[SOLVED] Windows Critical fail Kernel-power 41

[TheKittenSlay3r]

So recently I've been experiencing a lot of crashes, originally the crashes I linked to were from a cooler going up but just enough to slowly allow my cpu to overheat until crash, since then this has been resolved and most of the crashes have gone away. Now I'm still experiencing crashes after a entire fresh reboot of windows (clean drives and all). I've searched all over and so far I've tried a few solutions such as changing my power plan to performance, reseating my GPU as well as my CPU and checking my connectors for any bad or loose connections yet still no dice. So far the pattern has seemed to be my PC crashing after around being on for about 12 hours or more (most of this is idle with no serious loads besides the occasion when I play a game or something, no excessive temps on GPU or CPU). My PSU is a brand new EVGA 850 gold which I still haven't ruled out as a possible factor in the crashing problem however from every site I see that I plug my build into to see what PSU I should be running on my config they all recommend a 650 which is not 100% accurate as I'm also running a 3080 Ultra FTW which consumes a considerable amount more power. My thought so far is it could be a memory dump of some kind occurring as it's only happened when I play games at least since the reboot, before the reboot I had some very strange hangs where my screens would completely power off but I could still hear my friends and the pc would be completely on still.

*Side note for the memory dump I played Atlas recently where I experienced a majority of my crashes and I started running the game with no memory leaks and it almost completely fixed the issue as far as I can tell, the current crashes have only happened when playing Rainbow 6.

These are the logs from the detail on the error if it helps


- <Event xmlns="

http://schemas.microsoft.com/win/2004/08/events/event

">

- <System>

<Provider Name="Microsoft-Windows-Kernel-Power" Guid="{331c3b3a-2005-44c2-ac5e-77220c37d6b4}" />

<EventID>41</EventID>

<Version>8</Version>

<Level>1</Level>

<Task>63</Task>

<Opcode>0</Opcode>

<Keywords>0x8000400000000002</Keywords>

<TimeCreated SystemTime="2020-12-02T08:37:06.0940732Z" />

<EventRecordID>735</EventRecordID>

<Correlation />

<Execution ProcessID="4" ThreadID="8" />

<Channel>System</Channel>

<Computer>DESKTOP-RC094N2</Computer>

<Security UserID="S-1-5-18" />

</System>

- <EventData>

<Data Name="BugcheckCode">0</Data>

<Data Name="BugcheckParameter1">0x0</Data>

<Data Name="BugcheckParameter2">0x0</Data>

<Data Name="BugcheckParameter3">0x0</Data>

<Data Name="BugcheckParameter4">0x0</Data>

<Data Name="SleepInProgress">0</Data>

<Data Name="PowerButtonTimestamp">0</Data>

<Data Name="BootAppStatus">0</Data>

<Data Name="Checkpoint">0</Data>

<Data Name="ConnectedStandbyInProgress">false</Data>

<Data Name="SystemSleepTransitionsToOn">3</Data>

<Data Name="CsEntryScenarioInstanceId">0</Data>

<Data Name="BugcheckInfoFromEFI">false</Data>

<Data Name="CheckpointStatus">0</Data>

<Data Name="CsEntryScenarioInstanceIdV2">0</Data>

<Data Name="LongPowerButtonPressDetected">false</Data>

</EventData>

</Event>


Any help would be greatly appreciated!

 

[Colif]

kernel event 41 is a result of the crash, not a cause in itself. Windows is reporting it found the previous shutdown was unexpected.

What are full specs of the PC?

did you actually get a bsod or did it just restart?

I don't see a memory dump attached to post?

 

[TheKittenSlay3r]

kernel event 41 is a result of the crash, not a cause in itself. Windows is reporting it found the previous shutdown was unexpected.

What are full specs of the PC?

did you actually get a bsod or did it just restart?

I don't see a memory dump attached to post?

I7 9700k
EVGA 3080 Ulta FTW
32 GB Tridentz Ram
EVGA 850 Gold PSU
Icue 150i watercooler
Asus ROG STRIX Z370-E GAMING moba

Not Blue screens have happened just a straight restart, sometimes I'll get a blue windows screen when I relaunch and tells me windows ran into a problem but never an actual BSOD. Something else possibly worth noting when Benchmarking my PC does fine no crashes in any of my stress tests ect. however when I run OCCT's power test of my GPU I instantly crash.

kernel event 41 is a result of the crash, not a cause in itself. Windows is reporting it found the previous shutdown was unexpected.

