Black screen and crashes my PC

[private duksworth]

my advice is that when hinterlands next message comes out reply to them about the problem 

coz i don't know anything about computers so i can't help in any way

[JAFO]

7 hours ago, Thrasador said:

Jafo hates being beaten by a stupid box....

My reputation precedes me.. hehehe..

Ok.. so we know your drivers are up to date. But the latest drivers aren't always the best way to go. Sometimes they bring problems of their own, as I've seen a number of times with certain games. Nvidia in particular have a bit of a reputation in this regard. So I'd suggest trying a driver rollback, one or even two versions back, to see if it helps. I've known one person who had to go back 4 versions to get his game working again.. (he hadn't updated in quite a while, and when he did, it killed his game. He had to go back to what he was using before, and wait until the next release, which did in fact sort things out.)

 

So, since even a Win7 rollback didn't help, now we get to the hardware side of things.

7 hours ago, niklanidja said:

Self built

That's what I like to hear! A person after my own heart.

The next item on the list is overheating. Some games (and which games is not always obvious) can make computers work pretty hard, and may, if things are already marginal, trigger a thermal shutdown, which your symptoms do in fact resemble.

You mentioned that you get 10-15 minutes of play in windowed mode.. but what EXACTLY happens in full-screen mode? Insta-crash on launch? Do you get partway through, for instance, to where you load a save, or do you get a minute or two into the game before it dies?

When was the last time you opened the machine up and looked inside? Is everything clean and spotless? Or dusty as hell? Does the box sit on carpet or a hardwood floor? Carpet is notorious for rapidly clogging up the cooling system with fluff and dust. How many (if any) case fans did you fit? Or are you just relying on the CPU and PSU fans? Is your CPU air or liquid cooled? There are a number of excellent temperature monitoring programs out there.. do you use any? If so, what kind of temps are you typically seeing when gaming? If you don't have any such monitoring programs, please get one, check the temps, and report what you're seeing here. How old is the machine? If it's been a while,  the PSU itself may have gotten clogged with dust.

 

And that brings us to the subject of the power supply itself. The symptoms also resemble a power supply issue. How many watts is the PSU rated at? What brand is it? (Believe me, that can make a difference!) Go through the specs for all the components in the PC and total up their rated power draw. Post the numbers here for me, please. Is the total close to or more than the PSU can deliver? Possibly your PSU can't cope with the kind of peak load conditions some games induce.

Something that can cause problems with both overheating and power supply issues is - Overclocking. Of the CPU, the graphics card, or the RAM. Have you overclocked anything in your computer?

Sorry for all the questions, but diagnosis at a distance requires a lot of information at times. Hardware problems especially. Please be sure to answer every one of the questions I've asked.

[Carbon]

 I have the same machine as the OP with the only differences being a GTX 1070, no G-Sync but run at 120Hz, CPU @ 4.0 Ghz, 16GB @ 1866, Antec HCG 850m, ASUS z97 Pro and the game runs without a problem. Windows 10 (gaming mode disabled).

 Before anything, I would like to know the behavior of the system with other games. If you could post what other games you play and their behavior, that would be great. Also, what PSU (make/model/wattage) and mainboard are you using?

 My short list of common problems (drivers aside and very similar to JAFO's ideas):

1. Corrupt files. Uninstall/reinstall and verify files. Easy.

2. PSU failing. Borrow another or try yours on another machine. Easy unless you don't have access to another PSU (the paper clip test is useless), then hard.

3 . Heat. Clean everything; take the system apart. Easy, but time consuming.

 4. RAM failing / needs re-seating. (this causes so many weird problems it could fill a book). Re-seat modules (maybe place into different slots), run MemTest86. Easy, but time consuming. Take one stick out at a time and see what the system does; you may be able to isolate a bad/faulty stick. 24 GB...are your modules all the same make and density? How many slots do you have on your mainboard? What's the configuration (8x3?)?

