Overheating Issues(I think...)

[nighthex]

I don't know if this belongs in this particular section but I'm not sure what kind of problem I am having, so sorry if this topic is out of place.

I have a custom built computer that I bought from a friend. It has a AMD Duron 1.8 Applebred and a pretty powerful heatsink and fan with arctic silver.
I started getting strange crashes a few months after I bought the machine when I played World of Warcraft. My first solution was to reinstall Windows. Which worked for a few more months and then the crashes came back. Then I replaced my heatsink with a new one and bought some arctic silver. I had it overclocked and it worked for around two months. Then I underclocked it and it worked for another month. Now am at this point where I don't know what's causing the crashes. At full load my CPU temp never goes over 53-54c but when I'm playing any sort of game it can spike up to around 61c for no apparent reason. Whatever the problem is it getting worse. When I couldn't play World of Warcraft I'd play less hardware heavy games like Warcraft III and such. But after a while I'd crash while playing those games too. The only solution I have is replacing the CPU. I've read customer reviews on Newegg.com that say my particular CPU isn't made for gaming, which would explain my problem, but a friend said I should post my problem on this forum. Any second opinion would be greatly appreciated. Thanks.

 

[OI-Beowulf]

It is certainly an intriging issue. I have a couple questions first and I'll post a couple theories below.

At full load my CPU temp never goes over 53-54c but when I'm playing any sort of game it can spike up to around 61c for no apparent reason.

How did you test this?
If you were running something like prime95 to load down the CPU, this doesn't completely load down the whole system.
World of Warcraft (and other games) stress many components in the system and thus can cause a higher amount of heat in the system to be generated.

How are the fans situated in your PC?
There is a significant jump between 53C and 61C under different load conditions and something you might try checking is the way the fans in the case are moving air. It might also help to know what kind of case you have.

How strong is your power supply and what brand is it?
What kind of video card do you have? (GPU model)
Does it use an auxilary power connector?

What exactly does the PC do when you say it "crashes"?
Is there a blue screen?
Does it simply reboot? or...
Does the game or application simply lock up completely? Sound may or may not get stuck in short loop.

These are the bits I'd look at first:
Fans in the case. Idealy, there should be at least 2 case fans not counting the power supply. One down by the front should be fairly powerful fan pulling air front the front and blowing into the case. (Fans have arrows for rotation and airflow on the side, you might have to remove it to check.)

There should also be a fan on the back of the case blowing out but it need not be as strong as the front fan. In fact, if it is rotating slower, that is better because it prevents creating a small vacuum in the case above the CPU.

If the machine is stable otherwise, and plays World of Warcraft, I don't think there is anything "wrong" with your CPU and I don't think the problem will go away if you replace it. The Duron was not the quickest gaming processor but it is no less capable of running games than any other processor. The worst you can expect are performance issues, but let's get to the bottom of stability first.

From what you posted, it sounds like there is either an overheat situation affecting other parts of the system (61C is high for the CPU but not crash-prone.) If the system is not overheating, I suspect it might be a power supply issue. Check the fans though, that doesn't cost much... and post as many answers to the other questions as you can.

Thanks and good luck!

 

[nighthex]

Hello Beowulf and thank you for trying to help me, I apologize if this reply is a bit late but the website was having some weird errors for me when I tried to reply earlier.

I didn't really officially 'test' my CPU load, I only made an assumption. I just minimized World of Warcraft when I had my CPU temp program running and got the reading of 53-54c. Even if that isn't an accurate full load reading it's as much strain I'll ever put onto my PC. And I've never heard of prime95 or anything similar. I just made assumptions.

I have 4 fans in my pc. I have a CPU fan(blowing air straight out of a vent in the side of the case), a heatsink fan, and two others. One of my extra fans is at the back of the case below the heatsink fan, it blows outward away from the CPU heatsink. Then I have another fan that is in the lower front of the case blowing air to the back of the case into the hard drive.

Edit: I do believe that the two fans that I have are the same model and therefore same speed.

I have a Rosewill 400w power supply.
I have a ATI Radeon 9600 All-in-wonder(no fan or heatsink from what I can tell).
The graphics card is just plugged into the AGP slot, there is no external wires coming from the card.

When my PC crashes the game I have up(these crashes have only occured during games) will lock-up completely. I can't Alt-Tab, Ctrl-Alt-Del or anything of that sort. The sound also stops and turns to a high-pitched droning which may be a short loop.

I think that answers all your questions. Please share with me any theories you have.

 

[OI-Beowulf]

From experience, I'm going to call WoW and accurate enough test of CPU load... :lol:

One thing you mentioned that jumps right out at me is the fan on the side of the case... which way is that pointing? (It should be blowing into the case.)

But let me see if I understand the configuration:
Fan on the back of the case, blowing air out (hopefully).
Fan on the front, near the bottom, pulling air in, blowing it over the hard drive.
Fan on the side panel... unknown wind direction. ??
Fan on the CPU, more than likely, this actually blows down over the CPU, not up.

Aside from the fan on the side panel, the configuration sounds right. I am suspecting there is a hardware issue at stake however but to be sure, check this:
Go to Start->Control Panel->Administrative Tools->Event Viewer, and check the "System" log by clicking on it in the left-hand column. Look for any Red Circles with "X"s in them that say "Error". (Almost all systems have something that windows has reported, don't be alarmed if there are some there.) If you double-click on the error, it will bring up some details. Look for errors that deal with WoW (in some form) or that occured as close to the time of the last crash. (There really shouldn't be too many to sift through.)

