[SOLVED] Windows 10 BSOD loop + loss of video signal ?

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
There's been a lot of stuff happening to my PC so I apologize if this gets too long and messy.
Before I start describing what’s going on, I built it 6 months ago (it was my first build) and I never had an issue until now, these are the specs: https://pcpartpicker.com/list/RX3HMv

My problems started a week ago: I was playing a game when the screen suddenly turned black and I heard a buzzing sound from my headphones. I panicked so I pressed the power button and after this the PC was stuck in a BSOD loop giving me errors such as SYSTEM THREAD EXCEPTION NOT HANDLED or RESOURCE NOT OWNED.
Then I tried to boot it in safe mode but I had no video signal. I also tried to use the command prompt from the recovery options and the monitor suddenly turned black here as well. This always happened when I was pressing enter to execute certain instructions (one was bootrec related and the other one was "dir" as I was checking some folders).
Thinking that there was some issue with Windows I tried to repair/reinstall it using an USB but the monitor turned black yet again, usually after choosing the installation drive so when it was supposed to start the installation process.

At this point, I realized that maybe the issue was hardware related so I've focused on the pieces on the mobo to find the culprit.

The first thing I did was removing the SSD and trying to boot Ubuntu from an USB but yet another black screen greeted me as soon as the files verification process was done. The same thing happened if I removed the HDD.

Then I used memtest86.
The first time, after 2 hours, it froze after 2 passes and it only gave me a warning regarding vulnerability to "high frequency row hammer bit flips".
The second time it gave me 32 errors during the hammer test and it froze.
The third time I removed one of the RAMs and it gave me 16 errors during the same test and it froze yet again.
This is what the screen looked like after the three tests: View: https://imgur.com/a/Z7iJ0AC

Then I did 5-6 tests on the single RAMs by moving them around the slots and I got a pass every time (4 successful passes).
I also did a couple of tests with memtest86+ to see if anything changed and these passed as well.
After this I tried yet again to boot Ubuntu from that USB by using a single RAM each time but I had no success.

Now I'm stuck at this point as I'm not 100% certain it is the RAMs' fault, I don't know if it's normal for these to break both at once. But I think it would be even weirder for all the 4 slots to suddenly have issues.
I forgot to mention that every time I had no video signal the yellow (DRAM according to the manual) led stayed lit up.

I'm currently not able to do tests with other RAM sticks so I hope the information I gave is enough for a diagnosis.
I also got these two weird video glitches, each of them occurred only once, hopefully not because of the GPU: View: https://imgur.com/a/jZl7VFo

Speaking of the GPU, I couldn't test this one either. Theoretically the one from the PC I'm using right now should be compatible with the new mobo but because of a physical limitation I had to give up.

So in conclusion, since it’s not clear to me what the memtest86 results mean, I’d want to know if what I did so far would be enough for me to ask Amazon for a replacement of if there was anything else I could have overlooked (besides actually trying other parts).

Thank you.
 

Colif

Win 10 Master
Moderator
right now I don't know for sure if problem is ram or the gpu itself. as both could lead to a blank screen.

I would get pc looked at by a repair store, they should have known working parts they can swap in to isolate the cause.
display working in bios and memtest/prime doesn't necessarily mean its not GPU as they may run in low graphics modes, I know BIOS does.
 

Colif

Win 10 Master
Moderator
do you have latest bios for motherboard as it can help with ram problems

its odd for memtest to cause errors one time, and then not the next. Did you change XMP settings?

corsair LPX ram can be troublesome.
 

Colif

Win 10 Master
Moderator
Reactions: atom032

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
I've followed the instructions in this guide and let it run for 24 hours.
The only odd thing was that the screen would turn black (there would still be video signal) after 10/15 minutes but pressing any key would immediately show what was going on, so I assume it kept running with no errors for all this time.
Took a few pics just in case: View: https://imgur.com/a/mt3FNRu
 

Colif

Win 10 Master
Moderator
so cpu looks okay, and ram worked towards end, are you still unable to install windows though?

