BIOS issue P4SDX motherboard... Help please

  • Thread starter Thread starter Marcus
  • Start date Start date
M

Marcus

Hi All,

I have just tried updating the BIOS on a P4SDX system with BIOS 1003, the
windows based program to do this reported failure to verify. I tried to
flash back the saved 1002 with same result. Taking very big gamble I copied
the saved 1002 bios to a floppy and rebooted the machine with the support
CD, now all I get is a black screen, the HD light constantly on, so it looks
like the recovery from bad bios doesn't work as the book says.

Any help / pointers appreciated.

Marcus.
 
DaveW,

It is a new motherboard, purchased today! Was working fine until I tried
updating the BIOS.

Thanks anyways,

Marcus
 
"Marcus" said:
DaveW,

It is a new motherboard, purchased today! Was working fine until I tried
updating the BIOS.

Thanks anyways,

Marcus
Click to expand...

In the worst case, as long as the flash chip is socketed, you can
get a replacement chip programmed by www.badflash.com . There is
no need for a new motherboard (Thanks, DaveW :-)).

In terms of recovery procedures, you can find someone with an
EEPROM programmer (yeah, right), you can "hot-flash" by using
a motherboard that has the same kind of flash chip interface,
or even someone who owns a BIOS Savior compatible with the chip
type, could be used to bootstrap your system. I consider all those
procedures to have a low probability of availability, and don't
bother to write reams of details about them any more.

Asus motherboards have had various versions of "Crashfree" for
a while. The BIOS is divided into two sections, the "boot block"
and the main code. The "boot block" contains a mini-BIOS, with
just enough brains to allow flash recovery.

The problem is, up until very recently, many Asus flashing tools
erase both the boot block and the main code, as a precursor to
flashing the BIOS. That effectively neuters the "Crashfree" feature.
It seems Asus has finally figured out this tiny detail, and
there are some boards now that can survive a flash failure.
Thus, for those lucky few boards, and modern versions of the
flash tools, you can actually follow the procedure in the manual
and recover the BIOS.

Boot block code is generally pretty dumb, and doesn't have enough
code to drive an AGP video card. That is why the screen remains
blank, even if the Crashfree feature is working. If you follow
the Crashfree procedure, and you can hear the BIOS attempting to
access the floppy drive, that is a good sign. That means there
is a good chance that Crashfree will work. If no attempt is being
made to find the floppy drive, it is likely the boot block is
erased.

With all of the gotchas in the above procedures, I've pretty well
given up on the more esoteric methods, and simply point people
to www.badflash.com or any of the other small companies that
provide flash chip services. Even Asus will bail you out for
$25, depending on which country you live in. The $25 buys you
a new $3 chip plus the programming service. Some companies also
offer to take your old chip, and reprogram it, for a reduced
fee. If you can afford to wait that long, and the shipping cost
is not too high, that is an alternative that can save a few
dollars.

HTh,
Paul
 
Hi Paul,

In the worst case, as long as the flash chip is socketed, you can
get a replacement chip programmed by www.badflash.com . There is
no need for a new motherboard (Thanks, DaveW :-)).
Click to expand...

Well that is one source I may have to try.

In terms of recovery procedures, you can find someone with an
EEPROM programmer (yeah, right), you can "hot-flash" by using
a motherboard that has the same kind of flash chip interface,
or even someone who owns a BIOS Savior compatible with the chip
type, could be used to bootstrap your system. I consider all those
procedures to have a low probability of availability, and don't
bother to write reams of details about them any more.
Click to expand...

I am fortunate here, we have an EEPROM programmer at work, however wether it
is compatible with the EEPROM type I don't yet know as I am awaiting a reply
from the testbay lads! Here's hoping....
Asus motherboards have had various versions of "Crashfree" for
a while. The BIOS is divided into two sections, the "boot block"
and the main code. The "boot block" contains a mini-BIOS, with
just enough brains to allow flash recovery.

The problem is, up until very recently, many Asus flashing tools
erase both the boot block and the main code, as a precursor to
flashing the BIOS. That effectively neuters the "Crashfree" feature.
It seems Asus has finally figured out this tiny detail, and
there are some boards now that can survive a flash failure.
Thus, for those lucky few boards, and modern versions of the
flash tools, you can actually follow the procedure in the manual
and recover the BIOS.
Click to expand...

Ah Ok, so once idea, however a little more thought was required.
Boot block code is generally pretty dumb, and doesn't have enough
code to drive an AGP video card. That is why the screen remains
blank, even if the Crashfree feature is working. If you follow
the Crashfree procedure, and you can hear the BIOS attempting to
access the floppy drive, that is a good sign. That means there
is a good chance that Crashfree will work. If no attempt is being
made to find the floppy drive, it is likely the boot block is
erased.
Click to expand...

I know how dumb the boot block code is, so happen to keep a PCI S3 Virge 2mb
card for just this type of problem, and previously it has worked for friends
machines, but the no floopy is where I got a little stuck.
With all of the gotchas in the above procedures, I've pretty well
given up on the more esoteric methods, and simply point people
to www.badflash.com or any of the other small companies that
provide flash chip services. Even Asus will bail you out for
$25, depending on which country you live in. The $25 buys you
a new $3 chip plus the programming service. Some companies also
offer to take your old chip, and reprogram it, for a reduced
fee. If you can afford to wait that long, and the shipping cost
is not too high, that is an alternative that can save a few
dollars.
Click to expand...

Unfortunately I can't afford to wait to long, and may try the 'hot flash'
approach as I know another board here uses the same type of EEPROM however
wether it flashes in the same manner I don't know (not that techically adept
yet), but I guess the worse is it won't work.
HTh,
Paul
Click to expand...

Thanks for reply.

Marcus
 
"Marcus" said:
Unfortunately I can't afford to wait to long, and may try the 'hot flash'
approach as I know another board here uses the same type of EEPROM however
wether it flashes in the same manner I don't know (not that techically adept
yet), but I guess the worse is it won't work.


Thanks for reply.

Marcus
Click to expand...

There is one thing to watch for. Say you have two motherboards, and
they both have LPC flash chips. Maybe one board uses an Award BIOS
and the other uses an AMI BIOS.

The Asus flash tool is going to check the BIOS before flashing it.
Chances are, an Award flasher is going to like the looks of an
Award BIOS file. I doubt you can "mix and match" the flashing
programs.

The other detail is at the BIOS level. Some flasher programs call
a "BIOS hook" as part of the flashing operation. When you
"hot-flash", the original BIOS image is, AFAIK, shadowed into RAM.
That is why you can pull the original flash chip out of the
motherboard, and shove a blank in there.

Now, consider if the donor motherboard uses a different BIOS
company than the BIOS company that makes the BIOS file to be
flashed. The missing "BIOS hook" will cause the flash program
to fail.

Your best bet would be if both motherboards use the same flash
chip type, but also both boards should use a BIOS from the
same company. I.e. Both boards should be Award or both
boards should be AMI. That might improve the odds on the
"hot-flash".

HTH,
Paul
 
Paul,

Thanks for all the advice and assistance generally but so far I have been
unsucessful, I have one last place to turn before having to use the paid for
services. We have a EEPROM programmer at work and the testbay lads think it
will do my EEPROM.

Marcus
 
Back
Top