• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

CPU Heatsink Cleaning, your thoughts on this (long)

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Maximus Nickus

Senior Shaman
Joined
Oct 4, 2001
Location
Milton Keynes, U.K.
For a long time we have used Isopropyl Alcohol to clean our Heatsinks, thinking it evaporates and leaves no real traces or a film behind.

I thought this to be Sincere until recently an email from an ex Military Technician Mr. Jeff Leckemby regarding this, here is what he had to say:

Originally an Email...

-----Original Message-----
From: JM Leckemby
Sent: 15 September 2002 15:58
To: [email protected]
Subject: "How to Lap a Heat sink"

Maximus Nickus,

I read "subject" with great interest. Although I am not an over-clocker - my 2 GHz is just fine - your article was intriguing. Being an "old-school" computer technician with much experience on Main Frame systems I have never heard of lapping heat sinks. I have no information or statistic to back my following comments. Just my opinion - and yes, opinions are like "a**holes", everybody has one... here's mine.

First: "After you have finished, you may wish to clean the heat sink's surface with some Isopropyl Alcohol." You mention in your article that there Isopropyl should be used because you don't want anything depleting the object of the effort. In that case use Denatured Alcohol. Isopropyl leaves a film behind. Computer techs for as long as I can remember have used Denatured Alcohol, and Isopropyl only in a pinch and the other is not available. Denatured Alcohol is just as common, and can be purchased in your local Home Improvement center.

Second: It seems to me that Artic Silver is a better solution for heat transfer than older products such as Silicone Heat Sink compounds. Also, the lapping technique. This removes all the little ridges on the heat sink. These little ridges provide more surface area don't they? It's similar West Virginia, a very mountainous state. If all its mountains were flattened out it would be larger than Texas, but if we just bull-dozed the mountains away, it would be only as large as its state-to-state boundaries. Goes for the CPU too. I would leave the little ridges behind for the surface area they provide. Have you tested the Artic Silver compound without lapping?

Just my thoughts.
Jeff Leckemby


-----2nd Message-----

Thank you for your thoughts, they are most appreciated, and I must apologize for my late reply, I was on Holiday.

Firstly regarding Isopropyl Alcohol to the best of my knowledge and everyone else I know it evaporates very quickly, leaving no traces or films behind. Hence why most of us here use it, if I'm wrong some information would be highly useful to improve the efficiency of the cleaning process.

The ridges provide more surface area but less of the surface is touching the CPU, yes you are correct but when the ridges are gone the whole Heatsink can contact the CPU, not just the peak of the ridge.

Yes Arctic Silver (all versions + Arctic Alumina) have been tested before lapping, they perform well but it is more beneficial to have a flatter Heatsink simply because AS is yet another 'obstacle' for Heat to transfer through. AS3 is the latest revision and works very well with a Lapped Heatsink; it fills any Micro-Voids left but then scarpers out of the way, thusly making it more efficient.

Thankyou for your comments
Regards,
Nick 'Maximus Nickus'


-----3rd Message-----

I spent 20 years in the military as a computer maintenance technician. We were not permitted to use Isopropyl because it left a film behind. Denatured alcohol was the only electronic cleaner we commonly used.

Considering your comments in regards to lapping. Simply make more since to me to keep the surface area of the heat sink and chip as big as possible and use a good heat sink compound. I would like to see a properly conducted test before I went through the hassle and dropping of a few Celsius.

Regards.

Jeff Leckemby


-----3rd Message-----

It's a tried and tested method, it works.

I will certainly look into denatured alcohol, do I have your permission to use your emails as quotes and include your name within overclockers.com?


Thankyou for your information,
Nick 'Maximus Nickus'


-----4th Message-----

That's fine.

I've included the whole email to give you an idea of what we were discussing, so what are your thoughts.
I find this very interesting.

I will conduct some test when I can get hold of this 'Denatured' Alcohol.

M_N


P.S. I do have his permission to include his Email (edited version of email) but I will NOT give this out unless there is Good reason, I don't want him recieving a Phleora of emails. I Hope you understand.
 
I don't know about the alcohol, but I believe in lapping the HS. Twice I've installed sinks without lapping. On one ocasion, lapping dropped 2Cs from 38C to 36C idle. Since then, it rarely goes over 36C idle(no oc by the way).

I'm no expert nor do I claim to be. I'm just a novice that has done extensive readings on this subject and tested 6 different HS on three different computers. All with temp drops after lapping.
 
cmdrjay4 said:
i use miller lite:beer:

ROFLMFAO LOL HAHAHAHAHAHA!!!

MAN THATS THE MOST HILARIOUS REPLY IVE EVER SEEN, THNX FOR THE LAFF hehe!!!

Well, now into the subject... ive been searching for pure isopropyl alcohol in my city (like hardware stores and drug stores) and i can't damn find it, so i would be glad to try normal Denatured Ethilic alcohol (the one you use as an antiseptic) on my CPU and post results if somebody is 100% sure it wont fry the thing. I need to reaply paste on my secondary rig and im about to get a tube of AS3, so pls somebody reply...
 
I know that in cassette head cleaning fluid the ingredient is denatured alcohol, as is the case for contact cleaning fluid. And I work in the electronics industy.
 
The ridges provide more surface area but less of the surface is touching the CPU, yes you are correct but when the ridges are gone the whole Heatsink can contact the CPU, not just the peak of the ridge.

removing the ridges also lessens the chance that there will be air bubbles trapped in the "valleys" of those ridges.
 
Man,.... i just use a air compressor at about 200psi :) that tube isnt supposed to be growing like that is it? :)
 
well its simple chemistry the denaturated is better...but anyways i really disagree with him on lappng...I think you should lap your heatsink to the best...cause the surface area gain is minimal and you don't get as good contact with the core if not lapped.....
 
OK i will second him on the Denatured Alcohal thing, i know for a fact that it does work better, I am an electronics tech of sort...

Denatured alcohal tends to have a lot less free bodies in it then isopropyl alcohal. and i havent used isopropal for a long time.

Isopropyl is available at drugs stores. Denatured is available at hardware stores
 
Last edited:
I think part of the problem with not lapping is that like the state of W Virginia, with all the mountains, you're not going to get the most even heat distribution possible, which in turn reduces heat transfer. And when pushing hardware to the limit, every little bit helps:burn:

DWolf:cool:
 
About the Lapping part, understand that lapping is to flatten out the base of you chosen cooler & not really to polish it out. The concern of Jeff Leckemby are the micro peaks & valleys that are present on the surface of the contact area not the flatness of the whole base. There is an ongoing debate weather you should stop at 800grit or go all the way to 2000grit to get the polish look & its up to individual users which they think will benifit them the most.
 
LOL! He kept replying to the lapping side of things so I thought I should clear it up.

With your Isopropyl Alcohol do you use the 70% or 90% which would mean there is 30% to 10% of something else that can leave a residue unlike denatured alcohol.
 
it is very hard to get pure alcohol - infact it is nearlly impossable , using distillation you get a purity of around 97 -99.8 % no more - that 3 - 0.3% leaves the thin film when the alcohol evaportes
 
Cleaning heatsinks

How about the whole of the heatsink, not just the base... like copper becomes oxidized by the chemicals in our skin. Ever notice the dirty fingerprints on your heatsink? I use a weak acid bath to clean that crap off.
 
Back