Just tried, from within XP, to FLASH the BIOS in the AV8 motherboard in my desktop, using the Windows based ASUS update utility and the latest (Beta) update (A018.002) taken-off the ASUS UK website. The update program erased the old bios (A018.001) but stalled at the 30% mark when writing the new BIOS.

The ASUS update utility allows you to restart the "Write" process, without restarting the application. I tried this twice, with the same result. When I exited the update application, it wouldn't restart because it could find any BIOS code to replace in the BIOS hardware ! Brilliant !

I re-booted the system, and it POSTed and the displayed the usual error message saying it couldn't find the A8V.ROM file. I inserted the original motherboard CD and the system reported that it had found the file on CD and was Flashing the BIOS. When it completed the Flash, I re-booted and now there's no video signal, no motherboard logo and no POST.

I tried re-booting and turning the power off, but to no effect.

If the flash from CD was successful, why isn't it posting ?

My understanding is that when you turn on the system the processor memory pointer automatically sets to memory location F000:FFF0 in the ROM-BIOS. This location contains a jump instruction to the address of the BIOS initialisation routine (variable according to ROM). I thought this routine was permament code and unaffected by BIOS updates ?

Nick

Have you cleared the CMOS?

Usually achieved by shorting a jumper on the motherboard...

Phil

or try removing the cmos battery for 20 mins, reinsert then reboot

I'll have a go.

Nick

Battery out, pins shorted by moving jumper to cover pins 2 & 3 (Per manual). Manual says 10-15 secs with the jumper in this position is sufficient to clear CMOS. I left it longer.

Reversed after 3 mins. Re-start, problem, as before.

The thing's not even beeping. This suggests to me that the initial BIOS code which checks the processor and chipsets functions on the main board is not functioning/available - I thought that this code was permanent, a BIOS bootsrap, which was unaffected by BIOS updates ?

Nick

Battery out for second time, pins shorted-out for 10 mins. Re-start, nothing. CPU fan working.

My understanding is that if the CPU fan circuit is not completed. The CPU hardware won't execute the jump instruction to the BIOS "Bootblock" loader and instead immediately executes a shut-down instruction.

This is not happening here. The AV8 isn't a fanless design and in this case the fan is running and the board is powering-up, including peripherals.

However, it appears that the motherboard checking routines are not executing. This could be because the jump instruction is not being executed - but it worked the first time round 'cause it demanded sight of the AV8.Rom file, indicating that it had detected an incorrect checksum in the BIOS code caused by the absence of the updateable bit.

It is no longer demanding sight of this file, so, I must assume that the BIOS checksum is correct, that the updateable part of the BIOS is loaded in the BIOS's EEPROM, and that the instruction pointer has jumped to the "Bootblock" loader address.

The "Bootblock" loader initially checks on the operation of the processor and the chipsets and then, if these are OK, goes on to check the off-main board peripherals, including the screen, subsequently handing process control to the bootstrap loader on disk which precedes to load the operating system. As there is no video signal and there have been no beeps, I can only assume that the start-up process is stuck somewhere between processor and chipset checking and video port checking.

I suppose there is the possibility that the BIOS checksum is correct but the code referred to is inapplicable to the hardware or corrupt with a compensating error. I think if this has happened then I'm stuffed, unless there is some way of deleting the updateable part of the BIOS code from cold boot - is there a proprietry program to do this or a BIOS emulator ?

There's a possibility that the Mk 2 SATA HD I have got fitted is causing the problem - its currently set t run in MK 1 mode i.e. 150MB per sec. I believe the BIOS ROM supports it, but the VIA chipset only supports it up to MK 1 standard. Certainly the chipset drivers in Windows XP only supported it after Service pack 2.

Nick

Think I've discovered what's happening. The keyboard lights are not working, so I guess that start-up has stalled short of instruction 1-2-2-1.

Computer Post / Beep codes

I suspect that the jump instruction at 1-2-1-3 isn't accessing the correct address in the updateable part of the code.

So, i'll need to Flash erase the ROM BIOS and write it again. How the h*ll do I do that from a cold boot ?

Nick

Chip programmer or JTAG, but I'm guessing the board doesn't have the JTAG header.

