Sign in with
Sign up | Sign in
Your question

Help, I reapplied thermal paste but it ended up making it worse!

Tags:
  • Heat
  • Heatsinks
  • Thermal Compound
  • CPUs
Last response: in CPUs
Share
August 5, 2014 10:31:53 PM

This is what my temps were at before:



When I took off my heat sink the thermal paste covered the whole CPU pretty much.

So I cleaned it off and reapplied it like this:

Pre-heat sink:


After heat-sink:


Idle:


Load:

Why did my clock speed drop dramatically here as well^?

To me the way I applied the thermal paste seems like the perfect amount, but why the hell did my idle and load temps increase by 20C? Please help

More about : reapplied thermal paste ended making worse

August 5, 2014 10:45:10 PM

Before it's asked. I wiped it off with Isopropyl and coffee filters, for a lint free cloth.
m
0
l
a b à CPUs
August 5, 2014 10:53:23 PM

Well You never mentioned what cooler you were using on the CPU.
So i will take it as a stock Intel cooler that came boxed with cpu.

First of all to do with the thermal compound it should be spread in a thin layer all the way across the heat dissipation pad of the cpu in a thin layer.

Second is, when you fit the cooler, if stock intel.
The right way to fit it so it sits even.

Is top left push pin, bottom right next, bottom left, top right.

The sequence is important as if done in the wrong order you create too much tension on the cpu clips that pass through the holes of the board and create a slight lift of the cooler in one corner.
The result is the 20 c of extra heat. and if you notice one core will me much hotter by a few Oc. Other than 2c or 3c between each core the cpu has.
m
0
l
Related resources
August 5, 2014 10:55:56 PM

uhhhh, idk if this is the reason, but you might want to do the spread method, so that the paste is fully covering the cpu, not like partial. maybe that's the reason why it got worse, also, make sure you seat the heatsink properly.

edit: and you might want to put a little more then what you had, didn't look like enough, that was kind of a small amount.
m
0
l
a b à CPUs
August 6, 2014 2:47:59 AM

Try again, too much thermal paste can act as an insulator and reduce performance.
Only a small drop is necessary. It will spread out under heat and pressure.
If you ever see it overflowing the cpu you have used too much.
m
0
l
a c 165 à CPUs
August 6, 2014 2:51:52 AM

Going by the picture, you didnt apply enough, the coverage on the heatspreader is poor.
m
0
l
August 6, 2014 2:54:54 AM

but it didn't spread all over the cpu, and what i mean by more is basically relative cause i said it didn't look like enough to spread on the cpu, but if it could be spread all over the cpu then by all means it was enough. that's all i meant. i'm not disagreeing that too much is bad either, don't get me wrong.

i always prefer the spread method, works everytime, when you use the drop method you never know if it's enough to spread over the cpu. and i never had a problem with it overflowing as i always use just enough. and yes, i know about too much. but again, i was being relative as i'm sure he could have figured that out.

@robcrezz yeah, i agree, that's the point i was trying to make, but someone was assuming i meant a lot of thermal paste is good. i totally agree with you. +1 to that robcrezz.

but hey, learning can be a painful experience. nuff said.
m
0
l
a c 165 à CPUs
August 6, 2014 2:58:09 AM

I still wouldnt recommend the spread method, much harder to get right. I would just do it again, using about 2x the amount of the first go.
m
0
l
August 6, 2014 3:03:04 AM

that's totally not true, i always get it right. now i admit it takes a bit effort, but if you can't get it right, your just flat out lazy.
m
0
l
a c 165 à CPUs
August 6, 2014 3:30:13 AM

alienworkshop said:
that's totally not true, i always get it right. now i admit it takes a bit effort, but if you can't get it right, your just flat out lazy.


Its not a case of being lazy, its just an inferior method and much harder to get as good results.
m
0
l
August 6, 2014 4:04:12 AM

i beg to differ, i seen several videos that stress the spread method is better, i never heard anyone say that the drop method is better.
m
0
l
a b à CPUs
August 6, 2014 4:11:59 AM

alienworkshop said:
i beg to differ, i seen several videos that stress the spread method is better, i never heard anyone say that the drop method is better.


Read this article it might help you with an answer

http://archive.benchmarkreviews.com/index.php?option=co...
m
0
l
a c 165 à CPUs
August 6, 2014 4:12:39 AM

alienworkshop said:
i beg to differ, i seen several videos that stress the spread method is better, i never heard anyone say that the drop method is better.


Allow me to elaborate.

The whole point of TIM is to simply fill the gaps between heatsink and CPU heatspreader.

1) The drop/pea method naturally spreads the TIM exactly where its needed (assuming the correct amount is applied).

2) The spread method is generally less effective because its not possible to know where more or less of the TIM is needed (heat spreaders and heatsinks are never 100% flat).

http://www.tomshardware.com/reviews/cooling-air-pressur...
m
0
l
August 6, 2014 5:46:10 AM

It's actually a Corsair h100i, but I will reapply with more.
m
0
l
August 6, 2014 8:54:09 AM

I don't know at this point. I've tried applying it so many different ways and nothing is changing the temperature.
m
0
l
a c 165 à CPUs
August 6, 2014 8:58:16 AM

Are you cleaning it off properly? Try using a new paste, that paste just looks wrong and too thick.
m
0
l
August 6, 2014 9:28:20 AM

Yh I give up. Nothing I do, cross method, line method, dot method, bigger dot method, spread method.. nothing works. Going to order new paste and try again later.
m
0
l
August 9, 2014 7:22:59 PM

Just an update, I got Arctic MX-4 and applied and it did nothing. For now I am not playing heavy load games, so it only reaches 65C max during load, which I can live with till I can somehow find the solution.
m
0
l
a c 165 à CPUs
August 10, 2014 12:27:33 PM

Episodic said:
Yh I give up. Nothing I do, cross method, line method, dot method, bigger dot method, spread method.. nothing works. Going to order new paste and try again later.


Episodic said:
Just an update, I got Arctic MX-4 and applied and it did nothing. For now I am not playing heavy load games, so it only reaches 65C max during load, which I can live with till I can somehow find the solution.


What coverage are you seeing now when you pull off the heatsink? 65'c isnt bad under load.
m
0
l
August 11, 2014 6:44:29 AM

RobCrezz said:
Episodic said:
Yh I give up. Nothing I do, cross method, line method, dot method, bigger dot method, spread method.. nothing works. Going to order new paste and try again later.


Episodic said:
Just an update, I got Arctic MX-4 and applied and it did nothing. For now I am not playing heavy load games, so it only reaches 65C max during load, which I can live with till I can somehow find the solution.


What coverage are you seeing now when you pull off the heatsink? 65'c isnt bad under load.


I haven't pulled it off yet, but I will when I get home. Let me refine what I mean by 'load'. Load is playing Smite on lowest settings, because I am using Intel 4000 graphics (waiting for RMA of GPU to come back). So yes, it's under load but not a heavy load. Under prime95 load it goes to 100C
m
0
l
a c 165 à CPUs
August 11, 2014 8:17:17 AM

Yeah it sounds like its either still not mounting correctly or faulty.
m
0
l
!