WOW! I just lapped my TRUE, and boy is it cooler!

darkfires

n00b
Joined
Dec 13, 2008
Messages
7
Boy am I in awe...my already awesome temps just dropped 15C!

Honestly for anyone thinking about lapping, it's really easy, and really worth it!

My question is, would lapping the CPU at this point be of any benefit? I already have my idle temps 1.5-2 degrees above ambient...and I just cant see it getting any lower?
 

Matthew Kane

Supreme [H]ardness
Joined
Dec 1, 2007
Messages
4,233
Another way of lowing temps of the base surface of your TRUE120 is not flat, is by putting a small coin or washer in between the mounting and the heatsink, this pushes the TRUE more forward towards the CPU making it a better contact in between.

As for lapping your CPU, you would see max 3-4'C cooler but depends on how well and how much you lap off. I wouldn't lap my CPU unless I was going to overclock it and use it in a W/C setup, but thats just my personal experience/preference.
 

trick0502

Supreme [H]ardness
Joined
Apr 17, 2006
Messages
5,504
Boy am I in awe...my already awesome temps just dropped 15C!

Honestly for anyone thinking about lapping, it's really easy, and really worth it!

My question is, would lapping the CPU at this point be of any benefit? I already have my idle temps 1.5-2 degrees above ambient...and I just cant see it getting any lower?

15c temp drops by lapping? only two ways that is possible.
1) your cooler was so uneven it isnt even funny
2) your cooler wasnt installed properly
 

Skull_Angel

[H]ard|Gawd
Joined
May 31, 2010
Messages
1,664
Another way of lowing temps of the base surface of your TRUE120 is not flat, is by putting a small coin or washer in between the mounting and the heatsink, this pushes the TRUE more forward towards the CPU making it a better contact in between.

that really depends on the mounting kit used. for most of them it does work, but I've found it unnecessary for the 1366 rev. B kit (wasn't really necessary with the 1366 rev. A kit either).
 

sok0

Gawd
Joined
Jul 25, 2009
Messages
689
15C drop in temps was likely a combination of to much TIM, incorrect installation, and flatness of the bottom of the block. And I'm not sure about a TRUE, but some cpu blocks are bowed on the bottom, and lapping can screw that up.
 

darkfires

n00b
Joined
Dec 13, 2008
Messages
7
The bottom of the TRUE, was waaaaay uneven. I discovered this when i started lapping...the center was the thickest part, and I could also tell by how there was NO paste in the very center where it had pushed it all out to the sides probably the size of a dime because it was so uneven. It's nice and flat now and making great contact.

I think that's the majority of it, is just that it was horribly unlevel!
 

Mickey21

Limp Gawd
Joined
Oct 6, 2004
Messages
442
I have never seen a TRUE gain that much over just lapping as others are saying. Must have been REALLY screwed up which personally I cant believe you didnt notice in the first place... Just saying.
 
D

Deleted member 184142

Guest
The main reason for lapping is to have both mating surfaces as close to a match (in this case as flat) as possible. I have never seen 15C from just lapping a HS base, I HAVE seen that much by lapping both HS and CPU though, so my guess is your given combo of HS base and CPU IHS were a very bad match. I would never lap a HS without lapping the CPU, but it all depends on your goals. Once the CPU is lapped, you can forget about any kind of warranty, and some people worry about reselling also, so people will not even look at a lapped CPU while others (if lapped right) might be a plus.
 

Mickey21

Limp Gawd
Joined
Oct 6, 2004
Messages
442
If he did in fact gain 15C from lapping just the HS, then (@BlueFireIce) it would seem there are cases where you would want to lap just the HS and not worry about the CPU. And it would seem those cases would almost always stem from a very unflat HS surface to mate with the CPU. In the case of the very popular TRUE series, I have never heard of one being this bad, not saying it isnt possible but I would think their quality control would keep this from happening. Probably a very isolated event or we would have heard more horror stories.

When it comes down to it, assuming a CPU heatsink mounting is good enough is not plainly good enough. A bit of proactive behaviour on the part of the owner is usually best practice.
 
D

Deleted member 184142

Guest
If he did in fact gain 15C from lapping just the HS, then (@BlueFireIce) it would seem there are cases where you would want to lap just the HS and not worry about the CPU.

There will always be cases where you will see a better performance (from unlapped) with lapping just the HS (which has almost always been the case with the TRUE), but very, very few if any case where it will not further improve the mating surface to lap both. My guess is since Thermalright states that their HS bases are made slightly convex to match "some" Intel CPU's (I call this lazy BS though) the OP was lucky/unlucky enough to have a CPU with a more "flat" IHS and as such made the contact area of the convex base worse, which is why he has seen such a large improvement from lapping the HS. Either that or he had a TRUE with an much greater convex base than normal, or a combo of the two. I have never liked the excuse for the convex base and the "rough" finish they have used for their HS's, but with a little work, they can be made great.
 

Mickey21

Limp Gawd
Joined
Oct 6, 2004
Messages
442
Agreed, wasnt calling you out as wrong of course, just stating that there appears to be cases where this can help. Not the norm I agree, but appears to be in this case if the OP is indeed having this issue.
 
D

Deleted member 184142

Guest
Agreed, wasnt calling you out as wrong of course

Lol, I didn't take it as such, just explaining that he is seeing this because of the TRUE's piss poor base. I find it amazing these coolers (at the time) were the best on the market, I can only imagine what they could have done if they had come with a proper base.
 

BillParrish

Supreme [H]ardness
Joined
Aug 25, 2006
Messages
7,520
It is always best to lapp both parts. Not necessary but just best. At least do the "razor blade test" on the cpu the next time you have a chance.
 
Top