PDA

View Full Version : [too much powaaah] PC continually reboots while playing PCars2 in VR



Myrkabah
26-09-2017, 05:14
System Specs:

GeForce GTX1080 8GB
16 GB DDR3 RAM
i7 4790K 4.0 mHz
Game installed on SSD

The game continually causes my computer to reboot, especially while loading the start of a track. I've tried lowering graphics settings, but it doesn't seem to resolve. I've updated my graphics drivers to 385.69, but it hasn't helped either. What else can I do?

(Thanks in advance!)

Aldo Zampatti
26-09-2017, 05:18
I would think this is due to PCARS sucking too much juice from your PSU.

Are you overclocked? If so, please remove all overclock CPU/GPU and give it a go to see if this aliviates or eliminates the issue. My setup is almost the same as yours

Morgan Henstridge
26-09-2017, 05:36
If it is crashing during the loading, it may also be a corrupt file or 2.

Verify the Game in Steam to make sure everything is OK there.

Myrkabah
26-09-2017, 05:51
I would think this is due to PCARS sucking too much juice from your PSU.

Are you overclocked? If so, please remove all overclock CPU/GPU and give it a go to see if this aliviates or eliminates the issue. My setup is almost the same as yours

I don't have any particular overclock settings that I've set up. My PSU is 750 watts, if that helps. I'm fairly certain that's sufficient, but I'm open to feedback on that.


If it is crashing during the loading, it may also be a corrupt file or 2.

Verify the Game in Steam to make sure everything is OK there.

I verified game - everything reported back as valid.

Morgan Henstridge
26-09-2017, 05:54
If you just launch the game without VR does it then work as expected??

Myrkabah
26-09-2017, 06:09
If you just launch the game without VR does it then work as expected??

I haven't actually tried that yet. I bought it for VR. :)

I did lower the graphics settings some more, and I've been able to play a few tracks without crashing. I may have just underestimated how much I needed to reduce, hopefully. I'll continue playing and report back if I experience further crashes, or if it generally works now.

gandy
26-09-2017, 06:30
My PSU is 750 watts


if you have an unbranded PSU it could be only pushing out around 500 watts under full load and the rest could be lost in heat. You really need a good quality psu like EVGA, Corsair etc.. that is rated 80 Bronze or better for the watt's you choose. So many people still see the psu as an afterthought when building a pc and go cheap on them and then some time later they start to have issues. Another thing to remember is that power supplies degrade over time with use if you are using an old psu.

Myrkabah
26-09-2017, 06:55
if you have an unbranded PSU it could be only pushing out around 500 watts under full load and the rest could be lost in heat. You really need a good quality psu like EVGA, Corsair etc.. that is rated 80 Bronze or better for the watt's you choose. So many people still see the psu as an afterthought when building a pc and go cheap on them and then some time later they start to have issues. Another thing to remember is that power supplies degrade over time with use if you are using an old psu.

It's a Thermaltake PSU, which I believe to be a reputable brand.

Lower graphics settings helped for a while, but I just crashed again. I'm going to try lowering further, but I definitely feel like I'm already lower (Medium or lower settings, Trilinear filtering, etc.) than this configuration should be able to handle. Plus it shouldn't react to too high settings by restarting my PC. What else could be happening?

Shepard2603
26-09-2017, 06:58
It's a Thermaltake PSU, which I believe to be a reputable brand.

Lower graphics settings helped for a while, but I just crashed again. I'm going to try lowering further, but I definitely feel like I'm already lower (Medium or lower settings, Trilinear filtering, etc.) than this configuration should be able to handle. Plus it shouldn't react to too high settings by restarting my PC. What else could be happening?
Perhaps you can log temps, voltage, load of CPU and GPU etc. to see what's going on while playing. That might help to find if there is a hardware issue.

gandy
26-09-2017, 07:10
Also have a look in Event Viewer in the Custom Logs/Administrative Events section to check to see if anything stands out relating to drivers or hardware.

