Stacey Thater and Jim Twellman have indicated that it would be nice if I could document an alignment procedure for this scope. I can try. As you'll see, it's more of an art than a science. An art in that there are no metrics that can be used to do this job - kind of like any other telescope alignment job: You just move things around and look at the results. After sufficient experience, things kind of become intuitive and thus relatively easier. But this is a much more complicated optical assembly than your typical refractor, Newtonian or Catadioptric. So it'll take a little bit of time and effort to get to know this one. So here's my first shot at documenting this (it was written to address a question over on the Cloudy nights forum: whats the focal length) so check that out and look for improvements and corrections to this sorry excuse for an alignment procedure:
The short answer is: As in all optical trains, the better the construction and alignment, the better the image.
The long answer: For this design, there are essentially three things that affect image quality and the closer they approach nominal, the better.
1] Primary mirror focal length and figure. This is not to critical - the corrector elements are designed around this focal length but I'm told a +/- 1% error is ok - the spider distance is where any minor FL errors are rectified. The figure is spherical so that's really easy to get right. As always, the quality of the mirror is important. We have never measured or checked its' focal length or figure (see next for the list of opticians that have had their hands on this thing).
2] The corrector/diagonal assembly is the most critical item. The individual elements obviously need to be ground to the correct prescription. I have no idea how well this was done for our scope. But considering who had their hands in it (Bob Cox, All Sassenberg, Al Woods and Jim Melka to name a few), I have the greatest of confidence that they were done right. Assembly of the corrector system is also of great concern. Again, the closer it can be assembled to optimum, the better. We also have the advantage that the mechanical assembly was machined by Bill Davis. He's the kind of guy you see pictured when you look at the definition of perfectionist. Bill once described to me how he ground and machined the spacers and measured their concentricity, flatness, face parallelism and thickness. It's an amazing process that practically guaranteed perfection.
3] Finally we have spacing and alignment of the optical train. The most critical part is the alignment of the optical axis of the corrector/diagonal assembly with the optical axis of the primary AND the DISTANCE of the corrector/diagonal assembly from the primary. The optical axis of both the primary AND the corrector/diagonal assembly MUT BE COINCEDENT. The larger the deviation the greater the image distortion. Bill designed the Spider assembly to allow adjustments fore and aft and side to side. This effectively allows the corrector/diagonal assembly to be as precisely aligned to the primaries optical axis as possible. Normally, this would be relatively easy to do.
We have two compounding problems with item three here: That damned perforated primary coupled with the lack of precision movement of the spider assembly.
The first problem is that the perforated primary makes it damn near impossible to align the optics with any kind of tools. A laser would work great but it simply exits that three inch hole in the mirror and impacts on the mirror cell. We have several possible solutions to this but my experience says it would be simpler and less time consuming to just replace the mirror (it costs far, far less to do this today than it did back in the 60's & 70's - The cost of glass and the effort already invested in figuring this mirror are why Herb kept IT and CHANGED the optical design accordingly).
The other compounding problem (which IS a major problem) is the inability to "Dial in" the spider. The problem here is precisely positioning the spider assembly, holding the corrector/diagonal assembly, with respect to the optical axis and focal point of the primary. It's difficult to describe in words here without a ton of verbiage. Let's just say there's a ton of "slop" available in the spider vane mounts from which to position the corrector/diagonal assembly in free space as well as a large amount of fine movement available in the primary "collimating" screws. The problem is there's no way to make precision movements. When one vane is moved, the others inevitably move as well - you can't lock down all three and move the forth. The spider has to be moved as an assembly. So that's the real flaw here: Precision alignment ABILITY.
As far as positioning goes, it seems that the first thing that must be locked down is making sure the axes of the corrector/diagonal and the primary are coincidental. Misalignment here seems to cause the majority of the image distortion. This seems to be the most sensitive part of the deal. It doesn't take much of an error to make a huge difference in image quality. It's difficult to describe the resulting image when this adjustment is made. It just takes time playing with this assembly to get a feel for which way you need to go for better quality. The next issue is distance of the corrector/diagonal from the primary. The corrector has two functions: Spherical correction and image extension. It seems to me that spherical aberration is more sensitive to distance error than the magnification function. Moving the spider back and forth should show improvements or degradation in this parameter. Lastly, the focuser orientation can be set for best image. Star testing will show any residual mis-alignments and mis-collimations. Yeah, if you have to change something, you'll have to change all things.
The mirror collimating screws are relatively easier since you can make very small changes in the mirror pointing. But don't forget, when you make a collimation adjustment, you're moving the optical AXIS AND the focal point of the mirror in an arc - and that changes the distance of the mirror focal point from the focal point of the corrector/diagonal assembly as well as the relative alignments of the optical axes - change one and you're back to fiddling with the other ;). It's a bitch when you have multiple components that are dependent upon each other and which have to be "Just so" and you can't use tools to make the process more precise and efficient. Did I mention that replacing the mirror would be the ultimate solution to these problems?
Lastly, the focuser needs to have some axial travel and shimming ability in its mounting. This is just to get it's optical axis in line with that coming from the diagonal. So I think you can see that this is a fourth thing that needs to be done simultaneously with the other adjustments. But it's not that difficult. In fact, you use the focuser to assist in alignment: While aligning the primary/corrector assemblies, the focuser is just left loose. The loose focuser is actually used to check image quality by moving it around and gauging the resulting image. You'll be able to see the image improve and degrade as you move it around. This is the trick to making adjustments of the spider. With patience, an "acceptable image" can be zeroed in and then locking down the focuser completes the alignment.
This whole thing can be frustrating but it get's easier as you "feel your way around the adjustments. Bill and I have a couple of ideas on how to improve the mechanics of this but we just haven't gotten around to taking it to the next level.
Now that you've gotten this far, let me tell you a trick: Do this at night with an artificial star several hundred yards away. You'll be amazed at how fast you can get a feel for what degrades the image how. Once you know that, it's not that difficult to get the image quality where you want it. Also, use the motor drive to help re-center the artificial star. Turn it on but disable tracking so you can use the up/down & left/right slew controls. One last thing. As you get closer to perfect collimation, there will be a place in the image where you will get perfect diffraction rings. This is the area where all rays are in proper alignment. The "Sweet spot" as they say. The location in the Field of View shows where you are and where you have to go to get this image to the center of the FOV. When that's done, you're done.
So I hope this gives you a kick start. This configuration was designed way before the likes of Zambuto and Lockwood started grinding out 16 to 40 inch f2 & f3 mirrors using Paracorr-II correctors (but isn't a Paracorr nothing more than an inline correcting assembly like the JB?). This exact design and scope have been superseded by newer designs and materials so I guess they can be considered relics of the past. They did the job the optical engineer wanted using available materials and techniques. And the goal of this design was to yield a short tube, 16" f6 deep sky telescope. It satisfies that goal very well.