Jump to content


Photo

Rosin alternative to Baker reflector-corrector

  • Please log in to reply
19 replies to this topic

#1 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3234
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 10 February 2013 - 11:24 AM

Here's an alternative to the Baker reflector-corrector that's completely suitable for CCD cameras in reasonably affordable price ranges. This 20" f/4.5 Rosin camera covers a flat 1"x1" square, 1.4" diagonal format, with diffraction-limited imagery over the 0.4-0.95um spectrum. The primary is a hyperboloid figured 33.8% deeper than the equivalent 20" f/5 paraboloid.

Seymour Rosin was a contemporary of James Baker, and the two collaborated on correcting wide-field astrocameras. Baker solved the spherical aberration problem with a full-aperture aspheric plate, while Rosin proposed removing spherical aberration by overcorrecting the primary mirror. Baker's solution offers fairly good field correction over large film/plate formats, while Rosin's solution offers tack-sharp wideband imagery over smaller formats.

In CCD format sizes most of us can afford, the Rosin is the simpler, cheaper solution of the two. For large formats, a variation of the Baker or Wynne/Ross corrector is more appropriate.

Mike

Attached Files



#2 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 10 February 2013 - 11:44 AM

I'll take a look Mike but I'm more interested in smaller diameter scope. Taking the Baker and a 220mm F3 mirror I have got close to something sensible for photography with a huge field of view onto a full frame dslr using F4 and Bk7 at around F2.8 give or take some. Then lost it due to a problem with schimdt plates and oslo plus sliders - locks up if a ref ray doesn't hit what it should.

:grin: Big field means small scopes especially when a dslr is on the end. F15 needs a small scope too otherwise lots of the field is wasted.

John
-

#3 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 20 Jan 2011

Posted 10 February 2013 - 06:07 PM

Nice design, Mike! I believe Baker's corrector was intended to be removed so the telescope could be used as a regular paraboloid -- for whatever reason. A Rosin doesn't have that flexibility, but then it really doesn't need it either.

John, you can scale down Mike's design and end up with an even better performance than at 20" f/4.5. Scaling down is almost always advantageous. But why would you rather make the Baker achromatic corretcor, mount it outside the tube assembly, and then bother making an aspheric corrector plate as well instead of a simple parabolidal mirror with two simple meniscus lenses?

#4 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3234
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 10 February 2013 - 08:14 PM

Well John, does this design approach what you were looking for? 220mm f/3 paraboloid, corrected with a full-aperture aspheric plate and doublet field corrector. Nice spots and MTF over a 3.74º FOV, and 0.425-0.75µm spectral band. Notice I approximated the aperture blockage from a Canon 60D-a camera body ahead of the focal plane. From a pupil blockage standpoint, it looks like the camera body would obscure less of the pupil than folding the corrector elements and camera 90º with a really big diagonal mirror.
Notice that this design doesn't need the costly and fragile S-FPL53 glass, and uses the far more desirable N-FK5 and N-ZK7 pair.
Comments?
Mike

Attached Files



#5 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3234
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 10 February 2013 - 08:24 PM

Here's the layout and spots. Note the pupil vignetting by the 220mm aperture primary mirror at the extreme field angles. It amounts to only about 14% light loss, which can be flat-fielded out.
Mike

Attached Files



#6 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 11 February 2013 - 10:55 AM

That's neat Mike. Thanks. Cheaper glasses too.

To me using a coma corrector and taking out the sa with a schmidt type corrector makes a lot of sense. You might say it saves making a spider but wont arf attract the dew. Baker sort of gets round the camera blocking the light, mine is a MK1 5D, they are cheap now. His corrector is a cemented doublet which takes out 2 glass surfaces making 4 in total. That may well make coating less important especially on the doublet. His lens projects the image well away from it. The F5 main mirror finishes up at F4.5 along with a reduction in aperture to 17in so it seems that the extra power has to come from the lens. It's an absolute ##!!### to do anything with the lens as it has to be just optimised for coma only. Olso's optimisation isn't too good at just doing that and in my case it has to be reasonable with colour too. I use the wavefront analysis for that. Best so far is 0.4 p/v at 3 degrees and perfection in the centre or 0.4 p/v at 3 degrees and 0.25 p/v at the centre. LOL the 0.4 just wont go away. I don't want a semi of 3 degrees but that seems to help the ray error functions work more effectively.

It seems Baker went a bit further than Ross who compromised his design for sa and coma leaving a blurred circle of confusion that is still suitable for some uses. Baker pointed out that many optical systems have large amounts of coma some where in them so why not take it all out and fix SA another way.

