best way to lap a heatsink

Mac · Oct 12, 2003

just got an slk-947u and it needs some lapping, BADLY. I can feel, hear, and see (yah see) the ridges.

so whats the best way to do it.. circular motions, figure 8, or just straight a striaght line??

also is it better to use wet or dry sandpaper or are they both about teh same?

thanks

crull · Oct 12, 2003

Mac said:
just got an slk-947u and it needs some lapping, BADLY. I can feel, hear, and see (yah see) the ridges.

so whats the best way to do it.. circular motions, figure 8, or just straight a striaght line??

also is it better to use wet or dry sandpaper or are they both about teh same?

thanks

http://overclockersclub.com/heatsinklappingguide.php

=[8c0re8]= · Oct 12, 2003

awsome guide man

thanks for the link..

I.M.O.G. · Oct 12, 2003

I wouldn't follow that guide.

The deficiencies of commonly recommended lapping tequniques are becoming more and more widely known.

The best lapping is a straight forward, lift to starting position, repeat motion. Any rotary motion will give you a convex base as the front edges take the brunt of the force. When you rotate or go in a circle, the outside edges of the heatsink will be sanded the most, while the middle is sanded the least - this makes it harder to have good contact with the die.

This forward-lift-repeat method is recommended by BillA for the best flatness, and he is one of a very few chosen people who have the tools to observe flatness at such an acute degree.

In your situation, if the milling marks on the base are that bad, you may want to start with 200 grit instead of 400 like the guide suggests.

This guide also suggests an inferior method, but it is more detailed than the previos guide:

http://forum.oc-forums.com/vb/showthread.php?s=&threadid=170745&highlight=ultimate+guide

Pay attention to the posts by graystar and billa in that thread.

crull · Oct 12, 2003

IMOG said:
I wouldn't follow that guide.

The deficiencies of commonly recommended lapping tequniques are becoming more and more widely known.

The best lapping is a straight forward, lift to starting position, repeat motion. Any rotary motion will give you a concave base as the front edges take the brunt of the force. When you rotate or go in a circle, the outside edges of the heatsink will be sanded the most, while the middle is sanded the least - this makes it harder to have good contact with the die.

This forward-lift-repeat method is recommended by BillA for the best flatness, and he is one of a very few chosen people who have the tools to observe flatness at such an acute degree.

In your situation, if the milling marks on the base are that bad, you may want to start with 200 grit instead of 400 like the guide suggests.

This guide also suggests an inferior method, but it is more detailed than the previos guide:

http://forum.oc-forums.com/vb/showthread.php?s=&threadid=170745&highlight=ultimate+guide

Pay attention to the posts by graystar and billa in that thread.

So many people have been lapping that way I doubt it really makes that much of a difference, but if the pressure is not right I can see how that can effect the edges more then center. I use a piece of glass or mirror and I place the sand paper on it. Then I apply direct pressure with the palm of my hand and twist the heatsink in a circle back and forth.

I.M.O.G. · Oct 12, 2003

That's why I presented the link, I have left the choice up to the individual:

You can lap using methods that the majority of enthusiasts use, or you can lap using the methods that the most knowledgeable enthusiasts use.

I prefer the straigt-lift-repeat method because trusted, experienced, more knowledgeable people in the community often say it's better.

Not everyone agrees on this of course.

crull · Oct 12, 2003

IMOG said:
That's why I presented the link, I have left the choice up to the individual:

You can lap using methods that the majority of enthusiasts use, or you can lap using the methods that the most knowledgeable enthusiasts use.

I prefer the straigt-lift-repeat method because trusted, experienced, more knowledgeable people in the community often say it's better.

Not everyone agrees on this of course.

Can you please explain this method a little bit better, thanks?

I.M.O.G. · Oct 12, 2003

A better explanation is also in the link I presented:

Graystar said:
I believe that a lot of what this person had written is wrong.

Lapping....
First, it is well known in the tool-sharpening industry that using such a procedure as he described will produce a convex surface. The reason is that when the block moves over the paper, the front edge will grab more than the center or rear. He even noted that this would occur. However, he didn't make the connection that his surface was not flat. Woodworkers have long known that in order to get a plane perfectly flat, you must push it in one direction only, *then lift it off the paper* to return to the starting position, then push forward again. Anything else will not get you a flat surface.

Second, the procedure he described is not lapping. You do not lap with sandpaper. You lap with a lapping plate and lapping compound. The problem here is of not understanding what lapping is.

(pic of Lapping plate)
http://www.statecollegecentral.com/metallathe/MLA19.html

Lapping is a surface treatment. It is a polishing process that removes minute amounts of material in a manner that generates very little heat. This is important because heat can distort the surface to the point where heat distortion is worse than the surface finish. At this point, lapping is pointless.

So, the results you want to achieve really require a two step process. First you make the part flat with the sandpaper as I described (sometimes referred to as roughing.) Then you lap the surface to remove surface irregularities.

As for cleaning the surface, compressed air is very good at scooping out tiny, tiny particles off the surface. If you use water then you need high velocity, which gets messy. Otherwise, the water just rolls over minute particles trap in voids on the surface. Scrubbing doesn't help, as you have nothing fine enough to get into the voids. Like trying to comb your hair with a comb that has teeth 12 inches apart.

