[SOLVED] PSU was set to “multi-rail” instead of “single-rail”. Changing to single rail would have solved it just the same as adding a second PCI-E cable.
Original Post:
My system reboots (eventually) every time I fire up VR. It doesn’t always happen immediately and it sometimes happens several times within 20-30 minutes of playing. It’s getting unbearable. Now, I will say that I would occasionally have this is with my Vive, but it was somewhat rare. Now that I have my Pimax, I literally had to disable SteamVR Home because it crashes every single time if I’m in there more than 30 seconds. I haven’t tried going back to my Vive yet to see if that changes things. I may do that tomorrow.
I checked the error logs and reliability monitor and I don’t see any issues. I also checked my GPU temps and they are high 50s under load. I do not experience crashed when NOT in VR. The system runs just fine otherwise and I have never had an issue outside of VR.
Tried:
Updating NVidia drivers
Updating Windows
Different USB ports
Disabling extra monitors
Turning off all Brainwarp options including parallel projections and hidden mask area
System:
Pimax 5K+ (PiTool 1.0.1.103)
9900k (Stock)
2080Ti (Stock, Driver 418.81)
Windows 10 Home (1809)
Oh, and the icing on the cake is that within the last hour I’m getting a ‘Not Tracked’ message in PiTool even though the base stations are shown as connected. Tried unplugging/replugging power/usb. Rebooting system. Rebooting HMD. Different USB ports.
EDIT/UPDATE: (2/10)
More things I tried:
Updating BIOS and all MB Drivers
Disabling all startup services except MS
Removed all extraneous hardware
Physically disconnected extra monitors and move Pimax DP cable to another connector
I was able to force a crash outside of VR by running IntelBurnTest and FurMark at the same time. Neither crashed when running by themselves. I monitored the power usage during those tests and it got nowhere near my power supply’s max, but I’m still highly suspicious that it’s not behaving as it should. I’m going to order another one which should arrive in a day or so and I’ll report back.
EDIT/UPDATE 2: (2/10)
Hmmm… so oddly enough I saw someone suggest using two PCI-E cables from the power supply to the GPU instead of using one that is a single input into the power supply and splits into two on the GPU side. I just did that and was able to run for 30 minutes or so without a crash. I’ll report back if I have more success or failure. Also ran the same stress tests with no crashes.
Someone on the NVidia subreddit informed me that my PSU supports single and multi rail. Switching this option via the Corsair Link desktop software would have also solved my problem instead of having to use 2 cables. (Imgur: The magic of the Internet)
Im having games randomly crash on me too, Im using .91. I’ve never have so many issues until I started using the 5K+. My old Vive has been a workhorse and my Pimax 4K “occasionally” has issues. And like you I get untracked base stations and sometimes a controller. But then when I start SteamVr sometimes everything will work, but sometimes I have to reboot PC. Its frustrating
It sounds like it might be a power supply issue. If it is old, cheap, or right at the edge of its capacity, there’s a good chance that could be the problem. I bought a new one to go with my RTX 2080, just so I could avoid any potential issues.
Tip: I keep a spare power supply on hand, just in case any of the family computers needs a replacement. It’s come in mighty handy, more than once.
I’m assuming it’s not anything to do with the Pimax specifically, since you were having problems before. Another possibility would be the motherboard. Or maybe the USB port might not be able to handle the power drain.
Try plugging the headset into other USB ports or even a powered USB hub.
The same here after 5 days of testing & fightings with stuttering, driver crashes, artifacts, BSOD, black screens I decide to do RMA of my Asus Strix RTX 2080 TI.
Saddly it’s starts after 14 days of using this card so now I’m f… with long service period.
I was able to force a crash outside of VR by running IntelBurnTest and FurMark at the same time. Neither crashed when running by themselves. I monitored the power usage during those tests and it got nowhere near my power supply’s max, but I’m still highly suspicious that it’s not behaving as it should. I’m going to order another one which should arrive in a day or so and I’ll report back.
heh, yeah this is exacly what I did ( bought 1000W titanium PSU ) and…that not helped , diconnencted some drives , works 6 h and start crashing again… IMHO your RTX have the same issue like mine - faulty mikron DDR6 memory but good luck anyway
I don’t have any of the other issues you described though (black screens, artifacts, stuttering, driver crashes), so I’m hopeful. I have a 1080 I can put in there to test after the power supply.
Hmmm… so oddly enough I saw someone suggest using two PCI-E cables from the power supply to the GPU instead of using one that is a single input into the power supply and splits into two on the GPU side. I just did that and was able to run for 30 minutes or so without a crash. I’ll report back if I have more success or failure. Also ran the same stress tests with no crashes.
Someone on the NVidia subreddit informed me that my PSU supports single and multi rail. Switching this option via the Corsair Link desktop software would have also solved my problem instead of having to use 2 cables. (Imgur: The magic of the Internet)
I’m glad you were able to diagnose and fix your problem. Now that you mention that, I seem to remember reading that you should use 2 different cables from the power supply.
