig out the board...

ok thanks a lot for this answer. I will try to put compatible cache if i have some over or from my other 486.

But most of time, the pc freeze after memory count.

i cleaned the simm 1 a lot

I think we are looking at one of these...
https://theretroweb.com/motherboards/s/qdi-v4s471-472p

Now is there still a cache RAM in the tag socket that's hiding under the drive cage/cables in your pics? Between VLB socket and CPU.

If there is, it might be bad or misconfigured. Some boards can work with a tag only.

BitWrangler wrote on 2026-08-21, 03:17:

Typical causes are cache problems, in this instance maybe enabled on the board while chips absent, memory problems, since it is not changed by modules, something above 64kb (or POST should catch it) and below 4MB. (Otherwise 4MB might work fine) so one of the address lines either dirty in the simm sockets or a line scratched under the board or something.... ORRRRR the 14.318 reference oscillator has drifted off by a lot and floppy timing is screwed.... or timing to ISA bus FDC is disrupted.

Gonna see if I can dig out the board...

it seems to be thos one exactly except the bios chip. maybe someone change the bios chip:

QDI
V4S471/472P

Oh yeah, yours says Gigabyte on it, which kinda suggests it didn't come in a QDI. That might explain failures. So probably want to ensure it has the right BIOS in it before anything else. Before the system configuration screen comes up on boot, the prior POST screen, there's a string describing the BIOS at the bottom of the screen, should be same or similar to one of the ones under the BIOS tab on the retroweb page.

BitWrangler wrote on 2026-08-21, 03:37:

Oh yeah, yours says Gigabyte on it, which kinda suggests it didn't come in a QDI. That might explain failures. So probably want to ensure it has the right BIOS in it before anything else. Before the system configuration screen comes up on boot, the prior POST screen, there's a string describing the BIOS at the bottom of the screen, should be same or similar to one of the ones under the BIOS tab on the retroweb page.

see the string

edit: just froze after a: dir

it doesnt seem similar enough. i have nothing to write the bios to a chip. hmm.

I will check the cpu type jumpers. A lot of jumpers make the cpu type. The board has no cpu in it when i received it

i check the manual and the bios option are pretty much the same. it's looks like 95%+ similar.

But i think the fact that i have no cache chip is a problem.

I past the night on it. i removed the tag cache chip that was there. i removed all cache jumper. disabled every cache in bios. it loads faster. but same problem. i tried another multi i/o card, another floppy and ide cable, different floppy, same problem. i even prepare hdd with another computer , nothing still boot. and no possibility to read any files or them from floppy. boot disk never boot and maked strange caracters. 1 time on 20 it boots but froze and cant copy anything. i tried a 527mb hdd with 1023 16 63 and still does not work. and thid hdd was booting on another computer. I think I'm going to give up. i dont have cache for him. i dont have other bios rom. if you think the bios award chip is flashable with a program let me know. i dont want to buy stuff to flash it. Thanks if you have any idea

BitWrangler wrote on 2026-08-21, 03:28:

I think we are looking at one of these... https://theretroweb.com/motherboards/s/qdi-v4s471-472p […]
Show full quote

I think we are looking at one of these...
https://theretroweb.com/motherboards/s/qdi-v4s471-472p

Now is there still a cache RAM in the tag socket that's hiding under the drive cage/cables in your pics? Between VLB socket and CPU.

If there is, it might be bad or misconfigured. Some boards can work with a tag only.

it was done. read my last post. thanks

Hmmm, it might be possible to hot flash a new BIOS into that chip, but that would need another motherboard that takes that size of ROM and also allows flash, I think that other 486 board you posted doesn't. It seems that this one might not run long enough for a flash to complete to do a flash on this board.

However, the chip needs to be confirmed to be a flash part not an EPROM or a mask ROM, which means peeling back the sticker and reading the part number. Jumper 73 next to the BIOS seems to be set in "flash ROM" position, which may have been for original BIOS, not this one. This one seems capable of getting as far as it does because it's at least the same chipset. Anyway, a quick and random thing to try is if jumper 73 improves behavior by being put to the other setting, that's a just about worth trying thing.

Anyway, if that is not a flash chip that could be hot flashed in another board, seems only way forward is to get a flashable chip, either new or salvaged from a dead board, hot flash the QDI bios to it and change it into this board.

Maybe someone else will have better ideas of what to do though.

BitWrangler wrote on 2026-08-21, 15:13:

Hmmm, it might be possible to hot flash a new BIOS into that chip, but that would need another motherboard that takes that size […]
Show full quote

Hmmm, it might be possible to hot flash a new BIOS into that chip, but that would need another motherboard that takes that size of ROM and also allows flash, I think that other 486 board you posted doesn't. It seems that this one might not run long enough for a flash to complete to do a flash on this board.

