ABIT motherboard compatibility

[Haig]

Hi guys,

Just a quick update on this. So far, the 2x lockups wouldn't have anything to do where it was made. The only common factor so far is that the lockups are with the metal bladed fans that have 2 wires.

I found one of those cards today and still couldn't reproduce it.

I'll be getting the RMA'd board from BC and trying that out on our Abit boards. Hopefully that will reproduce the problem.

I know that once person has managed to fix this by upping the core voltage to 2.1. Any others had luck with this?

Haig

[Zypher]

Some boards (*cough newer abits with newer bioses cough*) let you play with the I/O voltages, from ~3.3V range. I wouldn't stray to far from 3.3V but who knows?

Core V would only hit cpu as far as I understand....but hey if it works, it works The I/O could perhaps affect the agp bus...just a thought.

BTW, 3 wire here. I wonder if theres a way I can read the rpms on the fan...thats what wire #3 is for AFAIK

AGPboard Monitor... anyone interested? I love Motherboard Monitor myself. We could contact the programmer as a group and see what he would try.

I didn't mean to slam so hard before, it just seems everyone is in the either 'All systems Go' or the 'All systems ****ed to hell' camp...no inbetweens, which is what usually happens with H/W problems.

[Haig]

Hi Zypher,

you are right. The 3rd wire is to read the rpm of the fan. I doubt that the 3rd wire has anything to do with this. If anything, the fans with the metal blades take more current than the fans with the plastic blades.

Haig

[Scytale]

So it may be worth trying alternate cooling with the on board fan disabled...

[Scytale]

Tried it. I underclocked the MAX and used an external fan. No change.

But in the process I seem to have persuaded my Quantum Atlas to transform into a lorry full of bricks. A firm kick in the groin forced it to mutate into a whining blade saw. That's better.

[Zypher]

Kicking computer = big sign of trouble


...so Haig, anyway us laypeople can get a readout on those rpms? If we can, this guy will very likely put it in the program...and maybe get readouts for other 3-wire cards too.

[Rob M.]

My MAX with the 2X crash is Canadian, 2 wire. serial number ABY78534. I picked it up from a local computer store in Toronto on Aug 27. The original one I had with the same crash came from the same store on Aug 19, serial number ABY78452. (Not sure if it had 2 or 3 wires, but it was Canadian made also)

[glarec]

hi.. i have a max with 2 wires too and i have trouble with 2x.. i am using a abit bp6. should i just go ahead and get it exchanged? i got it from the matrox online store. any advice would be appreciated

------------------
Glarec
------
bp6-(2)433(oc488) celerons, g400max, ibm ultra scsi 9.1g, sb-live value, hitachi ss-753 19", plextor 32x, etc, etc, etc...

[bcronin]

Haig, my board was shipped back to Matrox last night at 5PM and has already reached the Plattsburgh NY address I was given for RMA returns. So, you should have it soon.

------------------
Cheers,
BC

[Danfoss]

Haig: i have BX6 Cel300a@450 'canadian 2 wire fan g400max' and the famous agp2x lockups

i will try upping the voltage....

i'm curious though.. if it is a manufacturing problem will everybody get a new card???

[This message has been edited by Danfoss (edited 09-28-1999).]

[Danfoss]

ok.. upped the voltage..... i set the voltage at 2.5 and it works!!! had no lockups at agp2x i will start lowering it to see what is the 'critical' setting

one question: why does upping the cpu voltage affect the vidcard????

[Rob M.]

Tried the cpu at 2.1v, It lasted about 3 hours before I got the same crash again. Not totally sure on this, but it seemed to make a little difference. Lately it would hang only a minute or two after boot if I enabled agp 2x

[Dutch]

I have the Abit BX-6 rev 1, BIOS JL, and a Pentium II 400 CPU. Just installed my Matrox G400 MAX yesterday and I haven't had one lockup yet.

My card was made in Ireland, it has plastic blades on the fan.

Now that I posted this I bet it starts locking on me!

[dwright]

Haig: I have Abit BH6 + G400max(irq10 forced & alone) + sblive(irq5 forced & alone). On day 1 the G400max(directly from Matrox about 3 weeks ago) I had the lockup problems but only with opengl games right after the menu to enter the game (90% of lockups are just prior to actual 3d game screen). With Directx games there are no lockups. I am able to duplicate lockups with Q3test(current version). If I turn on opengl extensions I get a lockup eventually. If I turn it off I receive few lockups. Once I get a lockup the only out is a reset. It continues to lockup until I turn opengl extensions off. Then there is a good chance (about 75%) that I will not lockup. If I change the setting to opengl extensions on while in Q3test then it is accepted and most of the time there is no lockup (I confirmed this by receiving higher frame rates). When ther is a lockup during this recovery procedure I have observed the mouse pointer is corrupt during the menu selection. Here is the summary of what I have tried:
bh6 bios is up-to-date
nothing overclocked (the PIII450 does o/c to 500 no problem confirmed with with 3dmark99max)
format hd & reinstall win98se (seems to fix lockups for a while)
Sblive drivers @ 2.1
Matrox @ 5.25 (also tried icd 5.13.021) = no effect
force agp 1x = no effect
force agp 2x = no effect
warm or cool temps = no effect
updated G400max bios = no effect
restored win98se from fresh install backup = no effect
tried reset/reboot 5 times in a row after lockup = lockup every time (when I changed opengl extension to off I got a lockup but with a corrupt game screen ie:it just got into the game)
I set opengl extensions off - exit game - launch game no lockup. Then I turn opengl extensions on and no lockup (most of the time)
If I keep testing it locks up again at some point and I repeat the process with opengl extensions.
Once in the game I have not had a lockup (I don't say in the game long because it just a test for G400max).
Note: Other opengl games lockup on entry from menu if Q3test locks up.

Can others with the lockup try this and report back?
What do you think h/w or s/w or both?

[Haig]

Hi guys,

Here's an update so far:

Lock up problems aren't CPU dependant because a few on this thread and a few calls into tech supp today were with celerons and Abit boards.

So the problem is now with Abit and any CPU with a metal bladed fan/heatsink.

Upping the voltage has fixed the lock ups for a few but not all.

I need one more item from you guys if possible. For the clients with the lock up problems, can you list the type of power supply and its specs?

BC's board should be in tomorrow and things might look clearer.

Danfoss - If we did have a hardware defect, we would see this problem on all motherboards. If you can, try the card in a different mobo and it won't lock up.

Thanks,

Haig