 5. Other hardware failing. Hard to diagnose, but as the system draws more power for some programs, things get stressed. This is a real hail Mary and is vague enough not to really ever be wrong, but failing hard drives can cause crashes, mainboards failing, etc. Very hard to pinpoint what it is, but with patience...

[JAFO]

Very good points, @Carbon... some comments follow:

2 hours ago, Carbon said:

Before anything, I would like to know the behavior of the system with other games.

Good point, though I'm inclined to think other games are behaving fine. Generally if several games have problems, people tend to mention it.

 

2 hours ago, Carbon said:

RAM failing / needs re-seating. (this causes so many weird problems it could fill a book).

Well spotted. This is one that many people don't think of. Regarding the need for re-seating, this is something that used to be more common before clips to lock down the RAM sticks were introduced. Thermal cycling (the heat up and cool down cycles of using a computer on a daily basis) is the term used to describe it. Basically, as the RAM gets hot, it expands a little. As it cools down, it shrinks a little.. over time, this causes the RAM sticks to work their way up out of the slot, until the contacts no longer make a good connection. As I said, not as big a problem as it used to be, but there is a secondary reason re-seating can help.

And that's dust buildup in the slots themselves. Dust can get in between the connectors, leading to intermittent problems. So when re-seating RAM, always clean the contacts on the stick with a soft pencil eraser (and blow the bits of eraser away afterwards!), and blow the slots clean to remove any dust before putting the RAM back in.

You've also reminded me of another item I meant to include above.. overclocking of CPUs , graphics cards, or RAM. That can cause problems with some games, with overheating, and with overloading power supplies. I'll edit my post above to include this.

[Thrasador]

21 minutes ago, JAFO said:

Very good points, @Carbon... some comments follow:

Good point, though I'm inclined to think other games are behaving fine. Generally if several games have problems, people tend to mention it.

 

Well spotted. This is one that many people don't think of. Regarding the need for re-seating, this is something that used to be more common before clips to lock down the RAM sticks were introduced. Thermal cycling (the heat up and cool down cycles of using a computer on a daily basis) is the term used to describe it. Basically, as the RAM gets hot, it expands a little. As it cools down, it shrinks a little.. over time, this causes the RAM sticks to work their way up out of the slot, until the contacts no longer make a good connection. As I said, not as big a problem as it used to be, but there is a secondary reason re-seating can help.

And that's dust buildup in the slots themselves. Dust can get in between the connectors, leading to intermittent problems. So when re-seating RAM, always clean the contacts on the stick with a soft pencil eraser (and blow the bits of eraser away afterwards!), and blow the slots clean to remove any dust before putting the RAM back in.

You've also reminded me of another item I meant to include above.. overclocking of CPUs , graphics cards, or RAM. That can cause problems with some games, with overheating, and with overloading power supplies. I'll edit my post above to include this.

I second the RAM question also because of dual channel and having matching sticks running in parallel. I don't know if it's still the case, but historically when I built my owm PCs you wanted RAM in pairs for the dual channel aspect....having an odd stick out could be bad news....which is why I believe Carbon specifically mentioned 3 sticks at 8GB a piece for 24 GB...

Hopefully instead it's 2 sticks of 12GB or 4 sticks of 6GB.

However again I'm pretty sure an issue like that would affect multiple games, and he stated The Long Dark is the ONLY game giving him a problem right now.

That's actually a pretty big piece of information that rules out many common issues....

So whatever the issue is it has to do with a configuration that The Long Dark specifically doesn’t like.

The only new ideas I have thought of since sleeping are a clean install of The Long Dark, and right clicking on tld.exe and running it as an administrator....and then maybe try playing with different windows compatability mode choices via properties, after right clicking on tld.exe and selecting properties....

[JAFO]

2 minutes ago, Thrasador said:

right clicking on tld.exe and running it as an administrator....and then maybe try playing with different windows compatability mode choices via properties, after right clicking on tld.exe and selecting properties....