BillA said:
follow the procedure that Greystar described above
its the same that I use

one direction, lift and return - never rotate
close to impossible to actually get it flat by hand, I never have
(and I can, and do, inspect for such)

be cool

LtBlue14 · Oct 12, 2003

IMOG is exactly right
it drives me crazy to see people recommending rotating or figure-8 lapping techniques. (what we're doing isn't actually lapping anyway, you need a lapping plate for that). in the end, it may not make a difference since the cores on cpus are so small compared to the heatsink - the curvature is probably not that significant over such a small area. BUT, the method IMOG describes is still the proper one. Push straight across, lift, repeat, never rotating

crull · Oct 12, 2003

OK you guys are getting far too technical here. Lapping is just a term, instead of saying I'm gonna sand it flat. As far as the best way to get the heatsink flat I'm open to all suggestions. I will try the lift method next time, but I also think twisting the heatsink works well as long as your using even pressure over the center of the heatsink.

Mac · Oct 12, 2003

excellent guides there, ill be sure to straight - lift - repeat for optimum results, thanks.

I.M.O.G. · Oct 13, 2003

crull said:
OK you guys are getting far too technical here. Lapping is just a term, instead of saying I'm gonna sand it flat. As far as the best way to get the heatsink flat I'm open to all suggestions. I will try the lift method next time, but I also think twisting the heatsink works well as long as your using even pressure over the center of the heatsink.

Those references are kind of technical. And your statement as to the use of the word lapping in our applications is accurate.

Saying the point of those references like this makes it much simpler:

The point of those comments is that rotational, turning, or twisting motions, make it impossible to apply even pressure over the center of the heatsink.

The front edge of a surface always grips the sandpaper the hardest and will sand faster.

This is why the straight lift repeat method is the best - because other methods make even/flat sanding nearly impossible despite pressure application.

It is easier to get a HS base flatter by using the straight-lift-repeat method.

crull · Oct 13, 2003

IMOG said:
Those references are kind of technical. And your statement as to the use of the word lapping in our applications is accurate.

Saying the point of those references like this makes it much simpler:

The point of those comments is that rotational, turning, or twisting motions, make it impossible to apply even pressure over the center of the heatsink.

The front edge of a surface always grips the sandpaper the hardest and will sand faster.

This is why the straight lift repeat method is the best - because other methods make even/flat sanding nearly impossible despite pressure application.

It is easier to get a HS base flatter by using the straight-lift-repeat method.

In the first place I only posted a link to a lapping guide, "God forbid I used the word lapping". You people take everything and make it into some big major event. He asked a question about lapping, and he got a link on how to do it "Did I say lapping again...I meant sanding flat?" Let him look it over and then research it. That method has worked fine for many many people, now all of a sudden its not good enough. If he uses that method compared to your lift and tuck method what's he gonna lose 1-2 C at the most if that. Come on get your heads out of your lap.............."Oh God I said it again". Idiots!!!

Mac · Oct 13, 2003

in defense to imog, he was just saying that there is a better way, i'm not sure how he offended you in doing that. It's his opinion and many others.

thanks for the information imog

crull · Oct 13, 2003

He didn't offend me. I just think that some people have to take something simple and turn it into something complicated. You asked a nice simple question, but its turned into a long complicated topic about the definition of the word lapping. I have lapped heatsinks many many times and twisting the heatsink with even pressure at its center will make it pretty flat. I doubt it will work much better using the lift method but if you think it will then give it a shot.

Mac · Oct 14, 2003

crull said:
You asked a nice simple question, but its turned into a long complicated topic about the definition of the word lapping.

the definition of lapping was brought up in a quote of someone else in a different thread, it was never brought up here.

xtern · Oct 15, 2003

good info cheers, it IS simple

Susquehannock · Oct 15, 2003

crull said:
OK you guys are getting far too technical here. Lapping is just a term, instead of saying I'm gonna sand it flat. As far as the best way to get the heatsink flat I'm open to all suggestions. I will try the lift method next time, but I also think twisting the heatsink works well as long as your using even pressure over the center of the heatsink.

Agreement here. :clap:

No need to :argue:

over a simple technique. We should remember
that the CPU core only touches in the very middle of the bottom
contact surface.
As long as that small area is flat the edges make no difference. Besides, if you keep even pressure on the
sink only a very slight edge will be rounded. (provided you have a flat surface under the paper)

For the record ... I usually use a straight back & forth motion.
Sometimes I like to twist the heatsink a few times to get a
different pattern on the surface for reference. But in general
I use the back & forth motion while completely washing
both the HS & paper between every few passes.

crull · Oct 15, 2003

Seeing how were talking about sanding heatsinks, I thought I would pass along a tip on sanding small heatsinks such as for ram or mosfets. They are very hard to hold when sanding. So I know this sounds crazy but a potato works great. Cut a potato in half and stick the heat sink on the cut end. Use the potato to hold and sand the heatsink flat.

DAGO · Oct 15, 2003

I have used the figure eight method for years, not only for lapping heatsinks, but for assuring a flat surface on engine parts for aircraft engines such as the valve covers...

When you use the single direction method, you put ALL of the uneven sanding on the same edge EVERY time you make the pass over the sand paper. This seems like a sure fire way to get an uneven surface...no matter how careful you are, that one edge is going to get more material removed than the others...
Seems to me that you are letting yourself in for alot of problems by using the single direction method, but in reality whatever works for you is the method to use...My tried and true method is to use figure eights, so thats what I'll stick with...

As Sly and the Family Stone said, "Different strokes for different folks."

best way to lap a heatsink

Member

Member

Member

Glorious Leader

Member

Glorious Leader

Member

Glorious Leader

Member

Member

Member

Glorious Leader

Member

Member

Member

Member

Member

The Mad Smelter

Member

Curmudgeon

Similar threads