Question Multiple BSOD's (after installing 21H2 or changing RAM ??)

[meerkato]

Hi,

Today i've experienced three BSOD's which seem to have different causes while running Windows 10 Pro. The computer was running fine until yesterday. Since yesterday i did two things i changed RAM sticks from 2x16GB to 2x32GB to 2x8GB (all Corsair LPX pairs) and i installed the 21H2 update for Windows. I now wonder what the problem is the RAM or the Win update or something else?

Hardware:
Intel 9900K
ASUS Z390-F Gaming (BIOS 2004 is up to date)
iGPU UHD 630
Corsair Vengeance LPX CMK16GX4M2B3000C15 (runs at 2133)
PSU BeQuiet (can't remember/see what model exactly because of the case)

What i tested so far (currently 2x8GB RAM installed)

1. memTest86 (ran a full pass)
2. sfc /scannow
3. checkdisk
4. disabled fast startup
5. disabled XMP in BIOS

Everything looks good so far. Drivers look up to date too. No idea what other things i could test.

Any ideas?

 

[meerkato]

You seem to be avoiding a clean install of Windows. What is your C: drive and how much free space do you have on it?

Yes, because this is a DAW and a nightmare to reinstall. I also just installed it less than a year ago.

I've currently 2x32GB installed and Windows says the Pagefile is 9GB. My c drive is just 120GB total and 50GB free space left.

 

[DavidM012]

Well you couldn't rule out the low drive space as the source of a PAGE_FAULT_IN_NONPAGED_AREA (50). Windows says the page file is 9gb - while the system is idle.

It's far too small for 64gb of memory. When you open apps and load data into them anything open in the memory could be paged out to the disk & if you're using a lot of memory and the page file runs out of disk space or simply can't grow to write any more to it- it may crash.

Thing about the 64gb kit being technically strictly not compliant with the memory support list only muddies the picture more.

So open your apps and load some data - see if it crashes
Open task manager / performance - for an idea of what it's using.

You could buy a 500gb or 1tb drive and use disk genius to clone or migrate the drive without reinstalling with a bootable usb - and resize the partition (volume) after migrating so there's more free space.

Looks like your board supports pci-e 3.0 rather than 4.0 but you can use a pci-e 4.0 drive on it they are backward compatible, so it only depends on what's available at a good price.

Memtest86 - sit through 64gb ram benchmark test isn't much fun either. Your cooling system needs to be in order and good too because it stresses the cpu. Also need Rufus to create the bootable usb.

Looked at your board site & manual :


1 x M.2 Socket 3, with M key, type 2242/2260/2280 storage devices support (both SATA & PCIE mode)*1
Intel® Z390 Chipset :
1 x M.2 Socket 3, with M key, type 2242/2260/2280/22110 storage devices support (PCIE 3.0 x 4 mode)
6 x SATA 6Gb/s port(s)


However it still isn't 100% clear if you have 1 or 2 M.2 drive ports on since there are no diagrams in the manuals. There usually is so I don't know why. it's because you have a problem and need it.

1 that's also Sata compatible and 1 that is Nvme only.

So it could be the page file size basically since your drive is small. Do you know if it's Sata 2.5 ssd, or nvme or sata m.2?

Use hwinfo to identify your system components.

 

[DavidM012]

Some nvme reviews - these are the performance drives you could get something else or just a pci-e 3.0 drive like one of the other western digital sn series, for example.

Could also try monitoring page file usage in performance monitor - howto - but if you're getting a crash, don't see how you're going to read any informative data.

 

[meerkato]

@DavidM012

Thanks for your input/help. I'll have a look at the Pagefile. I may fix it to say 8GB. It should work fine that way. For now it's controlled by Windows. I need to regain trust in this computer first before manually adjusting the size of the Pagefile.

 

[meerkato]

I still get random freezes. I wonder why i don't get BSOD's anymore but freezes instead.

 

[Colif]

i don't think page file is the cause of a page fault in non paged area. Page fault in non paged area is a driver error. The non paged area is ram, not page file (its in the name, its non paged as drivers have to run in ram)

Unless he uses all 64gb of ram I doubt page file size is a problem. Windows only uses what it needs. I have 32gb of ram but page file is only 2gb. Since its not creating dumps, it doesn't need the 64gb on drive.


Freezing without BSOD means its likely to be hardware problem.

 

[meerkato]

Freezing without BSOD means its likely to be hardware problem.

But how do i trace that? In Windows events it says something like "unexpected shutdown", so i don't get a hint at was isn't working. The only thing i have now is the three different mini dumps when the BSOD's occurred.

 

[Colif]

Windows may not know either.

Do you still have the minidumps?

1. Open Windows File Explore
   
2. Navigate to C:\Windows\Minidump
   
3. Copy the mini-dump files out onto your Desktop
   
4. Do not use Winzip, use the built in facility in Windows
   
5. Select those files on your Desktop, right click them and choose 'Send to' - Compressed (zipped) folder
   
6. Upload the zip file to the Cloud (OneDrive, DropBox . . . etc.)
   
7. Then post a link here to the zip file, so we can take a look for you . . .
   

So did problem only start after swapping ram?