Aldo Zampatti
26-09-2017, 15:16
I haven't actually tried that yet. I bought it for VR. :)

I did lower the graphics settings some more, and I've been able to play a few tracks without crashing. I may have just underestimated how much I needed to reduce, hopefully. I'll continue playing and report back if I experience further crashes, or if it generally works now.


If with lower settings works and with higher it reboots, then I stand by my first guest about power draw. Keep an eye on it since you might have an issue that is not 100% evident. Glad that you could play though. :)

GTsimms
26-09-2017, 15:42
If with lower settings works and with higher it reboots, then I stand by my first guest about power draw. Keep an eye on it since you might have an issue that is not 100% evident. Glad that you could play though. :)

My #1 suggestion would be PSU. I have an EVGA , that I need to send back. It trips on load only! It will test fine with the plug in box, since its not under a load!

Myrkabah
27-09-2017, 05:27
My #1 suggestion would be PSU. I have an EVGA , that I need to send back. It trips on load only! It will test fine with the plug in box, since its not under a load!

Hi there! I was able to recreate with a hardware monitor open, but I'm not able to post the link due to forum restrictions. Would you be willing to please lift those for me so I can post a link to a pastebin with logs in it?

Thanks!

Aldo Zampatti
27-09-2017, 05:30
Hi there! I was able to recreate with a hardware monitor open, but I'm not able to post the link due to forum restrictions. Would you be willing to please lift those for me so I can post a link to a pastebin with logs in it?

Thanks!


PM to me and I'll post it here.

FriendCalledFive
27-09-2017, 05:35
Open the case and have a desk fan blow air in, if the problem stops you know it is an overheating issue.

Morgan Henstridge
27-09-2017, 05:56
If your PSU is a decent brand, and you are familiar with overclocking, you may need to just give a little more voltage to the GPU in Afterburner or similar.

My 780ti needed a little boost when PCARS 1 was released.

Aldo Zampatti
27-09-2017, 15:37
Hi there! I was able to recreate with a hardware monitor open, but I'm not able to post the link due to forum restrictions. Would you be willing to please lift those for me so I can post a link to a pastebin with logs in it?

Thanks!

Here's Myrkabah's link:

https://pastebin.com/HGESE84B

Aldo Zampatti
27-09-2017, 15:41
The one that's titled Temperature#6 worries me. ~97 degrees consistently. Is that CPU?

If so, that's definitely the cause. let me know

Myrkabah
27-09-2017, 16:35
If your PSU is a decent brand, and you are familiar with overclocking, you may need to just give a little more voltage to the GPU in Afterburner or similar.

My 780ti needed a little boost when PCARS 1 was released.

This seems to have helped - I set the power to 102%, and was able to play with more reasonable settings. I'll continue playing with this and report back if I see further crashes!

Aldo Zampatti
28-09-2017, 01:17
This seems to have helped - I set the power to 102%, and was able to play with more reasonable settings. I'll continue playing with this and report back if I see further crashes!
What are your temperatures now? specifically that "temperature6" you had before.

Muz_J
28-09-2017, 01:32
The one that's titled Temperature#6 worries me. ~97 degrees consistently. Is that CPU?
If so, that's definitely the cause. let me know

I had a look out of curiosity.
That's looks like a motherboard temp sensor to me (at a guess).
The CPU core temp's aren't actually that high (early to mid 60's) - higher than ideal, probably not the root cause.

The temp you are referencing is listed as:
"/lpc/nct6791d/temperature/6
Temperature #6"

And if you check column G (imported the file as a CSV into Excel) - Fan control #6 is permanently on 100%.
That's listed as:

/lpc/nct6791d/control/5
Fan Control #6


So I suspect it's an air flow / cooling problem with the motherboard (at a guess - without knowing where that sensor is located physically).

Nail down the sensor location and you have your root cause.

Myrkabah
28-09-2017, 02:46
The one that's titled Temperature#6 worries me. ~97 degrees consistently. Is that CPU?