I am not ruling out a power supply issue however... The symptoms line up and that could be the cause. Perhaps a friend might let you borrow a power supply to test? You could also purchase one from a place that allows returns. (Best Buy, Circuit City, etc.)
Under load, your system consumes more power and stresses the power supply by as much as two times or more what it deals with when your PC is idle, or browsing the web. I'm assuming you haven't had too many lockups outside of gaming?

PCs that hardlock in applications like that with little or no warning are a pain to diagnose but the steps to try almost always follow the same order:
1) Fans, cooling - adequate? (I think you are fine in this case. Those temps are not out of the ordinary, warm but not too bad.)
2) Power Supply - weak? poor construction? (Unfortuately, these are the most overlooked cause of many grieves. Thankfully, they aren't too expensive.)
3) RAM - Faulty RAM can also cause a miyrad of problems including lockups but I'm not sure this is your issue without testing the other pieces first. RAM tends to cause blue screens and many other things in addition to lock-ups.
4) Mainboard - As components age, the parts under the most stress tend too show signs of wear first. Power Supplies, Motherboards, they are up there with hard drives minus the predictable mechanical failure. Check the motherboard for crusty yellow and brown substances oozing from the capacitors around the board. (They usually look like little coke cans) This is an indication that the board may no longer be stable enough to run the PC. But notice, it is number 4... check the first three, first.

Let me know about the fan when you get a chance and if it is possible to test a different power supply. I'm curious to know how things go.

 

[nighthex]

First, I think you misunderstood my fan situation. I only have two standalone fans(Fans not built into anything else) one in the bottom front blowing air to the back of the case and one in the back of the case blowing air out of the case. So air is being sucked in from the front and side of the case and blown out of the back. Then I have my CPU heatsink and fan blowing air out of the side of the case and finally the power supply fan.

Second, I'll see what I can do about getting a new power supply or borrowing one to test. After I'm more confident that a power supply is my problem.

Third, This RAM is no older than the computer and I doubt it's damaged in any way, again I really have no way of testing it.

Fourth, I looked at my motherboard but didn't find anything since this computer is only about two and a half years old and was never used for anything high-stress until I bought it.

Also I looked at my event viewer and found quite a lot of errors(Almost all of which the source was either Service Control Manager and sptd) but I really didn't find anything that said World of Warcraft but all the errors labeled under sptd say "Driver detected an internal error in its data structures for ." when I open them it doen't say what it found an error in but I think that's pretty important.

You theory seems to make sense except that it doesn't explain how crashes happen more often and during games that don't put the CPU under full load(at least I think.) Could this be related to the GPU overheating or anything else?

 

[Doughbuy]

Driver update for your video card? What card do you have? If you have an ATI card, you can use ATI tool to check around your card settings, use RivaTuner with a NVidia card...

I doubt its anything with the CPU, graphics card might be suspect...

 

[Mondoman]

...Then I have my CPU heatsink and fan blowing air out of the side of the case...

If true, this is EXTREMELY unusual. I have never heard of such a hsf. How do you know your hsf is sucking air up and away from the MB and blowing it out the side of the case? The standard setup is to have a vent on the side of the case, perhaps with a shroud/tube on the inside, with the opening above the hsf area. The CPU hsf then blows air down onto the heatsink. The vent/shroud is intended to supply outside (hopefully cooler) air to the vicinity of the hsf for better cooling of the hs.

You said you had replaced the CPU hsf. What brand and model do you have in there now?

Second, I'll see what I can do about getting a new power supply or borrowing one to test. After I'm more confident that a power supply is my problem.

What model PS do you have (there are many models of Rosewills)?

It sounds like an overheating situation to me. Another possibility is a component slowly going bad. The power supply could be going bad. Also, since you've been overclocking the CPU, that could be going bad. It could also be the MB going bad (capacitors especially). What brand and model MB do you have?

 

[nighthex]

:lol: My bad I meant to say that my CPU fan sucks air from the side vent onto the heatsink. Oops!

Anyway here's my heatsink

My power supply model number is: LC-B400ATX

And all I know about my motherboard is that on the manual I have of it it says A7V600 and it's made by ASUS.

Doughboy:

I doubt its anything with the CPU, graphics card might be suspect...

I have an ATI and I did update the drivers and tweak around with Catalyst with no results.

mpilchfamily:

If you want to rule out the RAM then download memtest86 (google it). This will make you a bootable floppy tha will test the RAM for you. If any errors pop up durring the test then you know it's the RAM.
Another posibility is the HDD may be going bad. There may be bad sectors forming. Run chkdsk and see if it finds any problems.

Did both, no errors.

 

[shadowduck]

My Athlon XP ran that hot. Sounds about normal really. Odd you are getting crashes. I don't tihnk heat is your problem at 61C.

 

[dougie_boy]

My Athlon XP ran that hot. Sounds about normal really. Odd you are getting crashes. I don't tihnk heat is your problem at 61C.

second that op.

id def go with the mem test at the moment.
also a few questions of my own...

how much ram do you have??
how long has that install of windows been on the PC?
do you have all the correct drivers of the VGA card/Sound?
do you scan reg for spyware??
do u use a firewall/internet security program??

 

[nighthex]

I did a memtest and it passed twice.

I have 512 MB of RAM.
I last reinstalled Windows a few months ago.
I just updated them.
Yes.
Yes, but I disabled them all.

Edit: I think something everyone is misunderstanding here is that my CPU temps are only 61c when I check my temps right after a crash. Normally the temps I see are around 52-53c. For whatever reason my CPU temp spikes up at the time of a crash. I don't know what could be wrong, but I'm considering replacing the CPU.