Jump to content


Photo

Help with a Ross corrector - please

  • Please log in to reply
28 replies to this topic

#1 Ajohn

Ajohn

    Mariner 2

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

Posted 29 January 2013 - 06:05 PM

I've tried scaling a number of ross correctors and playing around with them but on a 220mm F3 mirror that has to be parabolic they generally finish up too close to the focal plain. I need at least 50mm. Also 2 cheap glasses that should be easy to obtain. Best I have come up with so far is based on one out of a book. It meets the old 20um limit but I feel it needs to be better than that. :bawling: Does anyone have a suitable design or can they improve this one. The last dimension is from the focal plane and signs are opposite those usually used. I'm not after perfection so haven't much interest in more complex designs. For an astrograph the spots can be bigger than the diffraction spot. Maybe the current 20um would be ok?

John
-

Attached Files



#2 Ajohn

Ajohn

    Mariner 2

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

Posted 30 January 2013 - 08:41 AM

Forgot to add that the relatively even spot size appeals to me as it's for an astrograph and not visual use. No point of even thinking about visual use at F3, probably not at F4 either in my view. The field is also pretty flat for such a simple set up.

John
-

#3 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3216
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 30 January 2013 - 09:27 AM

The truth - you're not going to get very good images with a 2-element all-spherical corrector and a paraboloid this fast. I dun tried it overnight - see attached OSLO file for one of the best of some 300,000,000 global solutions - not very impressive. Also, check the funky-looking spot UFOs at the edge of the FOV!

You're either going to have to add another element, or allow the primary to deviate from a conic, as in a Rosin astrograph. Which will it be?

Mike

Attached Files



#4 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3216
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 30 January 2013 - 09:51 AM

Just for grinzies, check out this Rosin design I did by allowing the primary to go hyperboloidal. The difference in the image quality is spectacular, with the Rosin almost giving diffraction-limited performance, and at f/2.5! Mr. Rosin was/remains an under-appreciated genius in optical design.

Mike

Attached Files



#5 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3216
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 30 January 2013 - 10:06 AM

For those without OSLO-EDU, here's the difference going to a Rosin hyperboloidal primary can make.
Mike
PS: Also, if you're interested, download OSLO-EDU!

Attached Files



#6 mark1234

mark1234

    Sputnik

  • -----
  • Posts: 31
  • Joined: 15 Apr 2008

Posted 30 January 2013 - 10:28 AM

If you want some feedback from someone who has made these little lenses due to saving money -
assume 2 arc secs is your best seeing,
5 microns is your pixel size.
I can guess the spacing to CCD is due to using a DSLR camera?
The focal length will determine your sampling.
For 650mm focal length, better than or about 8 micron spot sizes or it will be a loss of detail you could have got. GSO corrector from teleskop-express is cheap at £45 for a lot of precision work and very important AR coatings on all lenses. Which you can never DIY or have done at any reasonable cost.
Look at my post on Schmidt Newtonian - the corrector plate is easier than making all the little lenses.
It is intended to make coma corrected enough that it can be further corrected with commercial bought cc (fastest possible F2.65 mirror but with coma already reduced by Schmidt plate to allow F4 coma corrector some chance to work). Still will give plenty of rewarding work to do.

There are lots of 'paper telescopes' using ray-tracing programs on this forum and others - CNC factories and hand-figuring by experts could possibly and very expensively make them accurate enough to gain the benefits. DIY never. Look out for the amazing guy who has just finished optics for his Lynx DIY astrocamera - and note that the results are maybe 1/3 wave at best. Those hundreds of hours of work could have made a truly amazing simple design scope. And he is asking where to get AR coatings done now, at the end of the project? I can arrange this cheap for him through my contacts - we're talking thousands of pounds for all his glass surfaces. And no guarantee that all ghosts will be eliminated. Tolerance your design and see what tiny, tiny changes in radius does on those steep curves! And even if you had good technic and progress through each radius with corrections, get perfect polish (small lenses are really really difficult by hand compared to mirrors)you are rewarded with ghosts of bright stars which cannot easily Photoshop away. Assume all images posted on the internet have been doctored in some way, so it is impossible to make comparisons between designs and actual photos in detail. Nobody likes to hang out dirty washing on the line for all to see. Even more so if you made the optics yourself or spent a lot of money. Pre-computer manipulated silver chemistry images from telescopes - eg AOO David Malin, Sloane Sky Survey are not pretty when bright stars are in the FOV. I use very thick window glass to make all sizes of lenses for Camera Obscura peep shows - fine image on a table, but not astrophotography! So I encourage you to keep going with this project, but be prepared to develop the whole concept based on getting good images. Every professional project has a manager who decides on how to reconcile all the information available and arrives at a compromise. Some ideas get put aside. Use your time wisely and only buy what you cannot make. Manage your own project to meet the overall intended aims, unless your aim is just to kill time!

