Opinion on Star Test
Posted 19 November 2012 - 02:17 PM
There is such a lot of information here that it is impossible to absorb everything in one pass. However, the patient amongst us will be rewarded. Rarely is there one single factor at work in an assembly with many optical surfaces, yet we are lucky these days that good instruments are available and affordable.
Based on one or two other threads and personal experience, I note one further factor that can be responsible for questions over performance - accurate mechanical assembly. One can have "perfect" optics, yet with a Mak or SCT there are many degrees of freedom that, if not well controlled, can really upset an instrument's performance. I've seen this at work in my Mak 127mm and C8, and was able to identify and correct issues in each. There must be some unlucky owners out there who have been let down by compromised assemblies and simply either accept the received wisdom that "SCT's are soft", or pass the 'scope on.
Posted 19 November 2012 - 09:48 PM
Posted 20 November 2012 - 01:14 AM
how to proceed and when.
Dick Parker's testing optics eludes to KE testing from a star.
I always like KE test, you can see a lot about the surface instead of trying to interpret rings from a defocused star, although there is all the info there in a star test it's not very intuitive.
from this page
Star Spot test
The Spherical Spot test is wonderfully sensitive because it is a null test. This means that when you do the test every part of the mirror is (supposed to be) at the same zone. Any part of the mirror which is at a different zone stands right out. You can get the same results from a complete telescope by using a real star as the light source.
Assemble your telescope, wait for some great seeing, and point to a star. You will need a sturdy mount for your telescope. Keep the star centered and remove the eyepiece (dob users usually prefer Polaris). You can start with a knife edge cutting in the cone of light, move the focuser to find the focal point of the telescope. This is a 1D null test and it is very sensitive. Many people have done this before. Now merely replace the knife edge with a very small spot. Since the star is a true point source, you will be able to use a spot as small as 0.001" for greater sensitivity.
Bring the spot to the focal point (keep the star centered). There is a null point there which will show you the quality of the wavefront exiting the telescope in graphic detail. But be aware, this test is severe. It tests the entire optical system. Not just the mirrors and/or lenses, but also the air the light travels through. It might even be as brutal as the regular star test (with an eyepiece; it shows the change in the diffraction pattern due to image aberrations). Certainly they should be used together. The star test can tell if you have errors, the Star Spot test will show you most of those errors directly.
Posted 20 November 2012 - 01:20 AM
Lemme read up on it. Thank you, Danny.
Posted 20 November 2012 - 01:27 AM
Posted 20 November 2012 - 04:26 AM
I think if you are really interested in characterizing this thing (and it sure seems like you are) then - yes - the natural step would be to do a Roddier test. But when I tried Roddier on my C11 I did not get very consistent results. I think that shows how hard it is to estimate the aberrations present based on studying the star through focus - particularly when high order terms are involved.
In addition - since there is still uncertainty regarding any clipping done by baffles inside the scope - the through-focus behavior may not match any of the simulations in texts or software tools you are referring to - since they don't include it in the model. It's a classic case where simulation may not capture the true phenomenon and you need real data.
So - I would aim to capture digital images of a star at focus and going through focus. For Roddier you need two images on exactly opposite sides of focus, so the size of the spot is identical. I don't think the Roddier software can handle a situation where the secondary obstruction is at a different location from the aperture stop (as it would be if the baffle were the AS) but it would at least be something to try.
Either way - it would be nice to see actual images of the star spots on a linear scale, without saturation. Preferably where the pixel scale is known accurately so you can tell exactly how far you are from focus.
Posted 20 November 2012 - 05:34 AM
It is very difficult to quantify a star test accurately, IMO. It's been said, "the interferometer takes no prisoners." You might be able to say it's good, or bad, or very nice. But only after filtering out biases as well as aberrations, seeing, what have you. Even then, it's hard to get simulations to present the data accurately. Close enough, maybe, a ball park figure. So, basically we're left kind of guessing. Hopefully well, along with some experienced opinions, too.
I'd love some real data, heck even a bench test. Roddier or KE test data. Barring any quantitative data, though, what's left is subjective visual, anecdotal, and some actual tests from Mak samples (on Rohr's site, mostly.) The latter is kind of like figuring out T-Rex behavior based on how a Komodo dragon behaves.
Truth is, it seems to be at least "good" according to Suiter and from what I can gather from Amateur Telescope Optics. And pretty darn good according to Jupiter. I think all the "circumstantial evidence" is pretty consistent, that's good enough for the time being.
You seem to have some interest in the design as well. I hope not to miss your thread if you find the time to publish your thoughts. Should be a very learned discussion.
Posted 20 November 2012 - 06:15 AM
As for my own write up on maks - I will probably get back to it - but the discussion in the atm/optics section wasn't particularly satisfying since, ironically, it was hard to keep people focused. But my main conclusion about maks is the same - they are very hard to make using pure spheres and good results require final touch up in an autocollimator to aspherize one of the elements a bit. The faster maks today, and even the Meade 7", are built around a strongly aspheric element, and often include a large secondary obstruction. So it is incorrect to say they are inherently good performers because they are all-spherical - and it is often incorrect to say they are refractor-like because the secondary obstruction is small.
Posted 20 November 2012 - 07:02 AM
Yea, I wish your previous topic had picked some steam. We may disagree a bit on what it means to be good, but smaller spot Maks can be pretty good (better than diffraction limited correction) with all spheres. I see it, just gotta let you have a peek. Of course they are much better with aspherics. Still, when you get back to it, a ray trace would be interesting to study at some point in the future. If I knew all the radii of mine, I'd give you the data.
Danny sent me on a KE journey, after a break to just enjoy observing for a week. Maybe some metaguide images then, too.
Cheers, Frank...scope is cooling.