Windows 10 screen goes black for a second before returning to normal

[tedus987]

Ok, so before I begin this is more for the idea of just finding the faulty component.

The PC in question will most likely be replaced in 2021 and some of the parts in it now are the replacement parts. So please don't point out the 2070S overkill.

so this is my fiancée’s PC, she uses it for gaming and CAD drawing & Graphic design it's in her own room with 3 screens hooked up to it and mirrored.

Original specks

Motherboard - 990X-gaming-SLI

CPU - FX8350

Ram - 32GB dominator LXP DDR3 ram (short of pulling it out, I do not know the speed but it was a pre-build)

GPUs – 2X EVGA 1070 in SLI

OS drive – Samsung NVME M.2 PCIe 250GB drive

PSU – Corsair RMi 850X

Sound card – soundblaster ZXR

Then I upgraded it in preparation to swap out the MB, RAM, CPU & GPUs.

Prep upgrade

Motherboard - 990X-gaming-SLI

CPU - FX8350

Ram - 32GB dominator LXP DDR3 ram (short of pulling it out, I do not know the speed but it was a pre-build)

GPUs – 2X EVGA 1070 in SLI

OS drive – Samsung NVME PCIe 250GB drive

PSU – Corsair RMi 1200X

Sound card – soundblaster ZXR

(side note, I swapped out the case for a bigger one.)

Then I upgraded the GPUs early as a gift to her… and to snag hers for my system since subnautica chugs on a 970…

GPU upgrade

Motherboard - 990X-gaming-SLI

CPU - FX8350

Ram - 32GB dominator LXP DDR3 ram (short of pulling it out, I do not know the speed but it was a pre-build)

GPUs – 2X EVGA 2070s in NVlink

OS drive – Samsung NVME PCIe 250GB drive

PSU – Corsair RMi 1200X

Sound card – soundblaster ZXR

Other things to note about all 3 systems, they all had sata drives. A BlueRay RW & a Samsung evo 850 1TB that have been in all 3 systems. Also her first two iterations had 2 * 2TB Seagate drives and the current iteration has a 3TB drive.

So with all that aside I would again like to note, this is not the final build, I will be getting a new MB, CPU, RAM and OS drive for her. Basically rebuilding her system with the twin 2070s in mind.

Now, the faults. Firstly

Random lockups

Once, every blue moon, her screen will freeze on windows, no idea as to when or how it happens. It just does… it might happen a month from now or today. This issue has persisted from the first to the current iteration.

Random going black for a second.

Ok, so my fiancée informs me this was something that “just happens” and has happened since the first iteration, she only told me now as before it was once in a month maybe and is now once every few days since the 2070S upgrade. (original specs) her screen will just go black for about 3 seconds then come back on as if nothing had happened. This has happened on both sets of GPUS so I have run a few tests, further note that her old 1070s are in my machine and work perfectly fine.

1 - furmark – 30, 60, 240 min stress test – passed all, stayed below 90, cooled to 50 when done.

2 – super positron – passed

3 - did it happen to multiple screens? – yes – so not the cables

4 – windows disconnect sound – no – the GPUs aren’t disconnecting from windows. Windows makes a noise when something disconnects.

Since this has happened on multiple GPUs with multiple power supplies I'm ruling them out.

Bios freezes

I discovered this one today but the bios randomly freezes, I will either get into it and it will then freeze or it will freeze when I try to enter it or it'll work fine. PC boots to OS just fine and works fine until it locks up 3 months from now. But her bios does randomly freeze.

-end of faults-

So I have puzzled and pondered this for some time. Could be the sata ports, could be any number of things I do not know.

As I put before give me 4 months and she’ll get a new system with the current GPUs, tower & sound card.

But I originally planed for the remaining system to become a gaming / film system in the bedroom. However before I do that I would like to run a few tests to see if I can sus out what the issue is. I do have a feeling these faults are connected, maybe I just have a bad Motherboard or a faulty sata cable.

