Problems with A7N8X-VM/400

[Polishwonder74]

Ok, so I just installed this neat little A7N8X-VM/400 into an Antec Aria case (pretty tight fit), I've got Windows XP SP1 going, all drivers working properly except for USB, temperatures seem not too high (49 - 51C idle). Everything seems to be doing relatively kosher here, but when I try to copy large folders onto here from another computer on my network, (1GB or more), it tends to hang and then suddenly restart. I've got the latest BIOS installed, and my system specs are:

AthlonXP 2800 Barton
2 X 256 MB DDR400 Corsair Value Select (dual-channel)
80GB Western Digital WD800 (8mb cache, 7200 rpm)
Onboard video, onboard sound.

Can anyone think of why my new rig is doing this?

Thanks for reading! :beer:

 

[LiLithTecH]

What error(s) are you getting in the EVENT LOG?

 

[Smilin]

What nic driver and version are you using?

 

[FlyingPenguin]

First of all right click My Computer, click on Properties, Advanced tab, and click the Settings button under "Startup and Recovery". Under "System Failure" uncheck "Automatically Restart". This will prevent Windows from automatically rebooting upon a serious error, and should allow you to actually see an error message caused when this problem occurs. This error message may give you more of a clue as to what's going on (leaving this setting enabled by default is one of Microsoft's dumber ideas).

If it's not a driver issue (make sure you download the latest driver for the NIC) then it's almost certainly a PCI IRQ Channel Sharing issue. CLASSIC symptoms. NOTE: This is NOT the same as an IRQ conflict, there are four PCI IRQ channels labelled A - D and all your PCI and AGP devices must share them.

This problem is very common with NICs, some PCI sound cards (the SBLive in particular is VERY prone to it) and SCSI controller cards. Some of these cards do NOT like to share a PCI IRQ channel I explain the problem in detail here with a fix: http://soldcentralfl.com/quakecoop/glfaq5.htm#5_9

I am assuming this is the onboard NIC? If so then some other card is causing the problem. Make a note of which slot each card is currently installed in right now. Pull ALL your PCI cards except the video card and boot up. Make sure the system is working okay (sounds like you can easily reproduce the problem).

If the system seems stable shut down and install ONE CARD AT A TIME. If it acts up, move the card to a different slot until it's resolved.

Repeat this process with all the PCI cards starting with the NIC card next if there is one.

If the NIC is an add-on card then move the NIC to a different slot than it's in now.

Hope this helps...

 

[Polishwonder74]

Oh, I hadn't thought of that (doh!!). I have a couple spare NIC's laying around (nice 3Com ones). I'll throw one of those in there and see if it makes a difference.

I'll report back to y'all ASAP. Thanks again!!

 

[Polishwonder74]

Ok, I put in a PCI network card, and I disabled the onboard NIC (caused an even tighter fit inside this little case). I'm getting restarts every now and then, but they're seeming to be more random. I notice some graphic artifacts following the mouse around sometimes, I am thinking of trying to software underclock the video to see if it helps any. Any other ideas?

 

[Polishwonder74]

AHA! I just got a very big clue. Ok, I'm using the Omega Nvidia drivers, and I just tried running 3DMark '03 and crashed her right away. NVdispl.dll I think.

So far as I can tell these are the latest drivers. Usually Omega drivers work better than the normal ones. What do you guys think?

 

[imported_Phil]

Use the standard nVIDIA drivers and report back.

Also, you don't mention your PSU - I'm going to assume that it's a generic model? In this day and age, a decent PSU is pretty much a requisite of a stable and reliable system.

 

[FlyingPenguin]

Definately switch to the standard nVidia drivers. Omega usually work okay, but they ARE hacked drivers.

I still feel it's a PCI IRQ sharing issue. The NIC itelf is not the cause, but some other device that was sharing the same PCI channel with the NIC (and probably something else) is to blame. Here's my link on troubleshooting this again: http://soldcentralfl.com/quakecoop/glfaq5.htm#5_9

If there's any other cards on this system you should pull them and see if that fixes the problem, then try to narrow down which one is being a problem.

 

[Creston]

I agree it's most likely a PCI IRQ sharing problem, but I'd like to throw a second option in there for overheating.

You say everything's a very tight fit, which means you probably have close to zero airflow through your case. Try the same procedure but with the side of your case removed, see how it goes then.

The randomness of it is what leads me to believe heat. If your CPU gets too hot (and it does quite a bit of work during network traffic), it's going to shut down.

49-51C idle seems pretty high to me. If your motherboard is set to shutting your CPU down at 60C, it's VERY likely this is your issue.

Creston

 

[Polishwonder74]

I will try using the normal NVidia drivers and see how it goes. I've put in a new NIC and disabled the onboard one, and I still got the crash. I'll try running her with the sides off, also. She's got a 300 Watt PS with a 120mm fan, and I also installed the indluded PCI slot exhaust fan. Thanks for the suggestions, I'll report back tonight.

