Possible motherboard failure - Seemingly random black screens of death/crashes in games and web browser

[TheSillyMuffin]

Hello everyone

Firstly, I apologize if I miss anything regarding posting rules. I will happily provide any information that is missing/is needed.

Secondly, crash dumps are not present on my PC because for the entire duration of these issues, they've always been black screens and not blue screen issues. I havent managed to find any crash information on my PC, but if someone has a different method I could try in my ignorance, Id happily do so.

The actual issue starts here:

 

For the past year or so Ive had crashing/freezing issues with my semi-new gaming PC. A year is a long time, I know, but the problem is that 95% of games work flawlessly and the last 5% is riddled with issues. Approximately 500 hours of game time were spent across 50+ games with no issues at all, but in games like Battlefield 4, Titanfall 2, Tomb Raider 2013 and Vanquish, black screen crashes have been frequent. Some settings Ive fiddled with have made Battlefield 4 and Titanfall 2 relatively stable, but the other games continue to have issues.

 

Every time a crash occurs, both my monitors instantly turn off, theres 2-5 seconds of glitched sound, sounds that were just playing looping around and glitching. The PC then shuts down and instantly restarts itself.

AMDs control panel software informs me that default settings have been restored because of a crash, and thats it. No crash dumps or anything.

Similar issues have happened while browsing the internet, specifically multiple twitch.tv tabs at the same time, altough a few times a single tab has caused it.

 

Things Ive tried to resolve this issue:

Replacing RAM

RMA:ing the GPU, no issues found.

Replacing PSU

Undocking and redocking CPU

Previous GPU drivers

Overclocking, underclocking

Memory testing via a few different applications

Windows reinstall, PC reset

 

The only thing remaining in my mind is the motherboard, but testing that is virtually impossible for me, so Im out of options and ideas.

 

PC PARTS LIST:

CPU: Ryzen 7 3700x

GPU: RX 5700 XT Red Devil

RAM: 16GB 3200mhz - 2 x 8GB

PSU: ROG STRIX 850W

MOTHERBOARD: ROG STRIX X570-F

MEMORY: 500GB 970 EVO SSD + 2TB 860 QVO

OS: x64 Windows 10

 

 

So am I just shit out of luck when it comes to solutions or is the only chance for a fix disassembling the motherboard?

As a relatively ignorant PC noob, this has been incredibly time consuming, difficult and frustrating. Any and all help is appreciated.

 

EDIT: Added motherboard and OS, because Im stupid and I forgot.

[bob.blunderton]

Did you list the motherboard model on an issue describing what is likely a motherboard (or less-likely a PSU) issue?

[Aibakins]

Do you have your GPU overclocked in any sort of way? Sometimes some applications dont really like that and kill itself.

[TheSillyMuffin]

1 minute ago, bob.blunderton said:

Did you list the motherboard model on an issue describing what is likely a motherboard (or less-likely a PSU) issue?

I edited it in now. Sorry. This is what happens when you start your day with looking at this stuff instead of having some coffee

1 minute ago, Aibakins said:

Do you have your GPU overclocked in any sort of way? Sometimes some applications dont really like that and kill itself.

All of my settings are completely default for GPU.

Ive enabled RAM DOCP in BIOS, to make it run at its marketed values, and Ive slightly increased my CPU voltage because that was suggested by other people. Neither have had any effect on this particular issue.

[Aibakins]

10 minutes ago, TheSillyMuffin said:

Similar issues have happened while browsing the internet, specifically multiple twitch.tv tabs at the same time

Try using 1 stick of ram, If crash happens, use other stick of ram and try to replicate the issue. If so then it probably is CPU/MOBO issue

[TheSillyMuffin]

2 minutes ago, Aibakins said:

Try using 1 stick of ram, If crash happens, use other stick of ram and try to replicate the issue. If so then it probably is CPU/MOBO issue

Alright, which RAM slots do you suggest for this test? Does it matter at all?

[Aibakins]

Just now, TheSillyMuffin said:

Alright, which RAM slots do you suggest for this test? Does it matter at all?

The ones you are using right now, actually how do you have your ram setup like oxox or ooxx or xxoo or xoxo, or xoox or oxxo.

 

your motherboard manual recommends, ooox, or oxox, or xxxx. Try that first

[TheSillyMuffin]

1 minute ago, Aibakins said:

The ones you are using right now, actually how do you have your ram setup like oxox or ooxx or xxoo or xoxo, or xoox or oxxo.