#7 mark1234

mark1234

    Sputnik

  • -----
  • Posts: 31
  • Joined: 15 Apr 2008

Posted 30 January 2013 - 10:59 AM

Mike, what's the best you can do with BK7 and F2? I have loads of old big epidiascope air spaced triplet lenses with thick lumps of that in and tried to use at least one of the existing AR coated surfaces for a FF/CC design. Diddly squit UFO's was the result. 125mm diameter max blanks, some BK7 25mm centre thickness. Even though I have a diamond generator and have previously done 0.5mm edge and centre thickness lenses without too many breakages - still couldn't get a decent design solution. And OSLO Edu will not do ghost analysis!
It is very frustrating that the prescriptions for the commercial cc's are not available so that optimising is possible - you have shown what a hyperbolic mirror/cc combination can do very well.

#8 Ajohn

Ajohn

    Mariner 2

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

Posted 30 January 2013 - 12:08 PM

Thanks Mike. This scope is a "daft"? idea that started a long time ago. Eventually I had a blank hogged out for F3 and I've fine ground it.
The basic idea is a "cassegrain", astrograph and sort of newtonion all in one via interchanging bits fitted to the spider. The cassegrain will have to be focused via the 2ndry mirror which doesn't frighten me too much. The rosin solution looks great and curiously I look at an RC design. With 5x mag on the 2ndry it's not that much worse than a 5x straight cassegrain from the glass removal point of view. I can't get oslo to load the len file you have attached though. It V6.6.0 and I only downloaded it a couple of weeks ago. Do I sense exotic glass?

The newtonian idea needs something different. A sort of super barlow under the 2ndry mirror to get to F6 - F8.

I would like to have a look at the 2 designs. Maybe screen shots of the data spread sheets would be easier? I did wonder if the flat surface of the design I posted was for later figuring. One aspect of the spot size must relate to the pixel size of dslr's for me. On that basis I wonder if a spot size of much under 10um has any real advantage. 4 or 5um maybe but diffraction limits would be waisted.

My oslo has done one or two odd things of late but it is still loading any files I have created with it.

John
-

#9 Ajohn

Ajohn

    Mariner 2

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

Posted 30 January 2013 - 12:19 PM

Hi Mark. You seem to follow my thinking on spot size. One aspect of that is 4 pixels to gain colour. That leaves me thinking maybe a 2 pixel sized spot is ok. The 4 pixels are interpolated when it comes to generating an actual photo. There is no chance what so ever of me putting a large ccd astro camera on any scope I am likely to own. F3 would be pointless for visual use.

My other "intention" is to use commercial items for the astrograph aspect. Some years ago some one made very fast newtonians for that and used the Bader part. Trouble is that there is no way of knowing how well anything will work without it's prescription.

John
-

#10 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3216
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 30 January 2013 - 03:39 PM

Here's an improved 220mm Rosin with a wider spectral correction (0.436-0.656um) that's plenty sharp for astrophotography. This design brings violet light into focus as well as blue-green-red, giving sharper images.
Mike

Attached Files



#11 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2575
  • Joined: 20 Jan 2011

Posted 30 January 2013 - 06:08 PM

Mike, why is the light coming form the right? had me confused for a moment! :)

#12 Mike I. Jones

Mike I. Jones

    Gemini

  • *****
  • Posts: 3216
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 30 January 2013 - 06:36 PM

Look at surface 1 - it's a mirror that reverses the light path. I prefer to design left-to-right when possible - makes center and edge thickness boundary constraints easier to assign and track.
Mike

#13 mark1234

mark1234

    Sputnik

  • -----
  • Posts: 31
  • Joined: 15 Apr 2008

Posted 30 January 2013 - 07:00 PM

Mike, that is a nice corrector design. Perhaps the field correction falls off too steeply for a DSLR though - 28mm on the diagonal?