Those at least should REALLY not be necessary.. if they are, it indicates much bigger problems with the Windows installation itself, because TLD works just fine 'as-is' on most peoples machines.

[Thrasador]

2 minutes ago, JAFO said:

Those at least should REALLY not be necessary.. if they are, it indicates much bigger problems with the Windows installation itself, because TLD works just fine 'as-is' on most peoples machines.

It may not normally be necessary, but it's easy to do and something to try. I run mine as administrator....that's a very regularly brought up troubleshooting suggestion.

[Thrasador]

I'm still hung up on graphics settings, I know he messed around with the Nvidia control panel and settings in game, but there are sooooo many settings to play with, especially in the Nvidia control panel.

Like after setting it to default, I'd still go through them all and make sure everything makes sense.

I'd make sure anything that could be application controlled is set to application controlled.

Then the drop down where you select the Nvidia card or integrated graphics or Auto....I would change it to the Nvidia card to give it no choice....I believe it defaults to Auto.

I'd do the same with PhysX....set it to the Nvidia card.

In game even if you slide it to low graphics it probably leaves shadows on, but sets them to "LOW."

I play with shadows OFF. Under display somewhere at the top of the graphics settings there is an on/off for shadows, which completely turns them off....I'd try that too with all of the above....

Really at this point though like I said I'm pretty much out of ideas. You and Carbon gave him a bunch of other stuff to try....so he should try that.

[Thrasador]

What I CAN say for sure is having some sort of "black screen crash" with CERTAIN games with THAT model card seems to be a pretty prevalent issue.....

https://www.google.com/search?q=gtx+980+ti+black+screen+crash&sa=X&ved=0ahUKEwi-qZaV0_7VAhUD4iYKHXgPCxYQ1QIIQygA&biw=360&bih=616&dpr=3

[Carbon]

 Interestingly - and somewhat counter-intuitively - sometimes newer drivers for older cards aren't the best choice. This may be another option to test...rolling back the drivers to a more stable version and not so far from the date of the card in question. Yes, features and functions with newer games perhaps, but older drivers are surprisingly robust. Last ditch though.

 RAM clips have been around for a long time and I'm not sure that they completely solve the re-seating issue; I have been building and servicing systems since '98 (MS certified back then!) and RAM has always been a go-to and often correct diagnosis. Anyhow, faulty RAM makes a system very unpredictable and unstable...it almost mimics other problems many times.

 PSU as well; if I had a dollar for every time a dying PSU caused trouble, I'd be rich. This is one of the most often overlooked aspects of building a PC; people tend to go with cheaper power, which is a bad, bad idea. Lots of clean power is so important and the 980 draws a lot of juice (250w. TDP). My current setup has a HCG 850w. PSU - total overkill wattage-wise and not cheap, but I'm not going to skimp on watts or quality.

[JAFO]

3 minutes ago, Carbon said:

PSU as well; if I had a dollar for every time a dying PSU caused trouble, I'd be rich. This is one of the most often overlooked aspects of building a PC; people tend to go with cheaper power, which is a bad, bad idea. Lots of clean power is so important

Hell yes to this..  When it comes to PSU's, always go for a quality brand, price be damned.. and go for at least 20% more watts than you think the system will ever need.

[niklanidja]

11 hours ago, JAFO said:

My reputation precedes me.. hehehe..

Ok.. so we know your drivers are up to date. But the latest drivers aren't always the best way to go. Sometimes they bring problems of their own, as I've seen a number of times with certain games. Nvidia in particular have a bit of a reputation in this regard. So I'd suggest trying a driver rollback, one or even two versions back, to see if it helps. I've known one person who had to go back 4 versions to get his game working again.. (he hadn't updated in quite a while, and when he did, it killed his game. He had to go back to what he was using before, and wait until the next release, which did in fact sort things out.)

