paul5000

Reputable
Feb 28, 2015
27
1
4,535
I recently rebuilt my system: upgraded motherboard, processor, video card and RAM. Since the upgrade I've been getting lots of Blue Screen of Death crashes. Mostly on boot - often the system will attempt to boot once or twice and get a BSOD, and then start working ok on the 2nd or 3rd try. Occasional BSOD crashes during normal operation as well. I will list the BSOD screens that I've seen below.

My system:

ASRock B450M
Ryzen 5 2600
G Skill DDR4-2400 8Gb x2
GTX 1060 6Gb
Samsung 850 EVO SSD 256Gb
Seagate 1TB
Windows 10 Home 64 bit (upgrade from Win 7)

One thing I have noticed and do not know if it's related: My mobo seems to think my RAM is 2133. It's set to auto-detect. If I manually select 2400, my system will not boot at all. This makes me wonder if my RAM is a problem?

What I’ve tried, to no effect:
  • Updating Bios
  • Uninstalling & reinstalling video card driver
  • Windows memory diagnostic
  • Mem Test (only 1 pass though)
  • System File checker

I get all these BSODs:
KMODE EXCEPTION NOT HANDLED
MEMORY MANAGEMENT
BAD POOL HEADER
DRIVER IRQL NOT LESS OR EQUAL
SYSTEM SERVICE EXCEPTION
IRQL NOT LESS OR EQUAL
SYSTEM SERVICE EXCEPTION
PAGE FAULT IN NONPAGED AREA
ATTEMPTED WRITE TO READONLY MEMORY
PAGE FAULT IN NONPAGED AREA
KERNEL SECURITY CHECK FAILURE
KERNEL MODE HEAP CORRUPTION
CLOCK WATCHDOG TIMEOUT
BAD SYSTEM CONFIG

I’m usually ok at troubleshooting this stuff but this one’s beyond me.

Happy to post any additional info that would help.
Thanks for your help!!
 
Solution
Random and various BSOD errors can often mean issues with memory. This is further likely when the BSODs blame ntoskrnl (which means the OS doesn't know a specific driver/file to blame). Check event viewer or use NirSoft's bluescreenview to analyze the BSOD minidumps and view summary details on each crash.

Try running memtest for up to 8 passes. If you get errors, then try to isolate your issue to a specific memory stick or slot (e.g. test each stick separately, then if one stick shows errors try that in a different slot, etc.). Faulty RAM can sometimes pass memtest without errors, but it's a good place to start.

Ideally you can try just swapping out the memory and seeing if these issues persist. If they do then there might be an issue...

btmack

Commendable
Aug 27, 2016
6
1
1,515
Random and various BSOD errors can often mean issues with memory. This is further likely when the BSODs blame ntoskrnl (which means the OS doesn't know a specific driver/file to blame). Check event viewer or use NirSoft's bluescreenview to analyze the BSOD minidumps and view summary details on each crash.

Try running memtest for up to 8 passes. If you get errors, then try to isolate your issue to a specific memory stick or slot (e.g. test each stick separately, then if one stick shows errors try that in a different slot, etc.). Faulty RAM can sometimes pass memtest without errors, but it's a good place to start.

Ideally you can try just swapping out the memory and seeing if these issues persist. If they do then there might be an issue with the motherboard itself. Either way it might help to either isolate the problem to RAM or rule that out before you proceed.

Best of luck.
 
Solution

paul5000

Reputable
Feb 28, 2015
27
1
4,535
Does this help? (from BlueScreenViewer)

==================================================
Dump File : 031719-7468-01.dmp
Crash Time : 2019-03-17 12:05:07 PM
Bug Check String : CRITICAL_PROCESS_DIED
Bug Check Code : 0x000000ef
Parameter 1 : ffffac8fc86a6080 Parameter 2 : 0000000000000000
Parameter 3 : 0000000000000000 Parameter 4 : 0000000000000000
Caused By Driver : ntoskrnl.exe
Caused By Address : ntoskrnl.exe+1b35e0
File Description : NT Kernel & System
Product Name : Microsoft® Windows® Operating System
Company : Microsoft Corporation
File Version : 10.0.17763.379 (WinBuild.160101.0800)
Processor : x64
Crash Address : ntoskrnl.exe+1b35e0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\WINDOWS\Minidump\031719-7468-01.dmp
Processors Count : 12
Major Version : 15
Minor Version : 17763
Dump File Size : 716,084
Dump File Time : 2019-03-17 12:06:15 PM
==================================================

==================================================
Dump File : 031719-6265-01.dmp
Crash Time : 2019-03-17 12:01:16 PM
Bug Check String :
Bug Check Code : 0x00000139
Parameter 1 : 0000000000000003 Parameter 2 : fffffa0077b6b820
Parameter 3 : fffffa0077b6b778 Parameter 4 : 0000000000000000
Caused By Driver : Ntfs.sys
Caused By Address : Ntfs.sys+112c90
File Description : NT File System Driver
Product Name : Microsoft® Windows® Operating System
Company : Microsoft Corporation
File Version : 10.0.17763.1 (WinBuild.160101.0800)
Processor : x64
Crash Address : ntoskrnl.exe+1b35e0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\WINDOWS\Minidump\031719-6265-01.dmp
Processors Count : 12
Major Version : 15
Minor Version : 17763
Dump File Size : 983,652
Dump File Time : 2019-03-17 12:01:58 PM
==================================================