John, "daft" is how many non-astronomy friends would describe astrophotography in general - not looking at the images, which they admire, but the nocturnal inactivity on the coldest nights of the year nursing tangled wires, laptops and frosted telescopes in the dark! I would guess many ATMs would like to keep their telescopes open to different configurations, it is what all major observatories have to do in spite of their big budgets. I've got two big slumped mirrors and have abandoned a perfectly functioning corrected DK because DSLR pixels got much smaller in the last 10 years, not cheaper. 20 inch at f8 is way too long a focal length.
So I now have to consider how to use a fast mirror to keep the focal length as short as possible.
The steep converging light cone of an f3 will rule out simple Barlow arrangements for visual use. If you check out Hyperstar users (maybe track them down by an image search)- several have tried a Barlow on a fast mirror after the corrector. My impression is it gave ok results but a 4 micron pixel size camera did the same job.
If you use an integrated design with a spider which would accept a convex secondary and core the primary you could have a Dall-Kirkham which could swap out for a flat giving a slightly under-corrected Newtonian with CC for imaging. I've done that with a 165mm f3.8 mirror and it works fine. RC secondaries ARE difficult to get a smooth aspheric shape - the smaller they are the worse it gets. It takes time and limitless patience to keep putting the mirror back in the test rig and re-aligning everything after just a few minutes zonal figuring work.
It would be a heck of job to re-collimate an RC each time you swapped configuration as well.
Re-collimating a DK cassegrain is quick, easy and can be done initially without anything more than just looking down the eyepiece tube. Replacing the Newtonian flat is actually more tedious.

Mark

#14 Ajohn

Ajohn

    Mariner 2

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

Posted 31 January 2013 - 06:53 AM

A thought for you Mark. I have recently become a fan of micro 4/3. Started with a Pen at about 10mp and moved to an OM EM5 which is 16mp. All of the decent lenses I have were razor sharp on the Pen and 100% res crops looked fine on a PC screen.That isn't the case with the OM but it is if I reduce the size of the crop somewhat. Same with full frame shots. DSLR's are slowly getting like compacts - way way more pixels than can effectively be used due to the optics. Who cares as people seldom use full res shots and most finish up on a PC screen anyway - 100dpi if you are lucky.

Mike's corrector misses the boat for me. Exotic glass and too much power but there may not be any harm departing from a paraboloid. I read somewhere that correctors that reduce the F ratio are easier but most seem to have a power of around 1x. I looked at an RC design and changes in curvature at f3 to f15 aren't too dramatic. For the gain at F15 it's not worth doing but if it helped elsewhere it might be. The F3 mirror will be difficult anyway. I want to look at a gregorian version of both as well as I feel that may be easier to make. I'm happy things can be put back where they were but these scopes do need focusing by moving a mirror.

I played with several corrector designs and notice that positioning can be very critical. Wouldn't this be worse if the corrector makes the F ratio shorter?

I've finally got Mike's prescription to load one is showing F5.2 with 170 odd mm of back focal length. Newtonian? The other is more or less diffraction limited but 0.01mm change in the focusing plane doubles that. 1/2mm change in the position of the corrector doesn't make much difference when refocused. S-LAH60 and S-FPL53 sound expensive to break though? A conic of -1.8 on the mirror sounds too much for an RC design. It comes out at -1.02 or so. My main aim is the compound scope. Closely followed by the astrograph. I wonder if the extreme conic is down to the reduction in focal length?

I have a small but reasonably effective machine shop. Main items are a boxford and a dore westbury miller. I see the scope as having a rather rigid spider with some form of location facility, probably a none locking taper and also a rigid "tube"

John
-

#15 mark1234

mark1234

    Sputnik

  • -----
  • Posts: 31
  • Joined: 15 Apr 2008

Posted 31 January 2013 - 06:40 PM

John,
here's my prescription for my 20" CDK, I've made a few small changes after construction and the index of the glass needs checking - using a microscope with a proper graduated focus scale is the easiest technic. When I did this more than ten years ago I thought I'd invented the configuration - but there is a version with much thicker glass (expensive) in telescope-optics.com.
I used 6mm plate glass for the correctors. Big scope though and too heavily built. The seeing is so much worse than anything else the spot sizes were overkill. Its not fun sticking your head between a big scope mirror cell and a massive steel fork mounting to observe visually. The only thing to watch in setting up was the primary-secondary distance throwing the image further than it could be focussed and the centration of the baffle tube holding the lens cell. Being so far from the focal plane they need to be square on - easy to see and adjust with a spider as per big pro scopes. The lens distance can be changed to get the final compromise on SA correction. I have a Calver 10" 1878 mirror re-figured by Calver himself again and then re-figured by Linscott 1908 (probably with pads on a bumble bees feet). It is great to use for an indoor collimator. The mirror looks as perfect as anything I've seen it's figure is so smooth and as Calver commented on the engraved back 'Very Fine'. Shame I haven't got a New Mexico sky and 4k x 4k 25micron Site sensor camera. Next astrograph should be a better compromise - I know I can use an off-axis aperture mask to get an effective 300mm aperture much slower scope for seeing-limited visual use.
Oh, by the way take note of the spots at flat field and the wavelengths are near UV to IR.
Your idea for taper fittings with a keyway sounds good.

