Black screen reboots (crashes) while gaming: Exhaustive info and details inside. Help very much appreciated.

Zackalope

Commendable
Oct 1, 2016
10
0
1,510
System: i7-2600k Intel CPU w. 212 coolermaster

p8p67 Deluxe Asus Motherboard

Corsair 750 hx PSU

G.Skill G.SKILL Ripjaws X Series 16GB (2 x 8GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800)

EVGA GTX 1080 SC

Samsung EVO 850 256 gb SSD + an offbrand 1TB SSD for storage

1080p large format monitor on HDMI

Windows 10 64 pro

**Issue** 'black screen' sudden reboot when gaming, usually after 10-30 minutes after start. Sometimes much less. Confirmed in Overwatch, and in Ark survival evolved (ark seems much worse).

What I've investigated so far:

* I've checked Bios, Ram is clocked at 1600, CPU clock at stock (3.3-3.8boost) . CPU has never run OC'ed except for one testing session, where it ran stable at 1.3v @4.4 ghz with low temps. Owner doesn't need it OC'ed (family member) so he didn't want to shorten life of CPU possibly.

* I've repeatedly run Unengine Heaven on loop at 1980x1020 with max AA and medium tess. GPU temps stay below 70c at all times, card is not clocked higher than factory stock. Custom fan curve IS set through precision x for cooling. No reboots when looping Heaven, even for hours. No artifacts, no slowdowns. 3dmark firestrike and timespy also have been run 10+ times each, no artifacts or crashing.

* I've run a blend test in Prime95 and small FFT for nearly an hour each: CPU temps on blend never exceed 60 degrees C on any core, small fft test temps never break 65 degrees C. No crashes.

* Furmark and OCCT testing also produces no high temps and no crashes on either the GPU or CPU tests. PSU test passed clean.

* I've removed the PSU and tested the rails with a multimeter. Results: 12v rail-12.22v all, 5v rail - 5.04 v , 3v rail - 3.35v, -12v rail - -12.09v. These are all well within tolerances.

* **I've fully wiped and reinstalled Windows 10 from scratch**. All drivers and updates brought up to date, including Nvidia drivers. GeForce experience is NOT installed.

* I've updated the BIOS on the motherboard.

* I've run MEMTEST from a USB stick, no issues.

----------------------------

I'm at my wits end. I've just pulled the power supply, and stuck in an old Raidmax 630ss to see if there's any difference with that, he's currently trying to reproduce the crash with that installed to try and rule out PSU problems.
I've looked up the exact issues from the event viewer in windows, but It doesn't give me much: 0x8000400000000002 . A kernel event id 41 only means an unexpected power problem has occurred and is what happens with both an unexpected shutdown/restart.

I've poured over every forum I can think of, but I'm not getting anywhere there either.

**What the hell is going on!!!!!!!!!** Any ideas or help appreciated, because the only things I can think of to do now is swapping Motherboard, CPU, RAM, which I have some RAM to swap out, but I dont have a CPU I can throw in there, nor a spare MOBO. So I would have to buy those, and that's silly for an undiagnosed problem.




 
Solution
This is going to be a hard troubleshoot, as you already know. Keep in mind the kernel event id 41 fault could also mean that any one of your components is pulling or pushing the wrong amount of power. Have you tried swapping the gpu? Or changing the pci/pcie slots your cards are in? I have had one similar problem like this. At the end it was 1 pcie pin shorting my gpu, but for some reason only under load. Just keep in mind electricity can be a finicky thing, just because you are stable with 12v on 200 watts dose not mean you will be stable on 12v at 600 watts.

Jonathan Schiller

Distinguished
Aug 31, 2013
72
0
18,660
This is going to be a hard troubleshoot, as you already know. Keep in mind the kernel event id 41 fault could also mean that any one of your components is pulling or pushing the wrong amount of power. Have you tried swapping the gpu? Or changing the pci/pcie slots your cards are in? I have had one similar problem like this. At the end it was 1 pcie pin shorting my gpu, but for some reason only under load. Just keep in mind electricity can be a finicky thing, just because you are stable with 12v on 200 watts dose not mean you will be stable on 12v at 600 watts.
 
Solution

Zackalope

Commendable
Oct 1, 2016
10
0
1,510


Well, I got 'lucky'. The PSU was the problem: Ended up ordering a new EVGA P2 unit, and blam, no more problems. Was able to figure this out because an old raidmax thrown in there fixed all the problems.

You were correct, the load voltages on the dying PSU were nothing like idle. Thanks@!!!