Two Piece Lapping Technique - Overclock.net

TheStickyWicket · 02-17-08

I've noticed, through search, that this may be a foreign idea to most...

Lapping involves(at least the most common method) a flat surface, whether it be glass or marble, and either sandpaper or rubbing compound. You make your passes, whatever direction you choose (i'd suggest circular these days), but there will always be imperfections. And think of it this way, if you lap both your cpu and HSF, and they both have these slight imperfections, bringing you further away from contact than at stock. Example:

When using circular motions, your hands, no matter who you are, will always lean towards the outside of the motion, causing convex shaping of the surface, however slight it may be. But if the same method is used on the HSF, you are going to get the same result. Now, you basically have two spoons, with their backside put together, only touching in the very middle.

Two piece lapping involved placing both objects (in this case the cpu and HSF or Water Block) together and using a very slight lapping compound...

When placing the two together to grind synchronized, they both come together as you slowly remove imperfections from both sides. This technque is very taxing, but in the end both pieces have the most contact possible, as you formed them basically together in unison. Whether or not they are flat at this point is not the issue, as you have used each of them as a template for the other, maximizing contactof the two metals. A quick search of lapping compounds(make sure you get copper types, and make sure it is non-cutting after use and cleaning). Same as sandpaper, you want to go in incrememts, most often referred to as different size micron type compounds.

Example:

this lapping technique is more commonly use for seating valves in cylinder heads for cars and trucks, maximizing seal for both intake and exhaust, using the two parts to ensure the best contact. i hope I was able to get this across in a way that is understandable to most.

Please remember that glass in not perfectly flat, nor is marble. But when using the cpu and HSF together, neither item needs to be...

TheStickyWicket · 02-17-08

here is a great site for Diamond Lapping compounds. It gets a bit expensive, but from conventional lapping to this two piece method, I have noticed a 4C difference on the same batch e8400. Overall it takes roughly 1 hour per micron size, but is a beautiful result in the end. Hope this helps.

waqasr · 02-17-08

Thats something i didnt know but seriously a piece of glass is flat enough for out hsfs and cpus..and it gives us a decent drop in temps..were not seating valves or anything so i think "normal" lapping is fine.

tofumonster · 02-17-08

That's pretty tight.

TheStickyWicket · 02-17-08

Quote:

Originally Posted by waqasr

Thats something i didnt know but seriously a piece of glass is flat enough for out hsfs and cpus..and it gives us a decent drop in temps..were not seating valves or anything so i think "normal" lapping is fine.

Sure, and agreed. but not all glass is perfectly flat, and I'm trying to point out that "flat" is not exactly the main goal(in a way, hard to explain). it's the contact between the two. Even if they had a 2 degree slant, as long as they both have a 2 degree slant, the contact would be near perfect with this method as they would be basically interlocking pieces. but yes, I agree, glass is flat. the motions you make with your hands are the issue. there is no way to exactly replicate the motions and even strokes by hand, thus the reason for this method. I have noticed huge drops in temp with this method, I had just hoped more of the members looking for extreme improvement may be able to accomplish this.

**grunion** · 02-17-08

Interesting

I rotate whatever I'm lapping 90° every 6-8 strokes, kinda prevents overdoing one edge.

TheStickyWicket · 02-17-08

Quote:

Originally Posted by grunion

Interesting

I rotate whatever I'm lapping 90° every 6-8 strokes, kinda prevents overdoing one edge.

Yeah, i tried that method, and used it for about a year. Then it was explained to me that its life sharpening a knife. The last stroke on the stone is the direction the final apex of the cutting edge will lean. the other issue with two objects rubbing together is trying to grip the cpu and hsf in this method.