Lapping how to on overclockers

[pHaestus]

http://www.overclockers.com/articles892/index03.asp

Part 2 from someone who knows what they are talking about. Many links to read up on as a result.

[BillA]

excellent stuff
shoud put that issue to bed

now if we can just get access to N8's visual reference surfaces . . . .

[decodeddiesel]

Well it certianly puts a big shank in the "lapped my heatsink and achieved a 4C improvement" BS. I seriously doubt there are very many who have gone to the extent of work outlined there. I had heard the term "pitch lapping" before, but had no idea pitch meant pitch as in tar like substance. I thought pitch refered to some angle approach or something...totally off...

I am very currious to know how the similar the techniques outlined by the links in this article are to N8's methods.

[jaydee]

Interesting. This seems more like a article written for commercial block lapping. I would like to see something wrote up for DIY'er blocks. The copper I get from my suppliers takes about a good 30mins of hard hand lapping with 220grit just to get the factory defects out of it. Another good 10-15mins per grit size up to 600grit, By the time I get through with the 600grit the grit is usually wore out to a point where it makes the finish smooth and shiny. That's usually where I stop.

This quote below leads me to belive I would be wasting my time going any further than that.

This is a pointless exercise unless you have already done all you can to improve your cooling system. First, switch to water-cooling and get a copper block; if you are doing it yourself, use silver. If you are using a radiator fan combo, switch to evaporative cooling (See Here), use a chiller, and throw on one or two TECs, then put the evaporative cooler outdoors.

There are two reasons to polish a mating surface to within 100nm:

You are testing a new design and need to discount poor contact, or
You want bragging rights for having polished a 100nm surface yourself.

[BillA]

a 'system' approach will always yield better overall results than chasing a single nit out the third story window

[superart]

But what if your CPU doesnt have an IHS, or you dont want to use it even if the cpu does have one, you cant really use the die as a "tool".

[N8]

Uggh, good info but I have to give some quick responses.

Most people do not want to, nor should they lap their IHS or use their IHS as a 'tool.' The common computer user just wants a flat(ter) heatsink/coldplate/waterblock. Anytime you start using or modifying the CPU you void the warranty of one of the more expensive and important parts of your system. A heatsink is usually much cheaper and hence, modifiable (heh, is that word?)

The gist of the article is to get the best mating surfaces between the base of the heatsink and the cpu by lapping and polishing the two surfaces together, similar to valve seat lapping in a automobile cylinder head when you put some lapping compound on a valve and spin it in the valve seat to match the valve surface to the seat surface for the best seal.

"Theoretically," if you do this correctly, it wont matter if the surfaces are flat or not as they will be matched to each other.

This process described does NOT describe how to get a heatsink surface "flat," it describes how to get them to "mate" the best to each other.

In conclusion to this quick note, I have lapped and polished quite a few IHS and CPU surfaces, and this is not trivial, nor recommended for the common user. The article in question does have its merits, but I think for the extreme user only.

[superart]

since when did procooling or overclockers.com cater to the comon user?

[jaydee]

Originally posted by superart
since when did procooling or overclockers.com cater to the comon user?

Overclocker's is all about the average user anymore actually. IHS are designed to bend under pressure and level out. If you lap the top flat then it will not seat properly when under pressure. At least that how it used to be with p4's. That might have all changed. The problem here is you need to use the IHS itself as the other half of the lapping. Certainly needs to come off the CPU or you will kill the CPU trying to lap a block with it. But if you take the IHS off then there is not need to use it at all.

And all this is for what? An unmeasurable gain in temps? It isn't about catering but about common sence IMO. .5C is not worth killing your CPU IMO. The gains are not there.