Critical error sound, no error message

[typpeo]

Hi All,
I've been having some issues with my PC for the last couple weeks now. I've been noticing the problems mostly when playing Eve or tf2. I thought it might be temps since it's warm out so I monitored them but they looked to be stable. I then thought it might have been my nvidia driver update and rolled those back with no success. Every now and then while playing a game I get the critical error sound but with no message and nothing in event viewer.

Eve was having problems in windowed mode, the FPS dropped to like 6fps, so I turned it off and in fullscreen fps went up to 60 something. I use to run in windowed mode all the time with no problems. Also, during certain times in the game when on vent the audio cracks really bad and the game is jerky. tf2 has been crashing every now and then and it use to run great before.

I've checked spybot and that had one virus but I cleaned that out. Besides that I'm not sure what it is or what else I can test. Maybe I should run a stress test and see if anything comes out of that. Any help would be great, thanks.

Here are my system stats:
XP Sp3
Intel core2d 1.86 (overclocked to 2.45)
2x1gb ram
geforce 8800 gts (640mb)
Soundmax onboard sound

[Digerati]

Whenever troubleshooting stability problems, I recommend returning all clock speeds to the defaults, and don't OC until problem resolved. Sudden shutdowns with no entries in the Event Logs means the shutdown was indeed, sudden. This typically means hardware, and yes heat is often the cause. I would open the case and make sure it is clean of heat trapping dust. Make sure all fans work fine. Blast a desk fan in there while testing. Inspect for bulging or leaky capacitors. Ensure power supply is good. You can test RAM using one of the following programs. Both require you to create and boot to a bootable floppy disk or CD to run the diagnostics. Using the floppy method is generally easier (and another reason to include floppy drives in new builds). However, the CD method is just as effective at detecting RAM problems. Allow the diagnostics to run for several passes or even overnight. You should have no reported errors.

Windows Memory Diagnostic - see the easy to follow instructions under Quick Start Information.
or
MemTest86+ (for more advanced users) - an excellent how-to guide is available here.

Alternatively, you could install a single RAM module and try running with that to see if it fails. Repeat process with remaining modules, hopefully identifying the bad stick through a process of elimination.

[typpeo]

Originally Posted by Digerati

Whenever troubleshooting stability problems, I recommend returning all clock speeds to the defaults, and don't OC until problem resolved. Sudden shutdowns with no entries in the Event Logs means the shutdown was indeed, sudden. This typically means hardware, and yes heat is often the cause. I would open the case and make sure it is clean of heat trapping dust. Make sure all fans work fine. Blast a desk fan in there while testing. Inspect for bulging or leaky capacitors. Ensure power supply is good. You can test RAM using one of the following programs. Both require you to create and boot to a bootable floppy disk or CD to run the diagnostics. Using the floppy method is generally easier (and another reason to include floppy drives in new builds). However, the CD method is just as effective at detecting RAM problems. Allow the diagnostics to run for several passes or even overnight. You should have no reported errors.

Windows Memory Diagnostic - see the easy to follow instructions under Quick Start Information.
or
MemTest86+ (for more advanced users) - an excellent how-to guide is available here.

Alternatively, you could install a single RAM module and try running with that to see if it fails. Repeat process with remaining modules, hopefully identifying the bad stick through a process of elimination.

Thanks Digerati,

I ran memtest for about 8 hours with no errors and 21 passes. So it seems like my ram is stable, it is still overclocked so I will try and set the settings back and see if that makes a difference. Anything else I could try?

Thanks

[Digerati]

Anything else I could try?

Yes.

Originally Posted by Digerati

This typically means hardware, and yes heat is often the cause. I would open the case and make sure it is clean of heat trapping dust. Make sure all fans work fine. Blast a desk fan in there while testing. Inspect for bulging or leaky capacitors. Ensure power supply is good.

[typpeo]

Originally Posted by Digerati

Yes.

The cover has been off for about a week now, it's a bit dusty. I will give it a good cleaning and double check all connections. All fans seem to be working fine.

After running memtest and restarting I played tf2 for a bit this morning and got the critical error sound. tf2 continued to play fine, no error message and nothing in event viewer. The critical error sound only seems to happen once and after a reboot.

thanks again.

[Digerati]

Running with the cover off is not always a good idea. Cases are designed to facilitate front to back air flow. With the side panel off, there is no real flow - especially across other heat sensitive devices. That's why I suggested running with a desk fan blasting in there when the side panel cover is removed.

[typpeo]

Well, I determined the problem was heat related. I wasn't running the AC yet and it seemed when it was getting hot out that that was when the problem was happening. I started running AC and a fan in the room and that seemed to work.

However, yesterday I played TF2 for a couple hours and then I had a really hard crash. The sound kept repeating and everything froze. I restarted my machine and the video card wouldn't display anything. I had to unplug the machine and then when I plugged it back it it was able to boot up. This again happened yesterday after only 30mins of use.

