Page 1 of 2

Forcing boot block in corrupted award bios

Posted: Tue Jun 08, 2004 2:55 pm
by ruelnov
Any ideas on how I could force boot block to execute so I could recover my dead system?

I just need boot block to search the floppy, and I could take it from there even if there's no video. Please help. I'm willing to try any possible workaround before I throw this motherboard into the garbage can.

My W49F002U bios chip is soldered into mb, so I think we could not talk about "hot flash" or a "replacement bios".

But any ideas on the set of keys to hold down while booting to force boot block in AWARD BIOS? If not, does shorting 2 high address pins work with W49F002U bios chip to force a checksum error? If so, which pin numbers?

Problem details:
1. Correct image file of Award v6 was flashed successfully without overwriting boot block (using /sb switch in command line), but I simply forgot to put back the xgroup module into the bios file before reflashing.
2. No video, no beeps, no floppy seek on boot up.
3. Bios chip is soldered into motherboard.
4. System chipset = SIS 630E, Processor = Intel Pentium 3

Posted: Tue Jun 08, 2004 5:00 pm
by lucske74
You can try it with this parameter :/F
When performing a flash in this operation the use of the /F parameter is required. Using this switch will force the flash to go through(Award Bios Only). Also, this parameter must be used by itself. No other parameters can be used in conjunction with it.

Posted: Wed Jun 09, 2004 4:49 pm
by ruelnov
Sorry, but I could not use that /F switch because there's no floppy access happening. Again, no beeps, no video, and floppy LED never comes on. In short, boot block may not have been executed at all. Not even if I put in a specially-prepared floppy with the autoexec.bat file for automatic reflashing.

Posted: Wed Jun 09, 2004 5:06 pm
by Rainbow
Try this:
short two topmost address pins on the flash ROM chip together (e.g. using a screwdriver) and power up. If the bootblock activates, remove the short.

Posted: Fri Jun 11, 2004 3:16 pm
by ruelnov
Rainbow:
I was actually able to recover my system before you replied, and using exactly the same method you suggested. It did work, although I was pretty anxious that time I may damage my bios chip totally. But I felt I had to do that last tinkering before I throw my mainboard into the garbage can, that after experts from 2 other websites hosting this same kind of forum suggested that my system is hopeless.

That trick gave me more confidence to continue tinkering/tweaking my bios, and I don't have to worry anymore about not having a bios savior or a backup chip.

It took me several trials, though, to arrive at the correct pins. But when I got a copy of the chip datasheet, that made it easier to locate the correct high address pins.

Anyway, thanks for your ideas. It simply validated the technical basis of what I did.

Posted: Fri Jun 11, 2004 6:33 pm
by PeteV
Hi!

Congratulations for ruelnov, and great thanks for Rainbow!

Some day I also may need this hint ...

Best regards,

Pete V.

-------------------------------------------------

Hi!

Now I tried it, and it really works! Thanks.

I did make the trial on a 128kB chip, shorting the address inputs A16 and A15, which were the pins 2 and 3 on the DIP32 chip.

Best regards,

Pete V.

Posted: Sat Jun 12, 2004 9:21 pm
by ruelnov
That shorting trick should work if the boot block code is not corrupted, and it should not be if /sb switch is used when flashing the bios (instead of /wb switch).

Rainbow: Have you experienced a bad flash and using /wb switch before? And if so, were you still able to recover by forcing boot block through the shorting trick even if boot block became corrupted?

Or does anybody had this experience?

flashed w/ wrong bios, short worket

Posted: Thu Jul 08, 2004 4:16 am
by echz420
I incorrectly flashed my BIOS with the wrong flash and can also confirm that the shorting method works with a PLCC Atmel AT49F002NT. Like the reader above said, shorting pins 2 & 3 (counting to the right of the pin with the oval/half-moon) - A15 & A16 - This resulted in the boot-block code coming up, which did display with my PCI video card despite reading otherwise that it wouldn't.

After I realized that the awd822a flash program was failing, I then tried uniflash with /E and BANG like no other it flashed like a charm. To quote the uniflash docs, .... "Your program just saved the day for me"!!!

Thanks for the great posts and help all... 8O 8O :lol:

Posted: Thu Jul 22, 2004 12:07 pm
by Ritchie
RULENOV: Would you be able to please attempt to explain the location of the two pins on the flash ROM chip that you shorted to sucessfully force activation of bootblock?

