BSODs Event 41 kernal-power BugCheck 59, 340, 239 - SOLVED! (GPU Extension Cable)

DWD1961

[H]ard|Gawd
Joined
Nov 30, 2019
Messages
1,314
=========
SOLITION: Removed GPU extension cables. Plugged PSU cables directly into GPU.
=========

So I've been trying to track this down for a month or more now. I'm getting random BSODs in a game I play, Mech Warrior Online, but not when I play EVE, but also when I try to play Eve Vanguard, which is late BETA.

In all instances, I get a BSOD:
Event ID 41, Kernal-Power BugcheckCode 59 or BugcheckCode 340 or BugcheckCode 239.
Win 11 isn't writing any dump files, even when I set it to Kernel Memory Dump only. There is no dmp file and there are no minidump files.

I tried stress testing my rig using P95 Max everything with Furmark running at the same time. No problems.
I ran memtest86+ for 4 hours: nothing, but I'm going to run it all night today.

I've unplugged every USB device I have except a keyboard and mouse.

BugcheckCode 59
BugcheckParameter1 0xc0000006

BugcheckCode 340
BugcheckParameter1 0xffffc30225f87000

BugcheckCode 239
BugcheckParameter1 0xffffe304e6051140

More Information:

As stated below, I've also run:
DISM /Online /Cleanup-Image /CheckHealth
DISM /Online /Cleanup-Image /ScanHealth
DISM /Online /Cleanup-Image /RestoreHealth
sfc /scannow

Cinbench ran with no errors.

Memtest86+ 8.5 hours, no errors.

The WIndows BSOD reports:
Critical Process Died
System Service Exception (today 1st time)

Next up:

Driver Verification Manager

--

After all of that BSOD's persisted.
 
Last edited:
sig rig with the 550w psu? maybe your hitting its limits. since the codes are different i was first suspecting ram...

gpu says:
  • Recommended Power Supply
    650W
 
sig rig with the 550w psu? maybe your hitting its limits. since the codes are different i was first suspecting ram...

gpu says:
  • Recommended Power Supply
    650W
Naw with everything wide open, hitting like 300-320 watts. If you add up the theoretical, MB max, card, CPU, Pump and one fan in the top (One fan in my entire system, believe it or not), it's still far below 550 watts - 20%. I can run Furmark and P95 Max Heat at the same time wide open and no errors.

I should have added some additional information, which I will add the below to my original post:

Only happens when playing those games -- I don't play others. It happened ONCE with EVE and only after I first installed EVE, and then after a reboot never has happened again, not once. With MWO it's intermittent and if I don't use the MWO repair tool before playing, it will do it frequently.

It NEVER happens watching movies, browsing, etc. It only happens in a gaming situation, when my card is ramped up somewhat or processing information more than normal computing.

I let memtest run last night for 8.5 hours, no errors.

I'm going to run Cinebench now since it's DX instead of OpenGL. I think Furmark uses Open GL.

I also ran:
DISM /Online /Cleanup-Image /CheckHealth
DISM /Online /Cleanup-Image /ScanHealth
DISM /Online /Cleanup-Image /RestoreHealth


I haven't tried to BSOD it but I know EVE Vanguard will reliably create a BSOD.

I'll report back with both Cinebench and trying Vanguard again in a few

 
Last edited:
Hey guys, I posted more information in the first post above.

EVE Vanguard just BSOD even with the new stuff above. Threw a new BSOD. Rather than, Critical Process Died this time I was greeted with "System Service Exception." So, Windows is giving me more rewards!
 
Well, the (repeatable) kernel bug tells us it's happening during low-level kernel operations, possibly during a CPU power level/sleep state change, or maybe the gpu requested something from the cpu and that caused it to wake up.

This tells me it's either an actual kernel bug (possible but unlikely), or there's a hardware fault (faulty power rail(s), or unstable bus/cpu/soc), or there's insufficient power for the instantaneous load when the fault occured.

That's my two cents anyway.
 
Well, the (repeatable) kernel bug tells us it's happening during low-level kernel operations, possibly during a CPU power level/sleep state change, or maybe the gpu requested something from the cpu and that caused it to wake up.

This tells me it's either an actual kernel bug (possible but unlikely), or there's a hardware fault (faulty power rail(s), or unstable bus/cpu/soc), or there's insufficient power for the instantaneous load when the fault occured.

That's my two cents anyway.
If it's a power problem, why can I run Furmark/Cinbench + P95 MAX together and get no errors? That's max power.

The fact that Win 11 refuses to do a dump file hamstrings me.
 
If it's a power problem, why can I run Furmark/Cinbench + P95 MAX together and get no errors? That's max power.

The fact that Win 11 refuses to do a dump file hamstrings me.
Instantaneous power can spike much higher during state changes, and some games put the gpu under specific load conditions which can hammer a PSU much more than synthetic tests can. Combined, you could easily exceed the rated load of a PSU for a very brief period (which is usually okay).

But, if the voltage on the CPU's power rail drops too much because of it, it can cause all kinds of issues, and usually results in a bsod (if not immediate reboot).

Edit: here's a "makeuseof" article which provides a few possible solutions, but nothing sure-fire: https://www.makeuseof.com/tag/fix-windows-10-kernel-power-error/
 
Last edited:
Instantaneous power can spike much higher during state changes, and some games put the gpu under specific load conditions which can hammer a PSU much more than synthetic tests can. Combined, you could easily exceed the rated load of a PSU for a very brief period (which is usually okay).

But, if the voltage on the CPU's power rail drops too much because of it, it can cause all kinds of issues, and usually results in a bsod (if not immediate reboot).

Edit: here's a "makeuseof" article which provides a few possible solutions, but nothing sure-fire: https://www.makeuseof.com/tag/fix-windows-10-kernel-power-error/
It will run at 600 watts and still do 92% efficiency. That puts it stable power output at least 600 watts. Besides, if it were spiking enough to shut the unit down, it would have been doing it all along since I've been playing these games well before the problem started.
 
I may have found the problem. I ran EVE Vanguard twice after my probable fix, and it ran perfectly. I was using a Videocard extension cable. I plugged the card directly into the factory cables and had no problem now twice in a row. I have heard of that, but I don't see how a cable can suddenly go bad to the extent it causes a BSOD. Tehre is no cap in it and it is only about 8" long, which would drop the voltage very minimally.

I'll report back after more testing, and then I'll connect the extension cable again to see if I can reproduce the BSOD. EVE Vanguard was a gift because it will reliable produce the BSOD, whereas MWO isn't a sure thing and not instantaneous.
 
I may have found the problem. I ran EVE Vanguard twice after my probable fix, and it ran perfectly. I was using a Videocard extension cable. I plugged the card directly into the factory cables and had no problem now twice in a row. I have heard of that, but I don't see how a cable can suddenly go bad to the extent it causes a BSOD. Tehre is no cap in it and it is only about 8" long, which would drop the voltage very minimally.

I'll report back after more testing, and then I'll connect the extension cable again to see if I can reproduce the BSOD. EVE Vanguard was a gift because it will reliable produce the BSOD, whereas MWO isn't a sure thing and not instantaneous.
I had similar issues when using a GPU riser cable that came with my Thermaltake p3 case. Plugged directly in and problem was gone.
 
I had similar issues when using a GPU riser cable that came with my Thermaltake p3 case. Plugged directly in and problem was gone.
Strange you hear about extensions and risers with GPU but not so much CPU.
 
Back
Top