==================================================
Dump File : 031619-6531-01.dmp
Crash Time : 2019-03-16 11:24:51 AM
Bug Check String : CRITICAL_PROCESS_DIED
Bug Check Code : 0x000000ef
Parameter 1 : ffffaa8f602dd2c0 Parameter 2 : 0000000000000000
Parameter 3 : 0000000000000000 Parameter 4 : 0000000000000000
Caused By Driver : ntoskrnl.exe
Caused By Address : ntoskrnl.exe+1b35e0
File Description : NT Kernel & System
Product Name : Microsoft® Windows® Operating System
Company : Microsoft Corporation
File Version : 10.0.17763.379 (WinBuild.160101.0800)
Processor : x64
Crash Address : ntoskrnl.exe+1b35e0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\WINDOWS\Minidump\031619-6531-01.dmp
Processors Count : 12
Major Version : 15
Minor Version : 17763
Dump File Size : 731,548
Dump File Time : 2019-03-16 11:26:25 AM
==================================================

==================================================
Dump File : 031619-8718-01.dmp
Crash Time : 2019-03-16 9:25:11 AM
Bug Check String : PAGE_FAULT_IN_NONPAGED_AREA
Bug Check Code : 0x00000050
Parameter 1 : ffffa98a9e8b4ce8 Parameter 2 : 0000000000000000
Parameter 3 : fffff80704040880 Parameter 4 : 0000000000000002
Caused By Driver : ntoskrnl.exe
Caused By Address : ntoskrnl.exe+1b35e0
File Description : NT Kernel & System
Product Name : Microsoft® Windows® Operating System
Company : Microsoft Corporation
File Version : 10.0.17763.379 (WinBuild.160101.0800)
Processor : x64
Crash Address : ntoskrnl.exe+1b35e0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\WINDOWS\Minidump\031619-8718-01.dmp
Processors Count : 12
Major Version : 15
Minor Version : 17763
Dump File Size : 1,213,596
Dump File Time : 2019-03-16 9:26:09 AM
==================================================
 

btmack

Commendable
Aug 27, 2016
6
1
1,515
Hopefully someone else will chime in with a second opinion, but speaking from my own experience those dumps make me strongly suspect the memory.

I'd recommending purchasing new RAM, installing it, and seeing how the system performs after that. You can still give memtest a try to obtain further evidence of whether your current RAM is faulty. If the new RAM doesn't alleviate your issue then it may be the motherboard, or some other issue (you could return the newly purchased RAM in that case).

IIRC Ryzen boards are more finicky with RAM support/compatibility, so you may want to visit the support page from your motherboard and see if there's a list or pdf of supported/compatible memory SKUs. I'm not sure if that improved with the 2nd gen of Ryzen hardware.

Since you are using the same OS as your previous configuration, you may also need to update/change other drivers (I see that you did this with the graphics driver). You could also try booting into safe mode or using a clean boot to see if this might be driver-related.
 
Last edited:

paul5000

Reputable
Feb 28, 2015
27
1
4,535
update: Memtest came up with no errors.

I tried various combinations of swapping RAM sticks. What I discovered is that with two sticks of RAM installed (including swapped positions), I get the BSOD crashes. But if I have only one stick of RAM installed (either stick, either slot) the system was stable.

THEN, I got the genius idea to try it with 2 sticks of RAM in the alternate slots (the 2 slots I now assume you're only supposed to use when you have 4 sticks of RAM). Windows did not boot. I went back to the proper slots / 1 stick of RAM, but the screen is all glitched - green with horizontal bars and you can barely make out what's on the screen. Here's a photo. And I can't seem to get things stable again. Whoops!

Thanks for your help!

dmwfev.jpg
 

athguy

Honorable
May 7, 2015
58
1
10,535
I
update: Memtest came up with no errors.

I tried various combinations of swapping RAM sticks. What I discovered is that with two sticks of RAM installed (including swapped positions), I get the BSOD crashes. But if I have only one stick of RAM installed (either stick, either slot) the system was stable.

THEN, I got the genius idea to try it with 2 sticks of RAM in the alternate slots (the 2 slots I now assume you're only supposed to use when you have 4 sticks of RAM). Windows did not boot. I went back to the proper slots / 1 stick of RAM, but the screen is all glitched - green with horizontal bars and you can barely make out what's on the screen. Here's a photo. And I can't seem to get things stable again. Whoops!

Thanks for your help!

dmwfev.jpg
Could be wrong but this suggests a problem with the Mobo at the ram slots. And with that visual glitch maybe your on board gpu is getting affected too
 
It certainly sounds like something RAM related; the BSODs are the one's I'd typically expect from a RAM related problem.

Lets go back a few steps: Check the BIOS and make sure the RAM is configured properly (if all else fails, the XMP profile should work). It's possible the RAM is misconfigured and is causing problems when two sticks are used.
 

paul5000

Reputable
Feb 28, 2015
27
1
4,535
Thanks. I don't have the glitchy green screen anymore; seems to have fixed itself without me doing anything differently.

I did have an issue with the mobo in auto-detect mode putting the RAM at 2133 when it is 2400.

I enabled the XMP profile. With one stick of RAM, all was good. I put in both, and got a BSOD on boot. :( I can post a BIOS screenshot of settings if you want.

I notice that my BIOS version is 2.00 but 3.30 is available. Should I update? What else to try?

Thanks for the help!
 

paul5000

Reputable
Feb 28, 2015
27
1
4,535
Thanks everyone for all the help! Updating the BIOS to newly released versions reduced my crashes a lot, but not entirely. So I replaced the RAM about a month ago and the system has been 100% good since then.

I would like to sell the RAM that I replaced if I can. It never failed any diagnostic tests .... but how safe is it to assume the RAM is fine but there was just a mobo compatibility issue?
 
Memory Tests sometimes can't find errors that have a low probability of occurring; its possible, for instance that one of the sticks needs slightly more then the normal amount of voltage to reliably switch states, for example. But a mobo compatibility problem isn't uncommon either.