Question Heatsink soldered to CPU on old Packard Bell all-in-one?

Dec 15, 2019
5
0
10
Hello,

I tried to dismantle my 10 y.o Packard Bell OneTwo L5710 in order to clean the fans and change the CPU's thermal paste.

However, I cannot for the life of me get the heatsink off the CPU, which looks like this View: https://imgur.com/a/G6fMASF


IIRC it's an Intel Pentium Dual Core E-5400, and the socket is a LGA775.

Even with 6 screws unscrewed, the heatsink is still tightly attached to the CPU. Moving the little metal bar doesn't do much besides pushing down on the motherboard, since AFAIK it's meant to open the CPU socket, whose upwards movement is prevented by that unmovable heatsink.

Is there something I'm missing, or are they soldered together? I'm not very hardware-savvy, but from my Google searches, I was under the impression it was rare for desktop computers to have their heatsink soldered.

Thanks for the help!
 
Dec 15, 2019
5
0
10
Thanks for the answer richie,

Just shooting in the dark here: would it make sense to try to heat up the thermal paste / thermal epoxy and then try to pry it off gently? For example by turning the PC on for a while (80-90°C while working last time I checked with HWMonitor), turning it off and quickly giving it a try before the CPU has a chance to cool down.
 
Is there anyway you can loosen the other end of the copper heat pipes, where there should be some fins there that dissipate the heat? If you can, then carefully twist the loosened heat sink mounting after heating it. It doesn't take a lot of twisting, just a small amount to break the air-tight seal the paste forms between heatsink and CPU heat spreader.

Except, it's entirely possible theres no heat spreader but sits directly on top of the CPU die. It also looks like something else is under there, maybe GPU. If you don't know for certain, I'd be leery of prying it up.
 
Dec 15, 2019
5
0
10
The other end of the copper pipes (bottom-left of the picture) is loose, it was only maintained by a metal frame I had already taken off. The GPU's not there, it was on the right, out of the picture, there's a hole in the motherboard where its fan was. Are you referring to that smaller square at the top? I have no idea what's in there.

Unfortunately, I cannot try that heating technique anymore, because the screen doesn't even turn on now. Since I think I reconnected everything correctly, I'm afraid I may have made a "boo-boo" somewhere, probably when I pushed the CPU socket's metal handle a little too much.

Would I still be able to get to the BIOS with a damaged/dysfunctional CPU?
 
The other end of the copper pipes (bottom-left of the picture) is loose, it was only maintained by a metal frame I had already taken off. The GPU's not there, it was on the right, out of the picture, there's a hole in the motherboard where its fan was. Are you referring to that smaller square at the top? I have no idea what's in there.

Unfortunately, I cannot try that heating technique anymore, because the screen doesn't even turn on now. Since I think I reconnected everything correctly, I'm afraid I may have made a "boo-boo" somewhere, probably when I pushed the CPU socket's metal handle a little too much.

Would I still be able to get to the BIOS with a damaged/dysfunctional CPU?

If my suspicion is correct about it being a 'direct-to-die' interface with the heatsink it's possible you chipped a corner of the die with uneven pressure. It's a VERY easy thing to do with GPU's (also direct-to-die) when remounting the heatsink after replacing thermal paste. If that's the case you'd not likely be able to POST at all (so no BIOS display either).

It could also be a cracked solder ball underneath the chip, also susceptible to damage from unequal pressure topside. Or just something else somewhere else.

Blowing on the heatsink with a hair drier for a few minutes will also heat it up; then twist it slightly to dislodge it.

I was indeed thinking the other square was the GPU. It could be the another chip, possibly PCH or something.
 
Dec 15, 2019
5
0
10
Alright, so I have good news, and bad news (for me, at least).

The good news is that I managed to dislodge the heatsink by blowing on it with a hair drier (as was instructed), turns out the thermal paste was all dried up and acting as some kind of glue. Fortunately there was an IHS so I'm a bit less worried I might have damaged the CPU. It's now all cleaned up and got new thermal paste, delightful.

Now the bad news: still no display at all. Except this time, it's even worse, since the computer turns itself off roughly 10 seconds after I turn it on, while previously it didn't. I guess it means my components do not pass the POST? Is there a way I could somehow diagnose where exactly something goes wrong at start-up?

EDIT: I'm currently trying to follow the advice from those two links, I'll update once I'm done:

Computer powers on and then immediately turns off

POST troubleshooting steps
 
Last edited:
Dec 15, 2019
5
0
10
Alright, update time.

Recapitulative timeline:

  • Before I dismantled my computer, everything worked fine.
  • I dismantled it, couldn't pry the heatsink off the CPU, forced a little in the process, but to no avail. did change the GPU paste without trouble.
  • I tried to restart it again, but while the computer did turn on and kept running, all I got was a black screen, no beep, no "Packard Bell" screen, nothing.
  • I finally managed to pry the heatsink off the CPU by heating it with a hairdryer. I changed the CPU's thermal paste as well as the paste from that little square above it. I'm not sure if this is important, but the CPU's paste was grey and dry, while that little's square paste was blue-ish and rather spongy. Put some Arctic Silver on both of them.
  • I tried restarting the computer again, after having made sure no cables were loose. This time, still a black screen, except that now, the computer turns itself off roughly 5 seconds after being powered. The PSU fan works well, both the CPU and GPU's fans briefly start moving at start-up but then quickly stop.
  • I completely removed the CPU from the motherboard, and when I restarted the computer, this time it didn't turn itself off, but stayed on the black screen, as it used to do in a previous step.
Am I correct in my assumption that I must have somehow damaged the CPU, its socket and/or the motherboard when I tried to forcefully take the heatsink off? I tried cleaning the CPU's thermal paste and re-applying it a second time to make sure it doesn't overheat and triggers an automatic shutdown, but same results.
 

TRENDING THREADS