So I thought I’d throw this out there and ask anyone on testing ideas. Anything sensible of course, first and foremost I’d like the bios to run smoothly.

currently I am running base tests and trying ideas to test wether it might be a part or overheating. for all I know it could be the exibel mouse that's doing it.

my testing so far

so for right now I'm focusing on trying to find out when the screen goes black for a second. as I mentioned before it was a non-isue back when she first noticed it and it's only after the 2070s getting installed that it's become an issue. since it was a fault that was present with her 1070's but happened maybe 1 a month but is now happening with the 2070S's once a few days. and her 1070s are in my PC and working just fine. (2 hour furmark test, ran the same web apps and no issues.) I can rule out the GPUs as actually being at fault.

as I mentioned before Furmark, Super positron, using multiple monitors to test cables, checking to see if windows flags the DC sound when it happens. I've done those. I also ran a 30min stress test via CPU-Z with core temp. never went over 47c. so I know it's not the CPU. the sound card was blocking the front fan of GPU 1 so I moved it to give the GPUs more breathing room but that didn't solve it. I have run memtest to check for RAM faults. further I have checked the bios version and it is version F1, the only bios version gigabyte made for this board.

Samsung magician is listing the M.2 drive as running high in temprature (57c), so I have ordered a low profile heatsink since the M.2 drive is under GPU1, from what my fiancée said the issue only seems to happen when she's running youtube or google docs on mozilla so I've had her move to chrome. (some research pointed out that mozilla is SSD heavy so I thought I'd see if it happened less with another browser. don't know how true that is...) the heat sink is going in Wednesday

If I can’t resolve it I’ll salvage what I can and trade in my twin 970s towards a 1660s and just have that in a new & cheap MB, RAM and CPU.

Final notes, Malwarebytes and kasperski say all versions of the system are clean so no malware issues.

So, testing ideas. As many as you can. I’ll be putting them in to practice in 4 months time.

 

[Kshipper]

You clearly know what you are doing. Did you check the Event Viewer for any logs that might point you in the right direction? It also might not hurt to seek the latest chipset driver and maybe do a clean driver install and use the Nvidia driver (if you aren't using the Nvidia version). Windows version? Might be time to get her up to at least 1909 or maybe 2004. I have solved some weird issues doing that. I don't know why it worked but I had a persistent black screen problem on a machine that was a year, or so old and updating Windows from 1903 to 1909 cured it.

I have also found that black screen issues seem to be related to older monitors. How old are those 3 screens. Try a single...newer screen? Probably can't do that since this rig is her daily driver and she needs the real estate.

You are in a tough spot. I suggest you proceed with the upgrade, move all the new parts over but before you scrap the old board rebuild it once more and keep it simple (for testing purposes). That way it will be easier to sort what is causing this issue.

 

[tedus987]

We ran chrome last night and Samsung magician registered 51c, no black screen issue. if it's the same tomorrow I'm going to install the M.2 heat shield and see if that helps the temp lower.

something that I can't believe I'm saying is that the hunch that it might be heat related could be the route cause. her 1070s where turbine GPUs (one fan draws air in along the heat sink and out the back.) while her 2070S 's are blowers. (two fans that pull air in and it exhausts in every direction.)

since one of those directions is AT the M.2 and the 2070S generate more heat than the 1070s. (I think, could be wrong) given the fault was more prevalent after the install.

I'll check the chipset drivers. it's up to date on the windows front. whenever I install new GPUs I always do a clean install to avoid conflicts and I do keep the drivers u to date.

her monitor is around 5 years old, her TV is around a year old and her pen screen is a few months old.

one of the tests was to have all 3 on and see if it only happened to 1, it happened simultaneously across multiple screens.

side note: event viewer, what would I be looking for. currently on a refresher course but still can't get my head round it.

 

[Kshipper]

All good info Tedus987. You might be right about the heat issue. Those new graphics cards put a lot of heat into the case. What case are you using and how many fans do you have? Location of the fans? Size? You want at least 2 on the front and 2 or 3 on the back and top to get the heat out. I prefer 120mm fans.

In terms of the Event Viewer. I would think you would be looking for something related to the graphics driver crashing when it goes black screen. You can always Google all the red errors you see with a timestamp close to the time where the black screen occurs. So if it happens next again, make a note of the time of day and check the event viewer around that time period, for red errors. Click [Custom View] then click [Administrative Events]. They will be arranged by date and time.

