PDA

View Full Version : Bowed Apogee GTX + Lapped e6600?

daedality
08-27-2007, 01:42 PM
I poked around the forums a bit and I didn't find a really definitive answer to the question. Right now I'm running a C2D e6600 (non-lapped) with a bowed GTX. So here it is: Will the temps be better with a lapped e6600 and bowed GTX?

(And before it turns into another mixed metal argument, I plan on getting the copper top.)
nikhsub1
08-27-2007, 02:01 PM
Not likely to make any much of a difference at all.
Monk0101
08-27-2007, 02:01 PM
For as little as I know about it (still learning a LOT), the purpose of Bowing a block is to get it to mate with the CPU which is not flat. The purpose of lapping a CPU us to get it as flat as possible. It would seem that if you did both, Bow the block and Lap the CPU, you would be just as bad off as if you had an unBowed block and a unLapped CPU. Suggest that you go one way or the other, but not both!
daedality
08-27-2007, 02:24 PM
Well, from what I've seen (which hasn't been that much), isn't the case usually that the IHS bows outwards and not inwards? That doesn't make much sense, I know, seeing as how the idea of the bow would be to make it fit better, but I've seen pictures of people in the process of lapping the CPU and the sand paper, when the CPU is held down flat, only touches the center. And they have to sand it down quite a bit before it hits the corners.

I may be wrong, however. What about popping off the IHS cap? Would the GTX be able to touch the chip? I've popped the IHS off of an older CPU before, but never a C2D, so I'm only fairly confident in doing it.
Mekrel
08-27-2007, 02:31 PM
For as little as I know about it (still learning a LOT), the purpose of Bowing a block is to get it to mate with the CPU which is not flat. The purpose of lapping a CPU us to get it as flat as possible. It would seem that if you did both, Bow the block and Lap the CPU, you would be just as bad off as if you had an unBowed block and a unLapped CPU. Suggest that you go one way or the other, but not both!

Not quite, a bowed block is for better mounting pressure in the center of the IHS. Considering its all sides that are being pushed out by the thicker O-Ring, once the block is screwed down - the pressure point is circular.

Look at nikhsub1's thread on the anatomy of a bow, you will see how the TIM spreads on bowed vs unbowed.

The bow will still give the better contact in the center of the IHS on either unlapped or lapped processors.

At Swiftech CES showdown where Gabe/Danny and others tested their blocks - they lapped Fuggers QX6700.
Monk0101
08-27-2007, 02:35 PM
Thanks Mekrel, learning this one little baby step at a time!
Bail_w
08-27-2007, 02:45 PM
a bowed block may hurt the performances of a lapped quad cpu but it will benefit a dual core for sure.
daedality
08-27-2007, 03:04 PM
a bowed block may hurt the performances of a lapped quad cpu but it will benefit a dual core for sure.

Yes, It's a dual core I'm using. It will definitely help? How much of an improvement should I expect using the bowed GTX and a lapped C2D?
nikhsub1
08-27-2007, 03:14 PM
a bowed block may hurt the performances of a lapped quad cpu but it will benefit a dual core for sure.
Really? Where did you hear that?
Mekrel
08-27-2007, 03:16 PM
I recommend running it un-lapped, and if temperatures are good then keep the warranty on your processor.

If they are not good, all it means is undoing the block - cleaning off the TIM on both the IHS and block and lapping.

I don't think it will hurt performance on a Quad if you lapped either, as I said - I'm sure it was a QX6700 they used on the CES test bed and I think Gabe and Danny would have left it un-lapped if it handicapped the blocks in anyway.
Freddie123
08-27-2007, 03:28 PM
a bowed block may hurt the performances of a lapped quad cpu but it will benefit a dual core for sure.

Really? Where did you hear that?

When someone took the IHS off a quad they noticed that the components were much more spaced out then on dual cores meaning the heat also is. A bowed block wouldn't cool all the components like it would on a dual core.
nikhsub1
08-27-2007, 03:38 PM
A bowed block wouldn't cool all the components like it would on a dual core.
Proof?
Mekrel
08-27-2007, 04:07 PM
Bowed block on a quad is better than flat, in the case of ApogeeGTX/GT.

Please see one of the most detailed reviews of the ApogeeGT bowed on the net.
http://www.cooling-masters.com/articles-45-5.html

Pictures on that page say a thousand words (albeit French words ;)). Look at the thin (spread from more pressure) on the base of the bowed Apogee on the very right of the set of block bottoms.
ls206
08-27-2007, 04:27 PM
Well, from what I've seen (which hasn't been that much), isn't the case usually that the IHS bows outwards and not inwards? That doesn't make much sense, I know, seeing as how the idea of the bow would be to make it fit better, but I've seen pictures of people in the process of lapping the CPU and the sand paper, when the CPU is held down flat, only touches the center. And they have to sand it down quite a bit before it hits the corners.

I thought it was the other way round
http://www.christianlykke.net/Lasse/Computer/25-07-07/CAM_0005_resize.JPG
from here (*lots of pics*)
http://www.xtremesystems.org/forums/showthread.php?t=143916&highlight=lapped

the guide linked above also shows the concave CPU plate.