It wasn't as hot as it has been in the past when this happened. Anyway, what would be the best way to solved this problem? It seems to me that it's a video card over heating problem. Would it be best to just replace the card? I don't have problems with it being unable to run games at max settings but I'm not sure how else to solve this.

thanks

[Digerati]

I determined the problem was heat related. I wasn't running the AC yet

I feel your pain. I have central AC for the house but I had to add a window AC to my computer room because the two computers in there were keeping the room too hot. In the winter, I block the central heat ducts and the computers keep the room comfortable.

But back on point, I screwed up. I just noticed you are overclocking. DON'T! Always take your clock speeds back to normal when troubleshooting stability problems. Unless an expert at electronics heat management, don't overclock.

What are your temperatures? How are you monitoring them?
****

If still having problems once all clocks rates are back to normal, including the graphics card, and all fans are running normally, then it is the case's responsibility to remove heat from the case. So look there next. Does your case support more fans, and better yet, larger fans? Larger (120mm as opposed to commonly used 80mm) fans move massive amounts of air, but at a lower RPM so noise levels are kept down. Check your PSU to make sure you have necessary power connections. Do NOT overclock until this is completely resolved. And most importantly, keep the interior, to include fan blades, vents, and heatsinks, clean of heat trapping dust. I recommend every PC owner inspect monthly and clean as necessary.

If your case does not support more fans, then you have 3 choices,

1. Get a new case
2. Blast a desk fan into the open side when working the computer hard
3. Stop working the computer hard

I recommend a new case. A good case will keep your system cool, and protect it from accident kicks and knocks with the vacuum. And a good case will last years, through several complete upgrade cycles. A good case will offer at least two large (120mm or larger) fan options, typically one in back to exhaust hot air out, and one in front to draw cool air in, providing good front to back air flow. I like Antec cases - and I will NEVER buy another case that does not have a removable, washable air filter. If there are any IDE or floppy drives installed, replace any flat ribbon cables with round cables of the proper length - they come in various colors for easy tracing.

It seems to me that it's a video card over heating problem. Would it be best to just replace the card?

Were you overclocking the card too?

If your case is not expelling the hot air (or drawing in enough cool air), a new card is not likely to help - assuming you would get a card of equal or better specifications (read: BTU production). The geforce 8800 gts is a good card now. If you have excellent front to back air flow through the [clean] case now, I might consider one of the many 3rd party fans available for that card.

Is your PSU sized right, and working properly?

[typpeo]

Thanks Digerati.

I'll list my whole system here to make life easier.

Lian Li Case 2 Front fan 80mm fans, 1 top 80mm fan, and a rear 80mm fan. I really love this case, just wish the back could handle a larger fan.

Intel Core 2 Duo E6300 Over clocked to about 2.45. I will turn off the over clocking and reset my settings to the original settings. Been hesitant to do this, but I will give it a shot. I have a Thermalright HR-01 attached to it, this thing is a monster.

Asus P5B

EVGA 8800 GTS 640mb This is not over clocked, I never wanted to mess with it on my video card.

And OCZ Platinum Revision 2 2GB (2x1)

I forget exactly which PSU I bought, but I'm pretty sure it was 800watt pretty top of the line (at the time I bought it.)

I've been checking temps with speed fan and another program. All core temps seem fine, the video card was in the 70 degree area I believe when I was running games. But I haven't been able to see temps at the time that my machine crashes.

That's pretty much it, I've never really had any issues with the machine and was hoping to have it around for another year. (It's over 2 years old now.)

In any case, maybe I just need to stop being cheap and run the AC in that room whenever I decide to game for a while.

[Digerati]

All core temps seem fine

What does that mean? What are the temps? Pushing 60° for the E6300 would make me nervous.

The only complaints I have ever had or heard about Lian Li is their price. You have a fine case. If you have room, you may be able to put higher CFM 80mm fans in there - but that will increase the noise factor.

[typpeo]

Originally Posted by Digerati

What does that mean? What are the temps? Pushing 60° for the E6300 would make me nervous.

The only complaints I have ever had or heard about Lian Li is their price. You have a fine case. If you have room, you may be able to put higher CFM 80mm fans in there - but that will increase the noise factor.

High 30's for CPU and mainboard, 40s for the HDDs.


Yea, the Lian Li was expensive, but I like the look so it was worth it. Maybe I'll look into better 80mm fans and see if that helps as well. But first to turn off over clocking.

[Digerati]

But first to turn off over clocking.

Right.

[Drizzles]

Looks like yous are having fun here

Yes I'd agree with Digerati, I too play Eve and TF2 and had the same problem with my CPU OCd ... even only minimally and I'd have issues.