Anyone ever rip the CPU out of the socket while removing the heatsink?

Dayaks

[H]F Junkie
Joined
Feb 22, 2012
Messages
9,771
2700x still works. Put it under water. Maxes out at 55C now and seems to hold 4.2Ghz + no problem during TimeSpy Extreme with “stock” boosts.
Bent a pin.
It was “typical” non-conductive/non-capacitive thermal paste.

A plus is AMD’s resilent pins. Bend a pin on Intel’s socket and it’s rough.

333BE1AF-D8A0-4056-BF23-3F2B38C6FDC2.jpeg
 
2700x still works. Put it under water. Maxes out at 55C now and seems to hold 4.2Ghz + no problem during TimeSpy Extreme with “stock” boosts.
Bent a pin.
It was “typical” non-conductive/non-capacitive thermal paste.

A plus is AMD’s resilent pins. Bend a pin on Intel’s socket and it’s rough.
Just pop a mechanical pencil over that pin and straighten it up. I still pull out amd cpus from the socket if the heatsink has been on a long time and I didn't twist enough to loosen up the thermal paste, it's no problem.
 
Just pop a mechanical pencil over that pin and straighten it up. I still pull out amd cpus from the socket if the heatsink has been on a long time and I didn't twist enough to loosen up the thermal paste, it's no problem.

Yeah it still works. Will try to gently twist off next time! I used a blade to get it off the heatsink. It was on their really well to say the least.
 
Unfortunately happens fairly often on the cooler test rig depending on the shape and finish of the baseplate of the cooler. No ill effects or bent pins in the times that it has happened (knock on a thing), but you will never get used to that velcro ripping sound when it happens.
 
I've mentioned this happening on am4 multiple times. Trick is to twist the heatsink before lifting it to break the bond. Sometimes you even need to apply downward pressure so the weight of the heatsink doesn't pull the cpu out. It seems to happen on am4 more than on 940/754/939/am2/am3, but I could be imagining that.
 
I always run Prime95 before taking the HS off and twist it a little first. Worked well so far.
 
I always run Prime95 before taking the HS off and twist it a little first. Worked well so far.

I honestly thought about sticking it in the oven to help get the processor off the heatsink easier after it was removed from the mobo. I managed to get the blade under it though.

Yeah, definitely doing something like Prime next time although I hope there isn’t one. I have a rule I like to break to not touch it once it’s running well.

I have this guy over it now instead of the stock cooler:

4D598DCA-E412-483E-8417-E777412A09A9.jpeg


Main reason is so I can heat exchange with the pool and saves on AC costs. Win win. But it is nice having the processor at 4.2-4.3Ghz instead of 3.9-4.0.
 
Last edited:
Yep. Had an Athlon 64 that was giving me grief, just wouldn't run stable*. Went to reseat it, carefully following AMD's directions on how to remove the heatsink, it ripped out of the socket anyway. Gave up on it, threw out the CPU and mobo, built a pretty-new-at-the-time LGA 775 system. Swore off AMD unless and until they went LGA, where the processor is locked down.

With the new non-sticky graphite-pad TIM (e.g.) it might be an option again, next Big Rebuild.

And yeah, I know Threadripper is LGA, but it's also totally unjustifiable for me.



*Except in BeOS! It crashed all the time in Windows but ran BeOS just fine!
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
every single 939/940/AM2/AM3 and AM4 build ive ever had to build/repair/upgrade.... always at 1 time or another. ripped out of the socket. Heart skips a beat or 2 but as long as your force is straight up and out, you shouldnt have anything to worry about. Ive never hurt a cpu.
 
Basically what Meeho said. Most of the time it is much easier to remove cooler/waterblock from a warm/hot CPU than one at room temp. I have a tool with tapered blade I sometime slip between cooler base and CPU IHS to break the bond/seal between surfaces, but haven't had to use it in many years. Just don't try and yank it off too fast. Even fresh TiM forms a seal holding the 2 flat surfaces that needs to be broken slowly or CPU can be pull out of it's socket.
 
Last edited:
I think with the wraith style coolers as you unscrew the 4 corners it literally raises the CPU out of the socket. Fortunately it lifts straight up, but still kind of sketchy.
 
My heatsink is center mounted on CPU and rotate a little even when the screws are fully tightened. Guess it saves me some hassle when removing the heatsink -- just twist it a little and make sure it's not "glued" to the CPU :ROFLMAO:
 
who hasn't?! next time, warm it up first by sitting in bios for a few minutes to loosen up the paste.
 
Back
Top