I'm not much interest in anything other than the visible light range. IR in particular is more likely to be pollution where I live and UV doesn't figure highly either.

Off topic buy one source of confusion is glasses. Should the N types be used now? eg N-B7 rather than BK7 for instance.

:D I'm not giving up on Baker yet but that one is the 1st alternative of any interest. Both sound cheaper and much easier to make than a 6in apo. Not so good but cost more than makes up for that.

John
-

#7 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3234
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 11 February 2013 - 01:10 PM

I've never heard of or experienced anti-reflection coatings being dew magnets. Anyone else have that experience?

Glenn LeDrew and others have described simple anti-fogging heaters here that should keep the corrector plate clear in all but the worst of nights without significantly disrupting the seeing.

The corrector must be broadband AR (BBAR) coated, as there is an unavoidable ghost reflection from the corrector plano surface that permeates your entire focal plane. This can be eliminated in a zero-power plano-parallel window by just tilting the window a couple of degrees, which dumps the ghosts completely out of the focal plane. On a weakly powered window like this Schmidt-type curve, it could introduce astigmatism. I can have a look at that for you and see how much plate tilt the system can stand.

The corrector elements should be BBAR coated just to improve their transmission. Per ZEMAX analysis, this system has no appreciable image or pupil ghosts.

I think you're better off with the wider spectrum I used. If you use a modded DSLR or Canon 60Da, either allow a lot more near-IR to reach focus along with the H-alpha. If left uncorrected, you will get image bloat from the out-of-focus infrared energy.

The N-type glasses are the preferred glasses from Schott, and are more readily available. Getting older glasses would just have to depend on who has what and how much in inventory. I believe that the N-glasses have a lower lead content and a little more environment-friendly MSDS sheets, but am not 100% sure of it.

I do all design work in ZEMAX because of my 20+ year familiarity with it, and because of the huge range of optimization operands available. I only use OSLO-EDU when giving systems here so more people can open and examine the files. I'm learning Don Dilworth's SYNOPSYS code on the side, but am not ready to switch over to it, and no one here has tried discovering it yet (to my knowledge). Even with my beginner-level knowledge of SYNOPSYS, I find it to be far superior to OSLO-EDU.

Mike

#8 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 11 February 2013 - 02:25 PM

Not the coating Mike the corrector. It's a well known problem in the UK especially with SCT etc. An over rated problem in my view. People use it to knock sct's.

My best so far on the Baker is this one F3.2, 170mm with a 220mm mirror. Not entirely right yet due to back focal length problems. P/Vs are 0.23, 0.23 and 0.2 across the field which could do with being a touch better. Image dia 55mm at 3 degrees. Spots are a bit better with some SA left uncorrected. I suspect that's true of Baker's too.

I need to find out how to get at illumination levels across the field in Oslo EDU. The obstruction is 30% at the moment plus mounting. No losses at the mirror

John
-

Attached Files



#9 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3234
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 11 February 2013 - 02:35 PM

Oh - by "arf" in your post #5673822 above I thought you meant anti-reflective coatings, and that you thought they would somehow be responsible for attracting dew.

Question: If your target camera is a full-frame DSLR, why is your semi-field angle 3º? If your clear aperture is 170mm and your f/# is f/3.2, the system EFL is 544mm. The corner of a full-frame DSLR is 22mm off axis, giving a semi-field of ATN(22/544) = 2.32º, not 3º. Looks like you're designing to a format significantly larger than a 24x36mm full-frame DSLR, which makes it harder to optimize than it needs to be.
No?
Mike

#10 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 11 February 2013 - 04:00 PM

I'm trying to sort that out now Mike. The efl is coming out at 521.868mm and the apperture is 164mm now and the field 2 degrees which give a 37mm dia image. :rainbow: I haven't got round to looking at the actual frame size of a 5D yet. More interested in getting it to work. At 2 degrees I get a fully illuminated field with a 50% obstruction as per Baker except he seem ok about loosing some of that. The p/v is now an amazingly even 0.2 waves across the field and spots look like this. No chance of covering the same spectrum as you have. Filters are cheaper than dear glass though and the glass as it is will probably tweak a little more.

Still some to change but in some ways it's now practical.

John
-

Attached Files



#11 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 20 Jan 2011

Posted 11 February 2013 - 04:22 PM

Still some to change but in some ways it's now practical.

If this is intended for photography it's not just practical it's done. Probably much easier to design on paper than make in a shop. Good luck!