 

[Colif]

My PSU is a brand new EVGA 850 gold which I still haven't ruled out as a possible factor in the crashing problem however from every site I see that I plug my build into to see what PSU I should be running on my config they all recommend a 650 which is not 100% accurate as I'm also running a 3080 Ultra FTW which consumes a considerable amount more power.

This test stess everu psu line to the max. Maybe your PSU canot handle maximum amperage, that your computer can take. My old Tagan U15 530W died always after 20-30 min with 12V rail at 11.2V. This Seasonic takes this test without any issue.

OCCT PSU Test

Folks, how exactly does OCCT tests a rig's PSU? When I tried this, my rig shutdown within five minutes :banghead:

www.techpowerup.com

maybe it is that PSU after all? they aren't the easiest item to test in PC.

what power plan are you running
in settings/system/power, click on the Additional power settings link under related settings? Are you on high performance?

i see lots of pc that can cope with running full speed but crash at idle , it seems odd to me.

you might want to run HWINFO and log system performance to see if any sensors offer a clue.
download hwinfo - https://www.hwinfo.com/download/
when you run app, tick sensors only and click run
on following screen, on the bottom right there are a number of buttons.
click the button to right of the clock, it shows a chart with a + on it.

This button lets you log your sensors scores until you turn PC off - this opens file explorer and lets you name a log file.
You would need to run log every day until you crash.
you can read the results in Excel or Google docs. It shows too much info if you ask me, it can be hard to identify the actual cause.

 

[TheKittenSlay3r]

I'll give this a shot, I am running on high performance in my power options

 

[TheKittenSlay3r]

OCCT PSU Test

Folks, how exactly does OCCT tests a rig's PSU? When I tried this, my rig shutdown within five minutes :banghead:

www.techpowerup.com

maybe it is that PSU after all? they aren't the easiest item to test in PC.

what power plan are you running
in settings/system/power, click on the Additional power settings link under related settings? Are you on high performance?

i see lots of pc that can cope with running full speed but crash at idle , it seems odd to me.

you might want to run HWINFO and log system performance to see if any sensors offer a clue.
download hwinfo - https://www.hwinfo.com/download/
when you run app, tick sensors only and click run
on following screen, on the bottom right there are a number of buttons.
click the button to right of the clock, it shows a chart with a + on it.

This button lets you log your sensors scores until you turn PC off - this opens file explorer and lets you name a log file.
You would need to run log every day until you crash.
you can read the results in Excel or Google docs. It shows too much info if you ask me, it can be hard to identify the actual cause.

https://docs.google.com/spreadsheets/d/191tnx-G3cEvvRHLG-Hy_va84Fkzqyf_Y4tGDxAr0e2o/edit?usp=sharing <---- This is the log I had created after a crash Same Kernel 41 power error

 

[Colif]

I wish my gpu sat at 44c, lowest i get is 48c. Your aio coolant is cooler too, i guess mine is only 240mm with 2 fans and yours is 360mm with 3 fans.

I don't see anything there, as far as I can see power is rock solid on all rails. I can ask others to look as I might have missed something

Are you overclocked? I can see all cores running at 4.8ghz all the time. Base clock for cpu is 3.6ghz - https://ark.intel.com/content/www/u...9700k-processor-12m-cache-up-to-4-90-ghz.html

 

[Andzario]

I had a similiar problem that you describe. I'm not sure if it helps you, but doing these two steps helped me:

1) Power Options > Edit Plan Settings > Change advanced power settings > PCI Express > Link State Power Management > Set it to Off

2) Power Options > Choose what the power buttons do (on the left) > Turn off the Windows Fast Startup

 

[Vic 40]

See only part of the Hwinfo log, upload it onto a site which facilitates this like SaberCatHost this one i like since not so much (or any) spam comes with it and does not require a log in to get the info.

 

[TheKittenSlay3r]

I wish my gpu sat at 44c, lowest i get is 48c. Your aio coolant is cooler too, i guess mine is only 240mm with 2 fans and yours is 360mm with 3 fans.

I don't see anything there, as far as I can see power is rock solid on all rails. I can ask others to look as I might have missed something

Are you overclocked? I can see all cores running at 4.8ghz all the time. Base clock for cpu is 3.6ghz - https://ark.intel.com/content/www/u...9700k-processor-12m-cache-up-to-4-90-ghz.html

Nope no overlocking whatsoever I just restored this PC the other day to maybe clear some issues but I've never really had the need to overclock in the first place, I can check my bios to double check but I've never set anything.