I would have thought that if it was a ram error the installer would have thrown errors at you before you chose a drive, as up until it copies itself onto a drive and restarts, its running off a ram drive

This always happened when I was pressing enter to execute certain instructions (one was bootrec related and the other one was "dir" as I was checking some folders).
maybe we looking in wrong place

are you able to test the nvme in another pc? since i assume PC isn't working still - https://www.silicon-power.com/web/download-ToolBox

this scans hdd but i doubt its that as you don't install win 10 on hdd (or I wouldn't) - https://www.seagate.com/au/en/support/downloads/seatools/seatools-win-master/
 

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
are you able to test the nvme in another pc?
Nope, I only have an 8+ year old prebuilt and I doubt I can put that on it...
I did another test with an Ubuntu USB on the HDD and physically removed the SSD. After the file systems check there was no video signal and the yellow (orange) led on the mobo was blinking (RAM again?)
 

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
3615Mhz CPU and 3596Mhz RAM doesn't seem like stock speed. Is this system overclocked?
As I said above, the only thing I did was enabling D.O.C.P following the second part of this guide ( https://www.asus.com/support/FAQ/1042256/ ) and changed RAM frequency to 3600 as it was set to 2133 by default. That was 6 months ago when I built the PC.
My guess is that memtest86+ doesn't show precise values and that's why they're like that
 

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
try setting ram to stock speed (2133) and see if you can install
Tried to install Windows and I got no video signal when it was "getting files ready for installation", at about 2%.
I also tried to enable safe mode and after rebooting itself, the monitor showed a command prompt for less than a second and it showed me the boot settings screen once again.
 

Colif

Win 10 Master
Moderator
right now I don't know for sure if problem is ram or the gpu itself. as both could lead to a blank screen.

I would get pc looked at by a repair store, they should have known working parts they can swap in to isolate the cause.
display working in bios and memtest/prime doesn't necessarily mean its not GPU as they may run in low graphics modes, I know BIOS does.
 

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
I would get pc looked at by a repair store, they should have known working parts they can swap in to isolate the cause.
Yeah, it seems bootable programs to diagnose GPUs aren't really a thing so I guess I have no choice.
Anyway, thanks for the help, really appreciate it as I've learned some new things from this.

I'll update this post when/if I manage to fix it.
EDIT: Took a gamble and removed a few parts of my old GPU to make it fit, I'm having the same problems so no doubt it's the RAMs.
EDIT2: New RAMs didn't fix the issues, maybe it's the MOBO... I'm so tired of this
 
Last edited:
Reactions: Colif

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
right now I don't know for sure if problem is ram or the gpu itself. as both could lead to a blank screen.
Sorry for bumping this again but I just tried a new mobo (same model) and it didn't fix a thing, do you think it could be the CPU at fault despite prime95 not failing?
 
Last edited:
GPU is definite suspect. It probably has failed. Replace it.

Test ram with lowered frequency settings.
DRAM voltage 1.35V​
Command rate 1T​
Latency settings 18-22-22-42​
Frequency 2133mhz​
Run memtest. If good, then increase ram frequency in steps. Test after each step. 2400mhz,2666mhz,2933mhz,3000mhz,3200mhz,3400mhz,3600mhz.
If a test fails, then return to previous frequency settings.

2933mhz should be achievable. Higher frequencies than that are not guarantied.
 
Reactions: atom032

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
GPU is definite suspect. It probably has failed. Replace it.
Test ram with lowered frequency settings-
As I briefly said above, I've already replaced the GPU with an old one (an R7 240) and replaced RAM with other sticks (same model but 3200 frequency) and the situation didn't change
 

atom032

Reputable
Aug 26, 2016
12
1
4,515
0
Did you run memtest on your new ram modules?
Probably need to fine-tune memory settings.
As mentioned before - 2933mhz should be safe to use. Above that - not guarantied.
I didn't as I assumed they'd be fine since they're new; I had left the default settings (stock speed which should be 2133) on the new mobo.
Guess I'll still give memtest a try, thanks for the tip
EDIT: tested with the frequencies above and even with 3200 (the advertised one) and didn't give me any errors

EDIT2: It was the CPU, problem solved!
 
Last edited:

ASK THE COMMUNITY

TRENDING THREADS