Mladen

#12 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3234
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 11 February 2013 - 04:25 PM

According to the Canon-USA website, the Canon 5D Mark II format is 36mm x 24mm, the same as old 35mm film. This gives a diagonal of 43.267mm and thus a semi-field to the corner of 21.633mm. I'm no OSLO expert, but can't you just set the Gaussian image height to 21.633 mm during your optimizations and let OSLO worry what field angle that ends up, depending on EFL? The semi-field angle to the corner of the format with your current EFL is ATN(21.633/521.868) = 2.37º. But, that will change as your EFL changes, whereas the 21.633mm is (or should be) invariant during optimization.

#13 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 11 February 2013 - 07:07 PM

I'm not too worried about a precise field angle Mike and playing around with the coma correcting doublet is bad enough on it's own without adding any other constraints. The main thing that has to do is reduce coma and if the semi angle changes it's not really a problem as given the scope the set up has it's always likely to be adequate. I have a notional idea of more than 3 degrees total.

There are ideally several things I need to do. Move the lens for one but more importantly get to the bottom of what Oslo EDU's schmidt aspheric constants mean actually on the glass.

John
-

#14 kfrederick

kfrederick

    Gemini

  • *****
  • Posts: 3017
  • Joined: 01 Feb 2008

Posted 03 March 2013 - 09:17 AM

I have a 20 inch f5 newt what advantage would it be if it was converted to this Rosin ? Thanks for these cool designs :jump:

#15 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 03 March 2013 - 01:03 PM

I didn't really clear up my reason for looking at the baker design and it's pertinent to the recent post. The attraction is that it works with a conventional parabaloid. That way I can still make my conventional cassegrain and have the possibility of an interchangeable head. Change the conic of the primary that ideally needs to go down rather than up and things get difficult.

The problem with a 20in F5 newtonian conversion baker style is that the field diameter is just short of 5ins covering a semi angle of 1.75 degrees. There are much simpler ways of getting a good 35mm frame sized image out of an F5 newtonian. Using a 220mm mirror the baker arrangement offers the possibility of very wide angle coverage over a 35mm frame.

MKV's point about a simple 2 element corrector is interesting but so far I haven't found a design that doesn't introduce huge ammounts of SA wide field with a 220mm F3 mirror over a 35mm frame. That's what bakers design is all about. Things are distinctly different at 20in F5 as the field angles are reduced considerably over the same field size.

If anyone has a design for a short 2 element corrector that could cover a 50mm field on my mirror and uses moderately priced glasses I would be very interested. I don't fancy mounting the camera on the front of the scope so short is important. There is a rather expensive long 3 element corrector available commercially that should work out at F3 but on my little scope trying to use it would be a joke.

John
-

#16 kfrederick

kfrederick

    Gemini

  • *****
  • Posts: 3017
  • Joined: 01 Feb 2008

Posted 07 March 2013 - 09:40 AM

So with the Rosin and a over corrected primary is better than a regular newt mirror and a coma corrector ? Thanks

#17 Ed Jones

Ed Jones

    Mercury-Atlas

  • *****
  • Posts: 2710
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 09 March 2013 - 12:08 PM

You're wanting a lot with only 2 lenses. Maybe this design might work, it has 2 off-the-shelf lenses. The larger lens needs a curve generated on the flat side, however, and you'll need to make the middle lens.

Attached Files



#18 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 09 March 2013 - 12:38 PM

Are there any rules on rosin style CDK's in regard to the conic of the primary? I dug up the patent and it uses an under corrected mirror. I checked very quickly this morning and have fairly decent spots over a 1/2 degree semi on a 250mm F9.3 based on an F3.3 main mirror. The primary conic is -0.772. P/V's run 0.664,0.3666 and 0.01837, rms 0.158, at 1/2 degree. It just uses a 2 element K5 corrector.

It looks like the main mirror conic is used to trim the corrector so there isn't a specific rule for it.

John
-

Attached Files



#19 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 11 March 2013 - 01:30 PM

I've improved on that one a little at 220mm dia. It should go somewhat bigger. This one has a conic constant around -0.75, 2 N-K5 lenses and an optional further K5 lens for flattening for photo work if wanted. It improves the spots a little for that. Visually the best image is on a 900mm rad so not worth worrying about. Full spectrum doesn't upset it by much. It's basically a French design where the conic wasn't given. It may be possible to tweak the lens a little.

John
-

Attached Files



#20 Ajohn

Ajohn

    Mariner 2

  • *****
  • Posts: 283
  • Joined: 03 Dec 2007

Posted 11 March 2013 - 01:34 PM

Spots - better than the look in some ways. RMS is under 0.1 across the field. 80% in 20u at the edge of the field.

John
-

Attached Files








Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics