Overclock.net banner
Cancel
Create thread New Forums

[390] Black screen with furmark/games

Jump to Latest
1 - 20 of 20 Posts
T

thegamehhh

· Registered
Joined
·
16 Posts
Discussion starter · #1 ·
Hi, havent been using my pc for a few months, after the break i just went straight into some games. And here begins my issue: as long as i am playing not-quite heavy gpu stress games like fifa or hearthstone, nothing major happens. But once i start to play some heavier games like the witcher, black screens occurs almost immadiately. Same happens with furmark and occt PSU stress test.
List of what have i done:
got newest drivers (stable and beta)
tested cpu (occt for 2 hours)
memtest passed
changed psu
tested gpu on another pc, nothing happened for like an hour
tried to lower power limit in afterburner, seems like -30 works just fine, at -27 black screen occurs after few mins
Any idea how to get rid of it/what might be responsible? Few pins on cpu socket are sligthly curved (dont know how to spell it in english). Do not own another mobo for testing.

cpu: i5 4690k (stock)
gpu: r9 390
ram: 2x 4gb 2133+2x 4gb 2400
mb: z97-g43
psu: P1-550B-BEFX

thanks for any kind of help!
 
#2 ·
Stop using Furmark. It puts an unrealistic load on your GPU that'd it'd never really experience. NVIDIA cards will purposely cripple themselves if it detects Furmark is running to prevent damage and I'm surprised AMD hasn't done something similar. If you want to test your GPU, use a benchmark like 3DMark or Unigine Heaven.

It sounds like a PSU or a GPU issue. If the GPU was tested in another system and was fine then I'd blame the PSU ... but you said you already swapped that out.

What PSU was in it originally? Hopefully you didn't replace the PSU with the XFX unit because that PSU is several years old at this point.
 
Save
Reply Quote Rep+ Rep+
Discussion starter · #3 · (Edited)
I do know that using furmark isnt wise, but when pc crashes literally seconds after turning it on..its much faster than any game :)

PSU i replaced is some corsair cx 500. But i am not sure its about psu, since the psu in the second pc is some 6+year old bequiet 400 or 450w.
today i am gonna test the gpu a bit more with -30% power, and see if it gonna work
may i add that occt gpu test with error checking (less heat) works just as fine
 
#4 ·
Furmark is just a heat test, if you're shutting off due to a bad PSU it's probably not wise to demand even more power from it.
What sort of PSU did the working PC have? Is the GPU overclocked? Is it getting too hot? When did you start to notice this happen? Any changes before that? Are you clean installing the drivers with DDU or just installing one then the other? Have you tried a different PCIE slot (I've had this make a difference)?
I'd blame the PSU but wouldn't that shut the whole thing down? Unless you're using multi rail PSUs each time.
 
Save
Reply Quote Rep+ Rep+
Discussion starter · #5 ·
What sort of PSU did the working PC have?
some bequiet 400w/450w

Is the GPU overclocked?
no and never been

Is it getting too hot?
temp at other pc was around 80C (furmark)

When did you start to notice this happen?
after a few months, meanwhile pc havent been used

Any changes before that?
the only ones are the curved(?) pins on cpu socket

Are you clean installing the drivers with DDU or just installing one then the other?
always using ddu in safe mode

Have you tried a different PCIE slot
gpu is too big to use second slot

I'd blame the PSU but wouldn't that shut the whole thing down?
gpu crashes with the corsair one as well and i am 100% sure he is working fine. also gpu doesnt crash at second pc with bequeit
 
#6 · (Edited)
Ok I'm getting confused here just a little with the "other PC". By other PC I assume you mean the one where the GPU did work, correct? Can we just call that one #2 and the one in the OP #1?

So you had a cx500 in system #1 then you replaced it with the XFX 550w unit. Why did you do that? Because of the GPU issue or something else?

Somewhere along the line you bent the pins in your motherboard...how and why did this happen?

When you replaced the PSU did you also rewire it with the matching cables or use the old ones? Does this issue exist with both PSUs?

