re:Re: Renishaw, digitizers

> Date: Fri, 12 Jan 2001 13:03:12 -0700
> From: "Smoke" <Smoke@...>
>Subject: Re: Renishaw, digitizers

>If you use a slightly over sized (relative to the ball diameters) ball end
>mill fopr the pockets, the probe/ball assembly would tend to "self-center"
>and this would eliminate both the problems mentioned by Elliot as well as
>his solution...using slots. machining the ball pockets would be much
>simpler too.

>Smoke

I suggest trying the grooves first. Oversized holes will not help much.
The problem is that too many points coming into contact, and the system is
overconstrained. It will have friction that prevents good repeatability.

A ball in an oversize spherical hole will roll around on the bottom. This
is not a very precise way to locate things. Some people used to use a ball
in a conical hole, this gives a line contact. Better, but still not very
good.

OTOH if you don't have high precision requirements three balls in three
pockets could be made to work if well made.

mirror positioners have been built with the three groove principle for many
years, it replaced all the other methods of positioning because of its
precision and ease of manufacture.

The first time I built a system with the three groove principle a 1 kg
mirror repeated in position to 1 micron everytime. And that was with old
banged up ball bearing for the balls, in a dirty optical shop. It was very
easy to build.

Elliot Burke