Mark

Attached Files



#16 Ajohn

Ajohn

    Mariner 2

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

Posted 01 February 2013 - 01:48 PM

That's some scope Mark!!! Oh how I wish there was some task orientated info on oslo. I may be able to borrow full 6.4 of some one who doesn't know how to use it.

I had wondered about using plate and checking with a microscope. No shortage of those here. It's easy to get carried away at UK microscope meets. I assume the glass figures in the prescription could be used as a guide for playing around?

Not sure about a keyway in the taper. My lathe came with a myford slotting attachment that was made on the lathe so slots come out on centre. The design should be about on the web if some one needs something like that. It can be fabricated. Trouble with keyways is that they need clearance. Still thinking about that aspect. Ideally it needs to work like a taper - no clearance and a bit of wear doesn't matter. A flat above the taper and a piece that can be fastened down against it in the correct position on the holder might be a better solution as it should give better angular location than a keyway - longer location length and no need for accurate machining.

John
-

#17 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2575
  • Joined: 20 Jan 2011

Posted 01 February 2013 - 04:14 PM

That's some scope Mark!!

Ajohn, not taking anything away from mark1234's work, that design has been around for a while, as even Mark mentions. You were on another thread justr ecently when I posted pretty much an identical configuration (click here).

Mladen

#18 mark1234

mark1234

    Sputnik

  • -----
  • Posts: 31
  • Joined: 15 Apr 2008

Posted 01 February 2013 - 04:33 PM

John, the taper is good idea and just needs an index to maintain the same orientation when replaced. Keyway was what came to mind, since machine tools often have a slot and pin for a foolproof solution in this situation.

The scope is now dismantled to grind the primary down to f2.8 for a Schmidt Newtonian astrograph. There is also another thread running about 'why no more cassegrains?' - well outside of pro observatory research the long focal length is un-useable except by Damien Peach in the Bahamas and Lucky imaging. The 12 micron diffraction disc on a 3.5 metre focal length is very visible and I cannot afford the luxury of exotic wide-angle eyepieces simply to reduce the image scale. I was given some big blanks and made a kiln to slump them thinking along the lines of my experience with SLR silver chemistry imaging. I was also given a whole heap of CCD chips removed from research cameras, including a back-thinned SITE 1k x 1k 25 um pixel device. Trying to get driver electronics working was nigh on impossible - in any case these devices need cooling with liquid nitrogen to get a decent dark noise figure. The present, which I've yet to catch up with, is small pixel cameras - some of the best DSO images resolution-wise are from telescopes with about 1" seeing matching the pixel size - like Robert Gendler and co. The RC and CDK commercial scopes at F5.5/6.8 etc with up to 20" aperture use exotic glass corrector/flatteners/reducers in order to still meet this criteria approx.
I assume you have input the file into OSLO - so try a shorter f ratio primary and same radius secondary like one of the CDKs out there does in its updated 2013 version. I guess it won't be pretty spots as the field gets bigger, really need those bits of fancy glass. Might be ok for you though?

Attached Files



#19 mark1234

mark1234

    Sputnik

  • -----
  • Posts: 31
  • Joined: 15 Apr 2008

Posted 01 February 2013 - 04:55 PM

So sorry Mladen not to realise you had previously posted something similar in design on another thread. I hope John had a good look at that too since it is smaller in overall scale.

John, the OSLO EDU free download is more than good enough. To find out how to do a particular task just Google what it is you are attempting and there should be several results from college coursework and other explicit tutorials, apart from the OSLO manual. Just find a design a bit like what you want and adapt it first to keep the process under control. Autofocus function will give you instant feedback on any changes.

