Sign in with
Sign up | Sign in
Your question

Weird Intermittent black screen occurring when playing games, would a new and more powerful PSU fix this?

Tags:
  • Overheat
  • Graphics Cards
  • Power Supplies
  • Games
  • Components
Last response: in Components
Share
July 27, 2014 8:38:06 AM

So I have a problem in that when I play games that tax my system, the screen goes black while the game is still running, sound and everything else is fine. Then the screen will go back to normal, this happens until ultimately it's just basically flashing the black screen. I am aware this is probably a problem with my PSU or GPU but I need help to determine which one I need to replace, if not both. It might have to do with how hot some of my components get when playing games, here are my temperatures when the black screens appear: http://puu.sh/atdy3/1ebef66ff0.png

PC Specs:

AMD Phenom 2 X4 955

XFX Radeon 6870

8GB DDR3 Ram

Asus M5A78L-M

OCZ ModXStream-Pro 500W

Any help would be most grateful :D 

More about : weird intermittent black screen occurring playing games powerful psu fix

a b U Graphics card
a b ) Power supply
July 27, 2014 11:02:03 AM

Look at the temps of the gpu with this,
hwinfo32
install/open it,mark sensors=V and click run.
Look at the voltages of the psu too,the +rails should be between -5%<->+5%, for the +12voltrail=>11.4<->12.6V

Did you update drivers (recently)?

Looking at speedfan is 84°c high,but not too high.Does a higher fanspeed,changed via the ccc,help?
m
1
l
July 27, 2014 12:25:32 PM

Vic 40 said:
Look at the temps of the gpu with this,
hwinfo32
install/open it,mark sensors=V and click run.
Look at the voltages of the psu too,the +rails should be between -5%<->+5%, for the +12voltrail=>11.4<->12.6V

Did you update drivers (recently)?

Looking at speedfan is 84°c high,but not too high.Does a higher fanspeed,changed via the ccc,help?


I'm going to try the higher fan speed and also then try hwinfo, I updated drivers two days ago just to make sure. I'll update soon.
m
0
l
Related resources
July 27, 2014 12:50:10 PM

Vic 40 said:
Look at the temps of the gpu with this,
hwinfo32
install/open it,mark sensors=V and click run.
Look at the voltages of the psu too,the +rails should be between -5%<->+5%, for the +12voltrail=>11.4<->12.6V

Did you update drivers (recently)?

Looking at speedfan is 84°c high,but not too high.Does a higher fanspeed,changed via the ccc,help?
Wasn't too sure what you meant so I took a screenshot.
http://puu.sh/atvCE/327f5f9f58.png
http://puu.sh/atvBk/276884b0d4.png
m
0
l
a b U Graphics card
a b ) Power supply
July 27, 2014 1:10:29 PM

Looking at those screenshots is the +12voltrail way under spec with 10.544V/10.610V.Can you look into the bios to see if these values are the same there?
If this is the case do you need a new psu.

You didn't tax the gpu looking at the temps? Even at idle do these seem good,but there can be quite a difference between load and idle so doing a benchmark could give these load temps.
m
0
l
July 27, 2014 2:22:13 PM

Vic 40 said:
Looking at those screenshots is the +12voltrail way under spec with 10.544V/10.610V.Can you look into the bios to see if these values are the same there?
If this is the case do you need a new psu..
Ok, is this what you were looking for? http://puu.sh/atD75/4b4fc65f1d.jpg

Quote:
You didn't tax the gpu looking at the temps? Even at idle do these seem good,but there can be quite a difference between load and idle so doing a benchmark could give these load temps
The SpeedFan screenshot that I posted was of the temperatures when the black screen started to occur
m
0
l
July 27, 2014 5:15:14 PM

http://battlelog.battlefield.com/bf4/forum/threadview/2...

I have a similar issue, and i believe i know thats its the cpu or the mobo underperforming, not the 280x

After playing many different types of games, i've concluded that the cpu is underperforming (based on whether a game is gpu or cpu driven) Battlefield is quite a cpu intensive game (the way hit reg works, 64 players and all the stuff going on). I have an r9 280x, which should be able to play near ultra settings. And it could do that, until a month back (this pc is 4 months old) But no i play on medium, otherwise i don't get 60fps!!!

