I went on vacation and made sure my computer was off for a week. When I arrived home my computer was on but there was no video. Long story short, there was a bad storm, the box wasn't plugged to a surge protector and apparently a surge knocked out alot of my components.
Since then I've replaced my psu (Ulta 550w), mobo (ASUS A8V-XE), RAM and bought a heavy duty UPS so that I don't have to go through this crap again. Well my computer now boots up fine, but would randomly restart, and once it did, it would continue to restart before it could get through POST, which lead me to believe that something overheated and was still hot. I took temperature readings of my box and the CPU would run idle at around 52 degrees and while playing a game it would increase to 60 degrees before I killed it for fear of stuff getting to hot and ruined. I took the side of my case off and put 2 big fans aiming directly into my box for more cooling, which got me by.
Last night I decided to try and fix my heat issue and took off the heatsink, cleaned off the thermal gel and reapplied it. Sure enough my box runs at a cool 45 degrees (cpu that is) idle and 52 under load. My video card runs around 51 degrees idle (NVIDIA's temperature monitor utility). So my cpu no longer is having a heat issue, and the SpeedFan readings say my HDDs aren't overheating (running at a cool 32 degrees). However my comp still restarts (particularly when I play a game)
I've been on my comp for 2 hours and it hasn't restarted, however I havn't booted up a game. As soon as I start up a game, though, my comp goes into restart mode.
your processor might be bad too. download prime95 and run 2 instances of it. select large in place tests (for max heat) and run that. Make sure you don't get any errors. Your computer is probably restarting when you load a game because that is stressing out the processor.
In order to run dual instances of prime95, put two shortcuts on the desktop and open the properties on the second one. for the target add [space]-A0
on the other shortcut add [space]-A1
make sure to leave a space before the -A0
If you get any errors at all, you will have to replace your processor. Let it run for at lease 6-8 hours. The longer the better. I have a feeling that you will get an error right away though if it restarts as soon as you start a game. Let me know.
At least you learned from your mistakes. I have always used a battery backup plugged into a surge protector. Last year, I was asleep when a power surge hit my apartment. It took out the surge protector but saved my computer. They are worth every penny. On a side note, I have heard that even a surger protector won't protect your computer against a lightning strike. unplugging is still the only way to prevent that.
Well now I’m completely confused. Immediately after I posted, I booted up a computer game (WoW if that matters) and was able to play for a solid 6-7 hours without my comp restarting. When I decide I’ve had enough, the computer reboots itself, but I just said I would deal with it in the morning and went to bed. The next day I boot up my comp and try and play again and I get restarts right when the loading bar is at 99%.
It’s interesting to note that I can stay on my comp for extended periods of time and listen to music, i-m, cruise the internet, etc without my comp restarting on me (so far, knock on wood), but the moment I try and play a game my box becomes unstable.
Jedi940 – I downloaded Prime95 but couldn’t really figure out how to run the program. I went to options/torture test and ran that and it gave me some results, my comp didn’t crash, but I wasn’t sure what to make of the numbers. Should I be running the torture test or the benchmark? Or some other setting in the program?
Mike99 – I have a NVIDIA GeForce 7900 GT/GTO 256MB video card. Before my computer bit it, I had an Ultra 550w psu and it worked like a champ, so that’s why I went with the same brand/wattage.
So since it restarts regularly at the loading screens, and then all of a sudden has these random times of taking pitty on me and then letting me play for several hours without incident, could my video card be on the fritz? Since the game is obviously using moure resources then if I were typing, I would think it would narrow it down to the processor or the video card, but I could be wrong.
For Prime95, you should be running the torture test. also, since you have a dual core cpu, you need to run two instances of it. if you only run one, it will only stress each core at 50%. The numbers don't really mean anything for you. What you need to be concerned about is how long it runs without errors. As long as it is running, you are fine. When it come to an error, it will say something like
"torture test ran for 10 minutes. 0 warnings 1 error, rounding error, answer was 5, expected to be less than four. "
also, usually only 1 instance will error. The other instance will continue to tun. This will allow you to stress your processor, heat it up, and see if it is ok. It is the program I used when overclocking to make sure my proc was stable. I ran it for 12 hours without errors to make sure mine was stable. If you cant run it on your system for more than an hour without an error, your proc is probably bad.
It has some info on Prime95 with screenshots and how to read it. Hope it will help. Farther down the page, it tells you how to run 2 instances of it simultaneously in case you didn't understand how I explained it. I may not have made it very easy to understand. Sorry if that is the case.
I'm still leaning to your processor being bad, but it is a little odd that you were able to play WoW for 6 hours without problem.
Try running Memtest86 on your system, games use large amounts of memory, your new memory may have problem. Have you tried refitting the original memory? The problem you are having is usually PSU, memory or the video card.
i suppose that mike could be right, If part of your memory isn'tt being used until you load a game like WoW, that couold be your problem. However sinceyou already replaced your RAM, i am skeptical of that. Still, running memtest would eliminate that as a possibility. I don't think it is the graphics card because that usually doesn't cause restarts. Usually, you just get errors and your display may revert back to 640 X 480 and 16 colors. I have never had a graphics card cause my computer to restart. Still, the latest drivers are always something that you should have.
Alright, so you guys have given some great ideas and let me come to the table with some updates:
Jedi940 - I got Prime95 up and running and ran it for 6 hours, and it didn't detect any errors (during the test the hottest my comp got was 63 degrees), cause a crash or anything that would fall into the category of making my computer unstable, which rules out the processor (correct?) I'm going to let it run again before bed for about 12 hours tonight to verify everything, but the fact that it ran for so long and didn't crash my box makes me think it isn't the processor (correct?)
Mike99 - I ran memtest86 in the very beginning when I replaced the RAM and nothing came up that would lead me to believe that my RAM is bad (which rules out the RAM I think)
TechGuy911 - I uninstalled my old video drivers and I was prompted with the standard "to complete the installation, you computer needs to be restarted" so I clicked yes. The comp restarted and when it was on the Windows XP loading screen, the computer restarted, and then went into a vicious cycle of restarts to I killed the power and didn't power it back up for about 20 min. When I started it again, it booted right up and I installed the new drivers.
Since I stopped the Prime95 test, I've been playing a game for roughly 8 hours and my computer hasn't restarted. Now I'm not sure if me updating the video card driver was the miracle cure, but i'm going to run Prime95 through the night and see what I wake up to.
So the only thing I can think of (mainly because I haven't replaced it unfortunatly) is that something is wrong with my video card? Sure enough I powered on my box this morning, booted up a game and WHAM! restart. I've diagnosed every other hardware component I can think of. Is there any way to test purely the video card and not RAM/Processor/etc to see if the isolated problem is with that hardware?
The comment was also brought up regarding the PSU giving off flakey voltages. Is there something I can use to stress the PSU and see if it causes a restart?
Those are my only last two ideas and if you have any other last ditch suggestions, it'd make me a happy man
You can use ATITool. Even though you have an Nvidia card, it will work. open it up and let it scan for artifacts. it will be a small window with a tumbling fuzzy cube. It will stress the GPU only.
Here is a link: http://downloads.guru3d.com/download.php?det=725
You can try 3DMark 06. Considering the tests that it runs it should be pretty thorough in testing out the 2D and 3D aspects of your card. If it has something to do with your video drivers (maybe a specific part) then it should find something during testing.
What about DirectX? You can try installing the latest version and then run the dxdiag (DirectX Diagnostic Tool).
Just for kicks you can even try playing some videos. AVIs, MPGs, QuickTime, DVDs, anything you can get your hands on.