Increasing mounting pressure of a laptop's heatsink. What do you think?

[ignafiltro]

I wanted to show this modification I just made to my laptop's heatsink (the reason is explained at the end).

When I was changing the thermal paste in my laptop, I noticed the cooler had this springs to make pressure against the die, but they were pretty easy to press, reducing the mounting pressure. To test if the mounting pressure made a difference I tried pressing on the cooler while the running a stress test, it did make a difference, dropping from 60°C to around 55°C on the GPU for example.

I wanted to increase the mounting pressure so i though of this:

I removed some specific screws (I would had removed all if I had enough washers).

And added a washer.

I hope it doesn't bend too much...

I will post how it works out.

Why am I doing this?
I am doing this pimp my laptop project (just for fun) with my 2010 Samsung R430 and one of the modifications is to change the CPU (Core 2 duo T6600 2.2ghz) for something with a higher clock speed.
I already did the research and my laptop chipset is ready to accommodate even a Core 2 Extreme QX9300 quadcore... ...If I'm able to cool it.

So I only did temp testing before the heat sink modification (I was too lazy to assemble the laptop again) with the original hardware and new thermal paste (Arctic Silver 5). The computer barely reached 55°C on the CPU and 60°C on the GPU running Prime95 small FFTs and Furmark at the same time.
(CPU undervolted and graphics card (GT 310M) overclocked).

With such a good temps I got exited and I am thinking on getting the QX9300. What do you think? Should I get a dual core chip instead? I should be able to game with and external gpu if it works.

 

[Ralston18]

When experimenting the cardinal rule is to vary only one thing at a time.

And take multiple readings with respect to each "varience".

Overall your setup/system may be within the acceptable specs. However (if some test happens to go outside the acceptable specs) then the endgame could be costly.

Every system is different - even "identical" systems.

Just a matter of the risks and expense you are willing to take.

 

[ignafiltro]

When experimenting the cardinal rule is to vary only one thing at a time.

And take multiple readings with respect to each "varience".

Overall your setup/system may be within the acceptable specs. However (if some test happens to go outside the acceptable specs) then the endgame could be costly.

Every system is different - even "identical" systems.

Just a matter of the risks and expense you are willing to take.

Well this time I was really lazy while testing and I didn't take as many reading as usual. I will take your advice and now test my temps in exactly the same way, the only change being the mounting pressure. Then i will test the difference between GPU overclocked and not overclocked, then CPU undervolted and no undervolt.

Also any recommendations for thermal paste? Ive seen many, but all comparisons are from before 2014.

 

[Ralston18]

Good plan.

As for thermal paste I will defer to other's who may regularly work with the various products. Barring obviously sub-standard products I think the matter is more the application process than anything else. (I remember a recent posting where someone was using peanut butter. That was around Apriil 1st maybe.....)

Plus manufacturer's these days have a tendency to quietly change a product and all of a sudden something good is now bad.

Someone who overclocks may be able to recommend and support a specific brand/product.

 

[ignafiltro]

Dammit, charger died, the wire failed. I am soldering the connector back right now, but I have to wake up early tomorrow dammit.