PC restarts after leaving it on for rendering

[Frick]

Hi, MentalRay doesn't use any GPU unless you tell it to. As far as I know there is no way to tell it to do that. Maya is usually based on CPU and RAM unless you use something like VRAY and start using VFX which I think will then start pulling from the card otherwise it will just use the CPU if the Quadro or the Firepro isnt
detected? (I am not sure)

I meant what happens if you load the GPU with something? Just to exclude weird PSU issues.

 

[-1nf1n1ty-]

I meant what happens if you load the GPU with something? Just to exclude weird PSU issues.

Sure, Ill do that test tonight!

Update on whats happening: I left the computer rendering last night, woke up and no restarts or increased temps. PC has stayed around the same temps I shared with you all in the previous post 72c and lower. Havent been able to do any cleaning for it yet or memtest will do it soon! It is still rendering.

 

[LAN_deRf_HA]

In my experience restart/shutdown without a BSOD is almost always a power supply or a ram stability issue. Had two Corsair units do that to me.

 

[Norton]

In my experience restart/shutdown without a BSOD is almost always a power supply or a ram stability issue. Had two Corsair units do that to me.

This

What model psu do you have? Corsair CX units like to restart on thermal protection at a fairly low temp.

 

[-1nf1n1ty-]

This

What model psu do you have? Corsair CX units like to restart on thermal protection at a fairly low temp.

Mine? this one : http://www.corsair.com/~/media/corsair/product-photo-root/t/tx650-angled.png I got it in 2006-7? (justa guess I've had it for a really long time and has been through a lot of past builds. EDIT: was 2009!

Think it could JUST now start giving me problems?

 

[jaggerwild]

It's the middle of summer in Florida, temps in the high 90s out side, its in side being put under a load. TEMPS! A 10 year old power supply, Does it run all the time or just in use?

 

[-1nf1n1ty-]

It's the middle of summer in Florida, temps in the high 90s out side, its in side being put under a load. TEMPS! A 10 year old power supply, Does it run all the time or just in use?

Just whenever I decide to turn the computer on, I usually turn it off when I am about to sleep, but decided to render some stuff

 

[Jetster]

render what with what program?

 

[DeathtoGnomes]

Hey dont forget that Windows 10 update autorestart bullshit. Check it before you start to do any rendering.

 

[-1nf1n1ty-]

render what with what program?

Render a 3D scene in Maya and Mental Ray with VFX

Hey dont forget that Windows 10 update autorestart bullshit. Check it before you start to do any rendering.

Good idea! I think that may have been just one of the restarts.

Update: Cleaned out PC with this blower that I have (LOTS of dust and dog hair, thank you dog). Installed new fans (Noctua 140mm) for the front of the case. Idles dropped a couple of degrees. Used optimized defaults in bios. About to run Memtest when I leave the house. Things seem to be running pretty well and at good temps on CPU side. GPU is about as well as it normally is 47c so nothing really changed there. Other than memtest anything else I need to do?

 

[DeathtoGnomes]

Paint your case Blue?

 

[jaggerwild]

Memtest is for people hunting ghosts, you had no memory issue's prior.

 

[eidairaman1]

is automatic restart disabled?


make sure ram voltage is set properly too

 

[-1nf1n1ty-]

Paint your case Blue?

Haha, That would be interesting!

is automatic restart disabled?


make sure ram voltage is set properly too

Yeah all of that is disabled and reset to defaults.

 

[Mussels]

just a thought seeing those temps - get a program like intel burn test (IBT), set it to very high and let it run its 10 passes (or more, if you have the time)

While it will do some basic ram and stability testing, what i'm wanting you to look for here is consistent scores (gflops results) - if its overheating enough to be unstable, the results should get slower and slower as time goes on, as the heat builds up.

 

[eidairaman1]

Haha, That would be interesting!



Yeah all of that is disabled and reset to defaults.

i ensure my ram is set manually for voltage.

