Question PC bluescreens and sometimes crashes without BSOD

[VitusB]

Hello.

For the past few months i've experienced a number of crashes and BSOD happening daily.
Most of the BSOD have the driver ntoskrnl.exe as the cause, which i've heard could be because of a hardware problem.
I've tried almost everything people have suggested, but nothing has helped.
Some things I've tried:

• Clean install
• MemTest86 (and tried another set of RAM)
• Installing another graphics card
• Stress tested CPU
• sfc /scannow
• Reinstalled graphics card drivers

If anyone have any ideas of how to solve it please let me know.

Some minidumps (I've attached a minidump from a driver verifier I just ran): https://drive.google.com/drive/folders/1gyb6GCE4dMGL_-ytXLeVucALeFgPLO5d?usp=sharing
Speccy: http://speccy.piriform.com/results/zlRGXMB1KWFtidMWtp1dO4Z

 

[VitusB]

Finally got a blue screen: https://drive.google.com/file/d/1Qekvg7aFZ4o-QeqLvLanbfNyz37AB5Ob/view?usp=sharing
(only the last one is from today, the rest are from the 18th)

I researched a bit on the internet and found this thread: https://answers.microsoft.com/en-us...1c0-13f3-46db-84ef-efabea965b08?auth=1&page=5
Seems like he had sort of a similar problem and was able to pass stress test (like me), but it ended up being the CPU which was the cause

 

[Colif]

i think Gardenman will look at it later on.

Love to get an answer on this thread, its getting tooo long.

 

[gardenman]

I ran the dump files through the debugger and got the following information: https://jsfiddle.net/toL6a2md/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.

Spoiler

File information:	111820-48125-01.dmp (Nov 18 2020 - 08:37:48)
Bugcheck:	ATTEMPTED_EXECUTE_OF_NOEXECUTE_MEMORY (FC)
Driver warnings:	*** WARNING: Unable to verify timestamp for amdkmdag.sys
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 23 Hour(s), 38 Min(s), and 55 Sec(s)

File information:	111820-44531-01.dmp (Nov 18 2020 - 12:05:15)
Bugcheck:	KMODE_EXCEPTION_NOT_HANDLED (1E)
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 3 Hour(s), 23 Min(s), and 48 Sec(s)

File information:	111820-41468-01.dmp (Nov 18 2020 - 13:57:00)
Bugcheck:	DRIVER_IRQL_NOT_LESS_OR_EQUAL (D1)
Driver warnings:	*** WARNING: Unable to verify timestamp for vgk.sys
Probably caused by:	memory_corruption (Process: vgc.exe)
Uptime:	0 Day(s), 1 Hour(s), 51 Min(s), and 07 Sec(s)

File information:	111820-398718-01.dmp (Nov 18 2020 - 16:50:02)
Bugcheck:	DRIVER_IRQL_NOT_LESS_OR_EQUAL (D1)
Driver warnings:	*** WARNING: Unable to verify timestamp for amdkmdag.sys
Probably caused by:	memory_corruption (Process: VALORANT-Win64-Shipping.exe)
Uptime:	0 Day(s), 2 Hour(s), 52 Min(s), and 25 Sec(s)

File information:	112220-345609-01.dmp (Nov 22 2020 - 16:28:28)
Bugcheck:	UNEXPECTED_KERNEL_MODE_TRAP (7F)
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 1 Hour(s), 20 Min(s), and 55 Sec(s)

Comment: Vanguard is still installed.

This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.

 

[VitusB]

Comment: Vanguard is still installed.

Isn't it only in the dumps from the 18th? I uninstalled Valorant shortly after

 

[gardenman]

Yes, 18th and 22nd. Ok, good to hear that it's uninstalled now. If you keep crashing upload more dumps.

 

[VitusB]

Actually I'm gonna send my pc to a repair shop today. But thanks for the help everyone. I'm not gonna close the thread yet if for some reason they can't fix it

 

[Darkbreeze]

👍

 

[VitusB]

Update: my pc still crashes. I'm really getting tired of this.

