Thanks Wavytone for the advice, but I did make my move yesterday on the Mangin. One single "wet" with 302 1/2 emery brought the RC back to what it should be and then I finished it off with 2 wets each of 1200 and 1600 emery.
I'm about 3 hours into polishing now and the RC is holding nicely.
One of the problems with the first polishing was that we had a heat wave and the ambient temperature was about 28 degrees - way too high for Acculap Standard hardness. Now I'm polishing at 22 degrees and another heatwave is upon us. Must finish it today or wait for the cool change.
I am not going to worry about 15nm of residual astigmatism - it is completely insignificant and quite small considering the complexity of the instrument and the lack of collimation adjustments.
Yup you have SA with a touch of astigmatism, the astigmatic axis runs from 10 o'clock to 4 o'clock. Could have told you that from the image, never mind the roddier analysis.
[snip]
You know what you have to do... though to remove a high zone of 5 microns I'd be practicing the gentle art of hand figuring with pitch and rouge, pressing with petal-shaped card templates to control where the contact zones are... forget fine grinding - far too drastic, and even cerium oxide is too drastic. Rouge is far more gentle when it comes to figuring.
The residual spherical aberration can be turned off in software. ALL residuals (including astigmatism) then amount to about 22nm RMS (less than 1/25 wave) which is excellent even by visual standards. Astigmatism itself is about 15nm (1/36 wave RMS), just about undetectable.
And tackling 5 microns (10 waves!) with rouge would take better part of one's life. And most likely result in scratches (too many sessions and Mr Murphy doesn't sleep). Rouge is way to slow to do anything, and best left for massaging in those last tenths of a wave. Utterly out of place for an astrograph with more than 50% obstruction.
BTW this "high zone" isn't so much on glass - most of this residual third order SA comes from having a wrong radius at a wrong (actually WORST) place. Most of it could have been sorted out by respacing the components (at a cost of some softening of off axis images); but Stefan's design is already been cast in stone (and everything, including spacings) set to within a micron or so, so having R4 back to spec is the only solution.
Seeing that Stefan has already reground and mostly repolished the Mangin (I wanted to suggest using pads for raw polish as they won't flow at all, all but guaranteeing the steady curve on R4), it is all now academic ...
Managed to finish the polishing of R4, yesterday, and the Ronchi figure looks much nicer than the first time round - most likely due to a more correct polishing temperature. The previous figure had slightly rolled down inner and outer edges.
Last night I put the OTA back together without aluminizing R4 to try recording some extra focal images. It was difficult to capture enough photons but during my tests I noticed that there was quite severe coma present, just like the previous time when I just ignored it. The SA looked definitely smaller but the coma should have been absent - clearly I had another problem to trace.
I got up at 6am this morning and frantically started measuring everything. It did not take long to discover that the small biconvex lens, which is almost symmetrical, was installed back to front. I took so much care to avoid this very problem and somehow still managed to screw up.
So by 8am the Mangin was in the vacuum chamber and by 9 it had a nice fresh aluminium coating. By 10am I had done the autocollimation test that shows nice, very gently curving Ronchi lines.
And now I ready to hit St Kilda beach with some good reading and try to survive the 42 degrees day.
Well, a lot has happened since my last post and not all good.
After reworking R4 and reversing the incorrectly installed lens, star testing showed that the coma was gone, but the Roddier test showed an increase of the astigmatism. Looks like the lens may not have been evenly supported during polishing. The instrument seems extremely sensitive to miscollimation, that shows up as astigmatism, and I don't know at this stage how much of the stig is due to what.
I also had an adventure with the AR coating of the corrector lenses.
After searching far and wide to find someone willing to do BBAR coatings, I gave up and decided to do my own single layer MgF2 coating. I had a larger than usual evaporation boat already installed, one that I had used for evaporating silicone monoxide, and decided to use it rather than swap it out for one that I was used to for MgF2. It turned out to be a big mistake. As I was adjusting the current, I managed to overheat the boat and the MgF2 started spattering and I ended up with two lens surfaces with tiny droplets of MgF2.
Doing the second sides of the lenses I was more careful and had no problem.