If so, that's definitely the cause. let me know

I'm pretty sure this temperature reading is bogus. It corresponds with AUXTIN3 in SpeedFan, which *would* correspond with the PSU temperature, if I were sensing that - but I don't believe I am, so it's returning a weird reading. My PSU is cool to the touch while the readout is currently reading 99 degrees centigrade, which is a good indicator that it isn't actually overheating. Info found here (http://www.tomshardware.com/answers/id-2375584/auxtin-high-temp.html).

I haven't experienced any further crashes since setting my power limit on my GPU to 102%, so yay! I'm about to drive for a few hours, so I'll know if it's well and truly fixed after that.

Aldo Zampatti
28-09-2017, 02:48
I'm pretty sure this temperature reading is bogus. It corresponds with AUXTIN3 in SpeedFan, which *would* correspond with the PSU temperature, if I were sensing that - but I don't believe I am, so it's returning a weird reading. My PSU is cool to the touch while the readout is currently reading 99 degrees centigrade, which is a good indicator that it isn't actually overheating. Info found here (http://www.tomshardware.com/answers/id-2375584/auxtin-high-temp.html).

I haven't experienced any further crashes since setting my PSU to 102%, so yay! I'm about to drive for a few hours, so I'll know if it's well and truly fixed after that.

Great, I'm glad is sorted (so far). If you're able to play for an extended period, then report back so we can be aware that is fixed :)

Morgan Henstridge
28-09-2017, 03:22
I'm pretty sure this temperature reading is bogus. It corresponds with AUXTIN3 in SpeedFan, which *would* correspond with the PSU temperature, if I were sensing that - but I don't believe I am, so it's returning a weird reading. My PSU is cool to the touch while the readout is currently reading 99 degrees centigrade, which is a good indicator that it isn't actually overheating. Info found here (http://www.tomshardware.com/answers/id-2375584/auxtin-high-temp.html).

I haven't experienced any further crashes since setting my power limit on my GPU to 102%, so yay! I'm about to drive for a few hours, so I'll know if it's well and truly fixed after that.

I run with my power limit maxed out. Let these modern cards take as much power as it wants. It will still throttle itself up and down in accordance with the Temp limit.

Myrkabah
28-09-2017, 07:30
I set the power limit to 108% and played for a few hours tonight - I had two shutdowns in that time.

I"m noticing that it tends to happen on menus - like when I'm tuning the car, or starting the race. I'm using a Logitech G920 to do menu selection, if that matters.

I'll try lowering some of the graphics settings back tomorrow - although that would be a shame, everything was looking quite nice in VR. :)

Siberian Tiger
28-09-2017, 07:52
You can additionally try to Cap the Frames.

Launch Options: - fpscap 60 <- You can choose the Number. (Max FPS)

In the Menu's without this Launch Option there is no FPS Cap, and this could lead that your GPU is sucking more than it has to...

Myrkabah
29-09-2017, 02:25
OK - working with some different options tonight. I'm still seeing reboots every 30-60 minutes or so. To recap, this is what I've done so far:

- Verified game files in Steam
- Installed latest graphics drivers
- Checked temperatures and voltages
- Lowered graphics settings to Low-Medium
- Increased maximum power availability to GPU by 8%
- Capped frames for PCars2 at 60FPS

Things we haven't tried yet:

- Replacing my PSU
- ???

Does anyone have any other options or tests that I could try? :)

Aldo Zampatti
29-09-2017, 02:37
OK - working with some different options tonight. I'm still seeing reboots every 30-60 minutes or so. To recap, this is what I've done so far:

- Verified game files in Steam
- Installed latest graphics drivers
- Checked temperatures and voltages
- Lowered graphics settings to Low-Medium
- Increased maximum power availability to GPU by 8%
- Capped frames for PCars2 at 60FPS

Things we haven't tried yet:

- Replacing my PSU
- ???

