I successfully hot flashed an Award BIOS chip the other day. The details are here -

FIC 486-PIO3 mobo - why's it so slow?

If you scroll down to the message that begins with the blue word Success!

Unfortunately I don't have another PCI socket3 board with AWARD BIOS (I just gave one away a few days ago) - all I have left are no-name VLB boards with AMI BIOSes... moreover, I peeled off the label from the ruined BIOS, and it's the "look-at-the-chip-through-a-small-"window"" type - aren't those UV-erasable (and therefore maybe definitely ruined this time, through exposure to light)?

Maybe it's a good idea to do some careful research before performing an operation on hard to replace hardware before even attempting something like a bios flash.

Last time (about 3 months ago) i destroyed this card by running by accident video chipset info menu in diag software - few blicking screens and since then i got broken image (sort like a bad memory chip). Good is that i got another same card (only without gold platted memory chips), that still works. So never try to run modern diagnostic sofware checking resolutions of graphic cards on something like a czech hercules card using bulgarian motorola chip clone 😁

iulianv, you could give these guys a try:

http://www.biosman.com/

Since you have the ROM and the BIOS chip you could send it to them and they could program it for you.

Unfortunately I don't have another PCI socket3 board with AWARD BIOS

It doesn't have to be a PCI Socket 3 board to do a hotflash... it can be any kind of board that uses a similar type of flash chip. Sometimes it can be done with the boot ROM socket on a network card too, if you can find a NIC with the write-protect disabled.

And as for the window on the EPROM chips, those things require some pretty high intensity UV of a specific wavelength in order to erase 'em... ambient light isn't going to affect 'em much, it takes weeks even in direct sunlight. Just put a sticker back on the chip, and you're good.

I stuffed up one of my SS7 boards ages ago. Was really upset, but then I noticed the huge BIOS chip. And another board had the same kind of chip!

So I swapped the chips, booted the (previously) dead board, swapped them again while running, flashed the BIOS and the world was ok again 🤣

That day I felt like I went from hero to zero and to hero again 🤣

Was up all night and really happy that the board wasn't stuffed.

I think my only decent Socket 370 board (Gigabyte GA-6VXE7+) committed suicide a while back... and now I can't fix it now that I have the materials to do so.

About a year ago the network card was acting up so I was going to reboot it after installing updated drivers. After that it froze in the BIOS once, then after that just dumped me to the Award BootBlock BIOS, asking for a boot disk to flash the board. I tried for a while to get a disk to boot but didn't have any luck (it would load the flashing program, but then it wanted keyboard input and without a real BIOS it couldn't give me any) so I put the board aside for a while.

Recently I finally found what I think is a working program to flash the BIOS back onto it and the proper command line switches to push it along, which I didn't know before. So I pulled out the board and hooked it up to a spare PSU and monitor. Now it won't even start up... just a blank screen. Tried a bunch of video cards in every single slot. Hoping that I just need to mess with the jumpers or something to kickstart it. It's going to cost me way too much to replace this with another decent board through eBay, and my only other Socket 370 boards are both Intel 810s that I'd never be able to make into decent gaming machines.

What about the myth of "use an ISA video card if you ruined your BIOS"? It didn't work on my socket3 board...

AFAIK the floppy boot still works. When my BIOS was dead, it did boot of the floppy (flying blind).

Unfortunately there is no FDD activity on my board...

I feel that you don't need to bid adieu to your Socket 3 board just yet.

As daunting and dead-end they may seem, BIOS errors are actually very salvagable and resurrectable.

If you really want your board working, hot flashing or bios replacement will be useful to your condition. Other members might be able to help to hot flash a bios chip for you, if you want.

I've been doing some reading today, which brought new questions...

First, my BIOS is a Texas Instruments 27C010A-15 EPROM, which seems to be an "UV light erasable, electrically programmable ROM", and I've read either that you can use UniFlash on these, or that you need a special device (an external programmer) to program it. Which of these is correct?