 

your motherboard manual recommends, ooox, or oxox, or xxxx. Try that first

I have my RAM in slots 2 and 4, or oxox

And alright, I'll try this.

[bob.blunderton]

Definitely get the latest BIOS.  Make sure your RAM is on the QVL for the motherboard, goto ASUS site for that.  Make sure to use the right slots, AMD boards differ from intel boards.  Use the correct layout per the manual.

Do the games you're finding that fail out / crash have higher average CPU load than others?  Games will ferret out a cpu-to-memory issue faster than others the harder they use the processor.

Run your ram at the JEDEC spec speed like you'd get at first boot (not the XMP / DOCP speed), and try it with one stick if you still get crashes.  Yes it will be a bit pokey but it'll show you if you have an issue there.

You may have to add a touch of voltage to the memory (try adding 0.01v to it, if it's 1.35 make it 1.36 or 1.365) to get it stable at it's recommended speed or loosen the timings a tiny bit, my first-gen triple-channel intel system was like that with OCZ memory.  Did work fine after that though, and didn't have to do that anymore with a westmere CPU later down the road.

 

You may have to enable crash dump reporting if you are getting an instant reset on your system.  Some systems have it disabled, but getting a black screen before the software can catch it usually means something went wrong in CPU or memory/motherboard land before the OS could catch it.

[TheSillyMuffin]

2 minutes ago, bob.blunderton said:

Definitely get the latest BIOS.  Make sure your RAM is on the QVL for the motherboard, goto ASUS site for that.  Make sure to use the right slots, AMD boards differ from intel boards.  Use the correct layout per the manual.

Do the games you're finding that fail out / crash have higher average CPU load than others?  Games will ferret out a cpu-to-memory issue faster than others the harder they use the processor.

Run your ram at the JEDEC spec speed like you'd get at first boot (not the XMP / DOCP speed), and try it with one stick if you still get crashes.  Yes it will be a bit pokey but it'll show you if you have an issue there.

You may have to add a touch of voltage to the memory (try adding 0.01v to it, if it's 1.35 make it 1.36 or 1.365) to get it stable at it's recommended speed or loosen the timings a tiny bit, my first-gen triple-channel intel system was like that with OCZ memory.  Did work fine after that though, and didn't have to do that anymore with a westmere CPU later down the road.

 

You may have to enable crash dump reporting if you are getting an instant reset on your system.  Some systems have it disabled, but getting a black screen before the software can catch it usually means something went wrong in CPU or memory/motherboard land before the OS could catch it.

>Definitely get the latest BIOS

There is an update that I dont have but I have updated the BIOS before to no effect. Like I said, this issue has been happening for a long time.

I will get the newest update though, thanks.

>Make sure your RAM is on the QVL for the motherboard

It seems that my specific RAM serial code IS NOT on the list. Is this a definitive break? Ive never considered this before

>game CPU usage

All of them use quite a bit more CPU than others I think. Especially Battlefield and Vanquish since they have to render hundreds of projectiles at once

>Run your ram at the JEDEC spec speed like you'd get at first boot (not the XMP / DOCP speed), and try it with one stick if you still get crashes.

Ive used it for months at its default speeds (2400mhz I believe) and these issues persisted even then. I dont believe this to be a fix. I will however try the one stick thing

>You may have to add a touch of voltage to the memory (try adding 0.01v to it, if it's 1.35 make it 1.36 or 1.365) to get it stable at it's recommended speed or loosen the timings a tiny bit, my first-gen triple-channel intel system was like that with OCZ memory.

Alright, I'll try this as well

 

Do you think the RAM not being on the QVL is a big thing? Should I seek to replace it immediately?

[TheSillyMuffin]

38 minutes ago, bob.blunderton said:

Definitely get the latest BIOS.  Make sure your RAM is on the QVL for the motherboard, goto ASUS site for that.  Make sure to use the right slots, AMD boards differ from intel boards.  Use the correct layout per the manual.

Do the games you're finding that fail out / crash have higher average CPU load than others?  Games will ferret out a cpu-to-memory issue faster than others the harder they use the processor.

Run your ram at the JEDEC spec speed like you'd get at first boot (not the XMP / DOCP speed), and try it with one stick if you still get crashes.  Yes it will be a bit pokey but it'll show you if you have an issue there.

You may have to add a touch of voltage to the memory (try adding 0.01v to it, if it's 1.35 make it 1.36 or 1.365) to get it stable at it's recommended speed or loosen the timings a tiny bit, my first-gen triple-channel intel system was like that with OCZ memory.  Did work fine after that though, and didn't have to do that anymore with a westmere CPU later down the road.

