[SOLVED] Thermal paste turning liquid and CPU running slow after a month or so

[junkhardware]

I repasted my Lenovo E480 for the first time earlier on this year using "Cooler Master Mastergel Pro V2 " paste.
I could immediately see a huge speed increase, which was unbelievable.
A couple of months later though, my PC started running slow again and I had to crack my laptop open again!
When removing the cooling system, I was surprised to find an oily residual atop the processor (surrounded by some left over paste).
I though I might have not applied enough paste the first time around and I therefore applied a new slightly thicker layer of paste (after thorough cleaning of course).
Not even a couple of months later, here I am again with my PC getting sluggish!

Am I doing something wrong?
If not, what could be problem? Could it be that the heatsink needs replacement? I am wondering if the heat may have damaged it. Maybe it is not anymore perfectly flat (although I cannot tell by eye), preventing the paste to be homogeneously applied onto the processor.
Is this specific paste the problem? Should I use something else instead?

EDIT:

 

[lvt]

That's weird, thermal paste should be hardened over the months, not liquidified.

 

[thx1138v2]

On the back of the package it says ".,.non-curing and non-electrical conducting compound..." I assume the non-curing part is what is preventing the paste from hardening. Try a different thermal paste: https://www.amazon.de/gp/product/B07TTPZ1D8/ref=ppx_yo_dt_b_asin_title_o03_s00?ie=UTF8&psc=1

Hover your mouse over the photo of the back of the package to read it.

 

[junkhardware]

Thanks a lot @thx1138v2, it is very much appreciated.
So, why are they recommending this paste for this sole purpose ?
Would you know what is the best paste on the market? I recently read about these pads you can stick atop the processor instead of using normal paste.
Not sure if they are good, but maybe I should give it a try.

 

[iPeekYou]

Thanks a lot @thx1138v2, it is very much appreciated.
So, why are they recommending this paste for this sole purpose ?
Would you know what is the best paste on the market? I recently read about these pads you can stick atop the processor instead of using normal paste.
Not sure if they are good, but maybe I should give it a try.

Laptops present a different challenge than that of overclocked desktop processors; they tend to have lower mounting pressure on the heatsinks, and laptops have direct die contact from chip to heatsink. The latter pertains mostly to the utmost importance of total paste coverage --easy to solve. The former, however, means you often need thicker thermal interface to deal with.

If you want to delve deep into it, look at the discussions over at notebookreviews. They're constantly testing and discussing thermal pastes specially on, naturally, laptops. The main gist from my own research is the following:

• Good: Thermalright TFX and TF8, IC Diamond, Mastergel Maker Nano (the one with diamonds in it --like ICD), KPx (especially older formula, it's thicker), Arctic Silver 5, Gelid GC Extreme, Thermal Grizzly Hydronaut. And of course, liquid metal.
• Meh/caveat emptor: Phobya Thermalgrease 3, Kryonaut (very good performance, bad longevity --too runny and degrades quick when exposed to 80 deg C), ZF-EX (TFX rebranded, but I got a very bad batch that's dry out of the tube)
• Bad: Noctua NT-H1, Arctic MX-4, basically the runny, easy-to-apply stuff. Bad for longevity basically, not so much for thermal performance.

Out of those mentioned, I recall AS5 and ICD is especially long-lasting according to users. I'd chalk it up due to their high viscosity more than anything.

I'm personally using Hydronaut for my laptop since TG is the only brand that's consistent in quality and availability in my region.
I have KPx on desktop but I do agree their new stuff is nowhere as thick as their old formula.
TF8 and TFX is also pretty good, my only gripe is that TFX seems to like some curing period and it's incredibly sticky afterwards. It is thick, though, especially after said curing period.
Gelid GCE is pretty good also but some users report inconsistent batch quality; I can attest to that since I have two syringes of GCE and one of them is crap, the other is as good as KPx/Kryo on desktop.

Thermal pads' main schtick is that they're easy to apply and re-use. Performance, as you expected, isn't as good as top-end thermal pastes but decent. I'd stick to pastes for laptops since I imagine a thicker paste is better for addressing mounting pressure compared to a thin layer of carbon. Plus I don't take apart my laptop to reapply paste often unlike on desktop where I often test pastes.

If you're talking about the sticky ones, like EKWB's pads or TG's Minus pads, they're nowhere as good as, say, Carbonaut or MX-4. They're for chipset/VRM/VRAM cooling where you need adhesive that works both as thermal interface material and works well for uneven surfaces. They do work wonders for solving poor mounting pressure, but you'd have to have a really poor mounting to see the need to resort to those.

