William Graves wrote:

The formula is a good bit more complicated than that. Take a look at the
ECCENT.ZIP archive on my website for a program that implements the correct
formula.
Regards, Marv


Thanks, Marv. although I was not able to derive a formula - I see that I
had completely missed the bit about the width of the chuck jaws. This gets
interesting as I assume the compensation is because the jaws are no longer
tangent to the workpiece.

It's because the *centers* of the "other two" jaws are no longer the points of
tangency to the workpiece.

In my case, the jaws have a slight curvature!

Depending on how precise your offset needs to be, it may be safe to ignore the
curvature.

-jc-