[SOLVED] Support Team Stumped

Nov 22, 2020
2
0
10
Hey Folks,
Long time Lurker here,
I have a system issue that has honestly stumped about every IT person I know, including the team that built the machine. I'm out of options other than replacing all the remaining components unless someone has seen and knows a solution to this problem.

What's going on:
About 4 weeks ago, the system started acting odd, fans would start going high speed and fluctuate rather quickly for.. about 10-15 minutes then auto reboot the system. Turned the CPU overclock off, issue resolved, for several hours, then the reboots started again, no fan issues this time.

Many attempts later, (where I am currently) the system will boot up, reach the Bios, but will not load or install Windows without failure. Any attempt to install windows results in failure. The system is doing a boot cycle after every restart that completely destroys Windows 10's attempts to install or update (I have gotten the system to install windows previously, but it would corrupt due to the restarts very quickly.

Attempted solutions made thus far:
Windows update roll back - Negative
Virus Scans (multiple programs) - Negative.
Disconnecting all non-essential Devices - Negative.
Used Multiple independent drives to install windows and attempt to boot - Negative (2 SSD's and 1 HDD attempted)
Ram adjusted, re-seated, re-slotted, replaced. - Negative
Memtest performed, System fails on first test due to restart.
Temp checks before the system corrupted - Negative (all temperatures were found to be in acceptable margins.)
Power Supply Checked - Negative, RMA'd and replaced to no avail.
Bios? - Negative, bios updated to most recent update with zero changes.
Windows solutions: Startup Repair: Failed. Windows Reinstalled multiple times with multiple flash drives and one DVD to no long term success, issue is still happening.
SFC /scannow has failed to resolve or give any meaningful information.
Graphics Card has been removed. - Negative results there, issue has remained.

System information:
CPU: I7-8700K
GPU: GTX 1070-TI
PSU: EVGA SuperNOVA 1000 G3, 80 Plus Gold 1000W
Motherboard: ROG Strix 370-H Gaming
Ram: 32 Gigs of G-Skill Trident Z
Hard Disk: Evo 960, 250 Gig
360 Frostbyte Closed System Cooling attached to CPU.
OS: Windows 10 Pro.

Any ideas would be appreciated, I'm about to RMA the motherboard at this point before the warranty expires, not sure that explains the issues, that's what I can't get my head around. I cannot fathom what could be causing this issue. To allow it to run for hours at one point, and then for 10 minutes another, an issue that causes it to reboot an additional time when it restarts (thus corruption Windows installer/Update).
I have had the system for almost 3 years without any issue until now.

Thanks for any help or advice, I'm at my wit's end with this thing.
 
Solution
I would put money on the motherboard. Erratic fan behavior and sounds like something is corrupting your windows installation constantly. That last one is possible if a chip on the motherboard is bad.

I forgot which one, I think it was north or south bridge chip is bad it’s capable of corrupting drives. I’m sure someone will chime in on this as this area is out of my expertise and I only encountered this once before.

Use a read only disc or usb stick containing your OS of choice and start stress testing components to help narrow your culprits. Use 1 stick of ram at a time to help reduce variables. If what I think is wrong is the issue it wouldn’t surprise me if the memory and SSD are throwing errors, but that part is really...
What might have changed since 4 weeks ago when all was well?

If you can't run memtest86, try running with only one stick mounted in the preferred slot for your motherboard.
If no joy, try a different stick.
Theory being that there is a ram problem.

Past that, I would agree with a RMA for the motherboard.
While unusual, sometimes parts fail.
Usually it is infant mortality, but on occasions, failure can happen after a time period.
 
What might have changed since 4 weeks ago when all was well?

If you can't run memtest86, try running with only one stick mounted in the preferred slot for your motherboard.
If no joy, try a different stick.
Theory being that there is a ram problem.

Past that, I would agree with a RMA for the motherboard.
While unusual, sometimes parts fail.
Usually it is infant mortality, but on occasions, failure can happen after a time period.
I've had bad sticks of RAM and it causes Windows to reboot. But never have I had a stick of RAM cause memtest to reboot and that was with a stick that didn't make it past the 1st stage it was so bad. That said it is free and not much of a hassle to try with single RAM DIMMs.
 
Nov 22, 2020
2
0
10
Thank you Jeremyj and Geofelt for your responses.
Concerning the Ram, I can actually answer that one. I have 3 sticks at present, 2 that came with the system and a third that I bought to test the system with in the past 4 weeks. I have moved the sticks around, and tried different combinations (and solo starts) to no avail. The new stick didn't provide any changes during the failed memtest, or result in any improvement.

As to what changed, that's the problem . Nothing changed, it was gaming along just fine, and then issues started. To make matters worse, I haven't been playing system intensive games in the past 6 weeks. On the memtest, it almost always fails at 25% on the first test, failure caused by a reboot.

I am hoping it is the motherboard, and it is encouraging to hear others think likewise.
 

assasin32

Distinguished
Apr 23, 2008
1,356
22
19,515
I would put money on the motherboard. Erratic fan behavior and sounds like something is corrupting your windows installation constantly. That last one is possible if a chip on the motherboard is bad.

I forgot which one, I think it was north or south bridge chip is bad it’s capable of corrupting drives. I’m sure someone will chime in on this as this area is out of my expertise and I only encountered this once before.

Use a read only disc or usb stick containing your OS of choice and start stress testing components to help narrow your culprits. Use 1 stick of ram at a time to help reduce variables. If what I think is wrong is the issue it wouldn’t surprise me if the memory and SSD are throwing errors, but that part is really speculation on my part as to how it would react.
 
Solution