Computer crashes for no reason...

[Guest]

Hey,

I have a custom made computer (http://www.shockcannon.com/journal/?p=18 for the parts), and I've been using it harmoniously for 5 months. One day, my friend installed VMWare, failed to use it, then uninstalled it, and when I got back from work one day, the machine was off. I turned it on, but the machine shut itself down after 4 seconds. After many attempts and cleaning of dust bunnies, I got the thing working again, but when I get it working on Windows, the machine will freeze up and shut itself down randomly (when I play Warcraft III, browse movies on the internet, etc...).

Is there something I should look into or do I need to replace a part? Thanks.

 

[Guest]

I still don't know how that prevents the computer from loading the BIOs....

 

[croc]

Misleading subject... First get rid of the offending sloppyware, then see if its misbehaving for 'no reason'.

With VM ware, unless you are a registry pro, this might be a bit difficult. As this was a custom built system, one would suppose that the builders would've provided you with three good tools:

a: a recovery disk.
b: a support number.
c: a warranty.

Good luck...

 

[Guest]

Well, I've already reformatted my partition entirely, and running on a new installation. It shuts down less often, but it still does it once in a while (let's say when I'm playing a movie or running a youtube file. Don't know what would happen when I cracked Warcraft III again).

I'm trying to figure out if there's something I should replace or not. But for a system that's half a year old, I'm not sure if it's really a hardware problem though...

 

[croc]

Obviously option a: wasn't available?

What about options b: and c:?

 

[Surferosa]

I have done jobs on three pcs that had this problem; each had a different cause.

1) Windows was corrupt. Needed a full wipe & fresh install. If your PC is / was doing this in bios post- then this isnt your problem.

2) The motherboard & gpu was covered in a 2mm thick layer of dust. A good clean & it stopped happening (what a surprise!)

3) PSU was shot. This took me a while to diagnose. It was a generic brand that had seemingly worked fine for a year or so- and it was only when I swapped it out the problem went away.

It could be a lot of things though.. mobo, ram.. anything really. If you have some replacement parts to test with it helps. Try taking everything out of your mobo & reinstalling (ie gpu, ram, power cables). Make sure everyting fits properly.

 

[TabrisDarkPeace]

4) CPU shutdown temp in BIOS is set way to low (eg: Under 65 C is pointless) - In fact turn it off and turn the alarm on (set to 75C instead).

Thus if CPU gets hot it'll sound an alarm instead of shutting down.

This may help isolate the fault.

 

[Guest]

I was going crazy on the dust issue. I cleaned out all the dust inside my heatsink, and it worked fine for a bit. But now it either turns on for four seconds and shuts off (without a beep from the motherboard) or boots up and shuts down randomly when runnning windows (the music and games gets a little fuzzy as the CPU uses more power).

Is there a possibility that I may have to replace the power supply unit seeing as even the motherboard doesn't turn on in some cases?

 

[darkguset]

Obviously hardware problem. Isolate the issue by removing components and keep only the necessary ones. It sounds like a heat problem (heatsink not properly sitting on processor). Or could be your power supply or corrupt BIOS (try reseting CMOS). Isolate the faulty component and get rid of it. The issue with the VMware could be just coincidence.

Good luck!

 

[gse1]

Darkguset gives good advice. This definitely sounds like a hardware problem, although this doesn't sound like a heat problem to me. You've got a few options to try to identify the problem, try them *one at a time* (because if you do them all at once and fix the problem you don't know what fixed it).

1 - Remove all addon components (i.e. network card if you don't need it to run the machine or have onboard networking, sound card, graphics card if you have onboard graphics, TV cards, etc...)and run with a bare system. See if it still crashes.
2 - If you have multiple memory sticks in the machine, say for dual channel memory, take one of them out. See if the machine still crashes - if it does replace the one you took out and and take out a different one. Also run memtest like mpilchfamily suggested earlier.
3 - Check that all components are seated properly in their slots or plugs, press down on all connections to ensure a good contact

Can you clarify whether, when the machines crashes, it switches itself off or it reboots automatically? Also are there any other symptoms, like the hard drive light stays on just before it crashes?

 

[Guest]

Well, I reapplied the thermal grease and the machine works 50% better, but I may try it again because the computer shortly freezes up periodically. When I realized the symptom is getting worse (the fuzzy sounds from my music, periodic freezing of my videos, operating system widgets stalling up, etc...), I shut the thing off.

The next morning, I gave the machine another try and it's back to the ol' "starts up, and turns off after 4 seconds" routine. I may try the thermal grease routine one more time (seeing as how I did it wrong the first time), and if I still see the same problems, I may have to go through the "taking it apart one-by-one" routine and track down the problem that way (as Darkguset says).

The memtest test I ran before says the RAM is fine.

That's where I'm at right now.

 

[gse1]

This sounds like it could be intermittent hard drive or hard drive controller failure.

Can you try putting different hard drive(s) and/or an addon controller in?

 

[Guest]

How would it be a hard drive problem if the BIOs doesn't even boot up the operating system (in another words, doesn't the computer access the mobo first before going to the hard drive)?

 

[leo2kp]

Do you have onboard sound? If so, take it out and run the machine. If not, shut off the onboard sound from your BIOS. If it still happens, it's possible there's something wrong with your motherboard. Remove everything you don't need (including all but one stick of RAM) and run the machine for a while and see if it reboots randomly. If it still does, then the problem lies with one of the few components you have still hooked up, i.e. motherboard, processor, or the stick of RAM. Replace the RAM with the other module (if you have two or more) and run it. If it reboots again, then it's probably not RAM. The rest of it requires having a spare part to replace processor/motherboard. I'm affraid running Memtest and Prime95 won't reveal the exact cause since you'll probably get errors anyway weather its the processor or motherboard, but you have a much better idea of what's wrong this way.

The fact that you replaced the thermal compound and had to dust it out tells me that you may have fried something somewhere down the line, and it could be anything. First thing I'd suspect personally is PSU or voltage regulator on the motherboard.

 

[The_OGS]

I believe answers are to be found in your BIOS setup.
Ensure everything is individually set perfectly, especially your memory settings.
Check that the memory timings are all default/SPD (CPU-Z should show 5,5,5,15 @ 800MHz).
Check that the memory is getting at least ~2.0V - this is important!
Make sure everything is nice and cool and your CPU is not overheating. The temperature and voltage report within BIOS is often quite useful. Examine your voltages very carefully and watch them for a while (to determine if fluctuations).
There's something goin' on with your rig, before it even gets to OS...
Don't worry - you will find out what it is,
Regards

 

[weskurtz81]

Well,

First off, have you been overclocking, I skimmed the posts and did not see this question asked. Have you messed with any of the bios settings? Have you tried setting them back to defaults? If bios won't load, have you reset the jumper? Do you have a fluke to test the power coming out of the psu? To me, if the memory is fine, and assuming the motherboard is working properly, it sounds like the psu is putting out bad power, or some other issue with the psu. Have you checked the cpu temps?

wes

 

[waylander]

Have you tried resetting the BIOS by removing the CMOS battery and moving the jumper for a few minutes? Try that and see if it helps. I can't see how it could be a heat issue if you've already cleaned the dust and reset the cpu's hsf. I'd look at hardware too.

 

[leo2kp]

"Its not a Hard disk drive nor sound card failure. Those arent related with the problem your posting."

I had a friend whos faulty sound-card would cause his computer to randomly reboot. We installed a new sound card and it worked ever since. I think it's possible that the sound card could be faulty enough to cause a system to reboot, simply because it's plugged in to the PCI bus. But I agree that it's probably not HDD related.

 

[redwing]

I just want to clarify a few things.

1.When you say your PC sometimes shuts down after 4 seconds, is that 4 seconds after you press the power button, or 4 seconds after you load windows?

2.You mentionned that your PC "fails loading the BIOS" in an earlier post, can you clarify? Are you having trouble navigating the bios menu?

3.You say that your computer gets progressively slower as you run your OS, with music getting choppy, etc ... Could you monitor your CPU% usage as that happens? If your CPU% usage is close to 100 for medium/light tasks (web surfing, mp3 playing), then your issue could be software related (either a virus/spyware or some weird windows issue).

I've encountered many issues in my 10 years of building PCs. I had crashes that stemmed from something as typical as an unstable voltage to the CPU (weak PSU) to something as weird as driver incompatibility with am onboard soundcard for certain games. The sound issue was causing my computer to freeze up during games and was a pain to figure out/troubleshoot.

