PC is crashing...here are the symptoms, can you help?

[cptwhite_uk]

Okay the rig is as follows:

Athlon X2 4200+ Manchester @ stock (AC Freezer 64 Pro)
DFI Ultra-D Motherboard (passive heatsink on northbridge)
2Gb RAM (1x1Gb Super Talent and 1x1Gb Kingston Value)
X1950Pro standard cooling @ stock
Enermax Noisetaker 600W PSU
1x HDD (Seagate IDE)
1x Rom drive (brand unknown)

The general problem as of about 24 hours ago is that when running a 3D app the PC will crash in the first 2/3 minutes, and struggle to boot immediately after, unless it is turned off for a few minutes to cool before trying to reboot. This problem seems to have first started when I switched from my North-Q 400W PSU to the Enermax. Note, the PC seems fine for the most part when at desktop!

The crashing isn't immediate. First the screen goes to standby, I can tell the PC is still responsive as the sound effects still respond to mouse commands. Around 5 seconds later the PC hard crashes and is unresponsive with the sound looping/frozen.

Before installing the Enermax a few weeks ago, with the North-Q i was getting random blue screens, which I put down to the memory, but after running Memtest for a good 2/3 hours on each stick and having 0 errors it seemed to suggest the memory was fine, hence trying a new PSU.

I have since switched back to the North-Q PSU and I'm getting the crashing as per the Enermax. I intend to do a full system strip and rebuild to see if that sorts the problem, but assuming it doesn't what are the likely causes for this kind of crashing.

I realise the symptoms point to overheating. I'm fairly confident it's not the CPU I have core temp and speed fan running, which suggest at full load the max temp it's reached is about 57C (full load), I'm not sure how to measure the X1950pro but I have VPU recover enabled and "prepare an error report" ricked in the ATI control panel, and a screen hasn't popped up asking me to send one. Thus I can only assume this 1950 is fine too (always was before anyway).

I have tried one stick of RAM with the same crashing, so that suggests the RAM isn't at fault either. The only other option is overheating on the motherboard? Can anyone shed any light onto this, or suggest a course of action / other things to consider.

Many thank guys

PS I also have a spare motherboard (ECS Nforce4-A939) but don't want to do a complete reformat. What do I need to uninstall to switch motherboards?

[GSte]

My first thought would be the graphics card overheating as you said, there are programmes that can monitor this, if you download ATI tool that should help, its fairly straightforward to use. The other thing that stands out is the new PSU. I don't know much about PSUs, but it seems reasonable to think that if it's dodgy, the extra power draw from the graphics card being in use could 'show' its instability...... have you tested the hdd at all? Nothing suggests that this could be the fault, other than they account for most computer problems I see day to day

[cptwhite_uk]

Any package out there to measure and log the rails on the PSU? Recommendations? I'll download ATI Tool, used it before but didn't install it on this PC...so far

Edit ATI Tool = All NVIDIA and ATI cards except for Radeon X1950 Pro are supported in this release.

[JimNastics]

X1950Pro standard cooling @ stock

There's your problem... you have the infamous "power saving mode" overheating issue that has plagued many an X1950pro. Here's an equally infamous thread that started it all:

http://www.sapphiretech.com/en/forum...ead.php?t=8522

The main culprits were originally the Sapphire cards, but I think others have shown similar problems. Search for X1950pro VRM and you'll get lots of results on lots of forums You probably don't want to hear this but the easiest fix really is buying a different card, and get a refund / RMA on that X1950pro if you can. I ditched my Sapphire X1950pro for an 8800GTS.