My Water block lapped.

[AnomalouS]

I have been getting some terrible temps from my Q6600, so I planned on lapping the Swiftech Apogee GT block and the cpu... But after installing the stock cooling and running a prime bench, I decided to RMA the chip (think the IHS is bad...)

3.2g 1.4v on water @ idle = 42
3.2g 1.4v on water @ load = 65
3.2g 1.4v on stock air @ idle = 52
3.2g 1.4v on stock air @ load = 89
STOCK defualt 2.4g 1.208v @ idle = 44
STOCK default 2.4g 1.208v @ load = 68

So I continued on to lap my block.

Before


My assortment of sanding paper, discs and the all mighty FLAT STOCK of ½” plex.


Starting out with 400 grit….
This was after about 2 minutes… You can tell how the water block was crowned.


15 minutes and 2.5 sheets of 400 grit later.


Up to 600 grit now, looking much better.


And on to 800


Smoothing out well @ 1000…



I went 1200 > 1500 > 2000 > 2500 but the pictures didn’t turn out so well…

And here is my finished product at 2500 and a nice polish.

 

[jbunch07]

nicely done.
what are your temps looking like now?

 

[SK-1]

Is that a Fuzion?

 

[Whilhelm]

Swiftech Apogee GT? or an Apogee. Nice Lap job makes me want to lap my swifty.

 

[jbunch07]

i wish i knew what how much of a difference in temps he is getting, seems like ive read many times that lapping a a wb or proc doesn't really make that much difference in temps but rather having a flatter surface.

 

[Whilhelm]

Smoother surface means better contact between the IHS and the waterblock and this allows better heat transfer. Ideally the waterblock should be lapped. It will probably lower load temps by about 3-5 degrees Celsius.

If both surfaces could be machined to the exact same tolerance, such that there could be no air gap on any area of both surfaces, you could theoretically run it without any thermal paste and still get nice temperatures.

Hmm, that gives me an idea...

 

[jbunch07]

Smoother surface means better contact between the IHS and the waterblock and this allows better heat transfer. Ideally the waterblock should be lapped. It will probably lower load temps by about 3-5 degrees Celsius.

If both surfaces could be machined to the exact same tolerance, such that there could be no air gap on any area of both surfaces, you could theoretically run it without any thermal paste and still get nice temperatures.

Hmm, that gives me an idea...

well thats what i was getting at. ive seen people lap their waterblock but because it wasn't necessarily a flatter surface or even though it was very smooth looking they used a poor method of sanding and caused the waterblock to not make good contact to the IHS so the temps were not improved.

oh and good luck with your "idea" i know what your thinking and i would like to see someone do that!

 

[AnomalouS]

well thats what i was getting at. ive seen people lap their waterblock but because it wasn't necessarily a flatter surface or even though it was very smooth looking they used a poor method of sanding and caused the waterblock to not make good contact to the IHS so the temps were not improved.

It would be very easy to unknowingly sand the block out of truely flat state and put a small angle on it...As long as the block is "FLAT" and the person installing it knows how to make adjustments to the block when it is installed (Via the screws and springs, in say 1/4 turn adjustments" the block should have better results then a concave or crowned block. Best advice is to take your time when sanding, do not apply much if any pressure to the part being sanded.

The above piece is a Swiftech Apogee GT block

I Cannot give you any temp examples just yet, as I am waiting for the CPU to be RMA'd... Should have it back together later this week... I question lapping in general myself, but all in all the theory is correct... "The smoother the 2 surfaces can be, the better contact with the minimal amount of TIM"

I had to do something while my pc was down..lol

 

[rangerone766]

you prolly shouldnt have lapped down a apogee svn, those blocks are meant to be bowed in the first place.

i still think you should have kept that e8400, with what you use it for, it would have been faster. or you could have traded me, i said you could have my q6600 for your e8400. oh well, just get on teamspeak tonight if you can.

coming over to intel overclocking is a learning curve from amd's. just like i'd have no cloue how to overclock an amd, but can oc intel in my sleep. just takes a bit of time to figure it out.

 

[AnomalouS]

I have yet to find anything about not lapping an apogee block... I am working a ton of OT & DT right now, so I wont be on teamspeak too much. for a while..

 

[kyle2020]

you can see a temperature drop anywhere up to 5 degrees after lapping

 

[AnomalouS]

if i see any drop in temps it is worth it...
if the temps were to rise then I could always scuff if back to a stock sh1t crowned block.

 

[DonInKansas]

mmmm....Woodchuck......

+1 For a tasty lapping and a tasty beverage!

 

[newtekie1]

Oh, so your are one of those people that return a perfectly good CPU just because it doesn't overclock like you expected? You are the reason computer hardware is expensive.:shadedshu

 

[ZenZimZaliben]

Now just lap the IHS on your chip and your good to go. I saw about a 5c drop with both lapped.

Make sure your chip is a pretty good oc'r before you lap, cause once you do that its yours forever.

 

[AnomalouS]

Oh, so your are one of those people that return a perfectly good CPU just because it doesn't overclock like you expected? You are the reason computer hardware is expensive.:shadedshu

are you referring to me?

STOCK defualt 2.4g 1.208v @ idle = 44
STOCK default 2.4g 1.208v @ load = 68

do you think that is acceptable?

If product failure was always because of the customer they would not allow returns.

 

[rangerone766]

read this post svn
http://www.xtremesystems.org/Forums/showthread.php?t=187408&highlight=lapping+apogee

btw "svn" is the ingame name of the OP, hes a freind of mine

 

[AnomalouS]

i read it...
It makes some sense. Basically saying the IHS of the cpu can benefit from the water/air block being crowned, as it can apply additional pressure to the IHS therefore keeping it good contact with the chip.

Although I think that 2 pure flat surfaces with minimal TIM should still provide the best heat transfer.

vs

pictures from W1zzard's post.

 

[ckydmk]

Oh, so your are one of those people that return a perfectly good CPU just because it doesn't overclock like you expected? You are the reason computer hardware is expensive.:shadedshu

Is this all you do?? Bitch and complain about other people?? I frequent the forums, and every post I see of yours is either "Nvidia is the greatest things since sliced bread" or "F*** ATI fanboys". You seriously need to know that no one wants to read your post that consist constantly of you putting others down because YOU dont agree. If you have something negative to say. don't

 

[AnomalouS]

well he Posts: (4.21/day)...

edit... nevermind i will refrain from making further comments.