PDA

View Full Version : Reapplying thermal paste?




kntgsp
Jul 30, 2009, 12:22 PM
Just curious if anyone has tried replacing the stock thermal paste with Arctic Silver 5 or some other aftermarket product and if it produced discernible temp differences.

Just wondering because the stock thermal paste that PC manufacturers use is almost always crap and I generally replace it and see substantial improvements in temps.



TheRiseofTheFal
Jul 30, 2009, 12:59 PM
As far as I've heard on these forums, it seems to be a success in lowering temps, what MBA do you have? I haven't personally done it, but examining my own thermal paste its kind of disgraceful the slap dash way it was put on

kntgsp
Jul 30, 2009, 01:58 PM
As far as I've heard on these forums, it seems to be a success in lowering temps, what MBA do you have? I haven't personally done it, but examining my own thermal paste its kind of disgraceful the slap dash way it was put on

I have the Rev C 2.13 Ghz SSD. Seems like the higher 2.13Ghz could benefit more from reapplied thermal paste.

I heard the Rev. C paste application was much improved, but the stock thermal paste never, ever performs as well as AS5 or other good aftermarket ones.

But yea, if anyone has any info about anyone who has tried with a Rev. C and/or disassembly guide for the Rev. C, would be much appreciated. Just want to be extra careful since it's an Air and I haven't yanked one of these apart before, only the MBP and Powerbooks.

SeaneyC
Jul 30, 2009, 02:18 PM
Done it here, replaced the standard gunk on my Rev A 1.8ghz with AS5 and dropped about 5 centigrade, not amazing, but cooler must be better, right?!

Coolbook is still essential for any Rev. A owner, I'm not sure i'd also try re-applying paste given that it invalidates warranty if I was to try it again.

Choice is yours I guess :D

Oh yeh, and taking apart the Air is so ridiculously easy I wouldn't worry too much about the process, A and B were virtually the same, i'd be surprised if C was much different.

TheRiseofTheFal
Jul 30, 2009, 02:18 PM
though I could be quite wrong, I can't believe the disassemble process is much different from the Rev. A, and having taken apart the old school macbook and macbook pros, I can tell you with confidence it's a very simple process if you've been able to get into the innards of those guys!

McGilli
Jul 30, 2009, 03:05 PM
Search this forum - there are some really good posts about this topic where people have taken all kinds of pictures of exactly what you are asking about...

h1d
Jul 30, 2009, 08:20 PM
All revs come off the same. Done for both A and C.

kntgsp
Jul 31, 2009, 12:50 AM
Done it here, replaced the standard gunk on my Rev A 1.8ghz with AS5 and dropped about 5 centigrade, not amazing, but cooler must be better, right?!

Coolbook is still essential for any Rev. A owner, I'm not sure i'd also try re-applying paste given that it invalidates warranty if I was to try it again.

Choice is yours I guess :D

Oh yeh, and taking apart the Air is so ridiculously easy I wouldn't worry too much about the process, A and B were virtually the same, i'd be surprised if C was much different.

5 can be the difference between 1800 rpm and 4000 rpm on the MBA fans ;)

ayeying
Jul 31, 2009, 08:52 PM
5 can be the difference between 1800 rpm and 4000 rpm on the MBA fans ;)

Not entirely. If you're going solely on the CPU temps, yes, but the MBA uses temps from the enclosure and GPU to regulate fans also. You can have a 20 deg C CPU core but a 80 deg C GPU core and fans will be going at 6200rpm

kntgsp
Aug 1, 2009, 06:44 AM
Not entirely. If you're going solely on the CPU temps, yes, but the MBA uses temps from the enclosure and GPU to regulate fans also. You can have a 20 deg C CPU core but a 80 deg C GPU core and fans will be going at 6200rpm

You do realize thermal paste is applied to more than just a CPU.......

Man, you really mistook that.

ayeying
Aug 1, 2009, 11:53 AM
You do realize thermal paste is applied to more than just a CPU.......

Man, you really mistook that.

My point was, just a lower CPU temp doesn't entirely mean your fans won't be running at a higher RPM especially since you have no way of reading the GPU temp in OSX.

kntgsp
Aug 2, 2009, 05:29 AM
My point was, just a lower CPU temp doesn't entirely mean your fans won't be running at a higher RPM especially since you have no way of reading the GPU temp in OSX.

I never said just a lower CPU temp. I said a 5 drop can mean the difference between high fans and low fans. That drop would likely happen to both parts, as both utilize thermal paste as a heat transfer medium and both would see a similar drop in temps with reapplied AS5. Those are the two hottest parts in the MBA, hence why better thermal paste could easily contribute to lower fan speeds as both see a drop with AS5.

I've seen it countless times on countless PC GPUs and CPUs that I've reapplied paste to.

nph
Aug 2, 2009, 12:38 PM
Hi

I have seen a lot of reports showing the CPU temp dropped anywhere from 2-10 degrees Celsius. Did it have a negative (raising temperature) on the case itself underneath?

Thanks