 

You may have to enable crash dump reporting if you are getting an instant reset on your system.  Some systems have it disabled, but getting a black screen before the software can catch it usually means something went wrong in CPU or memory/motherboard land before the OS could catch it.

 

12 minutes ago, TheSillyMuffin said:

>Definitely get the latest BIOS

There is an update that I dont have but I have updated the BIOS before to no effect. Like I said, this issue has been happening for a long time.

I will get the newest update though, thanks.

>Make sure your RAM is on the QVL for the motherboard

It seems that my specific RAM serial code IS NOT on the list. Is this a definitive break? Ive never considered this before

>game CPU usage

All of them use quite a bit more CPU than others I think. Especially Battlefield and Vanquish since they have to render hundreds of projectiles at once

>Run your ram at the JEDEC spec speed like you'd get at first boot (not the XMP / DOCP speed), and try it with one stick if you still get crashes.

Ive used it for months at its default speeds (2400mhz I believe) and these issues persisted even then. I dont believe this to be a fix. I will however try the one stick thing

>You may have to add a touch of voltage to the memory (try adding 0.01v to it, if it's 1.35 make it 1.36 or 1.365) to get it stable at it's recommended speed or loosen the timings a tiny bit, my first-gen triple-channel intel system was like that with OCZ memory.

Alright, I'll try this as well

 

Do you think the RAM not being on the QVL is a big thing? Should I seek to replace it immediately?

To continue on this, both of my SSDs are also NOT on this list: https://dlcdnets.asus.com/pub/ASUS/mb/SocketAM4/ROG_STRIX_X570-F_GAMING/ROG_STRIX_X570-F_GAMING_Devices_report_20190715.pdf

 

Should I be worried about this? Could this cause instability?

[TheSillyMuffin]

7 hours ago, bob.blunderton said:

Definitely get the latest BIOS.  Make sure your RAM is on the QVL for the motherboard, goto ASUS site for that.  Make sure to use the right slots, AMD boards differ from intel boards.  Use the correct layout per the manual.

Do the games you're finding that fail out / crash have higher average CPU load than others?  Games will ferret out a cpu-to-memory issue faster than others the harder they use the processor.

Run your ram at the JEDEC spec speed like you'd get at first boot (not the XMP / DOCP speed), and try it with one stick if you still get crashes.  Yes it will be a bit pokey but it'll show you if you have an issue there.

You may have to add a touch of voltage to the memory (try adding 0.01v to it, if it's 1.35 make it 1.36 or 1.365) to get it stable at it's recommended speed or loosen the timings a tiny bit, my first-gen triple-channel intel system was like that with OCZ memory.  Did work fine after that though, and didn't have to do that anymore with a westmere CPU later down the road.

 

You may have to enable crash dump reporting if you are getting an instant reset on your system.  Some systems have it disabled, but getting a black screen before the software can catch it usually means something went wrong in CPU or memory/motherboard land before the OS could catch it.

I turned my RAM back down to its default speed of 2400Mhz, but I'm unable to change the DRAM voltage control. The option for it stays greyed out, and after I defaulted the RAM speed the voltage goes down from 1.35 anyway, so I wasnt able to do your other tip.

 

I also ran Memtest86 for 2.5 hours through all passes and 0 issues were found.

 

Any ideas?

[bob.blunderton]

8 hours ago, TheSillyMuffin said:

I turned my RAM back down to its default speed of 2400Mhz, but I'm unable to change the DRAM voltage control. The option for it stays greyed out, and after I defaulted the RAM speed the voltage goes down from 1.35 anyway, so I wasnt able to do your other tip.

 

I also ran Memtest86 for 2.5 hours through all passes and 0 issues were found.

 

Any ideas?

Get another model of power supply that's not based on the same platform as your PSU is, such as a Corsair unit, or Seasonic Platinum.

If this does not fix the issue, return it.

Oddly even though the PSU is an ASUS *BRAND* unit, it may have an issue with something in your system.

 

See note below (image link safe for work) cited from following page at the bottom of post. 

https://prnt.sc/1jkwhmp

EDIT: You may elect to pickup a reputable brand power supply locally for expediency sake, try one or two different manufacturer's units, if after the 2nd swap it isn't working, I'd say to try something else.  Make sure they have a good solid return policy.

 

PSU TIER LIST