 

[TheKittenSlay3r]

An interesting find, I opened and found my socket had a bent pin which I've since managed to save however the real find here was I left my CPU cooler unplugged by accident and fired up my pc and ran the OCCT power test and I did not crash during the test, I looked over at the end of the test and noticed the lights not on and realized what happened. I then plugged in my cooler and did the test once more and a crash in seconds came up. I'm now certain it's my PSU however everyone I see is telling me 850 is plenty for this build and according to PC part picker it should be perfectly fine as well, heres the list of 98% of what my PC is (motherboard and ram are newer models on here I believe) but still should be very similar in voltage

https://pcpartpicker.com/list/#compatibility_notes

Should it be safe to assume sending the PSU in for service? Or will I need to upgrade to a bigger one?

 

[Vic 40]

That is not an EVGA 850watt btu an 750watt Thermaltake (which is a good one in this case) in there and the 3080 can really spike when using power. So which psu do you really have?

 

[TheKittenSlay3r]

https://www.evga.com/products/product.aspx?pn=220-GA-0850-X1 I'm not sure what you're saying, this is the PSU inside my computer right now?

That is not an EVGA 850watt btu an 750watt Thermaltake (which is a good one in this case) in there and the 3080 can really spike when using power. So which psu do you really have?

 

[Vic 40]

https://www.evga.com/products/product.aspx?pn=220-GA-0850-X1 I'm not sure what you're saying, this is the PSU inside my computer right now?

Was just to make clear, the one in your pcpartpicker list is a different one.

Are you using all three pcie cables to power the gpu? Not just two where from one two connectors is used.

Which bios is used? Keep it at the stock bios. Could well be that such a bios pushes the power usage beyond what the psu might handle. Reviews give a very high (near 400watts) power usage for this card, where it even surpasses that on the oc bios.

 

[Colif]

note: I assume Vic is talking about bios on GPU when he asking which one is being used. That makes the most sense.

 

[Vic 40]

Yes was referring to gpu bios.

 

[TheKittenSlay3r]

Was just to make clear, the one in your pcpartpicker list is a different one.

Are you using all three pcie cables to power the gpu? Not just two where from one two connectors is used.

Which bios is used? Keep it at the stock bios. Could well be that such a bios pushes the power usage beyond what the psu might handle. Reviews give a very high (near 400watts) power usage for this card, where it even surpasses that on the oc bios.

I am using 2 cables one with 2 PCIE connectors and one solo, I'll add a third cable to the setup and see where we get, as far as my bios everything should be stock the only changes to my bios that have been made are to my fans which I disabled due to a cpu fan error and I use corsairs app to set my fans as all my case fans run through them. As a side note I did update the bios the other day as well.

 

[Vic 40]

As said above did i refer to the bios of the gpu which has a switch where you can go to OC mode, which i do not recommend looking at the already steep power usage with the stock bios.

 

[Colif]

GPU has a physical switch on it that lets you switch between bios, from what I can see its near the where those 2 cables go. I know there is a new beta bios for card that lets it use 450watts but its for overclockers and not recommended for normal users. You would have to download it so its not likely to be reason for above.

the only difference between both BIOS on the cards right now is a different fan curve - https://forums.evga.com/How-does-the-3080-FTW3-Ultra-OC-Bios-switch-work-m3115701.aspx

 

[Vic 40]

One review said that power usage still goes up when using that second bios.

 

[Colif]

its worth a try still. I was mainly just trying to locate the switch

 

[Vic 40]

Saw in that thread that one said that the power limit goes up as well and am still not sure that the psu will handle the spikes of this particular gpu, why i suggest to leave the switch alone for the increased power usage.

 

[TheKittenSlay3r]

Update It's been about 2 days since the last natural crash outside of me using occt power to force a crash, I downloaded GPUZ as well to monitor my power consumption and I'm having my doubts it was the PSU, Since I bent the pins back on the CPU socket I had one crash but it could be an outlier here as it was sometime right after but none since. I've also since switched from using two Molex powercables 1 with two 8 pin headers and went straight to three as for EVGA's requirements on the cards page. I've tried OCCT's again with three and was able to make it for a whopping 10 seconds this time as opposed to my normal crash as I pressed the button. Still possibly the PSU however I'm still monitoring my power usage spikes ect to find out exactly.

 

[Vic 40]

Might well be that with this particullar one with it's three 8pin pcie needs the 850watt just isn't enough to take care of the transient spikes the psu can give.