I got my motherboard replaced to a new one of the same model. But after installing Windows, drivers and some programs, it crashed.

That means that i've tried with different: RAM, GPU and Motherboard.
I've also tried stresstesting the CPU, where it passed everytime, but is there some way that it still could be faulty?

Here is a link to my dumpfiles: https://drive.google.com/file/d/1ViLZH1NQHXFBqxjs_Qepoa2kPn23816O/view?usp=sharing

 

[gardenman]

Here's the results for the 2 included minidumps: https://jsfiddle.net/s419r7L3/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.

Spoiler

File information:	123020-6046-01.dmp (Dec 30 2020 - 17:29:16)
Bugcheck:	IRQL_NOT_LESS_OR_EQUAL (A)
Driver warnings:	*** WARNING: Unable to verify timestamp for amdkmdag.sys
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 0 Hour(s), 24 Min(s), and 09 Sec(s)

File information:	123020-5500-01.dmp (Dec 30 2020 - 17:45:18)
Bugcheck:	KMODE_EXCEPTION_NOT_HANDLED (1E)
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 0 Hour(s), 08 Min(s), and 22 Sec(s)

This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.

 

[Darkbreeze]

So, try reinstalling Windows and DON'T install those "programs". Maybe one of those programs is what's causing the issue. Because there really isn't anything left that could be the problem other than the power supply or one of the programs or drivers you are installing.

The CPU, much as with anything, COULD be the problem, but it's incredibly rare that a CPU actually fails unless somebody has intentionally altered the voltage too high or too low, or has dropped the CPU at some point onto a (presumably) hard surface, or has bent one or more pins on the CPU, if it's an AMD model (Intel pins are on the motherboard). The fact that the CPU passes some tests doesn't necessarily mean anything though.

What "stess tests" did you run? WHICH specific tests IN the stress test utility? For how long?

 

[Darkbreeze]

Both of those dumps look to be related to memory corruption of some kind. Since you've run memory tests and tried different memory, as well as making sure the memory is installed in the correct slots, I doubt it actually IS the memory, so that pretty much leaves the power supply or something in Windows itself. Maybe a drive issue.

Any driver that loads into memory can trigger a dump related to memory, so it doesn't have to actually BE a memory issue to look that way.

 

[Colif]

top one is gpu drivers (again(as usual))
2nd not so clear
lots of bsod errors report memory corruption

So I assume repair store found nothing at all? Or did they suggest mb replacement?

i don't think its windows, thats for sure.
no valorant for a change either

things it isn't likely caused by -
MB - replaced. Still crashes
GPU - replaced
RAM - replaced
storage - installed on both ssd and hdd, crashes

CPU passes tests.
PSU?? https://www.anandtech.com/show/12070/the-corsair-tx550m-80plus-gold-psu-review/5#:~:text=Today we reviewed the least,an 80Plus Gold efficiency certification.

refresher -
CPU - Ryzen 3600
Mobo - B450 Tomahawk Max (replaced)
Ram - 16gb - replaced
GPU - RX 5700 XT (Card was replaced)
Storage - 1gb Seagate (tested with seatools for windows short test) installed win 10 on this, still crashes
250gb Samsung ssd
Other ssd (840 Pro)
PSU - Corsair TX550M, 550W PSU

PC crashes without BSOD - this shows its not windows. Hardware... but what?

 

[Darkbreeze]

Short test on Seagate drive tells us nothing. Need to run the long generic/Extended test. I would run the short DST/Quick test, for all drives, and then run the long generic/Extended, for all drives.

I'd also probably completely disconnect that 840 Pro from the system until you sort the problems out as I've seen some older 840 Samsung drives cause similar issues before in the past. It's old, so it's a prime suspect.

How old is that TX550m power supply?

How certain are we that the motherboard and graphics card were ACTUALLY replaced, rather than they just said they were "repaired" or "replaced"? I've seen this happen DOZENS of times where it's been said something was replaced but in reality it was the same hardware. And was also the problem.

