Help diagnosing a PC problem...

sektor

n00b
Joined
Jul 22, 2006
Messages
35
Ok...not exactly sure what is going on, so I was hoping to get some help here.

As of almost 2 weeks ago, during BF2142 gameplay, weird things have started to happen lately and I am trying to find out what the problem is.

Rig:

Gigabyte DS3 BIOS 7
E6600 NOT oc'd --Themalright CPU cooler, 120mm fan
2gigs DDR 800 memory, corsair
ATI x1900xt 512mb card -- Thermalright HR-03 VGA cooler with 90mm fan
Creative Xi-fi music sound card
Enhance PSU 500+ watts

Before two weeks ago, gameplay was flawless on 2142. I could play for hours with no problems.

Lately though, weird things have started to happen. A list of things that have happened.

-Game crashes to desktop for no apparent reason
-Game freezes up during gameplay...forced to hardboot.
-During gameplay, the game starts to stutter...then, it either freezes OR if I quickly quit and go to my desktop, it freezes there and I have to hard boot

Put on ATItool.
Side of the case is open currently.
Did a "Show 3D" test and temperatures were:

Idel GPU Temp: 41-43
Under load (From "Show 3D"): Highest it got was 75 C
Temperature of Temp chip: 51 C
Current Draw under load: 27.5A

At this point, I am not sure exactly what the problem is and am trying to narrow it down to what it could be.

I just got done playing about 40 minutes of Company of Heroes and it started to stutter, then it froze and locked my PC...therefore, I can rule out it being BF2142 and most likely, hardware problem.

I actually did a fresh install last week, to get everything 'clean' and 'tidy'.

Did a hard drive utility test from seagate and it came back with flying colors.

I am starting to think it is a heat issue. I don't think it is the CPU or the GPU, but could it be the mobo or a part of the motherboard? Maybe the Northbridge or something like that?

The trick is, narrowing it down...

I will say I have thought about replacing this particular motherboard (not a bad motherboard by any means) because it is not a full size (little smaller wide/tall) which is what I am used to...but it is a little bit smaller. I am wondering if the crampness is causing heat issues? Also, I only have two fan plugin slots on the motherboard. One for FAN, other for system. My other fans are plugged into the PSU via molex.

My theories are:

1.) My PSU is not powerful enough to power everything on my system and this is causing problems?

2.) A heat issue, in that the motherboard (possibly North bridge) is getting old and not cooled quick enough. Being that I have a large CPU cooler with 120mm fan and a HR-03 VGA cooler with 90mm fan, things are a little tight in there...hot air being trapped possibly?

Working on stressing my CPU.

Issue only happens when I game.

I appreciate the help.
 
A freind of mine had similar issues, and ruled it out to be potentially the PSU or the memory.

if you can, see if you can swap the power supply with something maybe a bit more beefier, ensuring you have dual 12V rails. name brand really helps as well.

In terms of the memory, if you can, test the memory. You can also further test by taking out one stick at a time by playing around with CoH, BF2142, etc.
 
It kind of sounds like you have a stick of ram going bad.

1.Try running the computer on one stick of ram if you have two. Try this with both of them.
2.Try different slots with the RAM.
3.Run Memtest overnight.
 
Going to do a quick test with memtest86...

RAM eh? I was thinking heat.

For my own knowledge, what are the indicators to lead you to believe a stick could be going bad.

This is corsair memory...doesn't it have a lifetime warranty?
 
Going to do a quick test with memtest86...

RAM eh? I was thinking heat.

For my own knowledge, what are the indicators to lead you to believe a stick could be going bad.

This is corsair memory...doesn't it have a lifetime warranty?

If that is what it says on the box or in the instructions.

Question: did you OC by any odd chance?

In order for you to RMA, as long as the memory is going bad NOT due to o/c'ing, you will have to call Corsair direct, and speak with a customer service represetnative.
 
Going to do a quick test with memtest86...

RAM eh? I was thinking heat.

For my own knowledge, what are the indicators to lead you to believe a stick could be going bad.

