WHEA logger events 19 and 20 and random reboots after BIOS update.

[Arcleon]

Hello

I have an issue with my motherboard and ASRock support have been unable to fix it so I am asking the community for help as a last resort before I buy a new board.

I have an ASRock 980DE3/U3S3 motherboard, I updated the BIOS from version 1.60 to 2.00 using the windows BIOS flashing tool from the ASRock page for the motherboard.

It said it was successful so I rebooted and shortly after rebooting my computer restarted it self.

Looking in the windows event viewer I found a few errors but these two consistently appeared when crashes occurred:
Warning 07/05/2016 17:30:35 Microsoft-Windows-WHEA-Logger 19 None "A corrected hardware error has occurred.
Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Cache Hierarchy Error
Processor ID: 2

Error 07/05/2016 17:30:35 Microsoft-Windows-WHEA-Logger 20 None "A fatal hardware error has occurred.
Component: AMD Northbridge
Error Source: Machine Check Exception
Error Type: 25
Processor ID: 0

I also found that this one would appear frequently:
Warning 07/05/2016 17:31:04 Microsoft-Windows-WHEA-Logger 19 None "A corrected hardware error has occurred.
Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Cache Hierarchy Error
Processor ID: 2

From what I have found by researching these tend to occur when there are voltage or wattage problem.
I have been using the same hardware for just under a couple of years so I do not think there is a wattage problem unless the update has caused an increase in power consumption.
I have not done any overclocking so the CPU voltage should only be stable/stock values.

What I have tried:
Flashing version 2.00 from the BIOS using the ROM file from the ASRock downloads page.
Going back to version 1.60 using the ROM file from the ASRock downloads page.
Flashing 2.00 using ROM file then removing the CMOS battery for 10 minutes.
Flashing 1.60 using ROM file then removing the CMOS battery for 10 minutes.
A fresh install of windows 7, it crashed partway through the installation.
CPU stress testing using bootable software (can't remember what it was called) which found no errors or crashes.
Memtestx86 which found no errors and did not crash.
Trying a 580W PSU instead of my current 550W PSU.
Trying a GPU which had about the same power consumption.
Loading the default values in the BIOS.
Setting the CPU voltage in the BIOS to the default value or 1.375V instead of "auto".
Contacting ASRock support: they advised checking the VRM temperature which was not overheating, Changing the processor voltage from auto to manual.
Contacted ASRocks RMA team and found that I am about 2 weeks out from the 2 year warranty end period of the original purchase date, I bought it from Amazon in June 2014 but it seems that Amazon bought it from ASRock on 12/5/2014.

The only things that seems to make a difference are:
Putting windows 7 into safe mode, I think this is just preventing full power usage or voltage changes.
Using a lower power GPU but I think this indicates a power problem with the motherboard and not a problem with the other 2 GPUs and 2 PSUs.

A DXDiag, system report from HWINFO64 and detailed views for the WHEA events are here

Basic system specs:
Motherboard: 980DE3/U3S3
CPU: AMD FX 8350
Memory: 16GB
Video Card: Gigabyte Nvidia GTX 960
SSD: CT250mx200ssd1 250GB
HDD: ST000DM 001-1ER164 2TB
Network card: TG-3468_V2
CD-ROM: DW-G120A,IDE
OS: Windows 7 64-bit

If anyone has any suggestions or ideas please let me know.

Thank you in advance.

 

[cin19]

Either 19 or 20 error related to the CPU/MB ( Northbridge https://upload.wikimedia.org/wikipedia/commons/thumb/5/51/Chipset_schematic.svg/300px-Chipset_schematic.svg.png ).

Because you used the fx8350 with the low power phase count MB, Asrock 80DE3/U3S3 ( 4+1) for 2 yrs. Even you don't oc the cpu and asrock did list the MB can support the fx8350, those power phases ( or VRMs) may have problem to keep up your fx8350 overtime. Also I don't know what cpu cooler you had, if you use the big top-down cooler ( not the tower or water cooler) and it will help.

