Yep, I finally grew the balls to lap both my Zalman 9700NT and my Q6600!!!!
Wet & Dry sandpaper
600 grit - 1 sheet
800 grit - 1 sheet
1200 grit - 1 sheet
1500 grit - 1 sheet
2000 grit - 1 sheet
Piece of 6mm glass (300mm x 200mm)
I taped the paper to the glass to prevent any movement and made a "backstop" (two nails hammered into my workbench) to stop the glass moving.
Here is the 9700NT stock base, all very nice and shiney but is it flat??
I started off with some 600grit wet and dry paper, the method I started using was 5 passes then rotate the base 90°, 5 passes, rotate 90°, five passes etc until all four axis have had the same amount of passes.
I used this method throughout the lapping, from here on in I'll no longer refer to the 90° rotation just the number of passes, please be aware that the rotation takes place everytime the number of passes in each direction has been completed. A "pass" for this operation is moving from one end of the sheet to the other in a single motion/direction and does not include a return pass, in the interest of technique I went in one direction only. The weight of the heatsink is sufficient downward pressure so there is no need to "lean on it" in an attempt to speed up the process, slow and steady is required.
So after 5 passes we see some results in the pic below. The cooler base is noticably concave with the corners being the high points, not really conducive to a good mating surface is it? the potential for air to be trapped in here is huge.
Given the amount of concavity I decided to increase the number of passes to 10 in an attempt at being finished sometime this week. Honestly, impatience is your worst enemy while doing this and even though I was fully aware of what is required I found myself becoming impatient. After 10 more passes we end up with the pic below, gonna be a long day.
10 more passes and we see the pic below, it was after this I stopped taking pics and continued until the chrome circle was the size of a fingernail using all grits until I was satisfied the base was "flat". To test for flatness a razor blade held on edge across the surface then held up to the light will display any irregularity, well good enough for the naked eye anyway.