I also have a 750w PSU and 2x4 gb 1600mhz RAM

the temps max @ 65-70 on the gpu and 50-60 on the cpu

But what is really interesting is that we both have the 970 chipset, just like the op in the above thread

I've had to replace the mobo already (first had faulty fan socket) I can't remember but i think that i could still play ultra with the second mobo.

I have had quite a few random BSODs (some caused by an external hard drive i had plugged in).

And i do remember that when i first built logged onto the pc i went to gigabyte's drivers page and there were a load there for my mobo. I didn't know which ones i should have used, so its possible that i'm using the wrong drivers.

[Specs:
Win 8.1 (supposedly better performance with bf4!!)
Gigabyte GA 970a-d3p
fx-6300
R9 280x
2x4 8gb 1600mhz
xfx 750w smartpower
samsung 840 evo (holds Win 8.1 and not much else
Seagate 1tb 7200rpm
m
0
l
July 27, 2014 5:16:06 PM

And i've encountered another problem. I'm getting the black screens, and a restart. No BSOD, and no overheating. (gpu doesn't go past 68, i haven't seen cpu go past 50)

Maybe the PSU is overheating, i don't know if there is a temp reading for that
m
0
l
July 28, 2014 1:21:37 AM

thorpejoe said:
http://battlelog.battlefield.com/bf4/forum/threadview/2...

I have a similar issue, and i believe i know thats its the cpu or the mobo underperforming, not the 280x

After playing many different types of games, i've concluded that the cpu is underperforming (based on whether a game is gpu or cpu driven) Battlefield is quite a cpu intensive game (the way hit reg works, 64 players and all the stuff going on). I have an r9 280x, which should be able to play near ultra settings. And it could do that, until a month back (this pc is 4 months old) But no i play on medium, otherwise i don't get 60fps!!!

I also have a 750w PSU and 2x4 gb 1600mhz RAM

the temps max @ 65-70 on the gpu and 50-60 on the cpu

But what is really interesting is that we both have the 970 chipset, just like the op in the above thread

I've had to replace the mobo already (first had faulty fan socket) I can't remember but i think that i could still play ultra with the second mobo.

I have had quite a few random BSODs (some caused by an external hard drive i had plugged in).

And i do remember that when i first built logged onto the pc i went to gigabyte's drivers page and there were a load there for my mobo. I didn't know which ones i should have used, so its possible that i'm using the wrong drivers.

[Specs:
Win 8.1 (supposedly better performance with bf4!!)
Gigabyte GA 970a-d3p
fx-6300
R9 280x
2x4 8gb 1600mhz
xfx 750w smartpower
samsung 840 evo (holds Win 8.1 and not much else
Seagate 1tb 7200rpm


Firstly, that link doesn't work, what was that linking to exactly? While it may be my CPU or Motherboard, did you replace either one to fix your problem or is the problem still apparent? I'll try to play some GPU and CPU intensive games to see if and where the problem occurs. Thanks for your comment :) 

EDIT: Also, Far Cry 3 (the game I've been having the problems on as well as Heros and Generals, DayZ and a couple others) is not very CPU intensive,I'm gonna try Witcher 2, a very GPU intensive game and see what happens.
m
0
l
July 28, 2014 3:06:13 AM

PieInTheSky said:
thorpejoe said:
http://battlelog.battlefield.com/bf4/forum/threadview/2...

I have a similar issue, and i believe i know thats its the cpu or the mobo underperforming, not the 280x

After playing many different types of games, i've concluded that the cpu is underperforming (based on whether a game is gpu or cpu driven) Battlefield is quite a cpu intensive game (the way hit reg works, 64 players and all the stuff going on). I have an r9 280x, which should be able to play near ultra settings. And it could do that, until a month back (this pc is 4 months old) But no i play on medium, otherwise i don't get 60fps!!!

I also have a 750w PSU and 2x4 gb 1600mhz RAM

the temps max @ 65-70 on the gpu and 50-60 on the cpu

But what is really interesting is that we both have the 970 chipset, just like the op in the above thread

I've had to replace the mobo already (first had faulty fan socket) I can't remember but i think that i could still play ultra with the second mobo.