 

[Fern]

49-51C idle seems pretty high to me

I agree with the above poster. Some, not all, XP chips will get unstable in 60c range. If you're at 51c, you may be going over 60c under load.

I don't know if thats the cause here, but even it weren't I'd be afraid of problems down the road. Check your vcore, blow out the dust bunnies, increase airflow etc.

Fern

 

[Polishwonder74]

Ok, I just loaded her right up to 100% with SETI@Home, and she settles to about 55 degrees (CPU) and 44 degrees in the case. Not too bad I think for such a tiny case. Now I'm working on the graphics problem. I hope I don't end up grabbing an AGP card off of ebay, that'll just add more heat to my case. She crashes as soon as she finishes loading the 3DMark '03 benchmark.

Edit: I should note that the Omega drivers didn't crash 'till the 1st benchmark was fully loaded. The newest Detonater drivers crashed about halfway through loading.

 

[Polishwonder74]

Ok, I have more news. In a desparate attempt to get to the bottom of this, I crammed the Radeon 9800 from my main rig into this little box and fired her up. I was pretty worried that I'd go over the 300 Watt PS capacity with that big-ass 9800, but I crossed my fingers, hit the power button, and she worked. I was able to run 3DMark '03 with no problems. I guess it's just a lousy onboard GPU. Everything works great until I try to do some 3D stuff.

I guess I'm just gonna end up getting a real AGP card for this rig. Damn, I was trying to save money by going for this mobo with onboard everything. I'll free up 32 MB of RAM by getting a separate card, so that'll be cool. Ebay, here I come. . .

Also, what is up with the BIOS on this motherboard? I went for this one 'cause I ASSumed that it would have the same killer BIOS that's on my A7N8X Deluxe. What a dirty trick Asus pulled on me! :|

 

[FlyingPenguin]

Just so you know, it might not be a flaky GPU but sensitivity to a flaky PSU as well.

All PSUs are not created equal. Cheap PSU = poor voltage regulation. Even a good PSU will have voltage regulation problems if it's not getting adequate cooling. Some hardware is more sensitive than others.

The cramp case and warmer temps, plus a PSU that may not be a high quality one (dunno what you have in there) could be resulting in some power fluctuations that the other card doesn't like, but the 9800 is handling better.

I'd test that other card in another rig first.

 

[Polishwonder74]

The other card is an onboard GPU, I can't really test it anywhere else. The power supply is an Antec (it's an Antec case).

 

[piasabird]

I dont think this particular motherboard has that good a reputation as a stable motherboard. I have seen some people have problems with it. You might just want to RMA the motherboard as fast as possible. Generally I use Asus Motherboards, but this one has had quite a few complaints.

 

[Polishwonder74]

Ok, I've got more bad news. I went ahead and started using this computer for work, and it's just getting worse and worse. I'm getting BSOD's for a lot more than just nv4disp.dll now. I'm getting errors writing to win32k, dxg, I've seen a couple RAM addresses come up, and a couple others that I'm not sure exactly what they are. When I first put this rig together, she ran non-stop for like 2 or 3 days without crashing, now she generally can't sit idle for more than a half an hour without spontaneously crashing (faster if I'm actually working). She crashes especially quickly when using MS Word XP. I might have to RMA this board to Asus. Do you fellas think there's a chance I could just get my money back from them, or am I gonna be stuck with another lemon A7N8X-VM/400?

 

[Polishwonder74]

OH SNAP!! Ok, I just read a review for this mobo at Newegg where a guy disabled video RAM pre-charge, so I tried it and I haven't crashed yet! It's been about 20 minutes, and I'm using MS Word, Photoshop and Firefox. I was also able to start up and run Futuremark '03 for the first time EVER with this onboard GPU!!

I've got my fingers crossed, this could be the answer to my prayers!

 

[yelo333]

Perhaps the integrated video is pushing the ram just a bit too hard...After all, the integtated video is the only part of the nforce2/AXP setup which really needs the dual channel bandwith. Have you tried relaxing ram timings/underclocking the FSB a bit? Also, at 40+ degrees C, your case temps are up there...Coupled with the passive heatsink on the Northbridge, perhaps temps are just getting way too hot? Have you tried touching the NB heatsink after it's on for awhile, and seeing if it's very warm?

Edit: Warm as in so hot it's uncomfortable to to touch.

 

[Polishwonder74]

I would have tried underclocking, but this BIOS is awful, I don't think I can. Why on EARTH would they use a TOTALLY different BIOS for this board than for the normal, full-sized A7N8X??

 

[Polishwonder74]

Looks like she's all set. I ran her all day, moderately heavy use, transferred a ton of files, Photoshop, Dreamweaver, Imageready, WS_FTP, Foobar and Firefox all at the same time, and she didn't crash once.

I guess the sun even shines on the dog's ass sometimes.

Thanks for the suggestions, fellas!!