System Overheating--Is it the Power Supply?

[cbuchach]

So I thought had my new 3770K system stable but started getting predicatable complete system restarts when I started playing some games. I can play May Payne 3 for 30-90 minutes and would randomly getting a quick complete system restart. I was able to reproduce the restart with OCCT stressing both the GPU and CPU which would restart in 5-10 minutes. I thought heat may be an issue so took the case cover off and put an extra fan to the power supply and with this setup the system would not restart.

So heat obviously is the issue and know that I need to improve the thermals of the system. As far as diagnosing which component of the system is overheating, I am suspecting the power supply. Is this possible or does anyone else have other suggestions?

I should add the my CPU temps are well within where they had been when stress testing only the cpu. The power supply is a FSP 600W model and the CPU is a 560Ti which is not overclocked.

 

[Ken g6]

The power supply is a FSP

That sounds like a possible problem. (Off brand PSU.) If you want to put off replacing it, how many rails does that thing have? And are they distributed among the GPU (takes two), CPU, and other powered components properly?

Just checking: Do you have software for monitoring CPU, GPU, and possibly chipset temperatures? Temps on the CPU should make it throttle rather than restart, so it's probably something else.

 

[lehtv]

Generally FSP is not bad and 600W should be more than enough for that system, but if it's an older unit that's seen some wear and tear or has weak +12V output as older units do, then it doesn't surprise me that it's failing to power your system.

 

[cbuchach]

The FSP is a decent unit, about a $100 when I purchased it 3-4 years ago.

My MB temps are around 40C and I have the fans set to max on the Asus Gene V when I was testing and it still restarted. The GPU temps get up near 85-90 C and the fan speed gets up to about 60-70%. The GPU is at stock speeds, and the fact the fan speed is not maxed out makes me assume the GPU temps are still within Nvidia specifications.

 

[Slugbait]

What make/model is the case?

 

[mfenn]

So heat obviously is the issue and know that I need to improve the thermals of the system. As far as diagnosing which component of the system is overheating, I am suspecting the power supply. Is this possible or does anyone else have other suggestions?

It is certainly possible that the PSU is overheating, but it could just as easily be another component. Slugbait's question is a good one, what sort of case do you have and how do you have the case fans arranged?

Does stressing either the GPU or CPU independently produce a crash? While this would not completely rule out a power supply fault, it would at least eliminate the other component from suspicion.

 

[cbuchach]

What make/model is the case?

It is the Lian li pc-a04a mATX case

 

[cbuchach]

What make/model is the case?

It is certainly possible that the PSU is overheating, but it could just as easily be another component. Slugbait's question is a good one, what sort of case do you have and how do you have the case fans arranged?

Does stressing either the GPU or CPU independently produce a crash? While this would not completely rule out a power supply fault, it would at least eliminate the other component from suspicion.

The case fan arrangement is stock as far as it was insta lled by Lian Li.

I haven't just tested the CPU for a crash which is a next good step.

 

[mfenn]

It is the Lian li pc-a04a mATX case

That case looks to come stock with a 120mm exhaust and 2 120mm intakes, so you should be OK from an airflow perspective (as long as all fans are spinning of course).

 

[cmdrdredd]

Have you tried setting the GPU fans to spin to 100% and try playing a game? Maybe the GPU is actually having issues when run at 90c. Worth a shot there.

Do you have another PSU you can try?

 

[robertdrumm]

When was the last time you cleaned the case? My desktop had built up dust causing the fans to run no stop and created a lot of heat.After blowing all the dust out the computer runs a far lower temp and the fans run half the amount of time. I know its a simple suggestion but it does happen.

 

[cbuchach]

Well, I installed a new power supply (a Silencer 600W model) into the system and it is perfectly stable even with torture testing the PSU with OCCT, which previously had restarted the system within 10-15 minutes. So it was the power supply that must have been overheating. Too bad, it was a 700W FSP modular model that I thought was a good unit.

 

[mfenn]

Well, I installed a new power supply (a Silencer 600W model) into the system and it is perfectly stable even with torture testing the PSU with OCCT, which previously had restarted the system within 10-15 minutes. So it was the power supply that must have been overheating. Too bad, it was a 700W FSP modular model that I thought was a good unit.

Glad to hear that you got it resolved!