Discussion Arctic Silver 5 Repaste (appears burnt out) after 26 months (incl pics) Noctua NHD-15S - Thoughts?

Achaios

Honorable
BANNED
May 28, 2013
225
23
10,695
So today I changed my CPU fan & HS to a Noctua NHD-15S. To do that I have also had to repaste, naturally.

I will get to the end result shortly.

The paste appeared burnt out and crumbled. Dunno how that is even possible, this is not my first time and I have done this many times before.

What do you think, is this normal?

99BBL9V.jpg


So the end result is 10 to 13 degrees COOLER than before at full load at 4.5GHz at 1.29V (4770k).

(Previously I'd get 63 degs at full load OC'd to 4.5 GHz, now I am getting 50-53).

I am at a complete loss to explain the results. The only thing I did different than all the other times is that I used Arctic Clean Remover and Arctic Clean Purifier on the IHS. Previously, I just used 99% isopropyl alcohol.

Note that my CPU is NOT delidded.

My previous cooler was Thermaltake Frio Overclock King, a 250W TDP cooler with 2X120 mm fans configured as intake exhaust as well as a big HS, although smaller than Noctua's.

It is unimaginable that Noctua has beaten a top-of-the-line cooler from Thermaltake so completely. Thoughts?

I forgot to add that AS5 has got a 200 hour long BURN IN (or setting) time, so the way I understand this I am to expect up to 5 degrees cooler temps after 200 hours of operation.

I can't believe this. OoOoOo
 
Last edited:
  • Like
Reactions: Banned by Net
Your residual paste photo looks normal.
You might have mounted the previous pump assembly a bit unevenly since the residual seems to be a bit offset.

Think about it.
Your liquid cooler had essentially two 120mm radiators and the NH-D15s has two radiators of 140mm.
The 140mm radiators are going to be better at cooling.
The only difference between liquid and air is where the radiators are located.

And... if you had mounted the aio radiators to exhaust out the top of your case, they would have been using marmer air from inside the case.

I have not noticed any important difference after paste curing.
So far as I can tell, AS% and the supplied noctua NH-1 work equally well.

All in all, a good change.
 

Achaios

Honorable
BANNED
May 28, 2013
225
23
10,695
Your residual paste photo looks normal.
You might have mounted the previous pump assembly a bit unevenly since the residual seems to be a bit offset.

Think about it.
Your liquid cooler had essentially two 120mm radiators and the NH-D15s has two radiators of 140mm.
The 140mm radiators are going to be better at cooling.
The only difference between liquid and air is where the radiators are located.

And... if you had mounted the aio radiators to exhaust out the top of your case, they would have been using marmer air from inside the case.

I have not noticed any important difference after paste curing.
So far as I can tell, AS% and the supplied noctua NH-1 work equally well.

All in all, a good change.

It was an air cooler.

I went from Best-That-China-Can-Do Air Cooling to EU Air Cooling.

It's a culture shock. Kinda like going from Bollywood to Hollywood.

Have a look.

https://www.thermaltake.com/products-model.aspx?id=C_00001646

And no, the 15S only has ONE FAN. The Austrians beat China with just 1 fan while China used 2 fans.

https://noctua.at/en/nh-d15s

Taiwan#1
 

Achaios

Honorable
BANNED
May 28, 2013
225
23
10,695
Not enough thermal paste was used and operating temperatures were too high (too much OC).

Yeah, you are half-right I think.

I used (and still do use) the recommended AS5 application method for Haswells:

sRdTwAu.jpg


You will note there's none at the edges where the crumbling and some of the yellowish stuff was.

However, I didn't take a pic of the CPU fan HS contact surface, and apparently most of the paste somehow chose to stay on the CPU fan HS contact surface rather than the CPU IHS.

So it might seem like I did use too little paste except that I didn't.

My temps were just fine with the previous application too, just not oustanding as they are now.