Mark

#20 Ajohn

Ajohn

    Mariner 2

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

Posted 01 February 2013 - 05:16 PM

The biggest problems I have with oslo edu is to get it to autofocus with slider movement and of late to get it to do ray traces at 3 wave lengths. Sometimes it would be better for the graphics window to update as that allows all of the aberrations to be seen as needed. No luck at all on googling for any of these settings. It took me a while to find out how to get the diffraction circle to show.

There doesn't seem to be a prescription for the smaller scope. Just a spot diagram.

John
-

#21 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2575
  • Joined: 20 Jan 2011

Posted 01 February 2013 - 09:23 PM

John, the slider, by definition, is intended for manual optimization, and autofocus isn't manual.

#22 mark1234

mark1234

    Sputnik

  • -----
  • Posts: 31
  • Joined: 15 Apr 2008

Posted 02 February 2013 - 10:02 AM

John,I just openend the attachment and all the specs lens spreadsheet were there? Did you use the top menu bar to actually open the lens spreadsheet, it may not do this automatically. Er, sorry if that sounds a bit patronising, but I admit I've sat there looking confused for a moment after opening an attachment expecting to see the spreadsheet.

Mark

#23 Ajohn

Ajohn

    Mariner 2

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

Posted 02 February 2013 - 10:16 AM

I have managed to find the manual this time. Might be what I call the web effect - mention something and people look and that moves things up the google search lists.

Only spot diagrams Mark so I can't look at scaling them.

The manual states that edu should default to 3 colour tracing and settings are in preferences. Mine defaulted to one as downloaded.

The absence of autofocus when using a slider makes the whole thing almost impossibly tedious. Life is too short. I did play with ?.? Premium a some time ago and persuaded the graphics window to update along with slider movement but can't remember if it also refocused. One thing for sure without a refocus the sliders are a bit of a joke.

One thing that surprises me is the absence of add on "macro's" on the web for edu. I assume some do write them but I have never managed to find any. Now I have the manual it seems to be possible to get autofocus via a slider with a ccl or star file.

:grin: A google of oslo filetype:ccl bought up 1 which at times seems to have come with it.Unfortunately .ccl is used elsewhere but various searches using it found nothing useful.

John
-
PS I started this post hours and hours ago so hope people don't mind that it may refer to ealier posts.
-

#24 mark1234

mark1234

    Sputnik

  • -----
  • Posts: 31
  • Joined: 15 Apr 2008

Posted 02 February 2013 - 10:17 AM

John,
if you autofocus while IMS thickness is on a slider, take note of the new thickness after autofocus on the spreadsheet and move slider until its value is in agreement (may need to change resolutions back and forth to do this).
The wavelengths are set using the menu bar which brings up a window. Now that should be in a demo in the tutorial - as other more experienced users suggest to me on the forum - use the tutorial.
Forget your own design while you learn, just play around with one of the examples so you can return to the original values at any time and know it works.
URL of a really good step by step college tutorial online.
http://www.amoptics.com/Resources.htm

Mark

#25 Ajohn

Ajohn

    Mariner 2

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

Posted 02 February 2013 - 12:55 PM

Thanks Mark. Might be some confusion on the len file. No problems downloading yours but MKV's just come up as jpg's of spot diagrams.

I found that tutorial link this morning and a number of other things. There seems to be a lot less .len files on the web these days but I found a few correctors to play with.

On sliders my edu came with an outline ccl file for use with the sliders.There is an auto focus command and it appears that commands can be included in ccl files so I will probably have a go at that shortly. It rings a bell from last time I used oslo -- ?.? premium. The callback in the slider dialogue could have numbers from 1 to 9 each of which caused different things to happen as a slider was moved. For instance if I remember correctly 9 updated all open graphics windows.

I don't find the usual tutorials a lot of use for what I often want to do. The diffraction spot for instance. Could be the same with wavelengths as I can't recollect seeing that mentioned in a number I have looked at. I have a feeling that winlens, also free, could prove useful for playing about in an educational fashion. There are video tutorials for that and I believe everything updates when sliders are moved. It's also possible to mix in standard Linos lenses. The optimising version isn't free of course. It has a much better interface than Oslo - things are even in logical places.

My edu shows one wavelength under the wavelength window as it came but the easiest answer to that may be to just use some ones files as a starting point because the data goes with the len file - or add it with an editor etc.

Seems that there was a Unix version of Oslo about. That would be better for me as I am running under linux and unix aps usually work without any problems other than x windows which tend to be a bit naff.

ohn
-






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics