How to "properly" apply AS5 on Q6600?

For intel quad cores a thin line left to right through the middle of the text on the IHS - horizontal not vertical... from much experience it does result in the best temperatures... you may be able to get away with 2 small blobs centered over the cores.

A blob in the center tends to work better with AMD CPUs or single core intels. (Tho I'd imagine that with the new AMD quads a similiar method to intel quads is most appropriate)
 
Last edited:
lines and smiley's definalty work better than thin spreading on intel heat spreaders, spread thin doesnt work well unless the cpu's heat spreader is lapped and perfectly flat because the thing is generally so warped a flat spread will leave quite big air pockets that even the thickest spread wont cover without airbubbles.

allowing compression to squeeze the paste allows it to be ultrathin where there is strong metal to metal contact, and thick where there would be an airpocket due to the concave processor.
 
Rroff said:
For intel quad cores a thin line left to right through the middle of the text on the IHS - horizontal not vertical... from much experience it does result in the best temperatures... you may be able to get away with 2 small blobs centered over the cores.

A blob in the center tends to work better with AMD CPUs or single core intels. (Tho I'd imagine that with the new AMD quads a similiar method to intel quads is most appropriate)
Click to expand...

So basically put the AS5 through the SLACR MALAY text? But extend to the edge of the chip?
 
You'll never get a definitive answer for this.
My current Q6600 (that i'm using right now) sits under a semi-lapped Freezer pro 7, with AS3 blobbed in the middle, overclocked to 2.9GHz and ALL CORES idling at 34°C.
That said, the rig upstairs has a Q6600 which i blobbed and spread with cling film all over the die, it sits under a stock intel HSF, overclocked to 3GHz, temps idle at around 40°C.
I'm positive that the diff. in temps is down to case/room ambient temps and different hardware, and NOT down to diff. in AS3 application.

If it works, it works! :)
 
well i used the spread method (finger in cling film) yesterday with AS5 my lowest temps on my i5 750 OC to 3.6ghz is 16c :eek:, idle is 25c ish and load is 61c ish
 
Last edited:
my q6700 is on 1.45v, 3.7ghz on a five year old lapped scythe ninja, idle cores are 38/39 and load no more than 70degC

using line method, vertical I think
 
The optimum way to do it is to lap both the chip and heatsink down to at least 2000 grit sandpaper that rests on a fresh pain of flat glass.

After sanding down clean with alcohol or another substance that totally evaporates after cleaning. DO NOT USE BOG ROLL! Use non-lint cloth or tissue. The best place to do this is in your bathroom after a shower but when the mirror is no longer fogged up, this is because the water will have taken a lot of the dust out of the air. Be careful though - you don't want the water sticking to your chip.

Put chip into socket

Now take the thermal paste (preferably one that is more stable and will last a long time like TX2) and apply as thin as possible with a fresh inside out food bag or a sergical glove (without the grains used for better traction). You only need to do this on the chip or the heatsink - not both. If you're applying to the sink use non-conductive paste. If using conductive paste be careful it won't leak over the edges.

Now apply heatsink with a bolt-through attachment.

There - wasn't that easy. All done.
 
Chris Beard said:
The optimum way to do it is to lap both the chip and heatsink down to at least 2000 grit sandpaper that rests on a fresh pain of flat glass.

After sanding down clean with alcohol or another substance that totally evaporates after cleaning. DO NOT USE BOG ROLL! Use non-lint cloth or tissue. The best place to do this is in your bathroom after a shower but when the mirror is no longer fogged up, this is because the water will have taken a lot of the dust out of the air. Be careful though - you don't want the water sticking to your chip.

Put chip into socket

Now take the thermal paste (preferably one that is more stable and will last a long time like TX2) and apply as thin as possible with a fresh inside out food bag or a sergical glove (without the grains used for better traction). You only need to do this on the chip or the heatsink - not both. If you're applying to the sink use non-conductive paste. If using conductive paste be careful it won't leak over the edges.

Now apply heatsink with a bolt-through attachment.

There - wasn't that easy. All done.
Click to expand...

I'd have said the window of opportunity where the level of dust in the air has dropped significantly and the moisture level has dropped enough to be safe to setup a CPU is too narrow to make this really feasible.

And manually spreading thermal paste on the IHS is a no no - it reduces the efficency of the natural spreading when the heatsink is applied resulting in ares with too much or too little paste and a high chance of air bubbles.
 
Only when both surfaces are not flat.

The idea is that the air is 'fresh'. The water has clung to the dust and then dropped out of the air leaving less time for dust to be re-absorbed and saturate the air. Obviously erring on the side of caution is advised.
 
i use the pea method but i swirl around the heatsink before fully installing it to get any trapped air bubbles out it works well for me;)
 
To be fair your method would have the best results under ideal circumstances - outside of a lab or without a huge amount of experience the horiztonal line method is most likely to bring results.
 
I tried both the line and spreading methods on a lapped q6600 and lapped TRUE (2000 grit on both), the line method tended to give slightly better results, 1-2c per core. Not much but every little helps.
 
I've always found the line method good for a consistant 2C decrease in max load temperatures on both IBT and OCCT.
 
You must log in or register to reply here.
Back
Top Bottom