Second, I've been investigating Old Thrashbarg's idea, about using a NIC for re-flashing the BIOS. I have a RTL8139C-based NIC at home (unfortunately it has a 28-pin ROM socket - I found a page describing where to connect the extra pins, however soldering is not my strongest skill), and also two or three RTL8139D-based NICs, and there is a RTFLASH.EXE on Realtek's site.

But I've also read that, unlink RTL8139A/B/C, the D version lacks this feature (to program the ROM), which might also be missing from NICs with A/B/C versions (due to cost reduction policies of the manufacturers).

Until I get home to try this, can someone share a success story about software-programming a 27C010, and one about successfully using a RTL8139D for ROM programming? 😀

I've used Uniflash on EPROMs as well as the usual 128k and 256k ROM chips for hotflashing. In all instances, hotflashing worked well. The Uniflash is really useful.

The only thing to practise is removing the bios chips carefully while the system is running, and making sure not to short circuit other components while doing it. If the chips were never removed before, it's helpful to remove and replace at least once before actually hotflashing, to make it easier to remove the bios chips when doing for real.

Also, while inserting back the chip, make sure to straighten out all pins on the chip carefully before re-inserting into the socket. The pins are fragile, more so for older ones.

(And I don't know anything about software programming. 😉)

By "software-programming" I meant using some utility (like UniFlash, AwdFlash or similar) with common PC hardware - that is, not having to use a "dedicated" programmer (like ALL-07 for example).

I don't have another mainboard with 32-pin BIOS around, so if I go for the flash-in-a-NIC idea, I guess I won't have to hot-flash...

One other question though: if I'm to use UniFlash to program my 27C010A, do I have to erase it first? That is, do I need a UV lamp? Old Thrashbarg seems to advise so in another thread, in the Marvin area...

Having to UV-erase it first would explain the "Erase flash fail" messages that I got with my attempts, but then what the hell was" Blanking flash ROM", and why did it succeed? 😀

Hold on, you can't flash upgrade an EPROM. You have to erase it first with UV light (if it has a window in it... if it doesn't, it's write-once and you have to replace the chip), and you would normally have to use a dedicated programmer to write to it. Uniflash and Awdflash shouldn't have even been able to touch that chip, they don't support the 27C series. Have you switched the jumper back to 'EPROM' and tried booting it again?

But if it is somehow screwed up, since the board supports flash chips, you can replace the EPROM with an EEPROM... a 28F010 should work. Those can be erased and programmed electrically, so it's just a matter of finding something with a 32-pin socket that you can use to flash it... a lot of PII/PIII boards used such chips.

Yes, it has a window through which the die is visible... so I can forget about the whole NIC thing.

Of course I tried switching the jumper back to EPROM, but the board still wouldn't POST. Maybe something else got "blanked" on it - I'll look closely when I get home, for some other chip that could resemble a flash memory...

I also have a few defective boards (different slots/sockets) - maybe I get lucky and one of them has a 32-pin EEPROM.

Well, unfortunately I don't have any 32-pin flash ROMs around, but I did find a 32-pin EPROM, labeled "ISA/PCI 486", so I put it on the board to see what happens.

What happens is that I have video output - after the BIOS screen of the video card immediately follows a screen by "Award BootBlock Bios v1.0" displaying "BIOS ROM checksum error", then booting from floppy is attempted (not successful - either non system disk or disk i/o error). This also happened to me (before breaking the BIOS) if I would put EDO RAM on the board.

I wonder what got actually destroyed and whether this gives me some chance to put that back...

Hm, that gives me an idea.

Try feeding it a bootable DOS floppy, with awdflash.exe, and an autoexec.bat containing the following:

@echo off
awdflash.exe xxxx.bin /py/sn/f/cc/r

Don't actually include a BIN file on the disk, though. I just want to see if it'll get far enough to load awdflash and attempt to flash the BIOS, don't want to actually write anything yet. You should also make sure the jumper is set to "EPROM" to be safe.

My idea is, if it'll actually get far enough that it tries to flash the BIOS, you could get a flash chip and hotflash it in that board, using the BIOS chip you found to kick off the process.