Have you tried the older sticks to see if it still occurs?

Thing about the 64gb kit being technically strictly not compliant with the memory support list only muddies the picture more.

this alone could be all it takes, ram might test fine but if its just a little bit off in specs, it can cause odd errors.

But if you have dumps I can at least look at them.

 

[meerkato]

Here are my mini dumps : https://c.web.de/@728170749533094581/s1ylxKQFQUCp8Sb7iofO6w

The freezes also occurred when i had another kit (2x8GB) installed. So i don't think the RAM is the compelling reason, but who knows.

@Colif Thanks for offering your help

 

[Colif]

Probably means its not ram causing freezes.

Conversion of dumps

report - Click run as fiddle to see report


File: 102222-5093-01.dmp (Oct 23 2022 - 02:57:37)
BugCheck: [SYSTEM_SERVICE_EXCEPTION (3B)]
Probably caused by: memory_corruption (Process: svchost.exe)
Uptime: 0 Day(s), 0 Hour(s), 01 Min(s), and 16 Sec(s)

File: 102222-5000-01.dmp (Oct 22 2022 - 23:40:37)
BugCheck: [DPC_WATCHDOG_VIOLATION (133)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 00 Min(s), and 49 Sec(s)

File: 102222-4890-01.dmp (Oct 23 2022 - 02:06:41)
BugCheck: [PAGE_FAULT_IN_NONPAGED_AREA (50)]
Probably caused by: memory_corruption (Process: csrss.exe)
Uptime: 0 Day(s), 0 Hour(s), 33 Min(s), and 23 Sec(s)


What USB drives do you have? external hdd?
2nd crash looks like an operation on an external drive.
mentions usbstor.sys, USBhub3,sys, USBXHCI.sys

So what webcam do you have?

Jun 26 2009

synusb64.sys

Steinberg Anti Piracy driver or Suyin USB Webcam 64-bit driver

as that driver isn't likely to be windows 10
or is part of

Oct 28 2020

iLokDrvr.sys

USB iLok Device Driver (PACE Anti-Piracy, Inc.)

 

[meerkato]

@Colif Thanks for having a look at it

The two devices are USB dongles (a Steinberg eLicencer and an iLok). They are for licensing purposes. For the iLok i found an update and installed it, but it didn't prevent freezes.

I have also have two external hdd (asmedia controller) enclosures connected, but they are turned of.

Edit: found an update for the eLicencer too and just installed it.

 

[meerkato]

I also wanted to add that this machine was turned off for 3-4 months before i started to mess with it again a few days ago. Mainly installing 21H2 update and changed the RAM like i described earlier.

 

[meerkato]

Something stange just happend. I had two Firefox windows open one showing a website and another playing a YT video to provoke a freeze. Then i turned the monitor off. When i turned the monitor on, both Firefox windows were closed. Is this normal behaviour? Never noticed something like this.

 

[Colif]

do you always have to have the USB attached? just curious if we could take it off just to test if it makes any difference

Browser windows shouldn't shut just because monitor is off.

there is a newer version of win 10 - 22H2 - but I doubt it change anything.

so you just using IGPU?

Try running this and see if it offers any new drivers - https://www.intel.com.au/content/www/au/en/support/intel-driver-support-assistant.html

 

[meerkato]

do you always have to have the USB attached? just curious if we could take it off just to test if it makes any difference

Yes, both are always attached. I removed them for now.

so you just using IGPU?

Yes, no other graphics card installed

Try running this and see if it offers any new drivers - https://www.intel.com.au/content/www/au/en/support/intel-driver-support-assistant.html

It says everything is up to date.

 

[meerkato]

If i only knew how to reproduce/provoke the freezes. Letting the computer run until it crashes is very cumbersome. I wonder why RAM tests and Prime95 didn't crash right away.

Edit: removing the dongles didn't help -> freezed again

 

[meerkato]

Just running memTest86 again, no errors so far. I'm preparing a USB stick for a fresh Windows 10 install wondering if i should go Windows 11 instead.

 

[meerkato]

I've just reinstalled Windows 10 on another SSD, but when running Windows Update (which installed 22H2) i got a BSOD again.

Here is the mini dump

https://c.web.de/@728170749533094581/ivzEybGGQAu8b9_LmqtAjA

 

[meerkato]

I did some further testing using OCCT. I let the CPU test run for a very short time, maybe 3 minutes each.

RAM at 2133 (XMP off)
-> immediately starts throwing "errors on physical core x y z"

RAM at 3200 (XMP on)
-> no errors

Then a full pass using Intel Processor Diagnostic Tool
-> no errors

Any ideas what that means?

Edit: My BIOS settings are quasi "optimized defaults". Just a few less important settings (ie audio off).
Edit 2: repeated the OCCT test, same result. Very stange since most people seem to have problems the other way around (XMP on).

