Lock ups

[Gibber]

Hi guys,
Having a few problems.
The only 2 games I play BF2 and counter-strike source keep locking up all the time. I did a prime95 run for 24 hours (bear in mind at stock settings) and it passed without a problem so the instability obviously doesn't lie with the memory or CPU.

I'm currently running a Sapphire X1800 XT (not overclocked)

I've checked the power rails and under load they all seem to be fine, the only thing that goes slightly lower is the 12v rail which goes down to 11.78v and the 3.3v which goes to 3.28v.

I've tried 2 PSU's, one is an unheard of make (yeah I know ) the other one is an ultra-x ATX 2.0 500watt psu. I did have a hiper 580 non modular for a a very short while but it crapped out after a day (kept cutting out) so I never had the chance to test if it cured the problem.

Reckon it's a PSU problem? I don't wanna have to fork out for a very expensive power supply only to find it didn't fix the problem.

 

[Jimbo Mahoney]

Fork out £5 for a cheap multimeter and measure the 12v and 5v rails from a molex connector when playing BF2 or whatever.

Motherboard voltage probes are notoriously bad at reporting the voltages.

e.g. my 12v line is being reported as 11.6v, my 5v is 4.8v and my 3.3v is 3.47v

I know for a fact the 12v is @ 12v, the 5v is @ 5v and the 3.3v is at 3.6v (OCZ Powerstream with adjustable rails).

 

[nO}{8]

get a 12mm fan directed onto the x1800, see if that helps, some of the mosfets may be overheating.

 

[Gibber]

Hmm will have to try this, thanks guys.

The GPU is idle at the moment and is @ 58c however this usually goes to 71-75c under load, it can sometimes go even higher than this but usually only when it's very hot weather. I'm not really sure on what the ideal temperature if for a GPU under load but it sounds to me like 75c is dangerously high.

I do have at the moment an 80mm fan on the side panel of the case blowing air out. Would it make any difference if this was turned the other way around? It's almost facing over the graphics card but not quite. I haven't felt how hot the mosfets get under load so i'll have a look in a moment.

Have got a 120mm zalman fan lying around here but nowhere to put as I haven't got the tools to mod this cheapo case to hold one.

 

[Hlebio]

75c isn't the best temperature but it is by no means high enough to freeze games I don't think.

Sounds like PSU or RAM to me.. what RAM and Motherboard you got?

 

[Gibber]

At the moment I'm using an Ultra-X 500watt PSU.

Memory is 4 sticks of 256mb TwinMOS running at stock timing and speeds but that's just a temporary measure until I get something better (think they use Winbond CH-5 chips) has ran perfectly fine on my old nforce2 motherboard @ 2T CMD rate playing CS & BF2.

Motherboard is a DFI Lanparty Ultra-D

To be sure of this I have run memtest over and over looping and have ran prime95 solid (using intense memory test) and also the in-place FFT tests without a problem.

 

[Hlebio]

4 sticks of 256mb you say? Hmm...

Tried taking 2 of them out and seeing how things go?

I had a problem similar to your's which I suspected to be RAM but memtest reported no errors. Still, I swapped my RAM for a different brand and the problem stopped..

So it could be a case of your mobo not liking the RAM brand/configuration..

 

[Indy500]

Looks like the PSU to me.

 

[Sir Random]

You could try underclocking the gfx card's gpu and ram by about 50Mhz and see if that cures the freezing. If it does, then you'll know it's the card that's causing the freezing.

 

[Gibber]

All good points guys

I'll give each one a go. It's just a process of elimination I guess. I'll report back with my findings once I have found the culprit!