i'm using a newtonian and i believe my mechanical collimation is pretty decent now .i tested it using ccdi real time, several runs, good seeing, 4000 stars, 1 min exposures, 1800mm f.l., and these ranged from 4 to 9 arc sec, averaged around 7 or so. the x/y tilt was consistently in the 0.1 to 0.3 range. can i be reasonably sure that a mirror adjustment is what is needed here and not a secondary adjustment or something else in the optical train?
thanks,
dave

Hi Dave,

CCDI measures the overall collimation of the system and can't distinguish between primary and secondary adjustments. If the overall collimation error measured by CCDI is low, then you can be sure that both are properly adjusted.

You will want to adjust the secondary using the traditional methods (cheshire, sight tube, etc.) When done, finish the adjustment by collimating the primary, using CCDI.

Your overall collimation error numbers sound good to me, so I doubt your secondary needs adjusting.

Regards,

-Paul

thanks paul

At last I found an answer to a question that's been bothering me for months! What exactly constitute "good" collimation measurements? I can get the error down to about 6" but that's it, no lower, tilt is also in the .2-.4 range. Is this good? Where would "bad" start?

that's probably ok. although since my post from last year i discovered that indeed my secondary alignment was off and once i got that straightened out i started to get readings in the 3-5 range.

Hi,

6" is pretty good for collimation with the multi-star method. The single defocused star method in CCDI 2 should allow you to get the error down to about 1-2 arcseconds with good seeing.

Tilt of .2 arcsecond isn't bad at all. What this value tells you is that there's a linear change of FWHM values from one side of the chip to another. Star sizes increase by 0.2 arcseconds FWHM. In other words, if the FWHM value on one side of the chip is 2 arcseconds, the FWHM value on the other side of the tilt will be 2.2 arcseconds. In most cases, this will not be noticeable. A tilt approaching 1 arcsecond (possibly as little as 0.7 arcsecond) may become visible, but that depends on seeing and overal system resolution.

Regards,

-Paul

quote:

Originally posted by k8jb:
At last I found an answer to a question that's been bothering me for months! What exactly constitute "good" collimation measurements? I can get the error down to about 6" but that's it, no lower, tilt is also in the .2-.4 range. Is this good? Where would "bad" start?