Random freezing problem :|

[HungryCakes]

I have a fairly decent system
Athlon 64 X2 4400+
MSI K8N Neo4 Platinum
2 sticks of Corsair XMS3200 1024MB RAM
Western Digital Caviar 320GB
BENQ LightScribe DW1655
SoundBlaster X-Fi Xtreme Audio
Antec Phantom 500w PSU
XFX GeForce 6800 XT

So the crashes go like this:
Every 45min-2 hours of use, the screen freezes, and I have to hard restart. Sometimes it restarts by itself, sometimes it doesn't. Sometimes there's a ?Windows has recovered from a serious error? message, sometimes there isn't. There tends to be a high pitched tone when iTunes has focus when it crashes, but this happens with no other application.
With more intensive applications (Photoshop, CS:S, most games), it freezes in 15-20 minutes. World of Warcraft is the exception. Strangely, OpenOffice is included in this category.
This tends to happen most frequently when I go afk, after 5-10 minutes of inactivity, it freezes. If I leave the computer on overnight, I wake up to a screen frozen with a view of something like 12:24 AM.

I generally torrent while I use the computer, upload capped at 80 kB/s, download ranging from 50-250 kB/s, dunno if this is a factor, might be with the harddrive.

I did a system format about 3 weeks ago, it helped nothing. 15 minutes into my new install ? froze.
I found my northbridge fan was not working about 4 weeks ago, I bought a replacement. Perhaps this is a symptom of a motherboard problem?
I'm using a Zalman CNPS9500 CPU fan, so I don't think processor cooling is an issue.
The crashes seem to be getting more and more frequent, possibly showing hardware degradation.
It can ? and does ? freeze sometimes 20 seconds upon booting. It crashed 3 times while writing this post, I decided to write it in OpenOffice.

What I want to know is this:
What is the MOST LIKELY CULPRIT? I am not averse to spending money on hardware, as I doubt this is a software issue.
What is the MOST PRUDENT course of action? I'd like to spend as little as possible on unnecessary upgrades.
I am willing to answer ANY questions, and will monitor this post closely.

Thanks for any help.

 

[TD77077]

Does sound like a hardware problem. I'd suspect RAM first. Could be timing - set too agressively or OCed? Or could be one stick not so good. Try running with just one stick at a time to test. Also check compatibility for MB and voltage settings. You could try increasing the voltage slightly.

Also check CPU temps after some use to make sure it's not overheating - the heatsink might not be seated well or the thermal conductivity might be bad. Also check PS voltages to make sure they are OK. Your PS should be sufficient but could have a prob.

Good luck and let us know what you find.

 

[HungryCakes]

Temps range from 30-40 C at most, I doubt it's the heatsink.

RAM could be the problem, I have 4 sticks but I'm only using 2. I've been switching them around, and it SEEMS to slightly affect the freezes, but that's purely anecdotal and not indicative of anything. My voltage is 2.40 to the RAM, that seems to have slightly improved stability. But again, that's anecdotal.

How can I check my PS voltages? What would be acceptable values?

 

[NHHokie]

One tip that might help figure the issue is to go to "Properties" of My Computers. Choose Advanced and then in the Start Up and Recovery section click on Settings. Under System Failure UNcheck the "Automatic Restart" box.

Then you will be able to read the error message on the blue screen. This allowed me to diagnose I had a consistent Page Fault error once. If you leave Automatic Restart checked, you never get to see the specific error message, even if you go to the Event Viewer after restart.

I agree it sounds like a hardware problem, but these can be the trickiest to resolve. Do you have the latest BIOS?

Best of luck

 

[HungryCakes]

Well, sometimes an error message doesn't show up, it's really a crapshoot when it comes down to it.

I'm not sure about my BIOS, and I'm a little apprehensive about flashing it, but I might give that a try.

 

[Xsorovan]

I was with TD77077 on the RAM issue, but if you've been moving those around and it doesn't fix anything I'm going to say that maybe your northbridge is damaged. If you noticed the fan was out who knows how long it was out, and if it might have gotten some heat damage during a marathon session of WoW.

I know of no simple way to test if the NorthBridge is burned out or not. If I remember correctly though that is where the memory controller resides. (subject to correction) So that could be the reason it looks like you're getting memory errors. The download site for your motherboard (http://www.msi.com.tw/program/support/s...swr/spt_swr_detail.php?UID=637&kind=1) includes a download of NTune which is NVidia's motherboard tweaker/ monitor. You might want to install that and run some diagnostic stuff and see what, if anything, you come up with.

Another suggestion would be to find a program that will "burn test" / "burn in" your various parts (CPU and RAM) to see which one locks up.

 

[HungryCakes]

Northbridge and/or memory controller damage seems likely, I doubt all 4 sticks of relatively high quality RAM all went bad, but the only way to fix that would be a new mobo, right?

That's a relatively cheap purchase, $150, right?

 

[Xsorovan]

Froogle-ing your mobo brings up anything from 80 something dollars to 200 dollars. I'm willing to bet it is the northbridge. (If it's got a fan on it, it's for a perfectly good reason.)

But yeah, just buy a new motherboard (the same one or another one, up to you) and put the old stuff on it and if that doesn't fix it then, well, I think you have another problem. In which case you can return the motherboard, get your money back and you can find out which part it is. But I seriously think it's the motherboard.

 

[HungryCakes]

I'm not sure I can return it if it's not compatible, but it seems like a promising avenue. Wish me luck!

 

[Xsorovan]

Well, make sure it's a store with a good return policy. "I no longer need it/ want it." should be a good enough answer for anyone. (Coupled with a receipt of course...)

Best of luck!

 

[Zelvek]

Be sure to buy he same motherboard as your current one or you may have to do a outright reinstall of windows. If you can't get the same one then buy one with the same chipset.

 

[HungryCakes]

Would this: http://www.msicomputer.com/product/p_sp...p?model=K8N_Neo4_Platinum/SLI&class=mb
likely require me to do a Windows reinstall? It's basically the same motherboard, but with SLI capabilities.

 

[DigitalCancer]

what about video card? I noticed no one mentioned that and you didn't mention which one you had. My wifes is starting to freeze after she's played Civ4 *snicker* for an hour or so, but i believe its the vid card getting ready to go. Its really dusty as well though so i need to take a can of AIR to it soon.

 

[WildHorse]

I'd say most likely culprits, in descending order, are:
vcore too low or too high
vdimm too low
ram timings off

The crashes seem to be getting more and more frequent, possibly showing hardware degradation.

...but probably not. Don't worry.

Suggest you set everything in bios to optimized defaults, then read up on what other guys with similar MSI rigs are using for bios settings.

 

[HungryCakes]

Right, I have a XFX GeForce 6800 XT, I'll edit my first post and put that in too.

I'll try setting the BIOS to optimized defaults, but I think most of the settings are already there.

 

[erickj92]

This may sound stupid but it may just work... Try getting a different version of your sound card driver......

 

[erickj92]

Try it, it could work

 

[python023]

Any replacement parts you could use to narrow down the culprit?
Another video card, compatible RAM, or mobo?
Possible even a PsU even though yours seems more than enough.

 

[GalvanizedYankee]

Given what has been said here, my money is on the PSU. Phantom? :thumbsdown:
Try your spare PSU and see how long it runs. Then buy a Corsair 520HX or a Silverstone 500W Element.
Never buy another Antec PSU. RMA the one you got if it has warranty left...Then sell it.

Your "serious error" is a voltage fluctuation that happens very quickly.