 

[junkhardware]

Thanks a lot @iPeekYou, I am going to pick one of these pastes you are recommending.
Any experience with the "Cooler Master Mastergel Pro V2 " though?
I thought I had bought something good there, but apparently not so? I am very surprised it does not last longer.
Any idea why it only works for slightly over a month?

 

[Phaaze88]

Any idea why it only works for slightly over a month?

https://www.techpowerup.com/forums/...-or-is-it-nonsense-lets-discuss.283046/page-2
See the very first post, second paragraph.

Some pastes don't handle constantly high temperatures, which laptops are well known for.
I figure this Mastergel Pro V2 falls in the same category as Kryonaut.

 

[geofelt]

First of all, understand the purpose of paste.
It is used to fill in microscopic pits in the mating surfaces that can trap air.
Air is a bad conductor of heat.
Paste is better.
But paste is not as good as metal to metal contact.
Too much paste is bad and can act as an insulator.
A small rice sized drop in the center will spread under heat and pressure.
I suspect you used too much.

That said, I suspect your paste was defective and I would change it out.
I have used as5, mx2 and nh1 all to equally good effect.

 

[junkhardware]

@geofelt, it is possible that I applied too much, but I really only applied a thin layer.
Maybe I should only put a tiny bit in the middle of the processor instead of covering the lot with paste (maybe that is what you meant by too much paste).
Hard to tell how much is enough as I cannot find any good videos online and the only one I find does not do anything different than I did.

 

[geofelt]

@geofelt, it is possible that I applied too much, but I really only applied a thin layer.
Maybe I should only put a tiny bit in the middle of the processor instead of covering the lot with paste (maybe that is what you meant by too much paste).
Hard to tell how much is enough as I cannot find any good videos online and the only one I find does not do anything different than I did.

I think a thin layer is too much.
Yes, the tiny bit should work fine. It really is hard to use too little.
When you remove the cooler, you should see a ring of paste that does not quite reach the edges of the cpu die.
If it overflows, you have used too much.
When cleaning the surfaces, use alcohol(70 to 91%)
I use a lint free paper coffee filter.

 

[junkhardware]

@geofelt, thanks. It overflows indeed. I wish I could paste a picture here, but it does not seem to be possible to attach files.

 

[Phaaze88]

@geofelt, thanks. It overflows indeed. I wish I could paste a picture here, but it does not seem to be possible to attach files.

imgur.com
postimages.org
imgbb.com

 

[junkhardware]

@geofelt and @iPeekYou, I have added a picture in my description.
What do you think about the "sweaty" part in the middle of the processor?
I understand I put too much paste, but why is there nothing left in the middle a month after applying fresh paste?

 

[geofelt]

I am more familiar with desktop coolers than laptop chips and coolers.
It looks to me that the mating surfaces are not smooth and do not meet really well.
But, I would not worry much since the absence of paste in the center says that the parts there are close to metal to metal contact.
That is good.
If you look closely, you can see a few rivers and pits which have been filled in with paste.
That is exactly what paste is designed to do.
If you are inclined to repaste, buy a good brand name paste. I use as5 ,mx4, or nh1 which comes with noctua coolers.
Opinions vary but differences in cooling may amount to a difference of 2c. or so.

 

[junkhardware]

@geofelt, thanks for the feedback.
The "Cooler Master Mastergel Pro V2 " paste I used had excellent reviews though. Very strange!
FWIW, I will try a different paste and see if that's better

 

[InvalidError]

But, I would not worry much since the absence of paste in the center says that the parts there are close to metal to metal contact.
That is good.

Unless the reason there is almost nothing left is because the thermal paste was too runny and ended up pumping all of the particles out of the interface such that only oil got left behind. This "gel" probably has too low viscosity to hold its particles in place.

 

[junkhardware]

@InvalidError, so are you saying this brand is probably bad after all?
Could the heatsink be too tight or damaged/curved due to the heat?

 

[InvalidError]

@InvalidError, so are you saying this brand is probably bad after all?
Could the heatsink be too tight or damaged/curved due to the heat?

Not necessarily bad, just apparently not appropriate for your specific application. Depending on how long the paste may have been sitting around, the reason it may have come out runnier than it is probably supposed to could be that it sat around long enough for a good part of solids to separate and not come out evenly, so you ended up getting more oil than anything else.

You would need temperatures well in excess of 100C to damage a heatsink and the CPU will trigger a thermal shutdown long before the heatsink gets anywhere near that hot, so heat damage to the heatsink shouldn't be a thing except for its plastic components if applicable.

 

[geofelt]

Oily separation smacks of old stock that has separated.
Good paste is cheap enough, and I agree that a new paste would be a good thing to try.

Too tight could be a problem if it warped the cooler or the base.
Uneven tightening could cause that.