ATM Afterthoughts of laser collimating

["Nils Olof Carlin" <nilsolof.carlin@swipnet.se>]

Answering a comment on laser collimators on the s.a.a, I did the
following arithmetics:
If you use a laser collimator, and you decenter the beam at the main
mirror, and then center the return beam exactly at the laser exit. The
error at the main mirror means an error type 1A of half this (zero at
COC, half at focus)
Take a mirror with f.l of 1500mm (60"), and you decenter by 2 mm at the
main mirror. The type 1A error is then 1 mm, not insignificant with a
fast mirror. The error type 1B is <5 arcmin, but this is quite
insignificant.

I have known of this potential problem with laser collimators, but I
haven't quite been conscious of the magnitude of this problem, and I
haven't discussed it at:
http://zebu.uoregon.edu/~mbartels/kolli/kolli.html   
(see here for an explanation of error types).

When collimating with a Cheshire eyepiece or any variation on this, the
above is simply no problem, as the 1A and 1B errors are (essentially)
independent.

To you who use lasers:
Have you ever experienced trouble that can be explained by this?
What precision do you estimate you have when centering the laser on the
main mirror? And what precision do you estimate for centering the
return beam at the laser?

Nils Olof