[SOLVED] Did I fry my CPU?

[mormoretti]

Hey all,

Today I started removing my ryzen 2600 stock cooler to replace it for a new/better one. When I unscrewed and disconnectrd the cooler and started pulling, it wouldn't come off. So applied a bit more force and it came off along with the processor stuck to it.

Did some research about how to remove a stuck processor from the heatsink and saw a website saying it was okay to soak the processor+heatsink in isopropyl alcohol. Did that, didn't work. Tried heating the heatsink with a hairblower, carefully so it wouldn't blow on the cpus capacitor, didn't work.

Then, I gave moving/twisting the cpu from the heatsink another try, and ended up bending one of the pins. My heart skipped a beat.

I carefully unbent the pin and gave up on trying to remove the cpu from the heatsink. I waited a few minutes for the alcohol to dry off before reconnecting. But when I put the cpu back in the motherboard, screwed and connected the stock cooler, my pc won't start. The components light up, the cpu stock cooler spins, but there's no image and the peripherals get no signal.

Was it completely stupid of me follow the instruction of soaking everything into alcohol? Or did I just not wait long enough for it to dry? I don't think it's the pins, cause the cpu fits right in the little wholes, no force needed.

Is there anything I could try to revive the cpu? Please help :/

 

[compprob237]

Is the EZ Debug LED for the CPU? If so, try reinstalling the CPU. Lift lever, remove CPU, reinstall CPU, close lever, latch lever in place, and then obviously re-install the cooler.
If the LED is for DRAM then your RAM is bad, something is wrong with the memory controller of the CPU, or something is wrong with the motherboard.
The other two are probably not coming on but if the VGA LED is lit then the GPU isn't seated properly (typically). The Boot LED just says there's no boot device.

I really wish I had access to your system since I could rule out the RAM and PSU. You really need to try those sticks individually in a known-working system.

 

[mormoretti]

Is the EZ Debug LED for the CPU? If so, try reinstalling the CPU. Lift lever, remove CPU, reinstall CPU, close lever, latch lever in place, and then obviously re-install the cooler.
If the LED is for DRAM then your RAM is bad, something is wrong with the memory controller of the CPU, or something is wrong with the motherboard.
The other two are probably not coming on but if the VGA LED is lit then the GPU isn't seated properly (typically). The Boot LED just says there's no boot device.

I really wish I had access to your system since I could rule out the RAM and PSU. You really need to try those sticks individually in a known-working system.

Yeah, sorry for the confusion. The Debug light now is just for the RAM. Is there any other test worth performing at this moment? I sent G-Skill an RMA request already.

 

[compprob237]

Yeah, sorry for the confusion. The Debug light now is just for the RAM. Is there any other test worth performing at this moment? I sent G-Skill an RMA request already.

Nope, that's pretty conclusive that the RAM's bad.
You mentioned that the PSU died. Well, I've had PSUs die and take RAM sticks with them so that may have been the cause.
Just remember when you're talking to G.Skill that you tell them that they just stopped working suddenly and after doing some testing and booting from individual sticks you've concluded that they must be bad. G.Skill's pretty good about their warranties (I RMA'd an 8 year old kit...) so it should be a pretty simple process. I sure hope they can help you from Brazil...

 

[InvalidError]

My guess is that the socket or board got damaged either when you ripped the CPU out of it or crushed it back in while the CPU was stuck to the HSF - the solder balls under the ZERO-insertion-force socket aren't meant to withstand the force needed to rip the CPU out of a locked socket. May have ripped apart some solder balls under the socket.

 

[mormoretti]

Nope, that's pretty conclusive that the RAM's bad.
You mentioned that the PSU died. Well, I've had PSUs die and take RAM sticks with them so that may have been the cause.
Just remember when you're talking to G.Skill that you tell them that they just stopped working suddenly and after doing some testing and booting from individual sticks you've concluded that they must be bad. G.Skill's pretty good about their warranties (I RMA'd an 8 year old kit...) so it should be a pretty simple process. I sure hope they can help you from Brazil...

Yeah, I said it just stopped working and it wouldn't boot unless I used a different ram module. I really hope they can help me.

Once again, mate, thank so much for the help. You've helped a complete stranger who was (still is, kinda) frustrated and clueless about what to do. I really wish someone helps you in a time of need like you've helped me.

Cheers!

 

[compprob237]

My guess is that the socket or board got damaged either when you ripped the CPU out of it or crushed it back in while the CPU was stuck to the HSF - the solder balls under the ZERO-insertion-force socket aren't meant to withstand the force needed to rip the CPU out of a locked socket. May have ripped apart some solder balls under the socket.

AMD really needs to come up with some form of CPU latching mechanism to hold the chip in the socket when it's latched. Even if it is something installed over the top of the socket that holds the corners down. They could also just move to LGA like Intel and be done with it.

 

[mormoretti]

My guess is that the socket or board got damaged either when you ripped the CPU out of it or crushed it back in while the CPU was stuck to the HSF - the solder balls under the ZERO-insertion-force socket aren't meant to withstand the force needed to rip the CPU out of a locked socket. May have ripped apart some solder balls under the socket.

Yeah, that makes sense. Would you say that caused some short circuit on the ram modules or something?

 

[mormoretti]

To be honest, I was really annoyed that the heatsink was stuck to the CPU in such a way that I simply could not undo. I mean, I've only had these parts for 6 months tops. I would understand if they were like 1~2 years old, but come on...

 

[InvalidError]

Yeah, that makes sense. Would you say that caused some short circuit on the ram modules or something?

My main suspicion is broken solder balls between the CPU socket and the motherboard.

As for why the stock HSF can be hard to remove, the pre-applied paste on many HSF has much higher viscosity (sometimes to the point of feeling like clay) than typical after-market paste. Trying to remove the HSF by pulling straight out forms a vacuum in the TIM and makes it very difficult if not impossible to pull off. The best way to remove the HSF in this case is to loosen the screws, turn the computer on for a few minuteswarm the CPU up which will soften the TIM, turn the PC off, remove the screws then give the HSF a little twist to break the paste bond without creating a vacuum.