Help needed - PC turns off during gaming

Soldato
Joined
29 Sep 2005
Posts
9,128
PC turns off during gaming, no warning or BSOD and then tries to restart but just hangs without posting. I can play certain games for small periods of time and others for longer, BF4 for example lasts much longer than Borderlands 2. Chivalry I can play for hours and nothing happens. It happened with my 560 SLI set up and still with new GPU. No crashes in windows, just after 15+ mins of gaming.

Specs:
a6nt.jpg


290 at stock. Temps check out okay on GPU (40-50% fan with temps at 65-80*c) and CPU seemed okay at 55-70*C max on load.

ogaz.jpg


i5 (which was at 4.6ghz from OCUK as a bundle), currently running at 3.3ghz stock as I lost the OcUK profile in a weird bios failure.

I've set everything in bios back to stock. PSU is a MODXSTREAM-PRO 700W modular power supply.

First thoughts:
1.) PSU on the way out, or overheating?
2.) Motherboard resisters overheat or are fried
3.) Bios voltage problems (I think it's at default though?) or restricting temps and shutting off
4.) CPU needs reseating with more thermal paste (still on original)
5.) RAM?

Really not sure what to start with, just seems like it could be so many different things it will takes ages to fix? It's be going on for some time now.
 
Last edited:
Missed this sorry..

I can rule a few things out.. It won't be a CPU overclock issue as you would BSOD.

Its also not a memory issue as it'd struggle to boot up in that situation..

I can only see it being one of two things (both you're not going to like).

1) You GPU may be pulling too much power and (overloading your 12v rail)..
Follow up question: What is your PSU?

2) Just general BF4 crashing, its a common bug.. It should be fixed by a patch [soon, if not already done].


Misread your post sorry...

I still think it could be a GPU or PSU issue though..

What vcore is your CPU pulling under 100% load and what temps does it hit?
 
Do you have any software like HWMonitor to check Psu voltages. I know it`s not completely accurate, but may give you some indication as to it`s efficiency.
 
Does this help?

2qm3.jpg



Originally the system was running at bundle 4.6ghz speed from Overclockers and then one day it just switched off and hasn't been the same since. This is what leads me to believe it's the PSU. But still doesn't make sense how some games (albeit older, less graphic intense) run okay for hours.
 
Put Max Payne 3 on ultra and got these readings..

xqu7.jpg


Seeing as I had these issues with 2 x 560 in SLI towards the last few months, and a single 560 and now a 290... I think I can rule out GPU?
 
Last edited:
Hi,

Before id suggest you go buy another psu costing you money , is there anybody with a decent system you could borrow a power supply to test maybe. ? I am personally leaning towards a psu problem, but if you could test another it should answer your problem hopefully then youll know if the outlay is needed on a PSU anyway.
 
Hi,

Before id suggest you go buy another psu costing you money , is there anybody with a decent system you could borrow a power supply to test maybe. ? I am personally leaning towards a psu problem, but if you could test another it should answer your problem hopefully then youll know if the outlay is needed on a PSU anyway.

Unfortunately I don't know anyone with a decent PSU. I mean I know the PSU should be capable of the power draw, but it is 2+ years old now and has been running i5 at 4.6ghz and 2 x 560's for that period.

Any recommendations on a replacement PSU? Am I able to send it back if I have the same problem with my system or is that against the returns policy? I have a feeling otherwise I'm throwing money at the problem for a bit of trial and error.

Was HW running during the game, it doesn't seem to have increased your CPU temps much atall...

Was about 5 mins of Max Payne 3 as I was off to bed. I'll try again this afternoon.

Thanks for the help so far all.
 
Last edited:
I'd say get into the BIOS and check voltages there, but if the machine is stable whilst idling then that prolly won't help you.

That PSU is rated for 46A on the 12V rails according to here:

http://ocz.com/consumer/psu/modxstream-pro-500w-700w-power-supply/specifications

That seems a little low for a 700W PSU, but that could just be me. I'm basing that off my 520W Corsair, which is 6 years old, and has 54A across the 12V rails. Surely with your previous SLI configuration AND a chip clocked to that extent it would be a close thing? I think the PSU is a safe bet as the culprit like others have said.

EDIT: check event viewer to see if there are any relevant logs in there, you never know.
 
Last edited:
I agree with the general feeling here , given that the restarts happen underload and crashes/hangs/restarts with no warning or PSU points towards a PSU issue to me.

Try running like IntelBurn to put extreme load on the CPU and then furmark for the GPU as this will put the PSU understress while keeping the GPU load high but "basic"

Should narrow down for sure if its a PSU or GPU issue
 
Does it happen during CPU stress testing?

Does it happen during GPU stress testing?

Your GPU will draw far more power than the CPU, so if it falls over with the latter and not the former it would also point to a power related problem, although this doesn't guarantee that it's the PSU to blame. It could be a fault with the GPU causing it to draw much more power than intended thus overloading the PSU.

The fact that the PC reboots though points at a power related issue. PSUs should have a number of protection features including over-current, over-voltage, over-power, temperature etc..I'm not sure how that exact PSU behaves during those fault conditions but one possible behaviour is a reset, cutting power to the system in the process, which of course will cause a reboot.
 
I will check the stress testing.

I went from 560 x 2, 560 alone and then to a 290. So I think the GPU is unlikely the issue.

I agree.

Let us know if you get the same issues with stress testing. It's clearly an issue that only happens with the PSU is loaded up. It would be useful to see how much load causes it to fall over.
 
Back
Top Bottom