I have had quite a few random BSODs (some caused by an external hard drive i had plugged in).

And i do remember that when i first built logged onto the pc i went to gigabyte's drivers page and there were a load there for my mobo. I didn't know which ones i should have used, so its possible that i'm using the wrong drivers.

[Specs:
Win 8.1 (supposedly better performance with bf4!!)
Gigabyte GA 970a-d3p
fx-6300
R9 280x
2x4 8gb 1600mhz
xfx 750w smartpower
samsung 840 evo (holds Win 8.1 and not much else
Seagate 1tb 7200rpm


Firstly, that link doesn't work, what was that linking to exactly? While it may be my CPU or Motherboard, did you replace either one to fix your problem or is the problem still apparent? I'll try to play some GPU and CPU intensive games to see if and where the problem occurs. Thanks for your comment :) 

EDIT: Also, Far Cry 3 (the game I've been having the problems on as well as Heros and Generals, DayZ and a couple others) is not very CPU intensive,I'm gonna try Witcher 2, a very GPU intensive game and see what happens.


I think i play mostly cpu intensive games, but they all still tax the GPU.

The only possible hardware problems i can think of are:
* I've already had to replace the mobo (bad fan header). When i applied the thermal paste, it was my first time doing so, and i was a little worried because the stuff didn't spread. I'd seen most people just put the CPU cooler on right away, but after 10 seconds or so the stuff was pretty dry, and i didn't think it would work (When you turn the PC on it'll heat up, but hey first time builder and everything!)

* Overheating/Faulty Power supply
* insufficient wattage, amperage, voltage
* The gpu may not be seated properly. I believe it is. but i also think it sits slightly off center (due to the weight of this bloody 280x) i've pushed it in again.
*drivers
* BIOS drivers: this seems likely, this is the second thing that i did no research on and had no idea about - BIOS updating. There were so many drivers and different.... things that i didn't know what i was doing
*Overheating: I know that the GPU hasn't gone past 70 Celsius (although these crashes happen in the middle of a game, i might not notice a spike past 70).
* The cpu could be overheating, but same as the reason above
*the PSU or Mobo could be overheating, i wouldn't know

the link was about : [http://battlelog.battlefield.com/bf4/forum/threadview/2...] but it corresponds to an issue cropping up with Gigabyte Ga970a mobos and FX-6300 and how badly they are performing due to issues that we haven't figured out.
m
0
l
July 28, 2014 3:09:14 AM

PieInTheSky said:
thorpejoe said:
http://battlelog.battlefield.com/bf4/forum/threadview/2...

I have a similar issue, and i believe i know thats its the cpu or the mobo underperforming, not the 280x

After playing many different types of games, i've concluded that the cpu is underperforming (based on whether a game is gpu or cpu driven) Battlefield is quite a cpu intensive game (the way hit reg works, 64 players and all the stuff going on). I have an r9 280x, which should be able to play near ultra settings. And it could do that, until a month back (this pc is 4 months old) But no i play on medium, otherwise i don't get 60fps!!!

I also have a 750w PSU and 2x4 gb 1600mhz RAM

the temps max @ 65-70 on the gpu and 50-60 on the cpu

But what is really interesting is that we both have the 970 chipset, just like the op in the above thread

I've had to replace the mobo already (first had faulty fan socket) I can't remember but i think that i could still play ultra with the second mobo.

I have had quite a few random BSODs (some caused by an external hard drive i had plugged in).

And i do remember that when i first built logged onto the pc i went to gigabyte's drivers page and there were a load there for my mobo. I didn't know which ones i should have used, so its possible that i'm using the wrong drivers.

[Specs:
Win 8.1 (supposedly better performance with bf4!!)
Gigabyte GA 970a-d3p
fx-6300
R9 280x
2x4 8gb 1600mhz
xfx 750w smartpower
samsung 840 evo (holds Win 8.1 and not much else
Seagate 1tb 7200rpm


Firstly, that link doesn't work, what was that linking to exactly? While it may be my CPU or Motherboard, did you replace either one to fix your problem or is the problem still apparent? I'll try to play some GPU and CPU intensive games to see if and where the problem occurs. Thanks for your comment :) 

EDIT: Also, Far Cry 3 (the game I've been having the problems on as well as Heros and Generals, DayZ and a couple others) is not very CPU intensive,I'm gonna try Witcher 2, a very GPU intensive game and see what happens.


And isn't Far Cry 3 an open world un scripted game, that would seem pretty CPU intensive to me, but i don't really /know, i've never played the game/
m
0
l
a b U Graphics card
a b ) Power supply
July 28, 2014 11:35:49 AM

PieInTheSky said:
Ok, is this what you were looking for? http://puu.sh/atD75/4b4fc65f1d.jpg


Yes this is and the bios should be right so hwinfo is probably making a faulty read out.

Do you have the opportunity to test the gpu in another pc? Maybe from a friend or ...
Or lend a psu ... ?


@thorpejoe,

Did you actually create your own thread instead of just hijacking other ones? If not is that a good "idea"!
m
0
l
July 28, 2014 3:05:58 PM

Vic 40 said:
PieInTheSky said:
Ok, is this what you were looking for? http://puu.sh/atD75/4b4fc65f1d.jpg


Yes this is and the bios should be right so hwinfo is probably making a faulty read out.

Do you have the opportunity to test the gpu in another pc? Maybe from a friend or ...
Or lend a psu ... ?


@thorpejoe,

Did you actually create your own thread instead of just hijacking other ones? If not is that a good "idea"!


I tried this but as it's such a infrequent occurring problem I couldn't recreate it with a different GPU in my existing system. So, when I used a new GPU and stress tested by playing some games for an extended period of time, it was fine. By playing today it was also fine, very annoying, but it still does happen.
m
0
l
July 28, 2014 4:16:36 PM

Vic 40 said:
PieInTheSky said:
Ok, is this what you were looking for? http://puu.sh/atD75/4b4fc65f1d.jpg


Yes this is and the bios should be right so hwinfo is probably making a faulty read out.

Do you have the opportunity to test the gpu in another pc? Maybe from a friend or ...
Or lend a psu ... ?


@thorpejoe,

Did you actually create your own thread instead of just hijacking other ones? If not is that a good "idea"!


Vic this is an issue affecting a truck load of people. the link i posted at the bottom is potential help. I'm just as desperate as anyone else to figure out whats behind this.
m
0
l
July 28, 2014 4:18:06 PM

PieInTheSky said:


So, the problem you were having was fixed by getting a new PSU?


I've made no changes to my hardware, but this seems to be the answer:

"-Update your chipset drivers to get sensor type drivers to load correctly.
Failure to load these drivers would not reboot your machine.

-A reset of the CPU without a BSOD in windows generally indicates a Power issue. A good power supply that detects that its own power is fluctuating will send a signal to the motherboard to reset the CPU.

The power supply may detect too low voltage on one line, or too much current being pulled by a device and tells the motherboard to shutdown to protect it. Take caution on if you swap out the power supply. I once swapped a power supply that was shutting down the system with a cheap one for testing. The cheap one would not shut down the motherboard. The problem ended up being a graphics card that went bad, the crappy power supply did not shut down and the motherboard/graphics card ended pulling too much current and catching fire.

I now only get good power supplies that actually check their own signals"
m
0
l
a c 929 U Graphics card
a c 2488 ) Power supply
July 28, 2014 4:22:26 PM

Your supply is a good one , my guess would be a card issue.
m
0
l
July 29, 2014 5:26:52 AM

SR-71 Blackbird said:
Your supply is a good one , my guess would be a card issue.


Are you referring to my Smartpower 750w or the OP's OCZmodstream?
m
0
l
a b U Graphics card
a b ) Power supply
July 29, 2014 11:13:51 AM

thorpejoe said:


Are you referring to my Smartpower 750w or the OP's OCZmodstream?

Exactly the question why you need to start your own thread.
m
0
l
July 29, 2014 11:45:35 AM

SR-71 Blackbird said:
Your supply is a good one , my guess would be a card issue.


Are you referring to mine or thorpejoe's?
m
0
l
July 29, 2014 4:28:33 PM

Vic 40 said:
thorpejoe said:


Are you referring to my Smartpower 750w or the OP's OCZmodstream?

Exactly the question why you need to start your own thread.


As soon as i wrote that i knew i should leave.

I'm new to the forums, and i'm sorry, but i'll still be following this thread and i hope you get it fixed Pie
m
0
l
a b U Graphics card
a b ) Power supply
July 30, 2014 12:35:13 PM

PieInTheSky said:
SR-71 Blackbird said:
Your supply is a good one , my guess would be a card issue.


Are you referring to mine or thorpejoe's?

I think/guess he's referring to your psu.

You could try this,

You can open hwinfo32,after that open the sensor window,then click at the bottom "Logging start" ,after that choose a name and place for the log.like Hwinfolog at your desktop.Leave it open and do game/stresstest for a while (30min.) or whatever you do until a crash.The log will end at that point but will be saved.
After the restart upload the log to a site without to much crap (!),for instance http://www.zippyshare.com/ , and leave a link to it here.You could also look at the log yourself with either excel or libre office calc.

I will look at it when having the time for it,That means it can take a little time.Maybe it tells a little more.
m
0
l
July 31, 2014 11:47:21 AM

Vic 40 said:
PieInTheSky said:
SR-71 Blackbird said:
Your supply is a good one , my guess would be a card issue.


Are you referring to mine or thorpejoe's?

I think/guess he's referring to you psu.

You could try this,

You can open hwinfo32,after that open the sensor window,then click at the bottom "Logging start" ,after that choose a name and place for the log.like Hwinfolog at your desktop.Leave it open and do game/stresstest for a while (30min.) or whatever you do until a crash.The log will end at that point but will be saved.
After the restart upload the log to a site without to much crap (!),for instance http://www.zippyshare.com/ , and leave a link to it here.You could also look at the log yourself with either excel or libre office calc.

I will look at it when having the time for it,That means it can take a little time.Maybe it tells a little more.


Ok, I've been doing this, will update soon.
m
0
l
a b U Graphics card
a b ) Power supply
July 31, 2014 12:12:42 PM

Okay.
m
0
l
a b U Graphics card
a b ) Power supply
August 2, 2014 12:58:49 AM

The log doesnt show anything out of the ordinary except the +12voltrails,but those are as said probablly faulty readings seeing the bios gives 12.259V.

Have looked again and the log is showing somewhat high temps for the cpu,can you clean the cooler (look if there's dust under the fan) or play with the sidepanel of to see if this will lower temps and gives positive results?

Can you otherwise,just to look,uninstall the current drivers and play with older drivers?
http://support.amd.com/en-us/download/desktop/previous?...
These are for windows 7/8/8.1->64bit,try something in the 13.X range.I assume you use the latest=14.4 !?
When installing the drivers do you get the option to install or to uninstall,choose the letter first to get rid of the current drivers,after a reboot install the 13.x driver you chose.
You can also use one of the following for the uninstalling,
http://www.guru3d.com/files_details/display_driver_unin... (use in safe mode)
or
http://treexy.com/products/driver-fusion (if it works in safe mode also to use in there)

Best to look at the temp=cpu thing first,it's the easiest one.

If all of this doesn't help is it going to come to swapping parts,maybe first gpu if you can lend one or the psu.

m
0
l
August 2, 2014 2:22:08 AM

Ok, Will clean my CPU on Monday as won't have space on Saturday and Sunday, will update soon.
m
0
l
a b U Graphics card
a b ) Power supply
August 2, 2014 2:28:17 AM

PieInTheSky said:
Ok, Will clean my CPU .

=Cooler!

And if you have fresh cooling paste could you even reseat the cooler with that.Clean the old paste of with a clean rag or some soft toilet paper and medical alcohol,then reaplly the fresh paste->just a really thin layer or a small dot (rice grain/pea size) if you want the cooler to spread it.
m
0
l
August 4, 2014 7:22:01 AM

Vic 40 said:
The log doesnt show anything out of the ordinary except the +12voltrails,but those are as said probablly faulty readings seeing the bios gives 12.259V.

Have looked again and the log is showing somewhat high temps for the cpu,can you clean the cooler (look if there's dust under the fan) or play with the sidepanel of to see if this will lower temps and gives positive results?

