I’ve been dealing with a really annoying technical problem on my PC for around a year now (maybe more) and I figured the smart technical minds of Qt3 would be better at diagnosing it than I’ve been.
For the last year or so, my PC has been randomly crashing back to desktop on a wide variety of games. Diablo III, Fallout 4, Rocket League, Starcraft II, it does’t matter the game, they all love to crash back to desktop in the same way: suddenly, no freezing or anything, just back to the desktop where I can immediately relaunch my game if I want to.
This has happened after I did a full reinstall of windows 8.1 on a brand new SSD, it’s continued to happen after I upgraded to windows 10. It’s a really strange (and unbelievably, holy shit unbelievably, annoying) problem in that the games never freeze up first, I never get a blue screen, the games just instantly quit and I find myself back on the desktop.
So the question is, what could be causing this? It clearly seems like something on my end, as it happens over a huge variety of game, but what could it be that’s persisted this long, over reformats and OS updates? Hardware doesn’t seem too likely, as it seems like those kind of problems lead to more hard crashes and freezes. Any help on this would be very, very appreciated.
Sounds like hardware to me. Possibly a heat problem, maybe your video card or CPU is getting too warm over time, then the game crashes and it cools off in a few seconds. Try running a temperature monitoring app. Could also be power supply if you’re running near the capacity of it. It’s strange that your system seems fine otherwise. Try to air-can dust out the whole thing, might help it run cooler.
It does sound like overheating or an inadequate PSU.
You could also check the Windows event log (on Windows 10, just right click the Start button and select Event Viewer). Check Windows Logs -> Application and System logs right after a CTD, you might see an error that could point you in the right direction.
So, I was just now playing Fallout 4, had been playing for about an hour, and sure enough, I crashed right back to the desktop suddenly. I took a look in the event viewer as suggested and looked at errors. I noticed that there have been 2 errors in the last hour, and they’re both listed under DeviceSetupManager. Does that point any fingers? I should note that one of the strange things about my computer is that I’m still running with an old SoundBlaster X-fi Xtreme Audio card. Any chance that could be messing with things?
Hmm, so I looked further into the DeviceSetupManager errors, and those happened at the time I started up my computer, not the time the game crashed. Looking into them they also seem like fairly benign errors, so I doubt that’s the issue. No event’s logged under “Applications and Services” in the event viewer, either.
I’ve wondered if the problem is my memory in the past. I suppose the one, possibly false, assumption I’ve been holding is that a memory problem would cause the computer to crash much harder than just dumping me back to the desktop. Is that a bad assumption? Regardless, I should test the memory out, any recommendations for a good memory testing program?
If it isn’t a BSOD, I highly doubt it is a hardware thing. CPU/GPU shorts/overheats generally result in major full system crashes. Overheats will shut off windows entirely to prevent damage, not just crash one program. Having dealt with a horrendous overheating/shorting problem in my PC over the last 6 months, if something is wrong with the hardware, it is big time wrong, and you will have whole system freezes and lockups. (Mobo and video card both broke, in my case)
This sounds like driver issue, particularly a sound card or video card driver issue.
Also, have you run antivirus lately? This could be malware. There are definitely viruses/programs that are built to randomly close applications, just to fuck with people. It is surprising that a full reinstall would cause problems still.
You did a fresh HD-wipe and disk reinstall? You might need to get a Windows 10 CD and do a complete system wipe.
Never in a million years would I have thought of that, but it’s stupid shit like that that always turns out to be the problem in my case. In one case, it was a stuck key, and it was messing with me for months until one day I happened to see that shift key tilted to one side.
I have had a nagging problem with my OC GTX 770 video card for over a year now that causes the video driver to give an error message that the display driver stopped responding and recovered or something like that. Some games it would recover after a long pause and keep running, but many times it would just close the game/application. This followed me from my old i7-920 computer to a brand new build with all new components (other than this same video card). I found that if I dialed back the OC to normal clock speed and reduced the power to 89% and top temp from 79 to 74, it would run stable and never do this again. I should probably RMA this thing, but I am stubborn and don’t want to wait days/weeks without one to use :).
I am not sure if this is something similar to what you are seeing, but I did have this happen in Diablo III in the spider lair area many times and used to happen like crazy in Archeage and I could never play Endless Legend when I bought it due to this annoying behavior. I figured I would add it just in case it may help the discussion. It took me like a year to finally figure out what I am pretty sure is the culprit now.
Well, I checked the F4 key and made sure it was clean and not acting strange, just to be sure. Seems like all is good on that front.
As for the display driver not responding, I actually saw a couple interesting occurrences while playing Legacy of the Void yesterday. Neither my video card or CPU are overclocked, as I’ve never found the decrease in stability to be worth it, however twice while playing Starcraft my display briefly went black for about a second, then came back up. The game didn’t crash, and I haven’t seen this behavior with any other games, but I figured maybe it’s a clue. I have a Acer 1440p G-sync monitor hooked up to my 970 via display port if that matters.
As far as the regular crashes, I had a couple on Fallout 4 last night. First one happened and I just booted the game back up. Second one happened and suddenly I wasn’t able to boot the game back up. The game failed to start right at that point where your monitor changes over to the game being fullscreen. I tried to get it going a couple times, then just restarted the computer. After restarting, I booted right back into the game and played for about 2 hours with no crashes. That does seem to be a trend. Sometimes I fire up my games and have no problems, other times they will chronically crash.
Like others have said, I really doubt it’s a memory issue, as that seems like it would lead to harder crashes, but I will run memtest 86. I’m really starting to become suspicious of some display issue though…
For what it is worth, that is exactly what would happen to me at times when I had my issue. Now that I remember, EQ Landmark was notorious for doing this to me, but in other games the behavior could be different.
Good luck finding the issue. I have no idea if mine is related, but wanted to mention it just in case.
To echo I have a similar issue from time to time in FF14 except it does lead to a game crash, most commonly in one of the (larger) new zones when it’s snowing. My theory was that the 970 was trying to call that half gig of slower vram and causing hell, but if you find a solution do share!
Try disabling your sound card (and if you have two, you should be disabling the on-board one anyway through BIOS) so you have absolutely no audio being used, just to eliminate that. If you are running a 970 NVidia it’s probably not a video driver issue, you’d get crash prompts from NVidia if it was. I agree with Jon, it doesn’t sound hardware related, but rather a driver issue.
I guess if it was me, at this point I’d be experimenting by (in addition to the suggestions above):
*If your video drivers are current, try rolling them back.
*Switching your connection from Display Port to DVI-D.
*I usually keep my last video card that I upgraded from around (not everyone does, so this may not work for you). If you’ve got your previous card around, try switching it out and see what happens.
*Have you already checked your GPU and CPU temps (especially under load)? Made sure all your fans are running when they should be? Note that fans in some newer video cards (like the 970) won’t come on unless the card is under load. You may need to remove the side panel and check all your fans (including the PSU) while under load.
*Make sure video card is fully seated in the slot. You might even try pulling it out and re-seating it.
*Make sure your memory modules are also fully seated in the slots.
*What PSU are you using? Everything plugged in good and tight? Everything, including your new SSD. Check all cables on both ends.
Mine are all pretty simple suggestions, because in my experience it is usually something simple that is the cause. Not always, but usually. I will say that despite the actual solution usually being simple, tracking down the problem is anything but. Diagnosing this kind of thing is a real bitch.
Would it be a good diagnostic to run prime95 while doing normal browsing just to see if it ever triggers an error? I know P95 icon used to turn yellow or whatever if you had issues with your memory. The theory is if it manages to go wrong, you can blame something other than the video card!
