How accurate should collimation be?

[E_ri_k]

Hey guys. Just wondering how accurate collimation should be. I have a little trouble collimating my 10" Newt. I'm using a Glatter laser, and Tubelug. First I center the spot on the primary by adjusting the secondary and then adjust the primary with the Tubelug, when I go back to check the center on the primary, it is always out by a little, but still within the center mark. Star tests always look pretty good.

How fussy should I be, and will it affect imaging greatly if there is a small error? I find my self spending maybe a good 20 minutes trying to get it perfect going back and forth?

Thanks.

[rowan46]

For visual near enough is good enough, but for imaging it needs to be pretty spot on any imperfections in your optical train are going to show up in your images

[umadog]

These are the axial alignment tolerances for high power viewing: http://www.catseyecollimation.com/Newtonian%20Axial%20Tolerances.pdf

If the seeing is good and you're pushing the limits of your scope then visual collimation becomes critical. You should iterate back and forth between primary and secondary adjustments until both look good. It may take 3 or 4 rounds.