|
Problem description: PC was sitting idle overnight, sat down this morning and it was stuck on the POST screen saying CPU Temperature Error. I let it cool down for an hour and booted into Windows, and watched the temp steadily climb until I tried to reboot. After reboot, it was stuck on POST again with the temperature error, and BIOS was reporting it had shot up past 80. Attempted fixes: I have a Corsair H100i all-in-one cooler. In BIOS, I could see the fan speeds at around 2700RPM, and the fans connected to the H100i are indeed spinning. I have never seen anything reported for the pump in BIOS. In Windows, the Corsair Link utility showed 0 rpm for both the fan and the pump, though the fans were definitely blowing at the time.It also showed a H100i temperature of 34 degrees, even while the CPU cores were going 75 to 90/ No errors in device manager, H100i is recognized in Devices and Printers as Corsair XRP (or something like that, can't check atm). Recent changes: I did do a format-reinstall of Windows on Friday night, and everything was looking good through the weekend until today. I hadn't done any gaming since the reinstall, but I did fire up my Oculus Rift for fifteen minutes Saturday and didn't see any temperature errors -- I figure that would be enough to spike it if it was having this issue on Saturday. Before this weekend, this PC had been in use since January and could handle 4k gaming without any trouble. -- Operating system: Win10 Pro x64 System specs: Intel 8700k, GeForce 1080 ti, Asus Z370-E motherboard, Corsair h100i v2 cooler. Location: Murica I have Googled and read the FAQ: Yes My thought is that the pump unit on the H100i must have failed, but I'd love to get a second opinion before I go out and spend a hundred bucks and an hour or two of fiddly work to replace it and see. I'm not sure how to test my theory without actually installing another cooler to see. Thanks! Phenotype fucked around with this message at 16:21 on Jul 16, 2018 |
#
?
Jul 16, 2018 16:00
|
|
|
|
| # ? May 4, 2024 15:57 |
|
|
Sounds like a failed pump not much else you can do but replace. Maybe try a different fan socket off the mobo. If you think there can be other factors in play You can test by buying a cheap 15 dollar cooler then return it after youve narrowed that to being the problem.
Killer Low Life fucked around with this message at 05:53 on Jul 17, 2018 |
|
#
?
Jul 17, 2018 05:51
|
|
|
It's almost certainly the pump. It's the most frequent point of failure with those CLCs. If you want to double-check it, you can hook it up to an external power supply: Just two pieces of wire and one of those 9-12V power bricks with a coaxial connector you get with a lot of devices will do fine. If it doesn't start running (and you should be able to hear it running at 12V), then it's dead. But that's not even necessary, really - the pump is very likely dead. Unless you want the looks of a closed loop liquid cooler, and if you don't want to go full custom loop, I recommend getting a good air cooler like the Noctua NH D14/15 or Dark Rock PRO 3/4. They cool just as well as most CLCs, but they don't have a pump that can fail.
|
|
#
?
Jul 17, 2018 11:24
|
|
|
Thanks for the confirmation, guys! New cooler seems to have done the trick, just booted up and temps are holding at around 28 degrees. Sure was a pain in the rear end though, I bought a slightly better model for the Best Buy price because I needed my PC up immediately, only to find out that despite the advertised measurements, it wouldn't QUITE fit in my case. So I had to go back the next day to get the exact same model, and didn't have time to install it anyway til tonight. Oh well!
|
|
#
?
Jul 19, 2018 05:19
|
|