Fortunately, the situation has been partially resolved since last posting. Following advice found on another site I was able to get the system to POST by the simple expedient of removing the second of the DDR DRAMS and clearing the CMOS - I think that removing the second card, forced the system to single channel memory mode.

So, now, it completes POST and hands-off to the selected OS of the two on the disk (XP Prof and Vista 64 ultimate).

However, it appears that BIOS information is still required to complete the load as both OS's fail to load completely - XP re-starts almost immediately, whilst Vista loads some of the drivers and then stops at a BSOD - I presume that the failure is caused by each OS accessing in appropriate information that was loaded by the original A8V.ROM file info that was loaded in to the updateable portion of the BIOS after the failed BIOS programming.

Further, I'm not sure whether the POST is achieved using the "Bootblock" rom only or bootblock + updateable bit. On the one hand, the system booted using an AGP video card. I believe that "Bootblock" rom code only supports ISA graphics, so that would sreem to suggest that the full ROM code is being used. However, as I said above, 2 -channel memory was blocking POST, suggesting that the full code to support this isn't available from the BIOS ROM.

Unfortunately, the only plug-in diagnostic system I have is ISA based and the board is PCI slots only.

I think the way foward now, as suggested by the article on the other site, is to format a floppy as an MS-DOS system disk, dump the ASUS DOS BIOS utility and the BIN file for the most recent BIOS update that worked (1018.001) onto it , and start the system using this disk and attempt to re-program the ROM BIOS from MS-DOS.

Only problem is my laptop has not got a floppy drive. The alternative is a USB pen drive, which the Desktop will boot from, but which the laptop will not write MS-DOS system to - either in DOS or XP (Option greyed-out).

Apparently, there was an HP program which would do this, but it seems to have disappeared off the inter-web.

So, in the absence of a suitable substitute, then next best bet is taking the board to a local dealer to see if they can remove the BIOS chip and re-program using either EEP or UV scrub and EEP - I suspect that the boards too old (5 years) for there to be stockholding of the BIOS ROM.

Nick

The reason for both O/S's failing to load is going to be driver related, especially if the HD is SATA. When reseting the cmos back to defaults, it may reset the transfer mode of the SATA bus. Go into the bios and check all the settings thoroughly (including setting the date/time). Pay particular attention to where the integrated devices are listed, and see if you have a SATA emulaiton option, if you have and its set to IDE try the other setting and see if the system boots.

there are loads of places online that can sort this out

bios-repair.co.uk | Flash BIOS Repair | Flash BIOS Replacement

PC BIOS reprogramming, replacement, recovery

etc

for the sake of 15-20 quid to reflash it....

or my other advice would be search ebay for an identical mobo and use its bios chip....

Many thanks for the advice.

I'll have some time free this evening, so will give it a try.

Nick.

Sure nothing else is working in there? Wouldn't be the first time I've been struggling to fix one problem only to find out a second fault had developed at the same time.

Had a bit of a play last night.

Vista seems to be hanging when it attempts to load an AGP.sys file ?

Even tried VGA mode at start-up by pressing F8, no change.

Corrected some of the user changeable BIOS values, but didn't seem to make any difference.

Tried to create a bootable MS-DOS bootable CD using CD roast, it will only act on floppies.

Any body know of another boot making program for MS-DOS?

I've got a bootable floppy running Dr DOS. Can Windows read and write to that.

Regards,

Nick

DR DOS will run fat 16 or fat 12 for the floppies so no problem

Sorted.

All it needed was the BIOS ROM re-flashing with the bin file that I had previously been using. All is sweetness and light now, almost.

The method.

Created an MS-DOS boot disk on a USB Flash drive using a downloaded demo of the "Active@Boot" program. Stuck copies of the MS-DOS version of the Award BIOS ROM flash program and the relevant bin file.

Booted the desktop, changed the default boot device to the USB drive. Connected the USB, re-booted the desktop into MS-DOS using the USB pen drive. Ran the Award re-flashing program from the MS-DOS command line. Done.

Removed USB drive, re-booted desktop, changed default drive in BIOS to HDD, re-booted. Hey presto Vista and XP available and load OK.

After booting into vista, the harddisk seemed to be constantly going for about 15 mins, as if it was re-mapping everything to the HDD and XP thew up a load of "Recovered from serious error messages". But alll seems well now. Phew !

Nick