 

[tedus987]

Fans front 2* 140 MM corsair LL in, base 2* 120MM corsair LL in, top 3* 120 mm LL out, rear 140mm corsair LL. all but the base are controlled by corsair link hub and the two bottom are run by the MB, also the CPU cooler is a Dark rock 3 with 1 135mm stock fan and 1 120mm additional fan.

I have a fear of water cooling so I specialize in setting up air cooled rigs. also there is no overclock since I want longevity out of my PCs. while the Motherboard, CPU, RAM, sound card and M.2 drive are all part of a pre-build, I moved it to it's new case which is a corsair 780T. (same as mine only white instead of black so we can differentiate.)

also I just installed the M.2 heat shield, 51C idle to 25c idle...

just hit the PC with super positron to get the GPUs to spin up, 42c after a 5 min blast...

a £10, low profile laptop 2280 heat sink cut the idle temp in half. what's more the case is now noticeably quieter..

ok, I'll tell my fiancée to jot down if she notices it happening. however since she moved to chrome where 1 day clean of the event.

side note :- if it turns out that the M.2 drive heating up was the issue the technician inside me will just loose it. after so many issues with this pre-build and individual part I swore I'd never buy Core PC parts (Motherboard, RAM, CPU, GPU) or a pre-build from that vendor again and this just cements it since they have custom logoed M.2 drive heat sinks but have no offer to put them on a system at pre-build. I only know about them because I bought an M.2 drive from them and they game me a complimentary sink that I lost.

 

[Kshipper]

Yes please keep us posted.....some of us deal with problems like this and would love to know the fixes that others have found! The true reason forums exist, like-minded individuals supporting each other. I recall a Jays Two Cents video where his crashes were related to the RGB portion of the RAM. Crazy eh!

 

[tedus987]

Well, 2 days in and no black screen, drive has been no higher than 42C, but that was a stress test, doing normal tasks it didn't go higher than 27c... gonna monitor it for a week since the fault is random. I've asked her to just keep doing the tasks that she normally does and if the fault happens to time stamp it.

 

[Kshipper]

Excellent =)

 

[tedus987]

Well it's been a week, she's used it for games, drawing and usual work loads, the M.2 has reached 54C but that was after a long session, two virus scans and she forgot to open a window to swap out air.

we haven't seen the screen issue once since before she swapped to chrome and I installed the heat shield.

since I use mozilla on multiple systems I know it's hard drive heavy and on M.2 cards that can heat them up so I'm going to note it as a lack of heat sink on the M.2 card.

in short if this Pre-Build system had been supplied with an M.2 heat shield then all of this could have been avoided.

so, SCAN relegated to hard drives and software purchases. this is another notch on why I no lnger by the bulk of my parts or pre-builds from them any more.

 

[Kshipper]

I have to say I have built many hundreds of systems with WD Black SN750 and no heat sink and only a handful of Samsung EVO and Pro drives. I have never had the trouble you had. Do keep the thread handy of anything changes. Good sleuthing! =)

 

[tedus987]

I spoke too soon, even happened today, a week after I installed the heat shield, twice in a row, fiancee was using chrome for both google docs and youtube.

fans were running quite high and M.2 drive is at 51c. trying to look now in event viewer for any alerts in administrative view but can't find any around the time stamp of the black screen.

if it starts happening more frequently I'll keep you posted.

however if it's once a week I might be on the right track...

but we can take from this the following.

1. not browser related
2. after reseating and applying a heat sink the error disappeared for a week.
3. nothing shows up in event viewer...

 

[Kshipper]

Bummer. I was hoping the Event Viewer would show the graphics driver being reset.

Since you were going to build new--I would go with that. Swap the pieces over and if the new one is good use the old parts to rebuild and make it as minimal of a build as you can for test purposes.

 

[tedus987]

I do still think it's heat related, the event only happens when the PC has been on for an extended length of time. due to the M.2 slot being behind the GPU I only put a heat sink that was designed for a laptop due to not knowing how much space I had behind there.

I am going to put a new M.2 in to the new build since this build will be a PC for films and connecting to either of our steam accounts over LAN.

when my fiancee has her new PC I am going to do a full reset of this one with the parts I have since it won't be hers then.

 

[Kshipper]

Sounds good =)