 

[Colif]

im not sure if the parts list is currently accurate, thats just the parts i seen in the last 6 pages

 

[VitusB]

What "stess tests" did you run? WHICH specific tests IN the stress test utility? For how long?

I ran Prime95 a while back, I think it was called large FFT's, here's the results: https://drive.google.com/file/d/1QWdOAY3-4uzhUF8-yanltN40bP_74B7c/view?usp=sharing
I also ran Memtest 86 and let it finish without errors.

So I assume repair store found nothing at all? Or did they suggest mb replacement?

yeah they suggested mb replacement

PSU?? https://www.anandtech.com/show/12070/the-corsair-tx550m-80plus-gold-psu-review/5#:~:text=Today we reviewed the least,an 80Plus Gold efficiency certification.

yup i think that's the one

CPU - Ryzen 3600
Mobo - B450 Tomahawk Max (replaced)
Ram - 16gb - replaced
GPU - RX 5700 XT (Card was replaced)
Storage - 1gb Seagate (tested with seatools for windows short test) installed win 10 on this, still crashes
250gb Samsung ssd
Other ssd (840 Pro)
PSU - Corsair TX550M, 550W PSU

yes that's my parts. I got the 250 gb samsung ssd for christmas, which is why it doesn't show in my earlier dump files.

Short test on Seagate drive tells us nothing. Need to run the long generic/Extended test. I would run the short DST/Quick test, for all drives, and then run the long generic/Extended, for all drives.

I'd also probably completely disconnect that 840 Pro from the system until you sort the problems out as I've seen some older 840 Samsung drives cause similar issues before in the past. It's old, so it's a prime suspect.

ok i will do that.

How old is that TX550m power supply?

I bought it december 2nd, 2019.

How certain are we that the motherboard and graphics card were ACTUALLY replaced, rather than they just said they were "repaired" or "replaced"? I've seen this happen DOZENS of times where it's been said something was replaced but in reality it was the same hardware. And was also the problem.

it's only my motherboard which have been replaced. I just swapped my graphics card with another one i had lying around to see if it would still crash, which it did, so I assumed it wasn't the graphics card causing the crashes, and therefore i put my original gpu back (AMD RX 5700 XT).
and the store says that they've replaced my motherboard, but really I can't be sure

Btw I had my pc be in safe mode overnight where it didn't crash. I don't know if it means something as the pc crashes comes at very random times

 

[VitusB]

Short test on Seagate drive tells us nothing. Need to run the long generic/Extended test. I would run the short DST/Quick test, for all drives, and then run the long generic/Extended, for all drives.

I'd also probably completely disconnect that 840 Pro from the system until you sort the problems out as I've seen some older 840 Samsung drives cause similar issues before in the past. It's old, so it's a prime suspect.

I removed the 840 pro ssd, and ran both Short DST and long generic test, which both drives passed.
Here's my latest crashes: https://drive.google.com/file/d/1Az3WbPL42IQRtji3076EVCCf_w3-FsTC/view?usp=sharing

 

[gardenman]

I ran the dump files through the debugger and got the following information: https://jsfiddle.net/h179jr84/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.

Spoiler

File information:	123120-6250-01.dmp (Dec 30 2020 - 19:14:18)
Bugcheck:	DRIVER_OVERRAN_STACK_BUFFER (F7)
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 0 Hour(s), 24 Min(s), and 17 Sec(s)

File information:	123120-5625-01.dmp (Dec 30 2020 - 18:22:18)
Bugcheck:	IRQL_NOT_LESS_OR_EQUAL (A)
Driver warnings:	*** WARNING: Unable to verify timestamp for amdkmdag.sys
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 0 Hour(s), 36 Min(s), and 28 Sec(s)

File information:	123120-5546-01.dmp (Dec 31 2020 - 08:54:44)
Bugcheck:	IRQL_NOT_LESS_OR_EQUAL (A)
Probably caused by:	memory_corruption (Process: Discord.exe)
Uptime:	0 Day(s), 2 Hour(s), 02 Min(s), and 34 Sec(s)

Comment: GPU was blamed for at least 1 of these dumps. This was also the case in the previous 2 batches of dump files. You say that you've tried another GPU, but it's possible that you have 2 bad GPUs, or just bad drivers from AMD. That's just my opinion and just a guess, but it's an educated guess.