However, the chip needs to be confirmed to be a flash part not an EPROM or a mask ROM, which means peeling back the sticker and reading the part number. Jumper 73 next to the BIOS seems to be set in "flash ROM" position, which may have been for original BIOS, not this one. This one seems capable of getting as far as it does because it's at least the same chipset. Anyway, a quick and random thing to try is if jumper 73 improves behavior by being put to the other setting, that's a just about worth trying thing.

Anyway, if that is not a flash chip that could be hot flashed in another board, seems only way forward is to get a flashable chip, either new or salvaged from a dead board, hot flash the QDI bios to it and change it into this board.

Maybe someone else will have better ideas of what to do though.

I got.it working correctly perfectly after 15h hours of testing. I'm not sure what's make the difference, i did three things. 1. I exchanged the dx-33 cpu for a sx-25 cpu and i changed all jumpers correctly. 2. After that i remove the multi i/o isa card from the last isa slot at the bottom ( i tried before the 2 other isa slot too ) and i put it on a isa track that is a vlb slot with the brown slot at the end. This is the first slot of the board, the first of the 3 vlb slot. 3. i did change the bios again because i removed the vattery before and the bios was reset. i put the safest setting i could. ( the 1.7gb wirks now, i even succeed to use the ontrack disk manager floppy disk to configure it correctly. i copy games and files from my floppy disk on the hdd and all games runs correctly on the hdd.

What did get it working ? Number 1, 2 or 3 ??? Thanks for answer. I fear touch anything now 🤣. I have a few cache chips over in my stock. could they could be use:

14 pins MCM6264CP20 X3, 14 pins MCM6264CP25 X4, 14pins W24257AK-15 95100 X1 and 11 pins IS61C66-20N 9138 ISSI X3. or i can take half of my other vlb 486 cache chip 4 of the 8 chip. and i need a tag chip

Because i suppose a 486 with no external cache should be very slow in games, doom 1+2, nfs, jazz jackrabbit, duke nukem 3d, keen 4 to 6...

Thanks for answers

Awesome, glad it's going enough to do something with. With no cache Doom and Duke3d will be awful, but the others should be okay, they only need 386 or fast 286.

I think I see the board the BIOS belongs to https://theretroweb.com/motherboards/s/chicony-ch-471b
Which as you can see has quite a different layout. Is is possible that some of the jumpers are more or less "software defined" by the firmware and thus do something different than what it is thought they do, because the other board used that input on the chipset for something else.

However, a flash could be a possibility on that board now it's running well enough.

I think the BIOS got lost somehow, corruption or chip removed and a previous owner found a chip off a gigabyte board, couldn't find details of the board, a few years before retroweb info was very scattered and hard to find, and guessed at this other SiS 471 board and tried that BIOS, and who knows if he even got as far as you did.

Some guesses why 33 maybe doesn't work.... possible flaky jumper pins or caps on the speed setting, if one or the other is making bad contact, CPU might be trying to jump to 40 or 50mhz and crashing. Possible wrong setting of the Vesa 33Mhz wait state jumper, which usually doesn't matter much until you've got a VLB card in, but maybe if the board is a bit marginal it does. Also the right feature settings for the DX might be mapped to the wrong pins by this firmware, SX CPU are very simple and like a default.

BitWrangler wrote on 2026-08-22, 03:42:

Awesome, glad it's going enough to do something with. With no cache Doom and Duke3d will be awful, but the others should be okay […]
Show full quote

Awesome, glad it's going enough to do something with. With no cache Doom and Duke3d will be awful, but the others should be okay, they only need 386 or fast 286.

I think I see the board the BIOS belongs to https://theretroweb.com/motherboards/s/chicony-ch-471b
Which as you can see has quite a different layout. Is is possible that some of the jumpers are more or less "software defined" by the firmware and thus do something different than what it is thought they do, because the other board used that input on the chipset for something else.

However, a flash could be a possibility on that board now it's running well enough.

I think the BIOS got lost somehow, corruption or chip removed and a previous owner found a chip off a gigabyte board, couldn't find details of the board, a few years before retroweb info was very scattered and hard to find, and guessed at this other SiS 471 board and tried that BIOS, and who knows if he even got as far as you did.

Some guesses why 33 maybe doesn't work.... possible flaky jumper pins or caps on the speed setting, if one or the other is making bad contact, CPU might be trying to jump to 40 or 50mhz and crashing. Possible wrong setting of the Vesa 33Mhz wait state jumper, which usually doesn't matter much until you've got a VLB card in, but maybe if the board is a bit marginal it does. Also the right feature settings for the DX might be mapped to the wrong pins by this firmware, SX CPU are very simple and like a default.

everything works great until now. i put all my games and software and utilities on the hdd with another pc and adapter. everything works so far even the sound card and i put my 32mb ram simm in it. ( i have 2 x 32mb, but 64 mb for a 486 25 or 33mhz is strange and i dont want problem with 64mb )

1. So do you think it's the cpu or the 2 bottom isa slot that was the problem or other things ? i can put the 33mhz back and change jumper again.