You may try to disable few cores in the BIOS to see the PC will run fine or not. Or if you can borrow one fx6 series cpu to test your MB. I think the problem is that MB has low power phage count.

 

[Arcleon]

Either 19 or 20 error related to the CPU/MB ( Northbridge https://upload.wikimedia.org/wikipedia/commons/thumb/5/51/Chipset_schematic.svg/300px-Chipset_schematic.svg.png ).

Because you used the fx8350 with the low power phase count MB, Asrock 80DE3/U3S3 ( 4+1) for 2 yrs. Even you don't oc the cpu and asrock did list the MB can support the fx8350, those power phases ( or VRMs) may have problem to keep up your fx8350 overtime. Also I don't know what cpu cooler you had, if you use the big top-down cooler ( not the tower or water cooler) and it will help.

You may try to disable few cores in the BIOS to see the PC will run fine or not. Or if you can borrow one fx6 series cpu to test your MB. I think the problem is that MB has low power phage count.

So you think that the motherboard is simply worn out due to the power usage and it failing after a BIOS update was a coincidence?
And the CPU fan I am using is a Cooler master Hyper TX3.

 

[cin19]

Yes, based on you used the low count VRMs MB with the fx8350. Without any tests I will not say the problem is from the MB, but It's quite highly possible. For more info you can check the link, like the section "What VRM cooling will do for you" , "Database of VRM failure incidents", and "What to do if you suspect your VRMs have failed" http://www.overclock.net/t/943109/about-vrms-mosfets-motherboard-safety-with-125w-tdp-processors
You may put back the stock cooler to try. Also try to disable the cores in the BIOS and like the "What to do if you suspect your VRMs have failed".

 

[Arcleon]

Yes, based on you used the low count VRMs MB with the fx8350. Without any tests I will not say the problem is from the MB, but It's quite highly possible. For more info you can check the link, like the section "What VRM cooling will do for you" , "Database of VRM failure incidents", and "What to do if you suspect your VRMs have failed" http://www.overclock.net/t/943109/about-vrms-mosfets-motherboard-safety-with-125w-tdp-processors
You may put back the stock cooler to try. Also try to disable the cores in the BIOS and like the "What to do if you suspect your VRMs have failed".

I set the CPU active core control to four cores from all and the problem still occured.
Turns out that the motherboard was just within 2 years from purchase so I got the motherboard replaced but the errors and crashes are still occurring with a new motherboard.

I am swapping out RAM now to see if that solves the issue, if it does not then the CPU must be the cause, a new motherboard having the same issue as a previous one is unlikely.

Swapped out RAM, still the same problem, must be CPU.

 

[cin19]

You may try use the this method: Bench it - Troubleshooting http://www.tomshardware.com/faq/id-1753671/bench-troubleshooting.html
Only use the cpu+stock cooler+ GPU+ one stick RAM or not RAM to boot the PC. If the PC does the same with one RAM, and the problem maybe from the cpu if you did test the RAM and GPU in other PC. If the PC can boot and has the no RAM error, and the problem maybe from the RAM. If the PC can't boot at all without the RAM, and the problem is from the cpu.

 

[Arcleon]

You may try use the this method: Bench it - Troubleshooting http://www.tomshardware.com/faq/id-1753671/bench-troubleshooting.html
Only use the cpu+stock cooler+ GPU+ one stick RAM or not RAM to boot the PC. If the PC does the same with one RAM, and the problem maybe from the cpu if you did test the RAM and GPU in other PC. If the PC can boot and has the no RAM error, and the problem maybe from the RAM. If the PC can't boot at all without the RAM, and the problem is from the cpu.

I have tried 2 different GPUs and 2 sticks of RAM, at this point either multiple components have developed faults simultaneously or its the CPU.
I am still within CPU warranty so I am sending it back to AMD for service.