 

So, since even a Win7 rollback didn't help, now we get to the hardware side of things.

That's what I like to hear! A person after my own heart.

The next item on the list is overheating. Some games (and which games is not always obvious) can make computers work pretty hard, and may, if things are already marginal, trigger a thermal shutdown, which your symptoms do in fact resemble.

You mentioned that you get 10-15 minutes of play in windowed mode.. but what EXACTLY happens in full-screen mode? Insta-crash on launch? Do you get partway through, for instance, to where you load a save, or do you get a minute or two into the game before it dies?

When was the last time you opened the machine up and looked inside? Is everything clean and spotless? Or dusty as hell? Does the box sit on carpet or a hardwood floor? Carpet is notorious for rapidly clogging up the cooling system with fluff and dust. How many (if any) case fans did you fit? Or are you just relying on the CPU and PSU fans? Is your CPU air or liquid cooled? There are a number of excellent temperature monitoring programs out there.. do you use any? If so, what kind of temps are you typically seeing when gaming? If you don't have any such monitoring programs, please get one, check the temps, and report what you're seeing here. How old is the machine? If it's been a while,  the PSU itself may have gotten clogged with dust.

 

And that brings us to the subject of the power supply itself. The symptoms also resemble a power supply issue. How many watts is the PSU rated at? What brand is it? (Believe me, that can make a difference!) Go through the specs for all the components in the PC and total up their rated power draw. Post the numbers here for me, please. Is the total close to or more than the PSU can deliver? Possibly your PSU can't cope with the kind of peak load conditions some games induce.

Something that can cause problems with both overheating and power supply issues is - Overclocking. Of the CPU, the graphics card, or the RAM. Have you overclocked anything in your computer?

Sorry for all the questions, but diagnosis at a distance requires a lot of information at times. Hardware problems especially. Please be sure to answer every one of the questions I've asked.

Okay lets answer this questions, i gave it a go with new(old) drivers, last  6 nvidia drivers to be exact still the same. Only for the first time i played in Borderless windowed mode i played for 15 minutes and after that crash it started crashing every few minutes, same thing happens in Fullscreen mode, so yes i get a few minutes in game after save loads. I clean my PC every single month, last time was under 10 days ago, i have 2 fans in my PC and 1 on CPU. CPU is air cooled and i cant even remember which cooler i bought since i threw the box it came in but its decent, way better than default one. When i game i get normal temps, never over 70! I have changed PSU last year actually, i got 750w gold certified, i dont remember the exact same model since i threw the box also haha, i know it has enough power since i used 600wats before and it wasn't even gold certified and had no problems  And i didnt OC, thanks for helping friend! 

[niklanidja]

10 hours ago, Carbon said:

 I have the same machine as the OP with the only differences being a GTX 1070, no G-Sync but run at 120Hz, CPU @ 4.0 Ghz, 16GB @ 1866, Antec HCG 850m, ASUS z97 Pro and the game runs without a problem. Windows 10 (gaming mode disabled).

 Before anything, I would like to know the behavior of the system with other games. If you could post what other games you play and their behavior, that would be great. Also, what PSU (make/model/wattage) and mainboard are you using?

 My short list of common problems (drivers aside and very similar to JAFO's ideas):

1. Corrupt files. Uninstall/reinstall and verify files. Easy.

2. PSU failing. Borrow another or try yours on another machine. Easy unless you don't have access to another PSU (the paper clip test is useless), then hard.

3 . Heat. Clean everything; take the system apart. Easy, but time consuming.

 4. RAM failing / needs re-seating. (this causes so many weird problems it could fill a book). Re-seat modules (maybe place into different slots), run MemTest86. Easy, but time consuming. Take one stick out at a time and see what the system does; you may be able to isolate a bad/faulty stick. 24 GB...are your modules all the same make and density? How many slots do you have on your mainboard? What's the configuration (8x3?)?

