Clevo P150 and P170 cooling systems neglected by Clevo! (solutions here)

[fantomasz]

how about to make 4 pieces and aply pressure on each corner
or buy old used heatsing,cut springs and use them as aditional spring for each corner

[fenryr423]

applying too much pressure to an uncapped GPU is going to trash your card...

[fantomasz]

possible to buy fan cable spliter for our laptop?

[Beefsticks]

I didn't try any of the other corners though.

Not to be disrespectful or condescending, but I must give you a /rolleyes


I repeatedly indicate in my tutorial that this is what happened on my notebook... I don't think you fully read what I wrote.

[Hurricane9]

No, I didn't really read it

[gcrain]

You won't distinguish the nuances in cooling performance without precise numbers on your Fan RPM and ambient temperatures

With all the threads on this subject this point can't be emphasized enough.

[Beefsticks]

So, I finally got my IC Diamond tube in the mail! Once I completely cleaned off the heatsink and GPU die of TIM, I noticed something: The poorly installed bent heat pipes of my heatsink assembly were holding my GPU heatsink up like a teeter-tauter over the GPU die. This is why I've had so many inconsistencies with cooling! One corner rested 1-2mm higher than the opposite and it required a lot of pressure to bend the heatsink to the GPU die. So, to fix this, I gradually bent the copper pipe at the right spot to where it could rest on the GPU die evenly without TIM, and I can say much less pressure is needed now for my heatsink to make even contact with the GPU die. Plus, the gap which required aluminum tape to seal is no longer there. So, after bending the pipes (not the heatsink tabs) I applied the ICD TIM after dipping the syringe in recently boiled water for 60 seconds:

New Test:

Kombustor @ 7 minutes with lid off: 78C

Kombustor @ 10 minutes with lid on: 88/89C

That 10-12C climb used to happen in less than 1 minute at a consistent fan RPM, and now it takes 3 minutes. Win! This is significantly longer to reach max temps at the same fan RPMs as my previous tests now, so I know its working better overall. Plus, this is even before the new IC Diamond application has cured. It has only had 20 minutes of heat. I am confident this system is working as efficiently as it can short of modifying the vent port on the laptop cover.


Summary for my notebook:

It seems Clevo did a bad job with aligning these copper pipes and just sloppily threw them in. In my notebook, this is the reason the metal tabs aren't strong enough. The pipes were resisting the GPU heatsink in the wrong direction. Now all thats left is to mess with the fan port sooner or later.

You won't distiguish the nuances in cooling performance without precise numbers on your Fan RPM and ambient temperatures and whether they are attibuted to cooling components (fan, heatsink, paste) or an EC issue (Fan RPM vs temperature thresholds...)

With all the threads on this subject this point can't be emphasized enough.

This isn't a forum for TIM lab tests. I'm just giving an easy method to check your heatsink/pipe assemblies. Also, a -10C drop in under 1 second is certainly not a nuance; the fan RPM doesn't even have time to change. Its an explosive drop in temperature.

[aduy]

So, I finally got my IC Diamond tube in the mail! Once I completely cleaned off the heatsink and GPU die of TIM, I noticed something: The poorly installed bent heat pipes of my heatsink assembly were holding my GPU heatsink up like a teeter-tauter over the GPU die. This is why I've had so many inconsistencies with cooling! One corner rested 1-2mm higher than the opposite and it required a lot of pressure to bend the heatsink to the GPU die. So, to fix this, I gradually bent the copper pipe at the right spot to where it could rest on the GPU die evenly without TIM, and I can say much less pressure is needed now for my heatsink to make even contact with the GPU die. Plus, the gap which required aluminum tape to seal is no longer there. So, after bending the pipes (not the heatsink tabs) I applied the ICD TIM after dipping the syringe in recently boiled water for 60 seconds:

New Test:

Kombustor @ 7 minutes with lid off: 78C

Kombustor @ 10 minutes with lid on: 88/89C

That 10-12C climb used to happen in less than 1 minute at a consistent fan RPM, and now it takes 3 minutes. Win! This is significantly longer to reach max temps at the same fan RPMs as my previous tests now, so I know its working better overall. Plus, this is even before the new IC Diamond application has cured. It has only had 20 minutes of heat. I am confident this system is working as efficiently as it can short of modifying the vent port on the laptop cover.


Summary for my notebook:

It seems Clevo did a bad job with aligning these copper pipes and just sloppily threw them in. In my notebook, this is the reason the metal tabs aren't strong enough. The pipes were resisting the GPU heatsink in the wrong direction. Now all thats left is to mess with the fan port sooner or later.





This isn't a forum for TIM lab tests. I'm just giving an easy method to check your heatsink/pipe assemblies. Also, a -10C drop in under 1 second is certainly not a nuance; the fan RPM doesn't even have time to change. Its an explosive drop in temperature.

wow i never thought of heating the thermal paste up before i applied it, when i get my copper tape and repaste ill be sure to do that, also i hooked my fans to the gpu fan using a splitter i made.

[hackness]

wow i never thought of heating the thermal paste up before i applied it, when i get my copper tape and repaste ill be sure to do that, also i hooked my fans to the gpu fan using a splitter i made.

Does it increase/decrease the RPM as the main GPU fan increases/decreases? Also the GPU fan is 5V, 2 x 12V extra fans is gonna be alright?

[aduy]

Does it increase/decrease the RPM as the main GPU fan increases/decreases? Also the GPU fan is 5V, 2 x 12V extra fans is gonna be alright?

they are 5v fans and they go along with the gpu fan, and when i press fn+1 they max out, so there is no problem supplying power to them. works perfectly. idles at 32c in a 26c room, and it goes down to 28c when everything is maxed.