If it were my PC and I've done everything you've tried I'd want to confirm that it's either the CPU, MOBO, PCIE, GPU or whatever. To do that I'd do whatever it takes to get to that point. Take the CPU back out and go at the pins with some sort of magnification, I have a macro cellphone lens I use for such things but coin collectors have a similar thing that you hold up to your eye so you can get really close to something. The pins could be responsible for communication across the PCIE and you're only having this issue when it uses one pin in particular. Not only do you need them in their original position but you also need them to be at the correct height to make good contact. Look under the CPU and make sure all the pads are clean so you get good contact, perhaps it's making enough contact to work sometimes but the resistance gets too high when it uses more power across that pin. Compare the finished socket to an image the same socket online, make sure no pins are missing.
If that doesn't work do whatever it takes to try the bottom PCIE, the easiest way would be to remove the PSU from the case but leave it hooked up so you can put the GPU there. The issue I had was with 9000 series GPU, it would work in my bottom slot but it wouldn't give me any display in the top slot where my 980 and current 1080 work just fine, as it so happens my board or CPU is stuck in 8x mode and I still have no idea how or why.
If that doesn't work I'd tear the whole thing down and build it outside the case to try to rule out any shorts.
Since it works in another PC I really thing it's motherboard or CPU related.

Another though is that perhaps this is software related since it's only a black screen instead of a crash but I really don't know. The only way to test that is to backup you C drive and do a clean install or use windows to-go with a flash drive.
 
Save
Reply Quote Rep+ Rep+
Discussion starter · #7 · (Edited)
Ok I'm getting confused here just a little with the "other PC". By other PC I assume you mean the one where the GPU did work, correct? Can we just call that one #2 and the one in the OP #1?
sure, thats easier to describe
just to make things clear:
1# PC: psu i been using all the time: xfx. psu i used for test with 1#pc unit: corsair
2# PC: psu: bequeit

bent the pins
thats the word i been searching for!

how and why did this happen?
changing the cooler, few months ago


When you replaced the PSU did you also rewire it with the matching cables or use the old ones?
each psu have their own wires

Does this issue exist with both PSUs?
yes

it's only a black screen instead of a crash
by crash u mean restarting the pc on his own?


thats the issue, have no way to test mobo nor cpu..gonna disassembly unit from the case and try second slot. about bent pins- i have tried straighten them, but they are still sligthly displaced
 
#10 ·
IMHo...
a bad PSU would generally cause a whole PC crash... Unless it was multiple rail, in which case the rail feeding the GPU could brownout and the rest of the system could continue
To me it sounds like the GPU crashing, but that would usually cause the program to crash and a windows message about the GPU being re initialised or something
So... I'm wondering if you are seeing something heat related causing part of the card to crap out.. like the memory or such like....
I wonder if it would be worth re pasting the card...Ive never needed to do it.... But a 390 has to be a few years old in the wild
 
Save
Reply Quote Rep+ Rep+
#11 ·
oh wait
you sure its the card and not the monitor/cable? ive had blanking issues trying to get 4k UHD with a cheap cable..
 
Save
Reply Quote Rep+ Rep+
#12 · (Edited)
Just to clarify
Black screen = Screen is dark but the PC is still running, Windows is still running. The game could also still be running.

Crash can mean a few things:
If it's restarting the PC that's a hard crash
Locking up the PC could be considered a soft crash
Locking up the game can be a driver crash but a driver crash can also lock up the entire PC.

Which one do you have?

Feel free to describe exactly what happens as that could help, I was under the impression that he PC was still running but the display stopped.

Check the pins again. If that's fine you really need to try it in the bottom slot as there could be an issue with the CPU communicating with the top slot. The pins could cause a similar issue I think. IMO you can rule out the GPU if it works in another PC.


