hi all,
so i've been having a series of random freezes lately--mostly during the load screens of gaming. the display would freeze, a USB disconnect sound would play, and a hard restart would be needed (alternately, the screen would sometimes go black and the computer would restart itself). in either case, the next boot would appear normal.
i combed the event viewer but mostly only found kernel power errors. (one did eventually turn up pointing to the GPU, which i've since disconnected and cleaned--it had quite a bit of dust in it.) SSD is brand new (samsung), and i ran a memory test as well. looking through posts here and elsewhere led me to consider the power supply (corsair TX850) as a possible culprit. as a result, i disconnected that too and tested it--first with the paperclip test, then with a multimeter. the paperclip test showed only that the fan did not start when the psu was on--but given that the TX850 has a variable speed fan, i wasn't sure if this meant anything. the multimeter showed all voltages normal.
with this in mind, i reconnected the PSU and booted the computer (working fine right now, though no gaming, and using the mobo GPU). i then ran prime95 small FFT to test the CPU. temps were really high (~100) so i reduced the OC to a pretty light amount. temps dropped to ~72. i examined the PSU a few times during the prime95 test. on the first occasion, the fan was spinning, so i thought--okay, it's not the PSU. the second time, though, i noticed the PSU fan was not spinning. this was with prime95 running, 100% load. it did seem to be trying to spin, though--a slight rocking or twitching motion, and when i tried a few times to give it a spin with a piece of paper, it started spinning on its own. a few minutes later, i looked, and it was stopped again.
so i wonder a few things:
so i've been having a series of random freezes lately--mostly during the load screens of gaming. the display would freeze, a USB disconnect sound would play, and a hard restart would be needed (alternately, the screen would sometimes go black and the computer would restart itself). in either case, the next boot would appear normal.
i combed the event viewer but mostly only found kernel power errors. (one did eventually turn up pointing to the GPU, which i've since disconnected and cleaned--it had quite a bit of dust in it.) SSD is brand new (samsung), and i ran a memory test as well. looking through posts here and elsewhere led me to consider the power supply (corsair TX850) as a possible culprit. as a result, i disconnected that too and tested it--first with the paperclip test, then with a multimeter. the paperclip test showed only that the fan did not start when the psu was on--but given that the TX850 has a variable speed fan, i wasn't sure if this meant anything. the multimeter showed all voltages normal.
with this in mind, i reconnected the PSU and booted the computer (working fine right now, though no gaming, and using the mobo GPU). i then ran prime95 small FFT to test the CPU. temps were really high (~100) so i reduced the OC to a pretty light amount. temps dropped to ~72. i examined the PSU a few times during the prime95 test. on the first occasion, the fan was spinning, so i thought--okay, it's not the PSU. the second time, though, i noticed the PSU fan was not spinning. this was with prime95 running, 100% load. it did seem to be trying to spin, though--a slight rocking or twitching motion, and when i tried a few times to give it a spin with a piece of paper, it started spinning on its own. a few minutes later, i looked, and it was stopped again.
so i wonder a few things:
- is this PSU faulty? if so, is it serviceable, or better off being replaced?
- does it seem likely that this kind of problem with the PSU was the cause of my crashes? if not, what else should i check?