Sudden Shut Down Issure

[Geister227]

I built my very first pc about two months ago. Since then I've had no real problems except for a single hard drive that died in the first few days. Now I'm having an odd issue with my computer powering off. On power up, everything appears fine, until I load something graphics intensive, such as a game. At that point, the computer turns off instantly (as if unplugged). When I press the power button I get nothing. If I completely remove power for about 20 minutes it boots normally. The computer will not stay on overnight if left idling in windows.

This issue started about a week ago, with infrequent shutdowns during games.


I'm wondering what you guys think the cause is. I'm leaning towards the PSU, but I was told it could also be the RAM.
If it is the PSU, do you think it could have caused the drive problems?



My main components are:
x58 motherboard
i7 920
2x gtx275's

I have a 1000w PSU


Thanks in advance.

 

[cmichael138]

You could download and run Memtest86 to see if it's a memory issue. You can also check your GPU temps with something like GPU-Z. It may indeed be a PSU issue, but check the other components if you don't have a spare PSU to substitute.

 

[Twoboxer]

Rapid shutdowns are nearly always heat or psu related.

Download RealTemp, run it, watch core temps at idle. What are they? Do they climb over time at idle? When you load the cpu, do they rise just before the PC shuts down? If so, we need to deal with heat. If not, we have "eliminated" heat as the cause.

Somewhat more likely, your psu fails to deliver what it promises. What make/model do you have?

 

[jsc]

Sounds like the PSU. In terms of rated capacity, it's certainly large enough. But there could be something wrong with it internally.

 

[outlw6669]

Yes, that sounds like it could be the PSU to me also.
Who makes your PSU and what is the model number?

It would not hurt to test some of your other hardware either.
Keep an eye on your temps using Realtemp or GPUz for the GPU's and Realtemp for the CPU.
Load test the CPU with Prime95 and check the RAM with Memtest.

 

[Geister227]

My PSU is a Corsair HX1000W

I haven't been able to keep it on long enough to get a current temperature reading, but about 3 weeks ago I had these temps at idle.
CPU 25c GPU1 50c GPU2 47c

While playing prototype (high settings), I had these temps.
CPU 35c GPU1 82c GPU2 72c

My processor, videocards, and RAM are have all completely unmodified settings.


I ran the OCCT memtest and had a shutdown in the first 6 seconds.
Is OCCT similar to memtest86?

 

[Twoboxer]

Since your PC was runing for a couple months, heat and psu are more likely than memory. And Prime95 will uncover more memory problems than memtest. So let's try a series of tests:

1) (Heat) Examine your heatsink/cooler. All four pins in correctly? Fan spinning on boot up? Were you poking around in there? Bump it? Dislodge it?

2) I presume you are running 3 sticks of 2GB.
- Run with one stick. If it runs, stress (run Prime95) and examine temps.
- Rinse/repeat with each of the other 2 sticks.
- If all three fail, odds are its your psu. (Try again removing one 275)
- If one fails, stress with two sticks.
- If two fail, you are horribly unlucky lol . . .