CPU stuck to the cooler? - Solved

[Norton]

what I meant is I thought the CPU sockets lever was designed to let it break free before the actual cpu Socket was was ripped from the board. So if for example what happened to the OP he wouldnt have to buy a new mobo because the whole cpu bed ripped lose from the mobo PCB, if such as this happened.

The lever is designed to hold the pins w/o damaging them (most of the time) so what happens is the cpu pulls out of the socket while the latch is still down if pulled up too hard.

The cpu pins may be damaged when this happens and should be checked but the socket is usually not damaged (depends on the socket type though)

Note that the same thing happened to @Chicken Patty with his Ryzen setup iirc

 

[HTC]

The lever is designed to hold the pins w/o damaging them (most of the time) so what happens is the cpu pulls out of the socket while the latch is still down if pulled up too hard.

The cpu pins may be damaged when this happens and should be checked but the socket is usually not damaged (depends on the socket type though)

Note that the same thing happened to @Chicken Patty with his Ryzen setup iirc

Really? Damn!

 

[Vya Domus]

Have any one tried using just water to clean off computer parts? Hearing if you let them dry then All should be good. Been wanting to try it but idk still

Problem is water can remain in nooks and crannies and not dry completely even after a while.

 

[hat]

It's happened to me before too. I recall using a butter knife to (gently) pry the CPU off the heatsink. It still worked afterwards.

@rk3066 I've done it many times. Usually with heatsinks and stuff, but I have used water to clean other components. Best thing to do is play it safe and make sure they're dry. As @Vya Domus says water can remain in nooks and crannies. Currently my method for this is to let any excess water run off, maybe give a little shake, and rinse whatever it was in 91% rubbing alcohol (which dries much faster than water). Then I use a hair dryer to try to dry it up. I've washed a motherboard and had it running again inside of an hour by doing this.

 

[Chicken Patty]

Yeah it did happen to me, and I spent about four hours trying to get the CPU back in the socket. Did you get things sorted out?

 

[ShurikN]

What surprised me was that the CPU came off from the socket while it was locked.

Happened to me once on an Athlon II 640 and a Gigabyte board. Nothing got damaged luckily.

 

[Bones]

Has happened to me plenty of times - I've done more damage getting the CPU separated from the cooler AFTER it was out than the CPU simply coming out of the socket with the cooler.
Some coolers I've noted will do that. Those I have that do it I tend to leave them in use with my daily driver machines.
The one on this machine is one that's notorious for doing it.

While the chip and cooler are still warm, giving the cooler a twist and sliding it towards the edge of the CPU lid will help but normally the retaining collar on the board won't let you go very far that way. As long as your pins are OK and you don't damage anything getting the two pieces separated after the CPU and cooler comes out it's fine.

 

[HTC]

Yeah it did happen to me, and I spent about four hours trying to get the CPU back in the socket. Did you get things sorted out?

Yes: it's sorted out. I edited the OP's title to reflect this!

It did scare the crap out of me, though: had never happened to me before! For a bit i thought i had screwed mobo + CPU.

 

[R00kie]

Had this happen to me numerous of times, with P4 CPU's and my FX8320, the FX was actually stuck so hard that it managed to puncture my finger and fly away a couple of feet.

 

[Outback Bronze]

Yep the good old Artic Silver with a P4 Northwood under water stuck like glue.

Soooooo many time's I had to realign the CPU pins.

 

[Shadowdust]

Same thing happened to me with one of my old Socket 754 setups. It was my first CPU with a heatspreader and the cooler suctioned to the CPU. I ripped it out and lost a few pins on the CPU. A lesson well learned! Now I just lightly wiggle until it begins to break free.

 

[eidairaman1]

Very common with 478 P4s, the 423s didn't do it...

 

[trparky]

One has to wonder if this is the reason why Intel went with the LGA setup. Intel's CPUs are locked down onto the motherboard by a sort of bracket when installed so technically this shouldn't happen in the Intel camp.

Unless you somehow manage to break the socket off of the motherboard but then you'd end up seriously breaking the motherboard.

