PC rebooting

[DanielLevesque]

Ive posted before about my pc rebooting, have since then ruled out Video card (baught new one) power seems to be fine as I have a 400W Psu.. its definitly hardware.. Im thinking its my board. While playing games my pc will just reboot at no specific time.. replaced my radeon 128 thinking nforce chipset might not agree with it.. which i made a grave mistake so I would like anyones help who can give. Thanks (btw no overclocking done)

-Dan

A7N266-c (Nforce Chipset)
1 Gig Cas 2.0 pc3000 OCZ ram (2 512 sticks)
Athlon XP 2200+ Tbred
100Gig WD Special Edition (8 meg of cache)
Nforce onboard audio
3Com Ethernet Card
40x samsung burner with 8 meg buffer (not problem baught it after)
Geforce 4 Ti 4200 64MB (Leadtek also new so not problem, Old card was Ati radeon 128mb)

 

[Ironstone]

Troubleshooting random events can be a real pain - you have my sympathy. :-(

When the same thing happened to me it WAS power, even though I had a 350W supply that was supposedly more than adequate for the job. I put an old 250W supply back in to test it and the reboots stopped. I eventually tracked it down to a problem with the 5v lines.

Can you borrow another one, or swap from another machine for a few days? It's fairly easy to rule out that way.

I have read that crappy power supplies are frequently responsible for random problems that we blame on Windows or whatever. I imagine that sags and brown-outs in your local electricity supply can also casue problems but that's harder to test without specialist gear. I built a little test rig and at certain times of the day our local supply drops and fluctuates quite badly. Apparently the better quality the power supply the more capable it is of handling variations. Big power rating doesn't mean big quality of course (and if you look at what they can actually deliver, some cheap power supplies are "imaginatively" rated!)

Of course, it might be nothing to doing with the power supply. And one more thing that might be worth trying, (as it's very easy to do) is pull the lead off the back of the reboot switch on your case. I once had one of those go feral on me too..

Good luck.

 

[DanielLevesque]

Sorry I forgot to mention Im also on a UPS

on my 5v line its usually around 4.8 the rest are all at par

-Dan

 

[Ironstone]

I've also got a UPS, but they aren't a foolproof solution. I'm on shaky ground here, but it's a percentage game rather than a 100% solution. As with power supplies, the quality of UPSs can vary considerably with regard to how much filtering they provide, and how well they respond generally to abnormal situations. (And of course, parts of them can fail too).

I had a long chat to a guy at the company I bought mine from (it's roughly a mid priced model) and he told me that the model that would give the most security and protection would cost me as much or more than the price of my computer

 

[DanielLevesque]

Yea, I agree.. reason im not thinking its power is the other PC in here has no problem... Im not sure if its just the Nforce chipset.. but ive tested both sticks of ram... video.. my question is.. if it was power wouldnt I see the flux in hw monitor? 5v+ never drops below 4.81 which i thought was kinda low but no one else seems to think it would cause for error... other than that all other power is close to exact?

-Dan

 

[phsstpok]

Have you tried different RAM with the nForce board?

<b>I have so many cookies I now have a FAT problem!</b>

 

[Ironstone]

My feeling about power is that it's more likely to be the power supply in the computer than the power supply from the mains (if it's power at all). I only say that because that's what happened to me - a power supply that I thought was reasonable quality, and which definitely had an adequate rating, did turn out to be the problem. And I spent a lot of time insisting that it was OK and checking other stuff instead.

When tested, the 5volts lines were still delivering the right numbers (which of course they usually will when you're dealing with random fault not a outright failure).

But when I looked closely, the fact that all the 5 volt lines had turned a little brown at the end was a strong clue!

I never found out why (maybe a big surge got past the filters in the UPS and damaged something, maybe not?).

I really only point the finger at the power supply because I had that trouble and changing the power supply fixed it. Probably a red herring, but they are pretty quick to swap over if you can borrow another (caution required with fitting and handling as always, of course).

 

[rubikian]

Could be the heat ventilation problem. Some PC restarts because of over heating. Do you have good air flow in your case?

 

[DanielLevesque]

I will order another PSU just to test it.. So no one else has experienced this problem with nforce specifically? using the 128bit controller?

I made the mistake of buying another video card thinking it was the problem.. because it only does it during 3d games

-Dan

 

[Ironstone]

Eek! Don't buy a new Power Supply on the basis of my ramblings! Unless you're using company money for an office machine of course...

I have several machines, so can usually swap things around "from stock" to test theories.

I'm not any kind of guru though, so my information (although correct as far as it goes) may be completely irrelevant to your situation. Can't you rip one out of somewhere else for a day?

 

[DanielLevesque]

I agree with you You suggestions are good.. and im trying to suck all the power I can with burnin test.. running every test at the max rate.. and I cant get the system to give 1 error over 15 minutes.. I did this due to a friend saying if it were power this is using all components so it would use all the power possible?

is this true?

-Dan

 

[Ironstone]

Dan,

A few weeks back - as part of a push to try and learn more about computers - I dismantled an old power supply and unsoldered it bit by bit, trying to understand the function of each component as I went.

At the end of the exercise I went back to my original theory of computer electricity - which is that the whole thing is run by tiny malevolent gnomes that live in all those little "houses" that we call capacitors, resistors etc.

When I troubleshoot I do a little thinking first and then I start by replacing the cheapest and easiest components (usually by nicking them out of other machines)until eventually I'm forced to take a punt and buy something.

Your test sounds like a good way of stressing the PSU, but it's quite hard to be sure that you really are stressing any component in the particular way that it needs to cause the failure. There's always some guy on a forum somewhere who will pop up and say "ah yes, but did you remember to remogrify your whifflers and uplink your arjgrunters - because if you didn't, and it was a Wednesday - that would would mean that"..... blah blah..

Maybe the power is a red herring - there are so many possibilities with computers...

I also spent a while trying to work out exactly which component uses how much power (the old "is 350w enough" question") and it's incredibly hard to find precise data. Some components simply don't list their usage and most have a range (E.g. hard drives use heaps to spin up but not so much to keep spinning or transfer data. Plus they use 5v and 12v lines).

When it gets down to specifics - i.e "when will everything actually be making the fullest demands all at once?" I get lost. So I take a conservative guess.

Good luck anyway..

Chris.

 

[Guest]

can you try to run the sticks of RAM one at a time to see if the problem disappears with one of them?

also you could try taking out the extras: network, sound etc. and see if the problem goes away with one piece of hardware.

After all that you have kinda narrowed it down to CPU or M/B. If you have buddies maybe you could borrow those items to see if trading one of those out makes the symptom disappear.