 

[-1nf1n1ty-]

Little update, everything seems to be running really well. I have found no problems. I even replaced the thermal compound. I did some tests on the fans, and it seems like the Corsair SP120 fans I have are not as good on the Rad as the NZXT FN V2 fans I've had for the case (came with the case). I've seen idle temps for the NZXT fans drop to 27c (load hit 70 and stayed low) while the Corsair is 30-32. I thought static pressure fans were great for liquid AIO coolers? Should I just use the NZXT fans? Thanks, other than this weird "problem" the PC seems to be running very well! So thanks everyone for all the help

 

[DeathtoGnomes]

Great to hear everything is good.

The Corsair fans top out at 2350/62 cfm while the NZXT fans are 1200rpom/45cfm. Its hard to believe the Corsair fans not performing better then the NZXT fans.

Assuming you are using push only config, I think the corsair fans are driving idle temps too low and forcing the PWM low causing the fans to cycle on and off, sounds like riding a rollercoaster.

If you try the Corsair fans again, find an app that lets you control speeds, whatever you can find that works for you. I use SpeedFan since my fans (push-pull) can be heard spinning up and down way too much

 

[Recon-UK]

I would look at your PSU, all your rig says is that you have a Corsair 650w PSU, not all Corsair unit's can withstand high loads (CS/CX)

 

[-1nf1n1ty-]

Great to hear everything is good.

The Corsair fans top out at 2350/62 cfm while the NZXT fans are 1200rpom/45cfm. Its hard to believe the Corsair fans not performing better then the NZXT fans.

Assuming you are using push only config, I think the corsair fans are driving idle temps too low and forcing the PWM low causing the fans to cycle on and off, sounds like riding a rollercoaster.

If you try the Corsair fans again, find an app that lets you control speeds, whatever you can find that works for you. I use SpeedFan since my fans (push-pull) can be heard spinning up and down way too much

I have the quiet edition that puts out 37.5 cfm and I use the Antec Chill control V (lol funny name) I also have the "speedfan" program and the fans are in push config

I would look at your PSU, all your rig says is that you have a Corsair 650w PSU, not all Corsair unit's can withstand high loads (CS/CX)

it's an old corsair PSU check the previous posts, I shared a picture of the exact model!

 

[DeathtoGnomes]

I have the quiet edition that puts out 37.5 cfm and I use the Antec Chill control V (lol funny name) I also have the "speedfan" program and the fans are in push config

*snip*

Use Speedfan > config > fan control, to set the fan speed percentage graph to maintain a speed that remains low until you actually do anything that will load up your CPU and raise the temp.

 

[-1nf1n1ty-]

Well I guess I spoke too soon, my PC started crashing and gave me the error of Thread stuck in device driver. After doing research I followed directions of uninstalling and reinstalling display drivers both for my onboard and card. Still getting the error. I wonder if this is what was causing my pc to restart.

I have used DDU to do the uninstall

EDIT: I fixed it

 

[alucasa]

80c shouldn't be an issue. I have had my rig shot up to 90c for more than 12+ hours during a long render and my rig withstood fine for years before I upgraded it.

I'd do an OS reinstall and do the render again with a clean OS. If it still reboots, then temp somewhere on the mobo or some other component is going over its limit.

 

[DeathtoGnomes]

If over heating still has you worried, you can change windows Power Options or even create a profile just for rendering jobs.

Change plan settings > change advanced power settings > scroll down to Processor Power Management > change these settings to see what works. You can prevent a meltdown with "Maximum processor state", you can adjust down in 1% steps.

Other than this you might need even more cooling power.

 

[peche]

dude try this…

#1, clean rad, fans and all hardware, also repaste with Goooooood stuff, like Arctic Cooling MX4
#2 download VALLEY BENCHMARK, open HWinfo and run sensors before running the bench crap…
#3 during this bench be sure to check temps while the benchs are running, to see if heavy loads are the responsible's of restarting you rig dude,
#4 If not, if you could also reach high temps but not reset with this app, try to set all your values to default on Bios, also if possible look for a newer version of your current Bios or flash it,
#5 try the program you use, if there is another restart after all this process, you might check or reinstall your render programs, also running the intel processor scan software that was mentioned in some posts ago it’s a great alternative,

Ram could be a problem, Older Corsair's units like TX are strong enough so I don’t think the problem should be there…

Regards,