If the dumps keep pointing to the GPU, then the GPU is one of the issues. BSODs can be (and often are) caused by multiple issues at once, so you can have other issues also. But I still feel the GPU is one of those issues.

This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.

 

[Darkbreeze]

I ran Prime95 a while back, I think it was called large FFT's

So, that's really not a CPU stability stress test.

Download and run Realbench. Choose the stress test option. Select HALF the amount of your installed memory in the memory option box on the stress test page. Run it for as long as you can, up to potentially 8 hours (So overnight often is a good choice). If it passes 8 hours of Realbench, then it's definitely, or at least very unlikely, not a CPU problem.

Alternatively you can, and probably should, also run the extended Windows memory diagnostic test, because Memtest does not always catch all memory issues. And if you want to do some additional testing that does a good job of testing both the CPU and memory, then run Prime95 and choose the Blend option. Not Large FFT, not Small FFT (Which is best used for thermal testing of the CPU) and not Smallest FFT. It might also be a good idea when running ANY of the Prime95 tests to be sure and disable all of the AVX options shown on the main options page when you open it.

Be sure to choose the Extended option in the Windows memory diagnostic test, not the standard option. Running this test will likely take a long time, much like running Memtest for four passes does, but it tests different things and in a different way and may catch memory problems not found by Memtest86.

Microsoft Archives

www.techrepublic.com

 

[VitusB]

Download and run Realbench. Choose the stress test option. Select HALF the amount of your installed memory in the memory option box on the stress test page. Run it for as long as you can, up to potentially 8 hours (So overnight often is a good choice). If it passes 8 hours of Realbench, then it's definitely, or at least very unlikely, not a CPU problem.

I ran Realbench stresstest for 8 hours with 8 gb of RAM, and was able to complete it without crashing.
I'll run Windows memory diagnostics now

Edit: I ran Windows memory extended diagnostics and it found no errors.

 

[VitusB]

My latest bsod's: https://drive.google.com/file/d/1JCHveVQ3xwUXhnG-GMTYPDkBfqbzFeAA/view?usp=sharing

 

[gardenman]

I ran the dump files through the debugger and got the following information: https://jsfiddle.net/nu6szv7b/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.

Spoiler

File information:	010221-5531-01.dmp (Jan 1 2021 - 17:59:28)
Bugcheck:	IRQL_NOT_LESS_OR_EQUAL (A)
Probably caused by:	memory_corruption (Process: Discord.exe)
Uptime:	0 Day(s), 0 Hour(s), 19 Min(s), and 42 Sec(s)

File information:	010121-5593-01.dmp (Jan 1 2021 - 15:13:28)
Bugcheck:	IRQL_NOT_LESS_OR_EQUAL (A)
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 3 Hour(s), 49 Min(s), and 30 Sec(s)

File information:	010121-5578-01.dmp (Jan 1 2021 - 16:08:06)
Bugcheck:	DRIVER_IRQL_NOT_LESS_OR_EQUAL (D1)
Driver warnings:	*** WARNING: Unable to verify timestamp for amdkmdag.sys
Probably caused by:	memory_corruption (Process: System)
Uptime:	0 Day(s), 0 Hour(s), 53 Min(s), and 56 Sec(s)

File information:	010121-5500-01.dmp (Jan 1 2021 - 11:22:51)
Bugcheck:	IRQL_NOT_LESS_OR_EQUAL (A)
Probably caused by:	memory_corruption (Process: System)
Uptime:	1 Day(s), 2 Hour(s), 27 Min(s), and 28 Sec(s)

File information:	010121-5593-02.dmp (Jan 1 2021 - 17:39:08)
Bugcheck:	IRQL_NOT_LESS_OR_EQUAL (A)
Probably caused by:	memory_corruption (Process: javaw.exe)
Uptime:	0 Day(s), 1 Hour(s), 30 Min(s), and 24 Sec(s)

This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.

