Help me diagnose my computer (please)

[Das Capitolin]

I am way too busy to keep pulling my hair out over this random problem, so hopefully you will be willing to help.

Last week I put together a new system:
MOBO: ABIT AB9-Pro I965 (latest BIOS flashed as of 11/16/06)
CPU: Intel E6600 @ default settings (27/34C with SI-128)
RAM: Corsair TWIN2X2048-6400C4 2x 1GB modules @ 800MHz (default)
SATA1: WD 36GB Raptor
SATA2: Seagate 400GB 7900.0
IDE: Toshiba 16x DVD-R/W
PSU: Antec TruePower Trio TP-650 650W
VGA: Foxconn GeForce 8800 GTS

Here are the symtoms:

While in Windows (XP Pro: fresh install with all updates as of 11/16/06), I will occasionally and randomly have the system freeze up solid. This is a very random occurance, sometimes while right-clicking to reach properties, or when I go to close a window, at login, or even when I was 10% into the CoH install.

What I have done to troubleshoot:

1) Ran MemTest86+ for 24 hours without any errors whatsoever.
2) Reloaded Windows from fresh install (MBR removed/full format) five times.
3) Run F@H on each core for over 6 hours without any problems.
4) Loaded all system drivers with the latest versions available.
5) Swapped out PSU, CPU, RAM, HDD, and VGA on the AB9-Pro, only to still have system freezes.

At this point, the problem points to the motherboard; but I have also reviewed the even log which indicated the DVD-R/W on the jMicron IDE did not respond properly. So this has me thinking that I could be having issues with this controller.

Please help. Links are appreciated.

[Kougar]

Yay for more motherboard issues! :roll:

Well, this is the only thing that is coming to mind for me: Have you double checked that your DVD-R/W drive is in Ultra DMA mode and not stuck in PIO mode? I believe the recent issues with that were due to poor 965P/Jmicron support issues, so it is worth checking. Goto your Device Manager, find the IDE channel -> Properties -> Advanced Settings tab, and be sure to check the right channel that the dvd-rw drive is using. :) Hopefully this will help...

[Tank]

Can't you just try and troubleshoot it then?? I mean you could just disconnect it and see if the problems happens again..
You could also try some older drivers for motherboard or the ones that came with it on CD, if you didn't do it already. Don't know if any of this helps. :(

[Das Capitolin]

Have you double checked that your DVD-R/W drive is in Ultra DMA mode and not stuck in PIO mode? I believe the recent issues with that were due to poor 965P/Jmicron support issues, so it is worth checking.

Sure enough, there was one device in PIO mode. I have no idea how it set itself that way, but I have changed it to DMA. I'm crossing my fingeres.

Tank, I would not have known about PIO mode and the Jmicron chipset unless I took plenty of time to research it. Right now I am so busy with business and reviews that I have absolutly no time for my own problems.

Thank you for your assistance Kougar!

[DragonMaster]

What about trying the whole setup with a different mobo?

[Das Capitolin]

What about trying the whole setup with a different mobo?

They all worked on a different motheboard. But this motherboard is probably fine, and the problem is most likely a setting or configuration issue.

EDIT:

Kougar, after changing to DMA mode, I noticed that I was able to load CoH without crashing, which is good. The bad side is that it took almost 30 full minutes, where it only took about three minutes before. Idea's? I then went back and looked in the device manager (after spending a whole lot of time researching the jmicron issues) and noticed that "Standard Dual Channel PCI IDE Controller" is the generic Windows XP driver. I updated it to the JMicron JMB36X driver 1.17.3.2.

I will try this out for a bit, but may move to the newer 1.17.4.0 beta driver next. http://www.station-drivers.com/page/jmicron.htm

[DragonMaster]

So the problem is with the mobo.

Looks like you'Re not alone:
http://forum.abit-usa.com/showthread.php?t=115396

[nam-ng]

They all worked on a different motheboard. But this motherboard is probably fine, and the problem is most likely a setting or configuration issue.

EDIT:

Kougar, after changing to DMA mode, I noticed that I was able to load CoH without crashing, which is good. The bad side is that it took almost 30 full minutes, where it only took about three minutes before. Idea's? I then went back and looked in the device manager (after spending a whole lot of time researching the jmicron issues) and noticed that "Standard Dual Channel PCI IDE Controller" is the generic Windows XP driver. I updated it to the JMicron JMB36X driver 1.17.3.2.

I will try this out for a bit, but may move to the newer 1.17.4.0 beta driver next. http://www.station-drivers.com/page/jmicron.htm

It can still be hardware problems. Have you tried a better power supply than the Antec TP3-650??

The Antec TP3-650 had inherent design weakness for 3.3V and 5V which internal and external IO devices relied upon.

The design weakness is shown HERE (http://www.antec.com/specs/TP3_650_spe.html).

[Das Capitolin]

Thanks Nam-ng. I have swapped the TP-650 with an Ultra X-Finity 550W, but the same results were seen at the time.

I reviewed the TP-650 for Bjorn3d a few months back, and found that it moved very little if anything at all when comparing no-load to idle, and idle to fully loaded. I even use this PSU for a few servers with no complaint.

Thanks again. So far the JMicron IDE settings seemed to be the culprit, and I really am not alone, since the web is cluttered with JMicron complaints describing my very same issue.

[nam-ng]

Thanks Nam-ng. I have swapped the TP-650 with an Ultra X-Finity 550W, but the same results were seen at the time.

I reviewed the TP-650 for Bjorn3d a few months back, and found that it moved very little if anything at all when comparing no-load to idle, and idle to fully loaded. I even use this PSU for a few servers with no complaint.

Thanks again. So far the JMicron IDE settings seemed to be the culprit, and I really am not alone, since the web is cluttered with JMicron complaints describing my very same issue.

Amazing... Most people including so called "experts" had no idea that 3.3V, 5V, and 12V1 outputs are directly inter-dependant to test for the inherent weakness.

[DragonMaster]

Heh, drivers...

Looks like you had a similar problem than me when I got my Adaptec 1205SA SATA adapter last year, I had to get the Silicon Image drivers and BIOS to get the card working, because Adaptec never updated them and they caused corruption on the HDD.