The VESA is set at <=33.

2. i have copied on the hdd the uniflash 1.40 and both firmware. the J73 is is not flash position. Can i put it to Flash and try to flash the firmaware or can i corrupt everything if the chip is not flashable ? and i dont know which to flash : mv4-v4.amw.amw or Q13G.ROM ? and if uniflash need patameter let me know too Thanks for answer.

Duke nukem 3d runs very bad 🤣
Doom runs bad enough worst than my 386 DX-33 with cache 🤣 ! I'm surprised it runs that bad with a cirrus logic vlb card 5428, i will put 2mb for better speed. I understand it's only a sx-25 with cache. I will try to put back the 33mhz

THANKS SO MUCH FOR ALL YOUR HELP. It is very appreciated !

It can't be flashed i think without special equipment, i removed the sticker, i saw a small windows on it, see the model i note it:

m27c512-I5fI B88AQ 933IC ST Singapore

I have a bios chip too check the picture, a AMI bios 486 chip, but it was on a board without vlb

BitWrangler wrote on 2026-08-22, 03:42:

Awesome, glad it's going enough to do something with. With no cache Doom and Duke3d will be awful, but the others should be okay […]
Show full quote

Awesome, glad it's going enough to do something with. With no cache Doom and Duke3d will be awful, but the others should be okay, they only need 386 or fast 286.

I think I see the board the BIOS belongs to https://theretroweb.com/motherboards/s/chicony-ch-471b
Which as you can see has quite a different layout. Is is possible that some of the jumpers are more or less "software defined" by the firmware and thus do something different than what it is thought they do, because the other board used that input on the chipset for something else.

However, a flash could be a possibility on that board now it's running well enough.

I think the BIOS got lost somehow, corruption or chip removed and a previous owner found a chip off a gigabyte board, couldn't find details of the board, a few years before retroweb info was very scattered and hard to find, and guessed at this other SiS 471 board and tried that BIOS, and who knows if he even got as far as you did.

Some guesses why 33 maybe doesn't work.... possible flaky jumper pins or caps on the speed setting, if one or the other is making bad contact, CPU might be trying to jump to 40 or 50mhz and crashing. Possible wrong setting of the Vesa 33Mhz wait state jumper, which usually doesn't matter much until you've got a VLB card in, but maybe if the board is a bit marginal it does. Also the right feature settings for the DX might be mapped to the wrong pins by this firmware, SX CPU are very simple and like a default.

I just found the problem. It doesn't work correctly with the 486 DX-33 CPU.
But i dont know if the the cpu is defective or the motherboard have a problrm with and the 33mhz things. The jumper are correct. The cpu has one leg a little bent, but its not broke and it enters easily.
Do you have a idea ? Thanks

I tried changing the vesa 33mhz and wait jumper and it changes nothing

But i tried running the 486 DX-33 CPU at 25mhz innstead of 33mhz with the DX jumper enabled as a 486 DX25 CPU and it seems to work perfect.
So is it a cpu problem or the motherboard doesnt want to run at 33mhz ? or bios problem because of bad chip

BitWrangler wrote on 2026-08-22, 03:42:

Awesome, glad it's going enough to do something with. With no cache Doom and Duke3d will be awful, but the others should be okay […]
Show full quote

Awesome, glad it's going enough to do something with. With no cache Doom and Duke3d will be awful, but the others should be okay, they only need 386 or fast 286.

I think I see the board the BIOS belongs to https://theretroweb.com/motherboards/s/chicony-ch-471b
Which as you can see has quite a different layout. Is is possible that some of the jumpers are more or less "software defined" by the firmware and thus do something different than what it is thought they do, because the other board used that input on the chipset for something else.

However, a flash could be a possibility on that board now it's running well enough.

I think the BIOS got lost somehow, corruption or chip removed and a previous owner found a chip off a gigabyte board, couldn't find details of the board, a few years before retroweb info was very scattered and hard to find, and guessed at this other SiS 471 board and tried that BIOS, and who knows if he even got as far as you did.

Some guesses why 33 maybe doesn't work.... possible flaky jumper pins or caps on the speed setting, if one or the other is making bad contact, CPU might be trying to jump to 40 or 50mhz and crashing. Possible wrong setting of the Vesa 33Mhz wait state jumper, which usually doesn't matter much until you've got a VLB card in, but maybe if the board is a bit marginal it does. Also the right feature settings for the DX might be mapped to the wrong pins by this firmware, SX CPU are very simple and like a default.

Problems resolved. I'm disappointed of myself. i build several 486 in my life...
The 486 DX-33 CPU was running at 3.45v instead of 5v. The manual was not clear about how to put it in 5V. I thought it was done automatically. i had to put jumper on U50 jumper 3-6 and 4-5 close. Thanks for everything and your support and help, it was very appreciated !

Now you can try 40MHz for fun 😀