 

[HTC]

This dude had the exact same thing happen to him (post # 716):

https://forums.anandtech.com/threads/threadripper-builders-thread.2512631/page-29#post-39056854

 

[Vario]

My friend had a Phenom II X6 1090T. Calls me up: "I Can't get this heatsink off bro what do I do?" I say : "I'm coming by don't do anything". By the time I get there hes managed to rip it out and tore and bent the pins on the 1090 and mangled the motherboard socket.

What always worked for me was to run prime 95 for 10 minutes then turn it off and unmount it before it cools down. Or if its a bricked system you use a hairdryer to heat it up and then twist it clockwise/counterclockwise while pushing down so it doesn't lift the cpu out of the socket.

 

[Space Lynx]

This thread has scared the crap out of me. I think I might use Arctic MX-4 in the future, since it seems to be rather... well I don't know, but in my experience with many pastes, temps are negligible as long as you use one of the popular ones, and MX-4 doesn't seem to have any issues as far as drying up a little or sticking.

 

[HTC]

This thread has scared the crap out of me. I think I might use Arctic MX-4 in the future, since it seems to be rather... well I don't know, but in my experience with many pastes, temps are negligible as long as you use one of the popular ones, and MX-4 doesn't seem to have any issues as far as drying up a little or sticking.

Note that the "original" stock paste was the one that came pre-applied with the stock cooler. After it's removal, i used Noctua's NH-T1 with the Noctua's cooler because i misplaced my MX3 (or is it 4?) paste: it was the only one i had available and it came with the cooler, meaning it's around 8 years old.

 

[Vario]

Doesn't matter what paste you have, this kind of thing will happen its not a big deal. In the future, run prime 95 get it real hot, then shut down and unbolt the cooler, then give it a twist back and forth a bit while pushing slightly down and it will come off as the paste bond shears off.

 

[Space Lynx]

Doesn't matter what paste you have, this kind of thing will happen its not a big deal. In the future, run prime 95 get it real hot, then shut down and unbolt the cooler, then give it a twist back and forth a bit while pushing slightly down and it will come off as the paste bond shears off.

I agree with you for most pastes, but if you have ever felt mx-2 or mx-4 paste, it really is... like more silky than other pastes, doesn't seem to solidify at all, its 1-2 degrees celsius warmer than Noctua paste i think, i personally use noctua paste, but if i go to ryzen i might switch over to mx-4, or if i switch over to any chip die that is bigger than intel lga 115x series, i think i will switch in general, just because the suction increases

 

[micropage7]

This thread has scared the crap out of me. I think I might use Arctic MX-4 in the future, since it seems to be rather... well I don't know, but in my experience with many pastes, temps are negligible as long as you use one of the popular ones, and MX-4 doesn't seem to have any issues as far as drying up a little or sticking.

Maybe it's the good reason for anyone, dont put too much paste there and make sure you spread it thin enough so it's safer when you remove the heatsink

 

[HTC]

Maybe it's the good reason for anyone, dont put too much paste there and make sure you spread it thin enough so it's safer when you remove the heatsink

The cooler that got stuck (Ryzen 1600 stock cooler) had the paste pre-applied.

I didn't change anything with the stock cooler until the time i tried to take it off: it was then that it got stuck. I had just turned down the PC prior to attempting this: the PC had been turned on for several hours but wasn't doing anything with high load intensity, so it wasn't "very hot", but not cold either.

 

[delshay]

This is how I lost my Opty 185 back in the day, when they were worth like $400.

You do realize the pins are soldered on. Look carefully at the pin base, just around it. Just solder on new pin(s). I did this without removing the base copper contact. Never tried removing the complete pin including the base, looks like to much work when I can just solder on a new pin on top of the old contact base.

 

[Skypup]

Ok...had same thing happen to old Prescott cpu during mobo recap...of course it was on motherboard for 18 years (GX270 tower)...this board was a real pita because of those stiff heatsink clips and the plastic tabs holding metal plate (Guess Dell was pretty confident their caps would last for decades)..