Losing display when gaming but temps seem fine?

jkelley9

Distinguished
Nov 15, 2011
85
0
18,630
I'm having an odd issue that I've searched for but I can't seem to find any info on it, but I've already started the troubleshooting process.

Most noticeably, I'm losing my display when playing WoW. Nothing beeps, no BSOD, everything on the PC appears to still be running, but I just don't have a display. Sound does stop too though, I believe.

About a month ago I switched from (2) 23.7" @ 1920x1080 displays to a single ultrawide 2560x1080 hoping that it was just too much juice for my car to handle. At first I thought it helped, I didn't have an issue for several days. Then it started happening again. Now it happens every time I play wow, about 20 minutes in. I also went from (2) DVI cables to those monitors to a single miniDP cable to the new monitor since that's the only input type that will support the maximum resolution. So it's apparently not a cable problem since I switched up cables.

I've checked CPU and GPU temps while gaming. I don't see anything abnormal. All well within safe limits.

First thing I did was run memtest86+ overnight. I went over to my computer in the morning and strangely enough the same thing happened, I lost the display signal but the computer appears to be running... I'm not sure if memtest86+ tests the video card memory though, I'm thinking maybe it does. Either way, it did the same thing. I watched it run for 30 minutes with zero errors and then went to bed, so I don't know when it crashed.

I ran Furmark on the GPU at 2560x1080 and within about 2 minutes it was at 95°C, so I shut it down. But I don't remember it being that high during gaming since I switched to windowed mode and watched my GPU temps via rainmeter GPU skin. Is this a concern? I was running is with MSAA 2x AA which is going to be harder on it, but I chose that option since I play WoW with MSA 2x,

I also ran prime95 on the CPU. Never saw a temp above 53°C. That test is still running now, but it's been running for a solid 30 minutes without any heating or any other issues. Everything has passed so far.

Last thing, when it "crashes" (or the monitor drops out at least) I have to shut it down from the power button. But if I go to try and restart it within 5-10 seconds, it will turn on, but I get 5 long beeps (about 0.75 seconds each maybe) and the monitor doesn't get signal. Everything else appears to be running. This leads me to some type of heat problem but I don't know? Could it be the GPU heat I was seeing in furmark? But again, it doesn't get that hot during WoW.

Hardware:
XFX HD6870 with an aftermarket arctic cooler heatsink and fan (NOT overclocked)
ASRock 970 Extreme 3
Phenom II x4 965 Black (NOT overclocked) but with an aftermarket cooler master heatsink and fan, I forget which one but it's the most common aftermarket one
Mushkin eco2 240GB SSD with a fresh copy of W7 installed by me about 2 months ago maybe
Patriot 2x4GB (8GB) 9-9-9-24 if I'm remembering correctly
RAIDMAX 750W PSU that is probably ~7 years old, all other components are < 4 years old

Air flow:
Top and back blows exhaust, front blows in. No side fan, but it is a large filter screen with good ventilation on the side.
I cleaned the PC with compressed air really well. Nothing is dirty or fouled, so that is not it. I gave it a very good cleaning.


My thoughts are that it's possibly the graphics card failing or the PSU. I'm trying to find a way to pinpoint the PSU if that's what it is.

Any help would be REALLY appreciated because BF is coming up and I should be able replace whatever part I need with a pretty good deal :)

Thanks!
 

MetalBreaker

Reputable
Nov 21, 2015
10
0
4,520
1) Did you check for loose connections? (display connector is not plugged in correctly)
2) Did you check that the card is locked in the PCI-E slot?
(The lock on the right of the PCI-E slot should be facing the right way. If it's going into the slot (going left), then the lock is unlocked. If that's the case, you can just push it to the right with your hand)
3) On the IO plate, is the screw screwed in tight?
4) Is the card pushed into the PCI-E slot all the way?

If the answers are:
1) Yes, and the display connector is plugged in correctly (goes in all the way)
2) Yes, the card is locked in the PCI-E slot.
3) Yes.
4) Yes.

[strike]Then what could be wrong is your display refresh rate. On my computer, GTA 5 was set to 75hz and the system display refresh rate was set to 60hz. Then, after about 20 minutes of gameplay, the screen would go black and the game would lose sound. I fixed it by synchronizing the two display refresh rates, that is, have GTA 5 to 75hz and have the system display refresh rate set to 75hz.[/strike]

If the display connector just won't go in all the way, here's what you can try, since that happened to me too (Also if the re-seating the graphics card didn't help, or you don't know how to do it):
NOTE: Wear an antistatic wristband or discharge yourself on your case or a metal radiator!
1) Unlock the PCI-E slot lock (remember the lever on the right side of the PCI-E slot?)
2) Unscrew the IO plate and any screws holding the graphics card in position
3) Remove the graphics card (put it somewhere temporarily, don't put it on anything metal)
4) Lay your case down in a position where the motherboard is on the part of the case that's touching the floor (or any material below it)
5) Insert the graphics card into the slot where it was before (grab it by both sides, for me, I had to push it in until I heard a click)
6) Line up the graphics card with the case's hole for the IO plate as best as you can (so there are no gaps)
7) Screw in the IO plate
8) If the graphics card didn't lock itself in with the PCI-E lock already, lock it in by pushing the lever to the right (in my case, there was no clicking sound.)
9) Place the case in the position it was in before

5 long beeps for ASRock 970 Extreme 3, by another Tom's Hardware thread, means that the CPU isn't seated correctly.. If the above solutions don't work, try re-seating the CPU and replacing the thermal paste.
For agarrity (Tom's Hardware contributor), it meant that his graphics card wasn't seated correctly. Quoting his post:
"The same happened to me, but it was actually my graphics card. It wasn't mounted properly, thus the 5 beeps. After mounting it properly, it worked and hasn't beeped since."

I would also check if there are any updates to the motherboard and/or graphics card BIOS (or UEFI, depends how old the equipment is), and if there is, I would update them.
You could also try updating your graphics driver if not already updated to the latest version.

It should be working now. Please, don't rush anything, take your time.
I hope this helped you.
Good luck.
If you have any questions to ask, go ahead :D

Excuse me if I misspelled something or if something I said doesn't make sense, I am 11, and English is not my native language.
If you are asking me how I know all this, I am highly interested in electronics, that is, more specifically, computers. :)
 

jkelley9

Distinguished
Nov 15, 2011
85
0
18,630


Sorry I didn't respond to this sooner, toms isn't emailing me when a reply hits I guess.

There are no loose connections. Like I said, I even switched cables (brand new cable) and slots on the card, same issue.

The card is definitely secured in there as it should be. I will try removing and re-inserting tonight I guess. The 5 long beeps being a CPU issue is puzzling because I ran a hard prime95 on the CPU and it didn't struggle in the least. Very low temps. I haven't touched the CPU in years except to clean off the fan with compressed air (very carefully). I haven't had this problem forever, only just recently.

the amd catalyst driver is the most recent available. so no updates are available.
 

MetalBreaker

Reputable
Nov 21, 2015
10
0
4,520


It's okay, but I was worried that you would never reply to me XD
Anyway, back to the topic,

By BIOS/UEFI, I didn't mean the drivers. The BIOS/UEFI of the graphics card and motherboard, I meant like the firmware of the graphics card and motherboard. Here, I found some stuff for you, try to flash them. You can look up tutorials on how to flash them. Don't worry, even if you have the latest BIOS/UEFI version, this will only overwrite it, and make sure it wasn't corrupted.
Note: Motherboard UEFI settings will be reset!
Graphics card (choose the appropriate model, core/mem (core/memory clock) and VRam amount (1024mb, 2048mb displayed on this site, try to get the one that matches your graphics card, but with the newest "version" field): http://www.techpowerup.com/vgabios/index.php?architecture=&manufacturer=XFX&model=HD+6870&interface=&memType=&memSize=
Motherboard (choose the ones with the date 8/7/2013, doesn't matter what UEFI installation method you use (Update BIOS Under)): http://www.asrock.com/mb/AMD/970%20Extreme3/?cat=Download&os=BIOS

Also, just tried my method of graphics card re-seating... Looks like I was wrong, but it doesn't have to be it, maybe it was just on my motherboard.
On my motherboard it was:
PCI-E lever lock facing outward = Unlocked
PCI-E lever lock facing inward = Locked
Also, when I pushed the graphics card in while the slot was unlocked, the graphics card automatically locked itself in. Smart tricks, ASUS :D

If none of these solutions work, you could try that graphics card in your friend's PC, and if it starts crashing, you can safely assume the graphics card is faulty. If it doesn't crash in your friend's computer, then it's not the graphics card. Now, let's step it up further (if the GPU isn't the cause), return your friend's graphics card into his computer and swap out the power supply for your power supply. If the system crashes, then you can safely assume it's the PSU. If it isn't either the GPU or the PSU, try testing your graphics card and the power supply together in your friend's system. If they don't work, the GPU is too powerful for your PSU for some reason, probably because the PSU is providing less wattage due to its age. However, if this works, your CPU or your motherboard could be faulty.
 

jkelley9

Distinguished
Nov 15, 2011
85
0
18,630

Since my graphics card and PSU are 7 and 4 years old respectively, I started just looking to replace both with some black friday deals. In my research, I'm showing that 2GB of VRAM isn't ideal for my resolution, and especialyl since my GPU is not as strong as the more recent models (it's a radeon HD 6-series...) and I run games with at least MSAA 2x anti-aliasing. This may be what's crashing my card. Tonight I'll try dropping the resolution in my game and not using AA and see what happens.
 

MetalBreaker

Reputable
Nov 21, 2015
10
0
4,520
That's probably the best thing to do. Also, I've heard that some antialiasing in games are causing issues with radeon cards. It may be it, or RIP HD 6870, you served us well or it could be the PSU, in which case, RIP Raidmax 750W, you served us well. I hope my answers to this topic helped you, and I wish you the best of luck. Be sure to post on this thread as you go so you can keep me informed ^.^
 

jkelley9

Distinguished
Nov 15, 2011
85
0
18,630


Unfortunately I just gave my wife my spare 430W corsair PSU for her build but even then it probably wouldn't have handled my rig so it wouldn't be able to tell me anything conclusive.

I went ahead and bought an R9 390 8GB 512-Bit video card and a 1000W PSU w/ Active PFC, continuous @ 40C, 80 PLUS Bronze, Semi-Mod. If it turns out to be memory or something CPU/motherboard related (which I doubt) I'll just upgrade those parts as well when the time comes. These 2 parts were pretty old anyways. I'm going to plug in the new PSU first and see if that fixes the issue... so that I can box and keep the HD 6870 as a spare :) and then replace the card with the 390 after that.

I'll keep you posted, since this seems to be an odd issue that there isn't much info about. Just a random screen drop that doesn't come back but the PC still runs is kind of weird.
 

MetalBreaker

Reputable
Nov 21, 2015
10
0
4,520


Anything new regarding this topic?
I'm sorry if I'm putting extra pressure on you :)

 

jkelley9

Distinguished
Nov 15, 2011
85
0
18,630


I literally just got done putting the new card and PSU in. I tinkered with the cables in the computer (unplugged the pcie ones on the card and plugged them back in), pulled the card out blew it off with compressed air and plugged it back in, and then took to some compressed air in the psu to blow out some dust. There wasn't a ton, but some. I then played WoW for ~2 hours with no problem lol. So not sure what happened. Oh well, I guess the gpu and psu upgrade came prematurely but whatever lol.
 

MDXX

Distinguished
Jan 17, 2011
1,405
0
19,360


Dude its freaking raid night and kazzak. You cant be having problems now!!!