Since then, I reground, polished and recoated one of the spattered lenses.
The second one can wait as it is not important at this stage.
Last night I sent Bratislav another set of test images for Roddier analysis.
Here's Bratislav's comment: "Yup, close enough. If stig is taken out in software, it is diffraction limited (1/22 wave RMS). PSF with and without astigmatism included in the snapshot. With stig components "on", Strehl drops to about 0.5 (L/8 RMS)."
I wanted to do some test imaging before doing any more work on the optics and last night the sky cleared with the moon putting on a low 35 degree altitude appearance.
Visually, with a 5x Barlow and 20mm eyepiece, giving 112.5x magnification, the moon looked nice and crisp with very well defined focus.
Then I captured a 1000 frame AVI with my unfiltered ASA120 mono cam.
So the attached image is wide spectrum, not even IR/UV cut filter used, and it should show a lot of atmospheric dispersion if I had used a colour camera.
I think that all I need to do is a bit of figuring to reduce the astigmatism and then I can move on to doing all the final coatings.
That moon shot looks good. Very close to the finished product now.
Thanks Mark, and you are absolutely right about being close with the optics: With Bratislav's help, I did a couple more Roddier tests and it turns out that no figuring is required.
The first one was done a few days ago and the only change from the last posted Roddier test is that I rotated the Mangin about 90 degrees. Interestingly the orientation of the astigmatism remained the same, suggesting that the problem may not be the Mangin.
I was planning to do some more testing to isolate the problem and, last night, as a first thing to try, I slackened the retaining ring of the front lens, and Bingo!, the astigmatism was gone. See the second attached test result.
Bratislav reckons that it is as good as astrographs get.
Stefan, when you made the carbon fibre tube. Did you use a single sheet of carbon fibre to wrap around the mandrel?
Cheers
Hi Troy,
I did it in several stages. First I used a strip of cloth that was long enough to make 3 layers. When the epoxy had set, I painted on a bit more epoxy, and when that set too I skimmed the surface on the lathe to even out irregularities. After that I laminated one more layer of carbon fibre, followed by more epoxy build-up and skimming. The wall thickness is about 1.5mm.
I had to take a bit of a detour, as I promised someone to refigure the secondary for an ODK10 and that was a big job that required the making of a precision test plate as well.
With that job completed, I got back onto this project a few days ago and I practically finished it. I did seven coating runs, three for magnesium fluoride AR coating, two for aluminium and two for silicon monoxide.
I made three spring loaded pins that live inside the periphery of the front lens cell and lock the dew shield in place when the later gets extended.
Also installed a string of resistors into the dew shield and made a connector block for a super expensive little connector socket that I bought from RS Components.
One very obvious mistake I made is that I did not pay attention to the helical pattern on the two carbon tubes and ended up with a mismatch. It won't be easy, but I will have to learn to live with that. I will probably wrap the dew shield with reflective material anyway.
All I need now is a dust cover and the heater cable, plus installing the guide scope.
Thanks guys for the kind words. I'm afraid there will be no full scale production because this would not be a viable commercial product, even if I could get the price down to an acceptable level.
Some of the reasons:
It can only cover APS size sensors.
Barely enough back focus to accommodate thin filter wheels.
Only manual focusing.
Some clear sky is predicted tomorrow, so may be able to have first light soon.
That is a fantastic looking piece of kit and I am sure it works very well. If I go to Vic South this year I hope you attend so I can take a closer look and drool. Incredible skills that you should consider utilising to make a business for these scopes. I am sure you would have a few line up.
As I alluded to before, this type of instrument would need to be at least one extra inch larger in aperture to be commercially viable.
However my next project will be very much in that category, but it will not be a Honders.
Last night I plugged in my QHY8 for the first time and I can report that the stars were sharp corner to corner. The polar alignment was not good and exposures longer than a few seconds were trailing so I don't have images to prove it. Also realised that I need to print a Bahtinov mask for it because it is very difficult to judge best focus with the tiny under sampled stars.
Paul,
Yes I will take it to VicSouth if the weather cooperates. I always make a last minute decision on attending, based on the weather forecast.