This is corsair memory...doesn't it have a lifetime warranty?

First of all, a quick test is useless, unless you get errors back immediately. Let it run for 24 hours minimum. I have seen it run fine for 18~20 hours straight, and then start getting errors out the ass.. let it run.

Crashing, locking everything you described can be caused by ram taking a shit... Not saying 100% for sure that's the problem, but it certainly can be, and is easy to isolate by running memtest for the correct amount of time... one or two passes doesn't mean anything...
 
No OC'ing of the RAM...normal use.

Alright...i'll start the process for a 24 hour burn and see what happens.

Any other things I can do to narrow down the process?

Thanks.
 
Ok....I have the memtest running on the computer right now...should I keep it going, or should I try taking out a stick and testing a single stick at a time?

Secondly, when doing the test, suggestions on the test type? Or should I just let the CD boot up and go to work?

thanks
 
Just let it run and see what happens. It could take awhile for an error to pop up so be patient.
 
Just let it run and see what happens. It could take awhile for an error to pop up so be patient.


Will do.

Lets say this comes back clean,with no errors (after 24 hour run period), what would that narrow it down to? Heat? Power?
 
I would check Power if you can go 24 hours with no errors on memtest.
 
UPDATE:

Ok...made progress here. Taking some suggestions above, I took out a stick of memory and did some testing.

I took out one stick, fired up BF2142 and besides long load times, some choppiness, the game froze up in 5 minutes. Freeze of the computer, hard boot down. Swapped out the sticks.

Funny thing. I swapped the sticks out and played BF2142 for the last hour. :eek: Was some choppiness (i did not turn the settings down...oh well...wanted to stress the crap out of the memory).

Anyways, I kept waiting for the system to freeze and it didn't. So I kept playing. Finally, decided to stop for the night and try it again tomorrow.

So it really does look like I have a bad stick.
I did run the memtest86 test for about 3 hours before I decided to play around, but no errors. However, that doesn't mean the sticks are good though...a full 24 hour test would most likely show errors like posted above..

Anywa, my plan:

Test again tomorrow, see what happens.


Question:

1.) Is it possible that the DIMM holder closest to the CPU slot could get hotter than the others, possibly causing a problem with that slot and stick of RAM? I ask because the stick that appears to be dying was in the slot closest to the CPU. The stick that appears to be working, I stuck in the farthest slot away from the CPU. Just thought i'd ask.

2.) Is their a motherboard that possibly spaces out the memory slots a bit farther away from the CPU unit than others? Or are they mostly universally the same distance from the CPU?

Well, been a frustrating day, but I feel like I have made progress.

Next steps?

Thanks for the help guys. I really thought it was a heat issue.

Cheers.
 
Looks like you found your culprit. Give corsair a call and get the RMA rolling.(Great Cu stormer service btw)

1. The Stick being closer to the memory should have nothing to do with the life of the DIMM.
2. All Mobos are different there are difference but I don't think that it really matters. If you are worried about it dedicate a fan to blow over your ram sticks. On my mobo the heat spreaders all but touch each other and I have had no problems the year I've had my rig.
 
#)*$&(*&)(*@&%)(#@*_)!*#($*&!*)!)~!!!!!


Did further testing last night...while using the so called "good stick", PC locked up on me twice last night. :confused:

Going to do additional testing tonight. Corsair wants me to change some voltages, run memtest for 3 passes (which really does not do anything...always passes).

any other suggestions?
Possible RAM is not being cooled enough?
Possible bad motherboard?

THanks.
 
Ok...hoping to get some feedback on ideas I have.

Ran orthos and ATITool for over 3 hours....zero problems, temps fine.

I decided to pull my Xi-Fi sound card out and test somethings.
Played a few games over an hour. ZERO problems.

Two things I am thinking of.

1.) Sound card died and is the problem. I found out creative does not always have good drivers.

2.) Possible sound card is pulling lots of juice from my PSU unit, causing the problem?

Thanks.

Sektor
 
Back
Top