[SOLVED] My CPU keeps getting stuck to my fan repeatedly, why?

brimah87

Distinguished
Jul 25, 2016
27
0
18,540
I’ve got AMD Phenom II, an AVC fan (AV-Z7UH40Q001-2215), and MSI 890FXA-GD70 motherboard. Using MX-4 Arctic thermal paste. I’m aware these are old. Other components are a Sandisk SSD, XFX R series GPU, and Redline Mushkin 2x4GB RAM. Sorry this post is so long.

A few months ago my fan stopped working and my CPU kept overheating. I tried to take it out and ran into that problem where the fan and CPU get stuck and I ended up ripping them out without releasing the lever. The paste had hardened, probably because my build is almost 10 years old. I got the pins straightened out and bought a new fan and new thermal paste, and cleaned everything off with alcohol. The system has been running fine for a few months. Today the PC just shut down (no error code on the motherboard LEDs, they’re blank/off, and no display on the monitor), I tried to inspect the CPU and found that it was stuck to the fan again, and I again couldn’t remove it without ripping the whole thing out and I bent some pins again. I don’t think the CPU was overheating this time, but it’s possible. This time the paste is still wet. I’m afraid that even if I get it running again, and even if I buy new components, the CPU and fan will get stuck. What’s making this happen, too much thermal paste? The fan mounting is too tight? Humidity? I don’t want to try to put it back together or buy newer components if they’re just going to break again in a few months. I had assumed that the paste hardening due to age was the issue, but clearly it isn’t because I used new paste this time and it was still wet when I opened up the case.

Also I guess I might as well ask: if the paste was still wet and the CPU (probably) wasn’t overheating, does it actually matter that the fan and CPU were stuck together while mounted in the motherboard? Did my PC shut down because of this, or was it more likely due to the processor being so old, or lingering damage to the pins or mounting bracket from the first time I ripped out the fan/CPU? Thanks in advance for the help!
 
Last edited:
Solution
Your past hardened paste woes are likely unrelated to current 'no post' issues.

Disconnect all drives SATA and PWR connectors, any unneeded RAID and sound cards, etc., and drop to just one RAM module in required slot (you can test each alternately if needed). Remove power from PSU, and remove battery from mainboard for 10 minutes, press/hold pwr switch for 20-30 seconds. Reinstall battery. Try powering back up with just mainboard/CPU, one RAM stick, and GPU. If still no luck, you can test your GPU elsewhere, or try another one.

Ultimately, with your rig hard powering off as it did, the mainboard or PSU are certainly suspects, and either is more likely than the CPU having a spontaneous internal catastrophe. If you have another...
It's common for the cooler to stick to the cpu. The easiest way Ive always found which I read on here about is firstly run a benchmark that warms up the cpu and paste then power off, then while they are both attached to the motherboard, slowly but firmly 'twist' the cooler off, it can be pretty stubborn but ive never failed to detach one this way with no bent pins etc.
 
Hey thanks for the reply, however that wasn’t really what I was asking. I know a couple ways to get the fan unstuck, I’m wondering why it happens so frequently to me and if it could be causing my PC to not start.
 
Your past hardened paste woes are likely unrelated to current 'no post' issues.

Disconnect all drives SATA and PWR connectors, any unneeded RAID and sound cards, etc., and drop to just one RAM module in required slot (you can test each alternately if needed). Remove power from PSU, and remove battery from mainboard for 10 minutes, press/hold pwr switch for 20-30 seconds. Reinstall battery. Try powering back up with just mainboard/CPU, one RAM stick, and GPU. If still no luck, you can test your GPU elsewhere, or try another one.

Ultimately, with your rig hard powering off as it did, the mainboard or PSU are certainly suspects, and either is more likely than the CPU having a spontaneous internal catastrophe. If you have another adequate PSU, or access to one, try a new one slaved in. (NOTE: do NOT ever mix modular cables from one PSU with another, as the pinouts at the PSU insertion points often differ!)
 
  • Like
Reactions: kurdtnz
Solution