I would like to attempt recovery of a board where I suspect all that might be wrong is that the BIOS is corrupted, quite possibly by a virus. It was originally giving the two tone beep that indicated a bootblock boot, and after initial attempts to troubleshoot the problem before I knew about bootblock, it has now deterioated to a similar condition that you reported in your original post, including no beeps and no video (even with an ISA video card).

The pins may be different since your board is probably much older (soldered chip) but they are probably the best starting point. Also what is the best way to short them - jam a small screwdriver into that point where the pins connect to the socket (in the case of an unsoldered chip)?

Posted: Thu Jul 22, 2004 2:29 pm
by ruelnov
Ritchie:

The 2 pins to short to force a checksum error varies from chip to chip. But these are usually the highest-numbered address pins (A10 and above).

These are the pins used by the system to read the System BIOS (original.bin for award v6), calculate the ROM checksum and see if it's valid before decompressing it into memory, and subsequently allow Bootblock POST to pass control over to the System BIOS.

You just have to fool the system into believing that the System BIOS is corrupt. This you do by giving your system a hard time reading the System BIOS by shorting the 2 high address pins. And when it could not read the System BIOS properly, ROM Checksum Error is detected "so to speak" and Bootblock recovery is activated.

Sometimes, any combination of the high address pins won't work to force a checksum error in some chips, like my Winbond W49F002U. But shorting the #WE pin with the highest-numbered address pin (A17) worked for this chip. You just have to be experimentative if you're not comfortable with "hot flashing" or "replacement BIOS".

But to avoid further damage to your chip if you're not sure which are the correct pins to short, measure the potential between the 2 pins by a voltmeter while the system is on. If the voltage reading is zero (or no potential at all), it is safe to short these pins.

But do not short the pins while the system is on. Instead, power down then do the short, then power up while still shorting. And as soon as you hear 3 beeps (1 long, 2 short), remove the short at once so that automatic reflashing from Drive A can proceed without errors (assuming you had autoexec.bat in it).

About how to do the shorting, the tip of a screwdriver would do. But with such minute pins on the PLCC chip, I'm pretty comfortable doing it with the tip of my multi-tester or voltmeter probe. Short the pins at the point where they come out of the chip.

Hope this helps, and do share with us the news (good or bad)!

Posted: Thu Aug 05, 2004 8:48 am
by Ritchie
Thanks rulenov.

Success!

See a summary of my result under VGA.

Posted: Mon Aug 09, 2004 4:59 pm
by ChameleonTPC
Does this apply to DIP package EEPROMs too! If so, which of the pins should I connect?

Posted: Mon Aug 09, 2004 5:38 pm
by ruelnov
It should apply as well. Pins to short would depend on your chip's datasheet. But more or less same pin designations as mentioned.

DIP or PLCC packaging doesn't alter the chip's intended functionality.

Did I really kill my Motherboard? Help!

Posted: Mon Aug 23, 2004 7:13 pm
by ripoli
I have a L7VMM2 motherboard.

It had the version 3.0 bios but the board is version 1.1. The bios is 32 pin welded on the board.

I tried to flash the bios with the correct one and it got dead.

Using the trick you teach about shortcutting pins 2 and 3 it respond to the Bootblock.

I tried to flash it again with Award Flash Utility (8.22a) it did not work and got dead again.

Then I got success in shortcutting the 2-3 pins and I was able to boot block again and this time I tried Uniflash. But after the 2nd beep I thought it was done and I power down the computer (jerk am I, didin´t read the how to).

Now, even if I take off the RAM module the board didn't beeps. Shortcutting does not work anymore.

Any advise to solve this?

Thank you in advance,

Libero

Posted: Tue Aug 24, 2004 3:04 am
by ruelnov
Had this same experience with my other board, shutting down PC prematurely thinking that flashing is finished.

My system did beep upon boot up, but it was an endless loop. I mean, my bootblock also got corrupted in the flashing process even when I did not use /wb switch. My flash rom is also soldered into the board, and I'm not able to recover this board up to now so far.

In your case, there are 2 suspects (your CPU and your bootbock). Try first your CPU in another working board and see if it beeps without RAM module. If it does, then your entire BIOS image is corrupted.