Can you otherwise,just to look,uninstall the current drivers and play with older drivers?
http://support.amd.com/en-us/download/desktop/previous?...
These are for windows 7/8/8.1->64bit,try something in the 13.X range.I assume you use the latest=14.4 !?
When installing the drivers do you get the option to install or to uninstall,choose the letter first to get rid of the current drivers,after a reboot install the 13.x driver you chose.
You can also use one of the following for the uninstalling,
http://www.guru3d.com/files_details/display_driver_unin... (use in safe mode)
or
http://treexy.com/products/driver-fusion (if it works in safe mode also to use in there)

Best to look at the temp=cpu thing first,it's the easiest one.

If all of this doesn't help is it going to come to swapping parts,maybe first gpu if you can lend one or the psu.



Ok, So I opened up my case and tried to remove the CPU cooler so as to better clean with some canned air, tried doing this, releasing the latch and unscrewing the screws that attack to the case itself, everything I needed to do. It wouldn't budge and I looked it up and it looks like the cooler has stuck to the CPU itself. Is this a big problem? Could this be causing my issue? Why is it like this?

I've also put the side panel back on and i'll see if it changes anything, getting real sick of my PC.
m
0
l
a b U Graphics card
a b ) Power supply
August 4, 2014 11:55:11 AM

First of all,do you have new cooling paste? Otherwise leave the cooler alone until you have this to replace the old.You should take the old of the cpu and cooler with some soft toilet paper or a clean rag and this dipped in some medical alcohol.

If/when you get this there'should be no need to unscrew anything.Taking the latches of should do the trick of loosening the cooler.If it's stuck try to do it after using the pc for a short while,the paste should be softer then.If it still wouldn't go could you try to turn the cooler on the cpu and gently pulling until it came loose.

Have you found dust under the fan of the cooler? Did you at least get rid of that?
m
0
l
August 5, 2014 7:23:33 AM

Vic 40 said:
First of all,do you have new cooling paste? Otherwise leave the cooler alone until you have this to replace the old.You should take the old of the cpu and cooler with some soft toilet paper or a clean rag and this dipped in some medical alcohol.

If/when you get this there'should be no need to unscrew anything.Taking the latches of should do the trick of loosening the cooler.If it's stuck try to do it after using the pc for a short while,the paste should be softer then.If it still wouldn't go could you try to turn the cooler on the cpu and gently pulling until it came loose.

Have you found dust under the fan of the cooler? Did you at least get rid of that?


No, I built the CPU three years ago and have never reapplied cooling paste, yeah I think I should put some new paste onto the CPU, I'll order some now. I did find some dust under the fan and I'm not sure it made any difference cleaning the small amount of dust that was there.
m
0
l
a b U Graphics card
a b ) Power supply
August 5, 2014 11:20:51 AM

PieInTheSky said:
So, when I used a new GPU and stress tested by playing some games for an extended period of time, it was fine.

I think i've read over this,but with another gpu you had no problems at all?

About the cpu temps,just see if new paste and a cleaned cooler give better temps and if this helps with the problem.Let me/us know when you've replaced it and how that works out.
m
0
l
August 6, 2014 4:01:50 AM

Vic 40 said:
PieInTheSky said:
So, when I used a new GPU and stress tested by playing some games for an extended period of time, it was fine.

I think i've read over this,but with another gpu you had no problems at all?

About the cpu temps,just see if new paste and a cleaned cooler give better temps and if this helps with the problem.Let me/us know when you've replaced it and how that works out.


As the problem only pops up here and then, it's hard to know if the GPU was helping, the other GPU was a lot less high in temperature but that might because it is a better GPU.

I'll keep you posted and see what happens.
m
0
l
a b U Graphics card
a b ) Power supply
August 6, 2014 11:43:14 AM

PieInTheSky said:

I'll keep you posted and see what happens.

Okay.
m
0
l
August 22, 2014 8:07:04 AM

Ok. I'm back, So here's the story, I had some problems with the thermal compound with some bent pins. Fixed them, then got my PC back working, I then played some games and it looks like the problem is pretty much fixed. If not, I've come into some money so I think I may just upgrade it with a new cpu, psu and gpu. I thank you for all your help and sorry for the extremely late reply. Thanks again.
m
0
l
a b U Graphics card
a b ) Power supply
August 22, 2014 8:51:44 AM

That's alright,thanks for letting us know.:D 
m
0
l
!