EDIT: Check under the hardware that mounts the cooler, check for scratches on the mobo traces. After I delidded my CPU I put my cooler back on and my PC would not boot, kept hanging on ram. Tried all sorts of things and feared I'd killed my PC, then I thought about he cooler. Took the standoffs out and sure enough...CoolerMasters brilliant solution for protecting the mother board was to put little black pieces of rubber under the standoffs with some very weak adhesive. One fell off and scratched the traces so I put some electrical tape down and it was good as new.
 
Save
Reply Quote Rep+ Rep+
Discussion starter · #13 ·
alanthecelt said:
oh wait
you sure its the card and not the monitor/cable? ive had blanking issues trying to get 4k UHD with a cheap cable..
Click to expand...
same monitor and cable has been used with PC#2


Feel free to describe exactly what happens as that could help, I was under the impression that he PC was still running but the display stopped.
screen goes black (information on the monitor: 'no connection' , sound is stucked for a while, and then nohting happens


Check the pins again. If that's fine you really need to try it in the bottom slot as there could be an issue with the CPU communicating with the top slot. The pins could cause a similar issue I think. IMO you can rule out the GPU if it works in another PC.
thats my first remark as well, have written this thread for hope it might be something else. i will try those tips as soon as i get back from work
 
#14 ·
I edited my last post to include something new.
When you get the black screen is it stuck this way or does it shut off? Good luck, I know the feeling.
 
Save
Reply Quote Rep+ Rep+
Discussion starter · #15 ·
check for scratches on the mobo traces.
mobo is as good as new. expect those pins

When you get the black screen is it stuck this way or does it shut off?
it is stuck for as long as i been waiting something to happen- few minutes
 
#16 ·
thegamehhh;27646392 When you get the black screen is it stuck this way or does it shut off? it is stuck for as long as i been waiting something to happen- few minutes[/QUOTE said:
re you saying you need to restart the PC?
Click to expand...
 
Save
Reply Quote Rep+ Rep+
#18 · (Edited)
Most more recent GPUs will all hit current limits even with maxed out power sliders in FurMark (even if you bypass any app specific limiters), but a working Hawaii part that doesn't have a cooling limitation shouldn't have any issues running FurMark at it's most stressful settings. Not good to do long term, however it can certainly be useful in revealing issues. Yes, it's an 'unrealistic' load and that's the entire point...often it will immediately reveal power delivery issues that could be intermittent and difficult to pin down with other tests.

Anyway, it could be a number of things, but the most likely culprit would seem to be the PSU and if at all possible I would test a different PSU without changing anything else. The fact the GPU runs in another system without issue strongly implies that it's not the problem.

The XFX P1-550B-BEFX is made by Seasonic, which is normally a solid OEM. However, I've run into some issues with some of their designs from this era that produced similar symptoms.
 
Save
Reply Quote Rep+ Rep+
Discussion starter · #19 ·
Latest news:
just got new coffe lake unit: ram, cpu and mobo. And guess what? blak screen occurs once again!
What i have forgot to write is the fact, that after the passed test with PC#2, furmark half-an-hour test with PC#1 have passed as well.
So here is the chronologicaly summary:
PC#1 with xfx psu: test failed
PC#2 with bequiet psu: test passed
PC#1 with xfx psu: test passed
PC#1 with xfx psu: test failed
PC#1 with corsair psu: test failed
PC#3 with xfx psu: test failed
that means my old mobo and cpus are probably fine, and most logical reason for the errors is gpu. i am gonna buy new gpu (wich i had on my mind for quite some time) and test it this final time. if i would put a bet, its 80% gpu and 20% psu
 
1 - 20 of 20 Posts
About this Discussion
19 Replies
6 Participants
Last post:
thegamehhh
Overclock.net
posts
28M
members
562K
Since
2004
The premier forum for overclocking experts and enthusiasts. Discuss hardware optimization, custom builds, benchmarking, cooling solutions, and pushing the boundaries of computing performance. From beginner guides to extreme overclocking, join our technical community to master system tuning.
Full Forum Listing
Explore Our Forums
Intel CPUs Hardware News NVIDIA PC Gaming Water Cooling