Phantom PC Crashing

Deformation

Reputable
Jul 5, 2014
100
0
4,710
A good friend of mine is having a pretty severe PC issue. I have been working with him for quite sometime trying to resolve this issue and at this point, i am pretty stumped. Below i will describe the issues that is happening and the steps we have taken thus far to resolve it.
_____________________________________________________________________________

The Issue:

Full PC Lockup, No mouse movement, No sound, no change of images, Additionally there are no changes in any peripherals such as his Logitec Keyboard locks down showing ram usage that he had it set too. The only way to pick up from this point is to hold power down, turn off, boot back up.

This has no specific application tied to it that I can tell, It occurs completely randomly but often. (4 - 10 times a day). Occurs when in use, or set idle with or without applications running. He and I use voice chat to communicate and when it does Crash, even that dies too.
_____________________________________________________________________________

Steps Taken for software ( In-Order, With reboots )


  • Disabled all Non Microsoft required Services
    Disable all Start-up Items
    Grabbed two hardware temp monitors ( all showed well within acceptable temps under load)
    Booted into Safemode and sat Idle ( error still occured )
    Wiped and reinstalled Windows
    Windows threw error with (dwm.exe) - Resolved with service pack 1
_____________________________________________________________________________

Software Summery

After doing all of that, the issue is still present on a fresh install of windows. So my thoughts move onto a hardware issue.
_________________________________________________________________________

Hardware Testing


  • Ran Memtest 4 times - Clean
    Ran sfc /scannow and DISM /Online /Cleanup-Image /CheckHealth - Clean
    Ran Prime95 to stress test CPU +Check temp - Clean
    Ran 3dMark to stress test GPU +Check Temp - Clean
__________________________________________________________________________

Final Summery

I have scanned, changed, altered anything i can think of potentially causing this problem except the PSU, I have just sent him out to grab either a PSU Tester or a Multi-Meter so we can individually test it. But I am unsure how to test it under stress, and due to the nature of how random this is, if a PSU Meter might even show anything worth while to help.

Normally i'd bypass most of these by simply swapping out parts but he does not have any spare parts around to be able to do so.

So I look to the Toms-Hardware Guru's for guidance on fixing this issue on anything i may have missed or that could still be tested.
_________________________________________________________________________

His PC


  • Windows 7
    AMD FX-9590
    EVGA GTX 970
    Asrock Fatality 990FX
    G.Skill Ripjaw X Series (4x8)
    Corsair HX 750W
 
+1^
The 9590 need a good 550 watt power supply by itself.
To get enough voltage on the 12v ATX-EPS connector.
The CX line is not recommended for gaming/ high usage systems. Much less a gaming system with a 9590 which draws 220watts by itself.
It is a cheap power supply designed for office/ web browsing computers.
If you look here it is a tier 4 unit.
http://www.tomshardware.co.uk/forum/id-2547993/psu-tier-list.html
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710



He has tested all Molex's and they're showing within the +/- 5% range. Had him sit and watch it for a while to check for any random spikes. Nothing alarming found there yet.

Voltages in Bios showing as :
Vcore : 1.424 +- 0.01
+12v 12.249
+5v 5.04 +- 0.02
+3.3v 3.2 +- 0.01

He stated his VCore is fluctuating slowly between 1.0 to 1.424, Seems a bit odd.
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710




Apologies, i actually got the PSU incorrect, its a HX750,
http://www.corsair.com/en-us/hx-series-hx750-power-supply-750-watt-80-plus-gold-certified-modular-psu

Are we still thinking the PSU still insufficient?
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710


No overclocking at all, and the 'Over Boost' function on the Mobo is also toggled to 'Off' - Would like to note the issue got worse when he flipped it on and booted once.
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710


Any particular brand/model - Wattage you would advise? Keeping in mind a little extra room for a few upgrades here and there?
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710




stay with a 750W? and Bronze? Isn't he nearing the 750w cap as it is? any upgrades i'd imagine would throw him over?

And unfortunately he doesn't have any spares to be able to swap, and his Local PC store will charge him an Arm and a Leg to do so, then another fortune to buy one.
 
You are right, I forgot how much an overclocked 8 core can draw. I re-calculated with a serious overclock on CPU and GPU and and came up with 843 watts. So that 750 is struggling at stock clocks. This one should be more than enough, considering future upgrades:

http://www.newegg.com/Product/Product.aspx?Item=N82E16817438010&nm_mc=AFC-C8Junction&cm_mmc=AFC-C8Junction-PCPartPicker,%20LLC-_-na-_-na-_-na&cm_sp=&AID=10446076&PID=3938566&SID=
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710


Well, he just ordered a Corsair RM1000X from the Tier 1 list provided in an earlier post. I will update this in the next few days for the solution or to continue the journey.
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710



PSU arrived, swapped out. Issue still happening. Next thoughts?
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710



Periphs checked on his fathers really low end pc, no issues there, No spare CPU or GPU to swap for testing. Gotta try to narrow it down further alternatively.

Since he is from AUS, replacing his mobo is 300$ and GPU is pushing 700$. Any other ideas or tests?
 
Is he water cooling the CPU?
If so then the VRM might be overheating and going out of spec on voltage.
Between the CPU and the rear I/O panel is a row of gold caps and a heatsink with "Killer" printed on it.
You definitely need good air flow over this area.
As a test hang, wire tie, zip tie tape, etc... a spare case fan blowing directly on the VRM heatsink and caps.
Normally with a stock or downdraft heatsink you get cooling from the CPU's heatsink fan for the VRM.
With water cooling that fan no longer exists and the VRM is in an area of low air flow.

 

Deformation

Reputable
Jul 5, 2014
100
0
4,710


No watercooling in the entire system. and temps are constantly monitored and there are no heats raising towards any threatening levels.

 
Have you tried running the system with 2 sticks of ram?
Also memtest86 at 4 passes is not quite a full test.
We normally run at least 7 passes or overnight.
Also if he is using a tall tower style cooler the VRM could still be overheating. Happened to me with a 212EVO. So I added a second fan for VRM cooling.
I had to slide the original fan up for ram clearance, so the VRM was not getting air flow. Adding a second fan slid down as far as possible solved the VRM overheating problem.

You state that it happens at Idle also so I would be more inclined towards memory.
Some times when running 4x8 sticks of ram you need ti increase the voltage slightly to get full stability.
 

Deformation

Reputable
Jul 5, 2014
100
0
4,710



Way ahead of you on that, I had him pull 2 of his 4 sticks out, It crashed again, So i had him swap the 2 sticks out for the other two and it continued. Im in process of having him try different ram slots instead of A1 A2 to try B1 and B2 to see if we get any changes.