Having said that, don't rule anything out from your possible causes. Since these causes are so numerous, in your situation I would try doing a fresh re-install of windows. I know this sounds like a lazy way out, but if you've been using your computer for 5 years (and still on your original windows install), you could have a seriously clogged up registry that could be costing you some performance. Re-installing windows + your apps and drivers will take you possibly a few hours, but this would scratch any software issues as possible causes off your list. If it still doesnt help, you can proceed with hardware troubleshooting and the costs that it involves.

 

[leo2kp]

^^

5 months <cough>

 

[The_OGS]

in your situation I would try doing a fresh re-install of windows

I stated to disagree, with agitation. No! No! His hardware is not solid, he's not ready for any OS yet!
However, it's not a new rig and of course you are correct that it's unclear what his problem is and the OS must be removed as a factor. I wouldn't troubleshoot his computer with a rotten Windows either, heheh.
But seriously, if you can't figure this problem out firehydra2k you gotta get your hardware solid first, and BIOS. Then, you get your new Windows. Then we see... sorry if not what you wanted to hear.
All good advice on stuff to check from everybody
L8R

 

[Guest]

http://www.newegg.com/Product/Product.asp?Item=N82E16813131022

That's the motherboard I got.

So far I reinstalled Windows to take care of the VMware thing (even though I now think it was a very dire path to take as it took forever to reinstall everything), and applied thermal grease a couple times due to the persistence of the problem (I was new to the thermal grease procedure at the time, but the online manual wasn't helping too much).

An hour ago, I had a friend supervise me on my 4th attempt on applying the thermal grease and everything. I bought alcohol and got some coffee filters to clean out the grease and reapplied it. However, he did say I did not put enough of it in there, and according the the Arctic silver website, they told me to put in at most 3/4 of a BB pellet. I ended using 1.5 of that amount and used an old credit card (sterilized with alcohol) to spread it out. After putting the heatsink back on, and pushed all the connections in and such.

...then he posed my video card for some reason. Took it out, and noticed that I missed that during my cleaning. Blasted it with canned air, and found a dead bug sitting behind the fan.

But yeah, I also took the CMOs battery out and reset the BIOs, and booted it backup again.

Now I currently have it blasting music constantly to test the degradation of the system. Before, either it turns on, and turns off BEFORE the mobo beeps (no OS or BIOS appears on the screen). And before i applied new thermal grease, it worked for a few hours, then noticed the computer freezing up a little bit at first, then more and more...until it got to a point where I got a headache from the music fuzzing up.

All this time, the computer was at 35 C at most, and the CPU usage did not show any abnormalities from usual use.

...and in maybe 3 hours I'm going to check and see if it's still up and running (the problems start arising maybe 3 or 4 hours after I do the fixes on my hardware).

Another thing I'm kinda curious about though, is that according to my mobo's software, all the voltage meters read a little less than what it should be (e.g. 11.75 V for 12 V, 3.22V for 4V, etc..). I don't know if that should be stable, and I have my motherboard on default factory configurations. Is that something I should worry about?

 

[waylander]

Another thing I'm kinda curious about though, is that according to my mobo's software, all the voltage meters read a little less than what it should be (e.g. 11.75 V for 12 V, 3.22V for 4V, etc..). I don't know if that should be stable, and I have my motherboard on default factory configurations. Is that something I should worry about?

It's nice that you got everything working again..

As to voltages, the +12v should be +/- 5% and all the others should be +/- 10%, which means you are ok for the +12v but not the 4v (I think this should be 5v right?).

 

[Guest]

Yeah, you're right. It's 5V. i kinda forgot that number.

Because I'm looking into how much resources all my parts take up from the power supply. I'm speculating that I may not have enough power generating from mine. I took out my second CD drive and floppy drive, and so far it's been running (I just left for dinner and came back an hour later.

I'll give it the 24 hour test and see if it starts flickering again. If that's the case, then I'll have to look into something that powers more than 500 W...

And I forgot to respond to another question. I don't overclock...

 

[Guest]

Ok, I stripped the computer down, i tested the rig with my friend's working power supply, and I am down to two components...the CPU or the motherboard.

I have one question. How am I supposed to test this? (Thanks everyone for your input!)