Does anyone have any other options or tests that I could try? :)

When you say "increase power availability to GPU by 8%" you mean like from MSI Afterburner giving 8% of additional power? 108%?

If so, don't! (for this testing). If the problem is de PSU, that would only aggravate things. Leave it at 100% (0% additional)

Morgan Henstridge
29-09-2017, 03:16
When you say "increase power availability to GPU by 8%" you mean like from MSI Afterburner giving 8% of additional power? 108%?

If so, don't! (for this testing). If the problem is de PSU, that would only aggravate things. Leave it at 100% (0% additional)

Or even to test the PSU theory, can you reduce the Power Limit to below 100%?? im not in front of a desktop PC at the moment to check this.

You could also set the resolution low, like 720p with Vsync set to half in the nvidia driver to really reduce the load and power draw on the GPU and see if this makes a difference.

Chellfire
30-09-2017, 04:47
I had the same issue. I tried a bunch of different fixes, including a new psu, no overclocking my 1060 (brand new also)
In the end, I ran an extension lead to another wall socket instead of running my pc/oculus/soundcard etc on a power board.
fixed it perfectly for me. I haven't had a crash since. My old PSU is probably fine too.

Aldo Zampatti
30-09-2017, 04:49
I had the same issue. I tried a bunch of different fixes, including a new psu, no overclocking my 1060 (brand new also)
In the end, I ran an extension lead to another wall socket instead of running my pc/oculus/soundcard etc on a power board.
fixed it perfectly for me. I haven't had a crash since. My old PSU is probably fine too.

Maybe signal noise from the outlet? Odd but it could...

Chellfire
30-09-2017, 04:53
Quite possibly. It didn't manifest until The new card and game started to draw a lot more power. I moved up from a 760 to the 1060 and added oculus. All good now though. :D

Myrkabah
01-10-2017, 17:20
I had the same issue. I tried a bunch of different fixes, including a new psu, no overclocking my 1060 (brand new also)
In the end, I ran an extension lead to another wall socket instead of running my pc/oculus/soundcard etc on a power board.
fixed it perfectly for me. I haven't had a crash since. My old PSU is probably fine too.

Interesting! I tried this late last night, and haven't experienced any further crashes. I'll have another play session this evening and report back on if this holds.

If this works - THANK YOU THANK YOU THANK YOU.

Mad Al
01-10-2017, 18:53
Quite possibly. It didn't manifest until The new card and game started to draw a lot more power. I moved up from a 760 to the 1060 and added oculus. All good now though. :D

Weird as a 1060 is supposed to have a lower maximum power draw than the 760... (120W vs 170W)

Myrkabah
02-10-2017, 19:46
Weird as a 1060 is supposed to have a lower maximum power draw than the 760... (120W vs 170W)

It's possible that it's more sensitive to power fluctuations despite having a lower power requirement.

That said - no crashes as of last night! I was able to increase my graphics settings back up, and things are working beautifully (except for some weird flashing/glitching every so often, which I'm willing to tolerate for now). Thanks again to everyone for their help! This one can be marked as officially solved. Thanks Chellfire for identifying the underlying cause. I probably never would have thought of that.

Stocky
02-10-2017, 22:40
Glad you solved your problem.

There's a thread about the flashing/glitches.

http://forum.projectcarsgame.com/showthread.php?53664-Known-Issue-Random-white-flashes-in-oculus-vr

Johngrim
02-10-2017, 22:52
Try underclocking your cpu.My 12 month stable overclock had to be reduced to stop full system crashes with project cars 2 .

Myrkabah
12-10-2017, 03:31
Hey all,

One last update! The problem came back (with a vengeance), but I actually fixed it this time!

It turns out it wasn't just PCars2. I had the same thing in Arktica.1. Once I discovered that, I looked at other more general help threads on the topic, and apparently something is up with the PCI-E slot I have my card in. Moving it to the other PCI-E x16 slot completely solved the problem.

In case it helps anyone else!