PC crashes when gaming when using a custom fan profile vs standard.

Hi everyone,

A week ago I upgraded 3 parts of my PC: CPU, MOBO and PSU. The rest of my parts GPU, RAM and storage are all carryovers from the previous build. Since upgrading I have run into intermittent black screens/hard freezes/bsod’s/reboots when playing games. Some of which would not even create a dump log, I have done by own initial troubleshooting. Making sure everything is up to date, BIOS, drivers etc with no luck. Replugging RAM, GPU, reseating cables still didn't have any effect. The one thing I noticed how I can pretty much guarantee to get it to crash is if I have the fan speed low in the BIOS it’ll crash when playing a game. I know what you are thinking and I did as well. It must be a temp issue as I moved to a SFF case. Many components are close together now so it would be reasonable that something was overheating. Sadly that isn't the case. I have used MSI afterburner and HW Temp and neither the CPU reaches higher than 83c and the GPU 70c. Testing in Forza Horizon 5 if I have the CPU and case fans at the “standard” mode (I have forgotten what it's called in the bios) my PC will not crash at least within 30mins - 1hr. Where as if I put my own fan curve dropping the speeds but still keeping the temps in check it'll crash within 10-15mins. This suggests to me something is overheating but im not seeing it in any monitoring software. I have done a fresh install of windows to the same problem. I do have a spare PSU which I will be trying at some point soon just to rule out the new PSU but other than that i'm bery much stuck as to what I can do next.

Honestly most likely the fan profile thing is just a coincidence that it'll crash when lowering it but still something I find abit strange.

The only settings I have changed other than the fan profiles are my CPU which I put in ECO mode and GPU I have decreased its power limit to 85%. The PC has crashed with both of these being stock so I'm pretty sure these are the culprits. Same with the ram XMP profile enabled/disabled makes no difference.

Thank you so much for reading my post. Any help would be massively appreciated.

OS Build: 19044.1387

Link to logs: EUANPC-(2021-12-07_20-44-42).zip - Google Drive

Hello, in my experience, there are two things that make a PC stop without reporting errors.
1. Temperature - which you seem to have ruled out.
2. Lack of power - which perhaps has not.
3. Edit - or a short, done that

Is you PSU providing enough juice?

Hi,

I believe so. It’s a SFX 600 platinum and I’m only running a 5600x and a 2060 super. So I shouldn’t be anywhere near a power limit problem. On power as I mentioned I’m going to try a known good PSU tomorrow using the exact same settings BIOS profiles etc as I’m using now to see if it crashes still. I hope it doesnt as this would be a easy fix compare to the alternative that it’s a CPU/MOBO problem.

As for temps I even got a infrared temp gun that I use on the BBQ just to measure surface temps so that I knew the software wasn’t glitchy 😅.

Edit: as for a short. I wouldn’t even know where to begin to search of one.

Euan385 said:

Hi,

I believe so. It’s a SFX 600 platinum and I’m only running a 5600x and a 2060 super. So I shouldn’t be anywhere near a power limit problem. On power as I mentioned I’m going to try a known good PSU tomorrow using the exact same settings BIOS profiles etc as I’m using now to see if it crashes still. I hope it doesnt as this would be a easy fix compare to the alternative that it’s a CPU/MOBO problem.

As for temps I even got a infrared temp gun that I use on the BBQ just to measure surface temps so that I knew the software wasn’t glitchy .

Edit: as for a short. I wouldn’t even know where to begin to search of one.

Check power supply next. Power supplies, just like everything else can have defects off the line. One machine off 1mm can mess up a whole batch. not enough to physically see, but enough to have strange behavior. The way its behaving, if your old power supply still works and can run this setup, use it to test. Otherwise, nicely explain whats happening and ask for an exchange on the power supply.

Easiest way to test for shorts if you want to is to pull all the parts out of the case. Assemble all parts on your desk (as long as its a wood desk). If not wood, place something insulating on top of the desk. Yes it will all run fine. To power it on, look in the manual for where the power switch hooks up. use a flat blade screwdriver or a pair of tweezers, but short (touch the metal) across those 2 pins and it will power on. Be careful with liquids and everything else. It will run fine like this but is very fragile and exposed. I ran mine like this for a week while waiting on amazon and my replacement case (see here to see it). If it runs fine with your fan curve like this, then a short is likely. Have to figure out what metal is touching where in the case.