 5. Other hardware failing. Hard to diagnose, but as the system draws more power for some programs, things get stressed. This is a real hail Mary and is vague enough not to really ever be wrong, but failing hard drives can cause crashes, mainboards failing, etc. Very hard to pinpoint what it is, but with patience...

I play league of legends, witcher 3, Rise of the tombraider, GTA V and no issues with those games.
Tried reinstalling a few times even changed my OS for this game ahhaha
Gigabyte z97x d3h i believe and PSU is i think Chieftec 750W, NAVITAS Series, 80+ Gold

Everything i have done even changed my ram slots and still the same issue! 
And last part im 120% sure none of my hardware is sketchy since this is literally the only game that causes issues 

 

[niklanidja]

7 hours ago, Thrasador said:

I second the RAM question also because of dual channel and having matching sticks running in parallel. I don't know if it's still the case, but historically when I built my owm PCs you wanted RAM in pairs for the dual channel aspect....having an odd stick out could be bad news....which is why I believe Carbon specifically mentioned 3 sticks at 8GB a piece for 24 GB...

Hopefully instead it's 2 sticks of 12GB or 4 sticks of 6GB.

However again I'm pretty sure an issue like that would affect multiple games, and he stated The Long Dark is the ONLY game giving him a problem right now.

That's actually a pretty big piece of information that rules out many common issues....

So whatever the issue is it has to do with a configuration that The Long Dark specifically doesn’t like.

The only new ideas I have thought of since sleeping are a clean install of The Long Dark, and right clicking on tld.exe and running it as an administrator....and then maybe try playing with different windows compatability mode choices via properties, after right clicking on tld.exe and selecting properties....

My ram is 2x8 and 2x4 and its set up correctly, even changed it this morning still the same 

[Patrick Carlson]

If you've tried all these steps below already, please note the information in there about contacting support@hinterlandgames.com with your appropriate logs and we'll see about getting to the bottom of the issue. 

 

[JAFO]

Many thanks @niklanidja for the patience and the detailed checking and reporting you've done. Good to see too that you have a good understanding of taking care of your PC. So many people treat them like an appliance that doesn't need caring for.

You've eliminated a heck of a lot of possibilities with those responses.. thanks. I'm going to give it some thought before attempting anything in-depth

 

However.. in the meantime, your power supply concerns me.

I'd like to preface my comments by noting that even an expensive PSU from a company with an excellent reputation may fail. Dud components occasionally happen to anyone. I had a brand-new excellent quality PSU fail after just a couple of months.

When a PSU fails, it can often do so very slowly.. and you'd need to have a digital voltmeter to know there were problems developing. (Speaking of which, are you up for that, if we ask you to?)

I looked up your PSU - in general, despite not being one of the better-known manufacturers, they seem to have a reasonable reputation, but some insist their gear has issues. So I went digging around some more.

47 minutes ago, niklanidja said:

PSU is i think Chieftec 750W, NAVITAS Series, 80+ Gold

And found this review. By people who know what they're doing, and do a proper teardown on it, and not just plug it in to see how it goes. I can highly recommend reading their "A Detailed Look Into PSUs" article, as well.

First of all, it's disappointing to see no ripple protection in the PSU. If excessive (50mV), voltage ripple can lead to subtle damage to other components in hard to diagnose ways. Voltage being too high or low can cause components (including RAM) to have errors. Abnormally high or low voltages can cause system shutdown and in extreme cases may burn out components.

Secondly, the quality of the hand-soldered (!) solder work in the review sample was shockingly poor. I honestly shuddered when I saw the pics. Just the sight of it made me want to reach for my soldering iron.

One of my other hats says "electronics nut" on it.. And I can assure you that bad solder joints can lead to all kinds of problems with a circuit, weird problems.. long before total failure occurs. Voltages being outside of acceptable limits seems a likely outcome.

