My laptop has leaf springs!

"ItalianJobR53 - now with added 'MERICA and unreliability" (italianjobr53) 08/10/2018 at 21:25 • Filed to: Laptoplopnik

1

17

Kinda sorta

See that thin metal bar with screws on both the ends

After I torqued down them screws:

The thin bar thingy flexes and acts like a spring (I presume to maintain contact between the CPU and the heat sink).

TheTurbochargedSquirrel > ItalianJobR53 - now with added 'MERICA and unreliability 08/10/2018 at 21:39

1

Yup, Intel CPUs are designed to have about 70-100 lbs of force applied to them by the heat sink (or other way of holding the CPU on the board/in the socket) . Springs are a common way of ensuring approximately the right amount of force is applied when the screws are tightened all the way.

Nom De Plume > ItalianJobR53 - now with added 'MERICA and unreliability 08/10/2018 at 22:07

1

Ha, I kinda sorta recently went overboard doing this to a ton of unused computers I had laying around. What did you use for thermal paste?  

LOREM IPSUM > ItalianJobR53 - now with added 'MERICA and unreliability 08/10/2018 at 22:10

1

What, no upgrade to coilovers?

ItalianJobR53 - now with added 'MERICA and unreliability > Nom De Plume 08/10/2018 at 22:31

1

Arctic MX4

I had IC diamond before this and it left scrathes/smudges on the chips :((((. Like this:

ItalianJobR53 - now with added 'MERICA and unreliability > TheTurbochargedSquirrel 08/10/2018 at 22:35

0

I don't think my springs will excert that much force.

Nom De Plume > ItalianJobR53 - now with added 'MERICA and unreliability 08/10/2018 at 22:53

0

Those smudges are exactly what you want. It means that by getting into those microscopic crevices it displaced any air that might have. 91% Isopropyl alcohol and a lint free cloth will remove this final layer so your reapplication of thermal paste is completely fresh. Instead of leaving a barrier of spent TIM between it and the processor.

Nom De Plume > ItalianJobR53 - now with added 'MERICA and unreliability 08/10/2018 at 22:57

0

On second glance those are amazingly clean other than the smudges. Every single laptop I pulled apart had almost no remaining thermal paste where it was needed and a considerably large pool of dried toothpaste all around it. Ended up using 3/16" thick plastic to chisel it away finally.

ItalianJobR53 - now with added 'MERICA and unreliability > Nom De Plume 08/10/2018 at 23:31

1

I used 99% isopropyl and a lot of tissues, that didn't remove them. So thats why I'm worried. Should I be?

ItalianJobR53 - now with added 'MERICA and unreliability > Nom De Plume 08/10/2018 at 23:33

1

Those photos are after I removed the old pasta. I had a layer of paste when i disassembled it.

Nom De Plume > ItalianJobR53 - now with added 'MERICA and unreliability 08/10/2018 at 23:56

1

Not sure if it reaches the level of worried. I would have put a few drops of alcohol on for 5-10 minutes and got on with it if they didn’t come up with light scrubbing. Maybe ran a plastic scraping tool over it. Once it cures some of this stuff can be pretty resistant to being removed. So it might run a degree warmer. Big deal.  

Tissues instead of a lens cloth might have been part of it.  You can actually rub or press into the surface you are preparing without it falling apart.  

ItalianJobR53 - now with added 'MERICA and unreliability > Nom De Plume 08/10/2018 at 23:59

0

Aah, I’ll remember that for next time.

For now, my CPU temps  have dropped by 13C (84 to 71 ) under now load (me gaming) so I'm happy.

Nom De Plume > ItalianJobR53 - now with added 'MERICA and unreliability 08/11/2018 at 00:07

1

Yep, high temps + long hours of use will do exactly what I saw written on there.  

ItalianJobR53 - now with added 'MERICA and unreliability > Nom De Plume 08/11/2018 at 00:31

0

I played CSGO 700-800 hrs in the last year,  400 hrs of other stuff and some FEM , so yeah, long hours and fairly gpu and cpu intensive :)

ItalianJobR53 - now with added 'MERICA and unreliability > LOREM IPSUM 08/11/2018 at 00:33

1

Hey, I’m in the land of  the free where the newest sportscar still uses transverse leafsprings.

facw > Nom De Plume 08/11/2018 at 14:40

0

Having little thermal paste remaining on the processor is not necessarily a bad thing. It’s meant to fill in the extremely small gaps on the mating surfaces, it doesn’t need to be thick at all. The stuff around the edges is probably just excess that got pushed out by the force of the clamp.

That said, I had work machine where we had to switch the heat sink due to a fan failure, and discovered that the builder had used no paste/TIM at all, which probably was not optimal (especially on something as hot as a dual-core P4).

Nom De Plume > facw 08/11/2018 at 15:27

0

Tinting, if we were discussing a fresh application using that technique it is a very good thing. Failing to replace a spent layer of thermal paste at the very highest point of heat transfer is a bad thing. You aren’t going to trouble of replacing it because it’s still working.

Were you aware thermal paste isn’t actually needed if the processor and heatsink mate properly? This is actually the ideal situation. The tolerances found in aftermarket factory made heatsink relaxed those that were in place on processors. Eventually they strayed far enough in fitment TIM became a necessity for establishing contact between the two plates. At that point consumers designs focused on lower quality material and cheaper construction we take for granted nowadays. V ery thin top plates on current Intel chips is a good example of this. You do not want to mount an older cooler that fits with considerably higher pressure anywhere near as tight as with older processors.

facw > Nom De Plume 08/11/2018 at 16:23

1

True enough. Despite the de-lidding craze for recent Intel desktop chips, I’m happy to have the heat spreader there. It probably does decrease heat transfer efficiency, but I’m old enough remember the original socket-A A thlons, which combined a very thin protective coat on the die, a slightly thicker package resulting in higher clamp tension, and a clamp design that resulted in sideways as well as downward pressure on the processor. I ended up chipping my die, but not enough that it ceased functioning, though I have to imagine it was close. I’ll take the current arrangement any day (though Intel can go fuck itself with its pushpins).