Release the heatsink latch anndddddd ....

Mehovoric · Aug 25, 2011

You can't pull the heatsink off because it's stuck to the CPU ...

(You have one of those CPUs where almost the entire chip is encased and you have to put thermal paste on that entire area. Like the AMD 6 core)

Now you can pull real hard and maybe have luck pulling the CPU right out of the locked socket WITHOUT busting any pins. But you don't necessarily want to do that.

So ... WHAT DO YOU DO?

redduc900 · Aug 25, 2011

Just gently twist the HSF back and forth, while pulling up on it slightly until it breaks free from the CPU.

RGone · Aug 25, 2011

redduc900 said:
Just gently twist the HSF back and forth, while pulling up on it slightly until it breaks free from the CPU.

He is for sure right about a twisting rather than just upward pull to release the HSF from the cpu.

I have had some success in heating bottom (toward bottom) of the heatsink with good hair dryer and then applying the twisting back and forth motion. Even the twisting back and forth motion may appear great before the HSF breaks loose. All I can say is care is better than haste.

onefstsnake · Aug 25, 2011

+1 on the twisting.

:thup:

mbentley · Aug 25, 2011

Twisting is most definitely the way to go, even if you can barely move it the tiniest bit in the bracket. I've also had some better luck with a system that has just been under a heavy load to provide some heat but some really old thermal paste will act like mortar no matter what and will require some decent twisting, that's for sure. It's never a good feeling to have pulled out a chip out from the socket when trying to detach the heatsink by straight pulling

Mehovoric · Aug 25, 2011

mbentley said:
Twisting is most definitely the way to go, even if you can barely move it the tiniest bit in the bracket.

I was just gonna say, what happens when the bracket doesn't allow the heatsink to move much if at all haha.
But yeah it seems like the last few times I've pulled a CPU out it stays stuck to the heatsink. I really freaked out when I went to pull out my new 6 core and it did that. Maybe now that its under a waterblock, I'll have a little bit better luck with it

But thanks guys! Any other methods this world should know about? This is the kind of thing that they don't teach in the A+ class just because they're too lazy or don't have a budget to be tearing computers apart all the time.

JKessell · Aug 25, 2011

I smiled when I saw the name of this thread, because when I went to remove my heatsink for the NH-D14 replacement, I experienced the exact same situation.

I sat there in wonder, pulling lightly at the HSF after unclipping it going, "Um... wtf...I'm pretty sure I unclipped it."

Little twisting finally did the job after much frustration.

Tech Tweaker · Aug 25, 2011

Mehovoric said:
I was just gonna say, what happens when the bracket doesn't allow the heatsink to move much if at all haha.

But thanks guys! Any other methods this world should know about? This is the kind of thing that they don't teach in the A+ class just because they're too lazy or don't have a budget to be tearing computers apart all the time.

I learned a little trick from working on socket 939 K8 platform motherboards (after pulling a CPU out of a locked socket once), that is to unscrew the mounting bracket from the backplate and just remove the bracket with the heatsink if you can't rotate the heatsink with the bracket screwed in place. Then there is nothing to prevent it from moving (unless it is glued down) and you can twist the heatsink all you need to in order to break it free.

Bisquit · Aug 26, 2011

Twist and shout, dude. Twist and shout.

On the bright side, your old contact between cpu and heatsink was done proper.

Mehovoric · Aug 26, 2011

Haha! never thought of it that way. If it doesn't come off then you did it right!

a c i d.f l y · Aug 30, 2011

Bringing the system online (go to BIOS) with the fan unplugged, sink unhinged and twisting the sink, should be enough heat to break the bond. It'll shutdown once it hits 80C or if you turn it off immediately after breaking the bond.

I've also tried placing the motherboard + cpu into a freezer -- that made it even easier to "break" the TIM from the CPU. Not recommended or the faint of heart.

Mehovoric · Aug 30, 2011

a c i d.f l y said:
Bringing the system online (go to BIOS) with the fan unplugged, sink unhinged and twisting the sink, should be enough heat to break the bond. It'll shutdown once it hits 80C or if you turn it off immediately after breaking the bond.

Isn't that inherently bad for the processor to let it reach that maximum hard shutdown temperature? Plus isn't it not a good idea to be working on powered system either? Even with a grounding strap you can still fry something.

RGone · Aug 30, 2011

People like "acid.fly" that have even used 'cold' to loosen the Heat Sink from the cpu, probably have the skill to do just exactly as he described about no fan and heating the cpu with Heat Sink loose to twist and remove.

That said, I used to unhook the fan to let temps rise while in bios where I could see how high the temps went and when it got close to temp shutdown, just shutdown manually. This set the heat sink compound immediately without having to wait the 200 hours for AS5 and the like. Now that said...I would just unhook the fan and let heat rise watching in bios and when it was just before cpu shutdown on its' own...heck I would turn off and twist and turn off.

Lots of ways to skin that old cat...old folks saying.

Tech Tweaker · Aug 31, 2011

RGone said:
That said, I used to unhook the fan to let temps rise while in bios where I could see how high the temps went and when it got close to temp shutdown, just shutdown manually. This set the heat sink compound immediately without having to wait the 200 hours for AS5 and the like. Now that said...I would just unhook the fan and let heat rise watching in bios and when it was just before cpu shutdown on its' own...heck I would turn off and twist and turn off.

Lots of ways to skin that old cat...old folks saying.

Hmm... Interesting idea, I may have to try that sometime, since I'm using AS5 in my machines at the moment (ran out of AC MX-2).

RGone · Sep 1, 2011

Tech Tweaker said:
Hmm... Interesting idea, I may have to try that sometime, since I'm using AS5 in my machines at the moment (ran out of AC MX-2).

It has been only in the last couple of years that we do not see compound setting times for whatever compound. I would guess that any compound has a setting/fully ready time for maximum effectiveness.

When I was doing a lot of benching and setting up 2 or 3 different cpus/boards in a week...well I did not have 20 hours, much less 200 hours to wait on heat sink compound to set. Once I was told about that pre-heating method...well the compound never did get better over time, it was good to go from pre-heat onward.

Release the heatsink latch anndddddd ....

Mehovoric

Member

redduc900

Inactive Moderator

RGone

Senior DFI Staff

onefstsnake

Member

mbentley

Gloriously Lead, Overclockix Chief Architect

Mehovoric

Member

JKessell

Member

Tech Tweaker

Member

Bisquit

Member

Mehovoric

Member

a c i d.f l y

Member

Mehovoric

Member

RGone

Senior DFI Staff

Tech Tweaker

Member

RGone

Senior DFI Staff

Similar threads