10" f/3.5 Astrograph with Rosin focal reducer
Posted 21 November 2012 - 08:52 AM
It could very well be similar to the ASA designs. I could probably get a good DL-ish design at f/2.8 over the same field and spectral band, but it would be a real bear to build and keep in alignment. f/3.5 is plenty fast for me.
Posted 22 November 2012 - 11:27 AM
A beautiful design, again. I found similiar ideas in my database. Post one of them only here. Dose this Ross corrector variant works on newtonian?
Posted 22 November 2012 - 04:22 PM
Loren, I hope you won't mind if I jump in on this. In order to get this level of correction, the primary mirorr must be hyperboloidal. A Ross corrector reduces coma in a Newtonian (i.e. paraboloidal primary), but at the cost of introducing spherical undercorrection - which is offset by hyperboloizing (overcorrecting) the primary. But then it's no longer a Newotnian.
Dose this Ross corrector variant works on newtonian?
Maybe modern glass melts offer a possibility of creating a Ross/Rosin type corrector that would correct for coma in a Newtonian without significant spherical residual. It may be worth looking into this. Mike, what say you?
Posted 22 November 2012 - 10:45 PM
Gobble 'til ya wobble,
Posted 23 November 2012 - 07:30 AM
Everyone is welcome. I understand what you said. For Newtonian, Wynne corrector gives excellent image but is diffcult to make. Ross corrector is simpler though causes spherical aberration. What I'm asking is will it be better to add one more element in Ross for newtonian?
Waiting for your comment! I'm playing a F/10 superachromat design in USP 2009/0296201. Take a look if you are interstring. I'll post it in another thread.
Posted 24 November 2012 - 07:07 AM
Hey Mike, I hope you had a nice Thanksgiving. It appears that your corrector's first surface (#4) is the "hairline trigger" that must be spot on to a very high precision - I mean a radius of curvature accurate to a few ten thousands of an inch. Vow!
Radii of the real element (#8 and 9) have more generous tolerances and could be used to counterbalance #4 with much greater radius tolerance.
But that means the radius of #4 must be known to better than 0.05%. Seems like this would be the first surface to be finished and then all other sufraces would be built around that one - again to very high precision.
I also find that perhaps thickness/spacings of the correctors elements must be in the range of ±0.0001 inch, and that the diffraction limited focal range is is in the order of only ±0.0005 inches.
As I said , this is just a preliminary look, using a slider wheel tool. Curious to see what your tolerance data will tell us.
Posted 24 November 2012 - 09:51 AM
Posted 24 November 2012 - 10:15 AM
Posted 24 November 2012 - 11:40 AM
Posted 24 November 2012 - 04:56 PM
We're still out on the road for holidays. If you'll post up Lurie's prescription, we can give it a look in ZEMAX and OSLO.
Posted 24 November 2012 - 09:09 PM
I discussed his design and one special case he intended for amateurs (using an ellipisoidal mirror and a plano-convex, plano-concave corrector) in The Best of Amateur Telescope Making Journal, Volume 2, 17-18.6 "Lurie Anastigmats", pp. 375-392, Willmann-Bell, 2003.
I have his 1975 JOSA article and will gladly post the equations - unless Mike thinks it's off topic.
Posted 25 November 2012 - 12:53 PM
That's interesting arrangement, positive element in front. still only four glass-to-air surfaces. When scaled up to 250mm f/3.5 it is not quite as good as Mike's color-wise, but that wouldn't cause any difference in practical terms. But why use FPL51, if FPL53 is (I think) cheaper? Here's what it looks like.
Posted 26 November 2012 - 03:25 AM
Speaking of tinkering on those cloudy nights, it would be interesting to do a direct theoretical comparison of this Mike I. Jones-Rosin corrected system as opposed to a 10-inch f/3.5 Newtonian with a Paracorr and a 10-inch f/3.5 Newotnian using a Wynne-type corrector (both of which are commercially available).
These are not quite what you want, but there is a coma corrector design with two cemented doublets in the Telescopes, Eyepieces, and Astrographs by Smith, Ceragioli and Berry. Orion Optics have spots for a set of F3.8 Newtonians with an Optimised Wynne corrector with the 10" spots here:
and the others being accessible from here (follow the red links in the table):
Posted 26 November 2012 - 05:34 AM
The design and the results of all of these are either public knowledge or can be obtained by raytracing, but not of Paracorr. My understanding is that it consists of two cemented achroamts. It would be interesting to see how Paracorr would compare in an f/3.5 Newtonian, one degree off-axis, against the other correctors.
Posted 26 November 2012 - 01:12 PM
Hi Loren. My point was that if you add a third element to a paraboloidal mirror you end up with a Wynne-type corrector.
You mean finally I'll get a Wynne corrector?
Posted 26 November 2012 - 04:22 PM
Paracorr's design is not public knowledge, but it is my understanding that it does not resemble the Wynne configuration. Rather I read somewhere that it consists of two sets of cemented achromats. Regardless, since it's a family secret of sorts, it is impossible to say how good it is on f/3 to f/3.5 paraboloids.
So far, all high performance correctors for very fast Newtonians resemble the Wynne corrector, having either three or four elements.
Posted 26 November 2012 - 09:04 PM
The article is in French (thank you, Lord, for Google Tranalate!). The author presents imaging evidence of the performance and concludes that the Wynne and Paracorr are abot the same for f/3 to f/3.5, but that the Wynne has less vignetting and covers larger image frames. At the same time, he concludes that Paracorr is the most cost-effective product.
The plot thickens.
PS I would also like to add that if the Paracorr is indeed a set of two achromatic doublets then it beats any other corrector thus far in manufacturing as well. The Wynne requires very thin, deeply curved lenses that look more like potato chips! Not an easy thing for an amateur.
The Rosin requires a pretty strong hyperbolic primary and a couple of meniscus lenses, with slightly more agreeable thickness, but still n ot easy. One must never forget that an f/3 or f/3.5 hyperboloid is not an easy thing to make especially because many of the available tests presnet a problem with full illumination of optics - and the steeper the curves the tighter the tolerances!
So, at least one, rather well documented test seems to favor Paracorr as the corrector of choice for Newtonians.
Hopefully we didn't drift too much off topic.
Posted 26 November 2012 - 09:38 PM
Posted 27 November 2012 - 12:35 AM
But why use FPL51, if FPL53 is (I think) cheaper? Here's what it looks like.
Why FPL51? The relative price of fpl51 to BSL7 is cheaper.
I do have several FPL53 designs. And you can even use ordinary glass. Here is OSLO file.
Posted 27 November 2012 - 03:10 AM
Paracorr (parabola corrector) uses 2 multi-coated, high index doublets, is completely color-free, center and edge, and installs like a Barlow. Coma is corrected so well, the diffraction limited field area of an f/4.5 Dob/Newt is increased 36 times!" -- Al Nagler
For spot sizes see:
Posted 27 November 2012 - 07:23 AM
As for Paracorr's vignetting, most folks with imaging devices today have digital processors of such size that worrying about vignetting hardly seems relevant.
Posted 27 November 2012 - 08:04 AM
The relative price of fpl51 to BSL7 is cheaper.
You are right. Not long ago I saw the pressing price table, and only remembered FPL53 looking relatively cheap (obviously, processing adds quite a bit).
Yes, common glasses can be used (in the arrangement with the cemented positive doublet in front as well), but they always have the problem of not been able to correct for longitudinal and lateral color at the same time. When scaled to 10-inch f/3.5, and with minimized longitudinal aberration, the 406-8300nm spot at 1 degree off axis is spread over more than three green Airy disc diameters, laterally (not much less at 0.7 degrees). It is still within 0.025mm circle, but would probably result in elongated star shapes. The far red end is not that important, since CCD sensitivity drops toward it, but it is still there.
Posted 27 November 2012 - 08:17 AM
Posted 27 November 2012 - 09:31 AM
Might be a good project to work further on with lower focal ratios. We could call it the Cloudynights Collective Coma Corrector!