Random Black Screens

[Paranoid Kami]

Have been getting black screens for a week now. The computer will still run and play sound but the screen goes off. Sometimes it goes fully off and I have to turn off by PSU and then on again. At first it was only in-game then I started getting them even when just watching Youtube videos and there is no overheating.

 

Thought it was my GPU overheating so I replaced the thermal paste. Didn't work. My temps are pretty high though. 60 idling and 80-90 in-game.

Tried changing the fan curve for lower temps and still blacked out.

When I set fan speed to 50% my idle temps dropped to around 40 and 70ish load. The GPU hot spot was in the mid 90's with default fan on load.

On Win 11 now but was on Win 10 before and happened there too. Clean install.

Got latest drivers.

Not my monitor or monitor cable.

Tried changing PSU cable going to GPU.

Ran memory test and there were no errors.

 

be quiet! Pure Power 11 700W cm, BN630, Modular, 80 Plus Gold, Power Supply

EVGA 970 GPU

Intel(R) Core(TM) i7-8700 CPU @ 3.20GHz   3.19 GHz

16 GB Ram

Thermaltake Kandalf case

120 mm fan front. 120 back. 2 smaller ones, 1 in back and 1 top

 

 

I think it's either my power supply or GPU. So either I go to Memory Express or a scalper. Help me decide pls.

[Bombastinator]

There is some info on the machine but it’s in no way complete.  I’m most interested in this situation in what case you have and how the fans are laid out.  The entire system will likely be needed though to rule out problems if nothing else.  There are pinned posts in the troubleshooting section that amongst other things list what is usually needed.  The pencil icon will allow you to edit your first post to add them.  
 

I’m also curious about the temps after changing the fan curve?  Might be nothing.  This is why machine specs are needed.  Speeds things up.

Normally the most common cause of crash after boot is memory issues, but there are a bunch more. Going to maybe need to monitor temps over time.  There are apps that can do that.

[Paranoid Kami]

3 hours ago, Bombastinator said:

There is some info on the machine but it’s in no way complete.  I’m most interested in this situation in what case you have and how the fans are laid out.  The entire system will likely be needed though to rule out problems if nothing else.  There are pinned posts in the troubleshooting section that amongst other things list what is usually needed.  The pencil icon will allow you to edit your first post to add them.  
 

I’m also curious about the temps after changing the fan curve?  Might be nothing.  This is why machine specs are needed.  Speeds things up.

Normally the most common cause of crash after boot is memory issues, but there are a bunch more. Going to maybe need to monitor temps over time.  There are apps that can do that.

Okay, updating.

[Bombastinator]

5 hours ago, Paranoid Kami said:

Okay, updating.

Yeah case could be making a heat issue.  I haven’t see a case that short of airflow in a long time.  It’s really long on front drive slots.  Do you use all of em?  A fairly simple test to see if it’s a heat issue is pull the side panel leaving the case open on one side and see if the problem goes away.

Edited October 15, 2021 by Bombastinator

[Paranoid Kami]

5 hours ago, Bombastinator said:

Yeah case could be making a heat issue.  I haven’t see a case that short of airflow in a long time.  It’s really long on front drive slots.  Do you use all of em?  A fairly simple test to see if it’s a heat issue is pull the side panel leaving the case open on one side and see if the problem goes away.

Already tried taking off side panel and didn't work. I tried setting my fan settings in Precision to 'stealth' and downloaded Hardware Monitor. Will record temps and update this post.

 

At idle:

Mobo: 26-30

CPU: 28-30

SSD: 27

GPU: 40

GPU Hot Spot: 54

 

At Load:

Mobo: 40

CPU: 50

SSD: 33

GPU: 62

GPU Hot Spot: 75

 

Room temp is 23-24

 

GPU sounds like a chainsaw XD

[Bombastinator]

6 hours ago, Paranoid Kami said:

Already tried taking off side panel and didn't work. I tried setting my fan settings in Precision to 'stealth' and downloaded Hardware Monitor. Will record temps and update this post.

 

At idle:

Mobo: 26-30

CPU: 28-30

SSD: 27

GPU: 40

GPU Hot Spot: 54

 

At Load:

Mobo: 40

CPU: 50

SSD: 33

GPU: 62

GPU Hot Spot: 75

 

Room temp is 23-24

 

GPU sounds like a chainsaw XD

If taking off the side panel didn’t work it’s not a heat issue, or at least not just a heat issue. Those temps are really low not really high.  Sort of the opposite of previous.  Maybe it’s with the open case. You’ve got enough wattage in that PSU.  How old is it?  970s and 8700s aren’t power spikey the way newer stuff is. My best guess is random memory buggering. Memtest86 is free, faster than older ways to do it, and will give you a good if not absolute best answer.  It isn’t super fast though.  Was designed for machines with less memory than they have today.  Couple full runs ought to be if not definitive, at least gud’nuff to make other stuff more likely.  If it fails you’ve found your problem.  If it passes memory goes to the back of the que and other stuff gets looked at.  The older school method is to run one stick at a time and see if there are errors. Theory is that with a multi stick kit only one stick is going to have problems.  

[Paranoid Kami]

Actually, I think I solved the problem. Played all day and not a single crash. Rather than just leave the fan on at 50% I put on a curve and that seemed to solve it. I never wanted to put a curve before because I would have to have EVGA Precision start with Windows and just keeping it at 50% doesn't. My first temps were just with the default fan curve.

 

I have to get a new card in a few weeks so I can run the new Battlefield above all low at 30 fps so I guess I'll look for something that keeps the card cool and has low fan noise. Especially the low fan noise part.

[Bombastinator]

48 minutes ago, Paranoid Kami said:

Actually, I think I solved the problem. Played all day and not a single crash. Rather than just leave the fan on at 50% I put on a curve and that seemed to solve it. I never wanted to put a curve before because I would have to have EVGA Precision start with Windows and just keeping it at 50% doesn't. My first temps were just with the default fan curve.

 

I have to get a new card in a few weeks so I can run the new Battlefield above all low at 30 fps so I guess I'll look for something that keeps the card cool and has low fan noise. Especially the low fan noise part.

So is a heat problem or isn’t a heat problem? I suppose if it works it doesn’t matter.  New case might also be worth it if you don’t need that big HD rack in the front.  I haven’t seen a case that restrictive in a long time.

Edited October 16, 2021 by Bombastinator

[Paranoid Kami]

12 hours ago, Bombastinator said:

So is a heat problem or isn’t a heat problem? I suppose if it works it doesn’t matter.  New case might also be worth it if you don’t need that big HD rack in the front.  I haven’t seen a case that restrictive in a long time.

Heat problem. I recently moved. Comp used to be in an open area that's around 18 degrees and is now in my room that is around 23 degrees. Think that was the problem as it worked fine on default fan speed for so many years. I thought of getting a new case but my current one is too sexy to give up. Looks way better than the pictures. Also, it's big enough that I can use it as a mini desk.

 

How much of a difference would a new case make? I was looking at the Fractal Design 7 or Be Quiet Base 601.

[Paranoid Kami]

On 10/15/2021 at 11:53 PM, Bombastinator said:

So is a heat problem or isn’t a heat problem? I suppose if it works it doesn’t matter.  New case might also be worth it if you don’t need that big HD rack in the front.  I haven’t seen a case that restrictive in a long time.

 

Have been getting crashes again and my temps were fine. It's not a heat problem.

 

Did Memtest and ram is fine.

 

Think it's a power supply issue.

[Paranoid Kami]

Not power supply. Tried another one.