In your shoes, I wouldn't trust that PSU until I'd opened it up and inspected the workmanship for myself. Is it still in warranty, or how old is it?

[niklanidja]

1 minute ago, JAFO said:

Many thanks for the patience and the detailed checking and reporting you've done. Good to see too that you have a good understanding of taking care of your PC. So many people treat them like an appliance that doesn't need caring for.

You've eliminated a heck of a lot of possibilities with those responses.. thanks. I'm going to give it some thought before attempting anything in-depth

 

However.. in the meantime, your power supply concerns me.

I'd like to preface my comments by noting that even an expensive PSU from a company with an excellent reputation may fail. Dud components occasionally happen to anyone. I had a brand-new excellent quality PSU fail after just a couple of months.

When a PSU fails, it can often do so very slowly.. and you'd need to have a digital voltmeter to know there were problems developing. (Speaking of which, are you up for that, if we ask you to?)

I looked up your PSU - in general, despite not being one of the better-known manufacturers, they seem to have a reasonable reputation, but some insist their gear has issues. So I went digging around some more.

And found this review. By people who know what they're doing, and do a proper teardown on it, and not just plug it in to see how it goes. I can highly recommend reading their "A Detailed Look Into PSUs" article, as well.

First of all, it's disappointing to see no ripple protection in the PSU. If excessive (50mV), voltage ripple can lead to subtle damage to other components in hard to diagnose ways. Voltage being too high or low can cause components (including RAM) to have errors. Abnormally high or low voltages can cause system shutdown and in extreme cases may burn out components.

Secondly, the quality of the hand-soldered (!) solder work in the review sample was shockingly poor. I honestly shuddered when I saw the pics. Just the sight of it made me want to reach for my soldering iron.

One of my other hats says "electronics nut" on it.. And I can assure you that bad solder joints can lead to all kinds of problems with a circuit, weird problems.. long before total failure occurs. Voltages being outside of acceptable limits seems a likely outcome.

In your shoes, I wouldn't trust that PSU until I'd opened it up and inspected the workmanship for myself. Is it still in warranty, or how old is it?

Its about a year old, still in warranty! 

[JAFO]

Just now, niklanidja said:

Its about a year old, still in warranty! 

In that case, tell them you think it's acting up, and get them to check the voltages under load (and I stress that), to see if it's failing.

[niklanidja]

2 minutes ago, JAFO said:

In that case, tell them you think it's acting up, and get them to check the voltages under load (and I stress that), to see if it's failing.

I will in a few days, thanks! 

[JAFO]

7 minutes ago, niklanidja said:

Its about a year old, still in warranty! 

Almost forgot.. do you have a digital voltmeter? Do you know how to check a PSUs voltages?

[niklanidja]

5 minutes ago, JAFO said:

Almost forgot.. do you have a digital voltmeter? Do you know how to check a PSUs voltages?

I dont own one and i think i know how to check, just by going in BIOS i believe? 

[JAFO]

Just now, niklanidja said:

I dont own one and i think i know how to check, just by going in BIOS i believe? 

That's better than nothing, but I prefer taking the readings directly off the PSUs connectors, when possible. But by all means, yes, let's see what it has to say.

[niklanidja]

1 minute ago, JAFO said:

That's better than nothing, but I prefer taking the readings directly off the PSUs connectors, when possible. But by all means, yes, let's see what it has to say.

CPU Vcore 1200v and DRAM Voltage 1.368V and thats all by default on auto 

[JAFO]

The review gets worse.. (I'm still reading) At 50% load, one of the lines dropped below acceptable voltage levels.. Quite a few other tests failed at various loads too.

[JAFO]

2 minutes ago, niklanidja said:

CPU Vcore 1200v and DRAM Voltage 1.368V and thats all by default on auto 

So nothing on the 12v, 5v or 3.3v lines? Those are the ones I'm interested in. Looks like only a voltmeter can tell us. So we leave it to the warranty guys to check.