Not quite dead TCT K7TA-H

[bizzybody]

I have a Taiwan Commtech K7TA-H that's hanging at the "Verifying DMI pool data" line.

I've already read http://www.computerhope.com/issues/ch000474.htm and I've unplugged everything (except for the power connector and power button) from the board and I've taken the CMOS battery out, used the CMOS clear jumper, loaded failsafe defaults, changed RAM and videocard, reset ESCD data too. (Would be nice if the setup had a function to clear/reset DMI pool.)

The CPU is an Athlon 900 and I've tried various PC133 128meg DIMMs. The CPU was previously on an MSI-6378 where it worked perfectly.

It has the latest BIOS version. I didn't flash it, I cannot flash it until it gets past "Verifying DMI pool data". It has to *try* to boot before the BIOS can be flashed from a floppy disk or other media. (So please nobody tell me to flash or re-flash the BIOS because it's not possible until the board can boot!)

The BIOS chip is a large DIP in a socket, so if there's some way to clear the DMI data by removing the chip, I can do that.

[cp]

1. connect a floppy drive
2. enable the floppy controller
3. enable 'seek floppy at boot'
4. set floppy drive to 1st boot device
5. set the proper FSB (no overclocking!)
6. disable system/video bios caching

is the bios activating (seek) the floppy on bootup?

[KachiWachi]

It sounds to me like the OP cannot boot the machine, and wants to re-flash the BIOS in order to fix this condition.

Is this the case?

Thanks.

[cp]

yeah, i know. most of the time when the mainboard hangs at "Verifying DMI pool data" the real problem is something else. the message is the last thing that is visible (aka printed on the screen). this doesn't mean that it is the last thing that the bios does.

[bizzybody]

It does the floppy seek, then hangs. Setting Resource Control to manual- still hangs.

96 - 1.build MP table
2.build and upate ESCD
3.set CMOS century to 20h or 19h
4.load CMOS time into DOS timer clock
5.build MSIRQ routing table

When I set it to reset ESCD, it reports that it did, then shows the Verifying DMI line- so that's at least getting past #2.

Just tried setting the year to 1999- still hangs.

The person who gave me the board said he plugged the ATX power connector in with the AC cord plugged in and it wouldn't do anything at all. He put the CPU, RAM and other pieces in other computers and they were all OK.

The CPU and RAM I have on this board came from other PCs where they were working fine.

I don't know if flashing the BIOS would do anything, but I can't until the board will boot. If it'd boot there would be no reason to flash it because it already has the latest BIOS.

[cp]

hmm, those post code cards are really sweet.
did you try resetting the cmos and then booting without entering the bios just loading the defaults?

you could btw. always hotflash the bios on another working mainboard.

[bizzybody]

Yup, tried popping the CMOS battery out with the AC unplugged and then just booting. Also tried booting without the battery.

Same thing every time.

I'll check the BIOS chip numbers and see if one of the other boards uses the same size/voltage. Do all or most of those large DIP EEPROMs use the same pinout?

Is the CMOS RAM in the BIOS chip? Where's the DMI data held?

[cp]

Do all or most of those large DIP EEPROMs use the same pinout?

all. there are differences in the programming voltage (not operating voltage) of some chips. some need +12V to do programming, some only need +5V. as a rule of thumb if the chip number starts with 28... it needs +12V and if it starts with 29... it needs +5V.
btw. programming a 29... chip with +12V will most probably destroy the chip :)

Is the CMOS RAM in the BIOS chip?

no. it is situated in the southbridge. or on very very old mainboards (early pentium and 486) in the real time clock.

oh, one last suggestion: disable the floppy controller, set the floppy drive to none and attach a cd-rom to boot from. if it still doesn't work you can try to disable the L1/L2 caches.

[bizzybody]

Re your one last suggestion...

That finally got a success message out of it on updating DMI, but it still won't boot.

I changed it to having the floppy as the only thing connected and set to be the only bootable device. It seeks the drive then just turns it on and runs it endlessly.

It's like it has lost whatever bit of code that tells it what to do *next* after accessing a bootable device.

The only other board I have with the same type of BIOS chip socket is extremely dead, so no hot flashing attempts.

About the only thing left I can try with parts I have is swapping the Duron 900 for an Athlon MP 1.2Ghz.

Anyone here got an EEPROM programmer? What would you charge me if I mailed you the chip and the latest ROM file?

The south bridge is VIA 82C686B, I don't suppose there's a "paperclip" trick to forcibly wipe the CMOS and DMI? (As can be done with many of the olde Dallas CMOS/clock chips to wipe passwords.)

I've never seen anything like what this board is doing in 25+ years of working on PCs and other computers.

[edwin]

$30 from http://www.badflash.com/

not sure what other people will charge you. You could try and dig up a working board with the same bios socket for less than $30 ofcourse...

[cp]

disable the floppy controller and boot from a cd rom. don't forget to set drive a to NONE.

i have several (e)eprom programmers. anyway, you can boot a board of your choice to dos, remove the bios chip while running, pop your bios chip in and fire up uniflash. keep in mind the different programming voltages and almost nothing can go wrong.

the cmos is cleared when you remove the battery or set the jumper to clear cmos. the cmos is some sram that only keeps its data if power if supplied. thus the battery. once the power source is gone, the data is gone, too.

[bizzybody]

disable the floppy controller and boot from a cd rom. don't forget to set drive a to NONE.

Hello? Are you the same "cp" that made the other posts in this thread?

I said two posts above that I tried that AND IT DID NOT WORK! And I've said I do not have any other boards the BIOS chip will fit.

This happens almost EVERY TIME on forums. People don't READ- then tell the person with the problem to do what they've said they've already tried that didn't work. Worse than that are the @#$@$ing annoying sorts who at the end of a long thread say "I fixed it!" without saying HOW.

I've given up on this off-brand Taiwan Commtech motherboard and swapped in an Intel SE440BX-2 with a 450Mhz Pentium II. (It works but insists on having a non-existent IRQ "conflict" between SCI and ACPI- disabling SCI under 98SE.)

[cp]

It seeks the drive then just turns it on and runs it endlessly.

sounds like a damaged floppy/floppy controller. that's why i was suggesting to use a cd-rom instead.

btw. there is no post from you that mentions that you tried a cd-rom as boot device deactivating the floppy controller. or i'm overlooking something.
anyway, the advice is for free. i'm (and other, too) giving my spare time to help others solving their problems with computer stuff. if you already knew that forum are that bad and not helpful..why did you even bother to post? why not paying someone instead?

look, there is bunch of things i'd try if i've a mainboard hanging at verifying dmi pools. they'd take 10 minutes if i'd the machine sitting right in front of me. but this is a forum, so one step could take hours, even days.