 

[DavidM012]

Is the bsod/error only occurring with the 2x64gb memory kit installed? Have you tried it with the 32gb or 16gb kit? Also what 'ssd' are you using? Could be Sata 3 2.5inch, Sata 3 M.2 or M.2 NVMe.

Old ropey sata 3 data cable? On the drive hosting the OS could cause some pain.

Going windows 11 won't solve a hardware problem.

Maybe that non-compliant with the support list 64gb kit is coming back to bite you. When things are out of spec eg. non compliant memory / drive too small for page file, it makes it harder to trace the error since it could be one or several things.

The cpu failing a stress test at default settings isn't a good sign when you're getting a bsod - to my mind a system freeze with no numlock or capslock response & mouse pointer freeze is a cpu stop.

Inspect CPU socket for any misaligned pins - post a photo here if you see anything suspicious.

What's your PSU exactly again? Bequiet something. PSU Tier list - why could it be relevant, voltages of course. Does hardware monitor show the 12v, 5v and 3.3v values are within +/- 5%

Also there a 4 tiers of bequiet PSU - Tier A Dark Power Tier B Pure Power Tier C Dark Power Pro Tier D (iGPU system only) Pure Power L8

Well whatever guess it should be OK since you are using iGPU only still, low quality isn't good quality regardless. Try to be concise, since details are important.

The exact power supply matters particularly if/when you decided to upgrade the gpu.

The system running quite barebones - cpu, mobo, mem, drive, power supply, case - not too much to go wrong. I suppose you are using the standard Intel cooler? Are you familiar with it's usual operating temperature range?

To my mind, one should always 'overkill' cpu cooling regardless of whether it's OK to have a mediocre cooler. The only way to know is to make certain.

*Attaching a cooler after overheating a cpu won't remediate damage. The CPU Is susceptible to volts and heat. Argue that it's ok to be mediocre about power delivery and cooling.

 

[meerkato]

Is the bsod/error only occurring with the 2x64gb memory kit installed? Have you tried it with the 32gb or 16gb kit?

BSOD's / freezing occured also when i had the 2x8GB kit installed. I tend to think RAM is not the problem.

What's your PSU exactly again? Bequiet something. PSU Tier list - why could it be relevant, voltages of course. Does hardware monitor show the 12v, 5v and 3.3v values are within +/- 5%

Also there a 4 tiers of bequiet PSU - Tier A Dark Power Tier B Pure Power Tier C Dark Power Pro Tier D (iGPU system only) Pure Power L8

I'd say it's a Straight Power. It's definitely not a Dark Power. So Tier B. I'll have a look at the voltages.

To my mind, one should always 'overkill' cpu cooling regardless of whether it's OK to have a mediocre cooler. The only way to know is to make certain.

Cooling is definitely not a problem. I've a Be Quiet Dark Rock Pro 4. I get 50-60°C max when stress testing.

 

[meerkato]

Had a look at voltages under stress. Looks good, no signifcant drops.

At this point of time i'd say it's not the PSU, RAM or OS. That leaves the CPU / MOBO as main suspects. When Intel Processor Diagnostic Tool is right, it's not the CPU.

I also wonder whether the BSOD's first started when i turned XMP off. I think i had it on before all the time. I unfortunately can't remember exactly.

 

[meerkato]

Some thoughts/oberservations

- My last BSOD happend yesterday amid installing the 22H2 update. It crashed when the update did a restart. After that 22H2 got installed successfully and i didn't get a BSOD since then.
-> just crashed, now testing with XMP enabled

- My machine never crashes when not in Windows. (for example running memTest86)

 

[Colif]

Tried running Prime 95?
it checks RAM, CPU & PSU
https://www.guru3d.com/files-details/prime95-download.html
Prime 95 Instructions - https://appuals.com/how-to-run-a-cpu-stress-test-using-prime95/

it takes about 17 hours unless it gets an error. Any errors will show as a text file in the folder its installed in.

a common cause is rounding, but wait and see if you get that error.

 

[meerkato]

Tried running Prime 95?
it checks RAM, CPU & PSU
https://www.guru3d.com/files-details/prime95-download.html
Prime 95 Instructions - https://appuals.com/how-to-run-a-cpu-stress-test-using-prime95/

it takes about 17 hours unless it gets an error. Any errors will show as a text file in the folder its installed in.

a common cause is rounding, but wait and see if you get that error.

I did run Prime95 a few times but nowhere near as long as 17 hours. Since i don't overclock in general i don't see the need to run those test for hours. If something is fundamentally wrong it should be obvious a lot earlier with all different kind of tests. I seem to have located an issue without running those tests for hours: OCCT CPU test throws errors on physical cores as soon as it's started when i have XMP off. The same test doesn't throw those errors when i enable XMP. Those results were gained with just running the test for a few minutes. I now know there is probably something wrong with the CPU and/or MOBO. I mean i dont want to say that it's not useful to run tests for hours, but i think in my case it may not be needed. Although i may run Prime95 again for a little longer.

To me it would be more useful to know what's behind those OCCT errors in relation to XMP on/off. It does look like there is the answer to my BSOD's in there somewhere. But i don't have a clue how to proceed from here.