Unexplained Shutdown - Possibly Wrong PSU?

Seinoir

Distinguished
Apr 25, 2011
5
0
18,510
I've had this computer running for about 2 months, and from day 1, I've had problems with it.

Motherboard: ASUS P5N-D
CPU: Intel Core 2 Duo E8500 3166 Mhz Wolfdale
GPU: NVidia GeForce GTS 450
RAM: 4x A-DATA 2GB DDR2 SDRAM (8gb total)

I'm running my system on all default/auto values in the BIOS (no overclocking). For the most part, the system appears stable. My CPU and GPU run at about 35C at idle; at load, the CPU goes to about 45-50C and the GPU goes to 65C. I have plenty of ventilation running through my case. The problem is that whenever I attempt to run games for extended periods of time (over 30 minutes), my computer tends to completely shut down without warning. I discovered that if I cooled the room down, the system can run some games (Portal 2, Vindictus, Fallout 3) without a problem. However, I would prefer not to have to bundle up to use my computer. I'm fairly certain that this problem is either due to dust buildup or a problem with the PSU. According to http://extreme.outervision.com/PSUEngine, my system has a power requirement of 300-350W. I checked my power supply, and it's listed as 550W. Could this be causing the problems I'm having, or is there something else I should consider? If this is the problem, will I have to replace the PSU?
 

Seinoir

Distinguished
Apr 25, 2011
5
0
18,510


PSU: X-Clio GreatPower X14S4P3

I'm not sure about the Amp output; where would I find that?
 

jfby

Distinguished
Jun 4, 2010
418
0
18,810
First, why would you wait 2 months? If it's a part issue, most direct sellers drop RMA support after 30 days.

It sounds like something is overheating, not PSU issue. Use CPUid and have it log your temps and auto-save a logfile as you play. If your temps exceed what is recommended for a 450, then that's your key. Also have a logfile for your CPU temp and see if it exceeds the E8500 operating temp maximum. This is where I'd start.
 

According to XClio's specifications:

+3.3V@30A
+5V@28A
+12V1@18A
+12V2@18A
+12V3@16A
-12V@0.8A
+5VSB@3.0A

Continuous total output power is 550W max.
The combined power of +5V and +3.3V is 140W max.

It's not a high quality power supply but I figure it should be able to supply up to 32 Amps on its three combined +12 Volt rails.

Can't speak to the DC power quality of each rail. High electrical noise and ripple on the DC output rails can cause spontaneous reboots or shutdowns and even damage components.

Download and run TechPowerUp's GPU-Z freeware utility. On the Sensors tab page you can turn on Log to file and Continue refreshing this screen while GPU-Z is in the background which may help you in determining if the graphics card may be encountering any temperature or voltage problems. Play one of the games that causes the problem while GPU-Z is running in the background. After the system spontaneously shuts down you should be able to look at the log file in a text editor after rebooting to determine if anything about the graphics card is running out of spec.
 

Seinoir

Distinguished
Apr 25, 2011
5
0
18,510
Well, I just crashed my system playing Assassin's Creed, and nothing seems out of the ordinary with the GPU; according to GPU-Z the last recorded temp was 62C, well within normal bounds, VDDC was 1.125V. I'd be glad to look at the CPU readings if someone could recommend a suitable program. I think I'd rather go with a more stable PSU if this one is causing my crashing problems.
 

Seinoir

Distinguished
Apr 25, 2011
5
0
18,510


I bought this PSU at least a year ago; this is a pure home-built computer, and it took me 2 years to gather all the parts. The RMA support was long gone by the time I turned it on for the first time.