Why are laser collimators not good enough?

A simple thin-beam laser is rather sensitive to decentering error and it is a single pass collimation tool which is rather imprecise in its own for fast scope (f/5 and below). A barlowed laser beam with a center donut or a holographic one are however much more precise and insensitive to decentering.

Besides, how do you check the collimation of your laser collimator and has it got provision to correct its tilt?

Final collimation and adjustment should be really be done either with the autocollimator or in-focus at high magnification.

With the typical spread of a collimated laser beam you won't be able to tell the difference with the tolerance required for a fast system on such a baseline. I use a baseline of 10 meters but then I use it only for initial centring. I use a V-block and slowly rotate the collimator on it while adjusting the bolts holding the laser in place.

Hi Andy
I've found that an artificial star is another useful 'cloudy night' tool for collimating the scope.  I use the Geoptik model.  Even with my 14 inch Dob, I can get focus with a Barlow at 60 yards (distance from garage to back wall of orchard). I can then check collimation in-focus and out-focus and compare that to the laser collimator.  It's the next level up from using a laser, and you dont have to worry about clouds or tracking the star! One chap on Stargazer Lounge mentioned he just sticks the artificial star on a tripod and walks it far enough down the street for to get focus. Apparently leads to some interesting conversations with neighbors. As with star-alignment, I've also found the artificial star useful for detecting sources and cures for thermal currents within the tube. 

Andrea - thanks for the Laser-Barlow tip - just googled it. Will give that a go as well!

Mike

Hi Andy,  I do exactly as you describe for my PDS200 . In my personal view that is perfectly adequate for an F 5.0 telescope.  The importance of perfect stars also varies with the object , the scale of imaging and the angular scale of  the detail that you are trying to resolve. and of course the seeing.   With a coma corrector (Baader III in my case) the stars are near round across the field  at  ~ 1.0 arcsec - certainly at normal exposure times and seeing.  But push it  down to < 0.5 arcsec , good seeing and short frames (lucky maging) and any imperfections in the set up will become more obvious - but then everything matters including the mount - imperfections in the movement of that at the subsecond scale can also cause distorted stars.  F 4.0 Newtonian ? – much more sensitive and difficult to get right.  Tim

The attached shows the level of tolerances required to get maximum performance from a coma corrector for PAE and FAE. The tolerances with the coma corrector are tighter, because the coma is lower (i.e., the definition of a defect is tighter):

https://www.catseyecollimation.com/Newtonian%20Axial%20Tolerances.pdf

PAE tolerances are in fractions of a millimeter for fast Newtonians, FAE tolerances on the order of 1 mm or so for the systems we use. Our analysis and measurement tools should have corresponding precision. Visual judgement of a reasonably large diameter laser spot is not very useful here even if collimated.

A lot of this also depends on what your definition of good collimation is. If your definition is round stars at excellent FWHM through the entirety of the frame when viewed at 100% or more, then laser collimators, which are prone to collimation error themselves, are certainly inadequate for reasons already discussed. Even Cheshires or barlowed lasers (for which laser collimation is irrelevant) will only correct  one of the two forms of collimation error (primary axial error). Focuser axial error will need to be corrected using tools such as autocollimators using the correct protocol. It takes owning and using a fast Newtonian and being picky about your stars to realize just how sensitive these systems are to collimation and tilt. This old but very nice thread goes though the complexity involved:

https://www.cloudynights.com/topic/251778-concise-thread-about-autocollimatorsimprovements/#entry3189737

See the second post by Jason that talks about the fundamentals of collimation.

From an optical engineering standpoint, alignment is achieved when the wavefront errors over the field are minimized.  Seeing is what puts a tolerance on how accurately the system needs to be aligned but in general, if you can align the system to be within about a quarter of a wave (+/- 1/8 wv) from the performance achieved by the design, the system will work well under any seeing conditions.  Aligning a simple Newtonian telescope is purely a geometric exercise so it is among  the easiest optical systems to align--provided that you have access to the required mechanical adjustments.  The diagrams in the CN thread that you referenced didn't show up when I checked it out but at first glance, that whole thread seems unnecessary complicated.  The challenge boils down to positioning the secondary mirror, making the axis of the parabola collinear with the axis of the focuser, and minimized sensor tilt.  Adding a coma corrector will almost certainly increase the challenge of getting properly everything aligned.  It's a shame that most of the laser collimators that are sold for optical alignment into the amateur market are so poorly made.  That's really just about all you need to properly align a Newtonian--along with some software to measure sensor tilt.

John