 

[Darkbreeze]

When the store supposedly "replaced" the motherboard, did THEY take it out of your case and put the new one in? Did you SEE them replace it, or did you do the replacement? Do you know for a fact that the motherboard came new out of a brand new, unopened box that was sealed? Because if not, I'd assume that they did NOT replace the motherboard and were lying. And yes, they do this all the time. I've personally seen supposedly reputable stores like Best Buy and other repair shops tell customers that they had replaced some piece of hardware, usually the motherboard, and yet when we checked afterwards the same markings that we had put on there to identify whether the board had been replaced FOR THIS EXACT REASON were still there.

It's very embarrassing for the store, I assure you, when you take it back and point it out to them that the only reason you had them do the replacement is because it was under warranty from them and they couldn't even be counted on to be honest about it.

In any case, I honestly don't know what else this could be at this point aside from that really. It really seems like it's got to either be a motherboard or graphics card issue, but certainly the PSU is always potentially to blame for anything that seems like, well, anything. You would think that a PSU or motherboard issue would result in a failure to run or pass any of the memory or stress tests you've run though, so, IDK.

If you haven't already I'd take the whole thing out of the case and bench test it. Maybe something is improperly installed. I'm out of ideas to be honest.

 

[VitusB]

When the store supposedly "replaced" the motherboard, did THEY take it out of your case and put the new one in? Did you SEE them replace it, or did you do the replacement?

I returned my old motherboard by mail, and got a new one sent to me which I then installed by myself. It looked like it was brand new, as there was plastic on it which I had to peel off.

In any case, I honestly don't know what else this could be at this point aside from that really. It really seems like it's got to either be a motherboard or graphics card issue, but certainly the PSU is always potentially to blame for anything that seems like, well, anything. You would think that a PSU or motherboard issue would result in a failure to run or pass any of the memory or stress tests you've run though, so, IDK.

If you haven't already I'd take the whole thing out of the case and bench test it. Maybe something is improperly installed. I'm out of ideas to be honest.

Well thanks for your help.
I'll try to reseat everything, and then since the PSU is the only thing i haven't tried replacing (apart from the CPU), I'll probably buy a new one.

 

[Colif]

i think top one is gpu drivers, based on previously having seen similar errors in past (I could say that about almost all errors though)
i can't tell with the 2nd one is
I don't need to look at 3rd. it tells me.
not sure about the last 2, 5th could be gpu again.

I wonder if this is another case of what John thinks is a bug in the bios/chipset drivers for B450 motherboards, as another long running BSOD post we have is same CPU/MB combination. It seems DB has 2 threads with the same CPU & MB combo and we been fighting a bug.

you might get this with a overheating problem but it seems that there is just some problem with the cpu and bios version on this chipset right now.
you might just lock down your cpu and wait for bios updates and chipset updates that AMD should release sometime this month.

you might lock down the cpu via
going into registry edit and changing this setting:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\amdppm
Start to have a value of 4

you should note the current value so you can later change it back.

link
Before you edit registry, create a backup - https://neosmart.net/wiki/backup-restore-registry/

 

[VitusB]

i think top one is gpu drivers, based on previously having seen similar errors in past (I could say that about almost all errors though)
i can't tell with the 2nd one is
I don't need to look at 3rd. it tells me.
not sure about the last 2, 5th could be gpu again.

I wonder if this is another case of what John thinks is a bug in the bios/chipset drivers for B450 motherboards, as another long running BSOD post we have is same CPU/MB combination. It seems DB has 2 threads with the same CPU & MB combo and we been fighting a bug.


link
Before you edit registry, create a backup - https://neosmart.net/wiki/backup-restore-registry/

Interesting.
I don't know much about the whole registry thing, but can you explain what it is I'm changing?