My troublesome P5W DH De Luxe boards

Well-spotted GMAN ! Reckon that means the BIOS/chip is not u/s (pun!).
- LS

Leachim Sredna wrote:
I can report some progress. I ordered a new BIOS chip from some Ebay
advertiser, and while wating for it to arrive I tried an experiment. Having
extracted the BIOS chips from both boards with ease using a couple of
toothpick-like ponted woodens sticks so as not to damage anything, I swapped
them over.

I now had the BIOS chip from the newer board in the older one, and vice
versa. And the old board, the 1.03G revision, now with the chip from the
newer board, booted up as it should! The POST screen showed the chip to
have the 2801 BIOS. I was able to connect a hard drive, optical disc, ram,
video card etc.However there is still a problem at the BIOS level. For some
reason, it would not let me enter a date for the month of May later than May
12 (we were the 19th), which suggested to me that this BIOS file is perhaps
slightly corrupt. However Windows XP et me enter the correct date for the
system...

The newer board will still not boot up at all. Guess I'll have to wait for
the new BIOS chip to arrive to check that one out further. It could be the
processor (I'l loath to keep swapping these around, especially because of
that lousy Intel heatsink/fan fixing arrangement).

The problem with the older board was therfore clearly with the BIOS. What is
not clear is why the BIOS file got so damaged as to be useless. Nor is it
clear why the BIOS in the newer board is apparently slightly corrupted, nor
why the newer board would not boot up at all with BIOS/chip. I can see
absolutely no explanation of wny the BIOS should get ocrrupted in this way.
I would be grateful for any suggestions as to the safest/most reliable
method of flashing the latest BIOS onto these two chip, bearing in mind that
I should soon have a third one arriving with, supposedly, the latest BIOS
revision already on it.

I'll check out these Port 80 debugger cards too, Paul.
Thanks in advance
LS

When it comes to updating the BIOS, some care is required.

First, check the web site, and look for "warnings" about broken BIOS
update methods. An Asus motherboard may have three techniques for
updating the BIOS, and sometimes, two of them are broken and will
result in the BIOS update failing. So first you want to check to see
if there are known issues.

The new web site format, doesn't make it possible to see any warnings!

http://support.asus.com/Download.asp...uxe&p=1&s=2 2

Asus has a forum web site, and you can look through the threads for your board,
and see if there are issues with flashing.

http://vip.asus.com/forum/view.aspx?...Language=en-us

Section 4.1 of the manual lists the methods.

1) Asus Update (Windows program - downloads the BIOS and flashes while in Windows)

2) EZ Flash 2 (press a key in the BIOS screen, to trigger updating the BIOS)
The BIOS has a copy of AFUDOS inside it, and will prompt for a file to use
for burning.

3) AFUDOS is an MSDOS program. Using a Win98 boot floppy, with the copy
of MSDOS on it, you include a copy of AFUDOS and the BIOS file.
Since a Win98 MSDOS floppy reads and understands FAT32, I can store
the file on the root of C: and then it doesn't have to be stored on
the floppy. Being able to use the hard drive to store the BIOS file,
means there is room to store the original BIOS (backup the BIOS first),
as well as store the new BIOS file. Otherwise, a 1440K floppy isn't big
enough to hold modern BIOS files, which can be large.

4) CrashFree BIOS 3 is the BIOS code triggered when the checksum of the
BIOS is bad. Inserting the motherboard CD, would have a properly named
BIOS file to use for recovery (check the root of the CD to see what
files are on it). But the BIOS file on the CD will be very old, and
can actual cause the motherboard to fail to start. So CrashFree is a
bit of a trap. If you end up stuck in CrashFree, then prepare your
recovery media carefully with a modern BIOS. Don't immediately leap
to using the motherboard CD as the source of the BIOS file, because
it may actually hinder recovery.

Of those methods, (2) is relatively easy to set up. At least, it required
no MSDOS boot floppy. I've done a few motherboards here with a method
similar to (3). I don't like (1) because of the danger the BIOS update
will half complete and fail. Windows OS is a poor place to be flashing
the BIOS.

In addition to the support web site, Asus also has the files on an FTP site.

ftp://ftp.asus.com/pub/ASUS/mb/socke...20DH%20Deluxe/

You can see the naming convention for the files, has changed over time. They're
all ZIP files.

As a test, I downloaded one. It is 724,175 bytes.

ftp://ftp.asus.com/pub/ASUS/mb/socke...eluxe-3002.zip

Extracting the file inside the ZIP, I get P5W-DH-ASUS-Deluxe-3002.ROM,
1,048,576 bytes. Always check that the BIOS file is a power_of_two in
size. That is exactly 1.00 Megabytes and equals 1024*1024. The EEPROM
on your board, is listed in the manual as "8 Mb Flash ROM", meaning
8 megabits which is equivalent to 1 megabyte. And so I know that I
have unzipped the file properly. As the size of the file 1048576 bytes,
matches the 1 megabyte EEPROM.

That file name is too long to be used with many flashing tools. Try
a 8.3 format file name instead.

P5W-DH-ASUS-Deluxe-3002.ROM -- P5WD3002.ROM

The tools may even be sensitive to the file extension.

Using your motherboard CD and looking with file Explorer at the
root of the CD, you may see .ROM files there. That will show
the convention used for proper naming, which you can verify.

There are plenty of other rules for flashing, but the above
is enough for a start.

The BIOS consists of two pieces. The "boot block" code is tiny
and in theory, should never need to be changed. If that rule were
followed, of never changing the boot block, it would be harder
to corrupt the BIOS sufficiently to prevent recovery. But in the
Real World, the BIOS tools frequently update both the Boot Block
and the main BIOS code block, and if the boot block is erased
and the flashing tool fails to write a new boot block, then the
motherboard cannot be recovered. Then, getting a 32PLCC chip
from someone with a programmer machine, is the recovery procedure.

Paul