Unstable new build continues to crash

[LJF3]

Gigabyte GA-970-Gaming SLI
AMD FX 8320E
2 x 4TB Patriot Viper RAM
Corsair CX650M PSU
EVGA geforce 210
Windows 10
Samsung 850 EVO 500GB - System
3 x 3TB WD Blue HDD running Spanned Volume across all 3.

Need some help. What started as a project to try and revive old parts turned into a fresh build. Everything is brand new, including the case. Installed Windows 10 to run as a Plex and P2P server. The OS continuously reboots itself or locks up.

When it reboots itself the Reliability Monitor shows the Windows shut down unexpectedly. When it freezes, the screen is locked up with no response from keyboard and mouse and requires a hard reset. Again the Reliability Monitor will show an unexpected shutdown.

I have uTorrent installed which was causing a crash; the Webhelper running file utorrentie.exe was causing a crash. To resolve, I closed uTorrent and do not have it running.

I run BitDefender for anti-virus and firewall.

Any ideas?

 

[SgtScream]

What are your cpu temps?

 

[LJF3]

What are your cpu temps?

Currently

24 C with a recent max of 36 C

I have CPUID HWMonitor and can copy text file if it would help.

 

[SgtScream]

I'd suggest testing with a known good power supply.

 

[LJF3]

I'd suggest testing with a known good power supply.

The PSU is brand new, been installed maybe 3 days. Should I buy a new one to test?

If its a temp problem I could add more fans.

 

[SgtScream]

Your temps look good. Did you monitor the temps under heavy load?

 

[LJF3]

Your temps look good. Did you monitor the temps under heavy load?

No, I can run Prime95 and do a stress test now though.

Also, anyway to test power supply with voltage readings or anything?

 

[SgtScream]

It sounds like it could be either a power supply issue or a heat issue. In my experience, the only way to test a power supply with 100% certainty is to completely swap it out. Power supply testers are inaccurate.

 

[LJF3]

It sounds like it could be either a power supply issue or a heat issue. In my experience, the only way to test a power supply with 100% certainty is to completely swap it out. Power supply testers are inaccurate.

I appreciate all the help and information. Currently running torture test with Prime95. May buy new power supply just to be absolutely certain. Will post the results in a couple hours.

I am open to any other ideas as well.

 

[LJF3]

I have run Prime95 Torture Test (Blend) for over an hour. CPU temp and GPU temp has not gone above 50 C.

I am going to get a new PSU just to rule that out. I think 650 is overkill and I can get away with 550W.

Reliability Monitor after crashes does not show any info related to hardware, drivers, or software; just unexpected shutdown. Are there any other logs to review?

 

[LJF3]

Need some more assistance. I thought I was in the clear but not quite yet.

Yesterday I replaced my PSU with an upgraded Corsair. I went from the CX line to the RMX line with 80 Plus Gold ceritifcation.

I then ran Prime95 Torture Test (Blend). The test ran for over 12 hours with no issues. No heat concerns. Nothing. All in the clear. This morning machine was still running the torture test. Ended the test, did a few minor things with no issues. I decided to do another Prime95 torture test this time using Small FFT mode. After an hour or so no issues, temps were all good with room to move. I was connected via RDP and monitoring closely the CPU and Graphics temps as well as everything else.

Suddenly the RDP session ended. I go to the machine itself and find the screen completely frozen, but not black. It is at the Windows 10 lock screen. Keyboard and mouse are unresponsive and unable to connect via RDP. Power button is unresponsive and the machine is just hanging. This has happened previously without any stress or activity on the computer.

I am not sure where to look next...

 

[LJF3]

I still have not identified the issue. 12 hour MemTest86+ came up with no issues. No cooling issues either. System is not overclocked and I restored Bios defaults. The log in Event Viewer indicates a n Event 41,Kernel-Power Critical Error with a Task 63. I have followed the troubleshooting steps in the Msft Support article related to the error.

I am wondering if the Samsung 850 EVO 500GB is the issue. For some reason this drive is not on the compatibility list for the MB even though the smaller 250GB is. I am at a loss for what to do next.

 

[LJF3]

I wanted to close the loop on this for future reference if needed. After weeks of troubleshooting I can only point to heat being the issue. Could also have been a bad sensor on motherboard. I replaced motherboard and CPU and have not had any issues (knock on wood). SSD was not the issue because when running diagnostics off thumb drive the issue continued and required constant restarts of diagnostic tests. Thanks for all the help with this.

 

[LJF3]

It sounds like it could be either a power supply issue or a heat issue. In my experience, the only way to test a power supply with 100% certainty is to completely swap it out. Power supply testers are inaccurate.

Most likely heat or bad sensors on MB. Posted additional details below. Thanks for all the help!