Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Classical Cassegrain Telescope from Ealing and Competition, Testing of...

cassegrain classic equipment imaging optics reflector ATM
  • Please log in to reply
39 replies to this topic

#26 junkbum35

junkbum35

    Sputnik

  • *****
  • topic starter
  • Posts: 28
  • Joined: 25 Aug 2012

Posted 09 October 2017 - 12:17 PM

I've seen that type of problem with other scopes, Mirzam/JimC, but I am rather sure that's not it. Neither a thin primary mirror nor a thin secondary mirror, and definitely not glued! This was pretty pricey stuff 40-50 years ago,

 

Guy



#27 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8575
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 09 October 2017 - 01:23 PM

 Guy,

     Looking forward to hearing or better yet seeing the results of the Hindle sphere test of the secondary if you choose to go that way. If you still see  the twist up results you'll  know for sure it is coming  in the figure of the secondary and it time to make a lap and start polishing. 

 

                 - Dave 



#28 junkbum35

junkbum35

    Sputnik

  • *****
  • topic starter
  • Posts: 28
  • Joined: 25 Aug 2012

Posted 09 October 2017 - 11:03 PM

Not sure what I'm going to do.

 

I forgot to mention something!

 

Bill B took a 6" surplus shed achromatic doublet (f/10? f/15? I don't remember) that I had purchased and put to one side, one day when I was off visiting grandchildren, and taped it in front of the secondary. The ensuing focal length was enormous, don't remember exactly what. He saw that described as a possibility for convex=surface testing in Malacara, iirc.

 

Now the ronchi lines from that were not as insanely hallucinatory as the entire primary/secondary/flat setup. They didn't conform to any sort of 'good' ronchigram I've seen bevoere, but at least it was very symmetrical. Straight lines near the center and tremendously turned towards the edge.

 

It was such a distant, tiny target that it was extremely difficult to photograph with my cellphone. Perhaps I should try an actual dslr with a telephoto lens again, mounted with a large ronchi grating and a little LED off to the side. (It gave me fits with the focusing and aiming before...)

 

I'll try to get over there tomorrow (tuesday) and see if I can pull it off. I'll post photos on my blog, since I don't see how to do it on here.

 

guy



#29 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8575
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 10 October 2017 - 09:09 AM

 Guy,

    To post a picture you use the " More Reply Options" at the bottom.  That brings up a text window at the bottom of the screen and "Choose File". You select the picture file and use "Attach This File" then "Add to post". The picture will then be attached to your post. 

  The Malacara test which is described in  "Advanced Telescope Making - Optics" is not that sensitive when you use a long focus lens. The beam exiting the lens is not very convergent  so the convex surface doesn't see the light at the steeper angle that it does when used with a fast mirror. To have the sensitivity you need a lens and test setup that produce a convergent beam that is similar to that  produced by the primary. 

 

                   - Dave 



#30 starman876

starman876

    Hubble

  • *****
  • Posts: 18654
  • Joined: 28 Apr 2008
  • Loc: VA

Posted 10 October 2017 - 10:33 AM

awesome to see two experts on optics talking about the finer points of optics  testing. bow.gif  waytogo.gif

 

I was wondering about the Hindle sphere.  Would not a primary of an SCT serve for that purpose?  


  • tim53 likes this

#31 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8575
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 10 October 2017 - 12:47 PM

  In the Hindle test, light is reflected off the secondary, onto the primary, back off the primary and back onto the secondary then back to focus behind the primary mirror   So your testing at the radius of curvature of the spherical mirror. So to have that happen you need a spherical mirror that has 1/2 the focal length of the primary mirror used in the telescope. So if wanted to test the secondary mirror that is used a say 8" f/16 Cass that has a primary that is f/4  ie 32" you would need to use a spherical mirror that has focal length of 16" and a radius of curvature of 32".  So you could use a SCT primary which is spherical and has 16" focal length ( 8" f/2) to test the secondary used in a cassegrain that uses a 8" f/4 parabolic primary.

 

                    - Dave 



#32 starman876

starman876

    Hubble

  • *****
  • Posts: 18654
  • Joined: 28 Apr 2008
  • Loc: VA

Posted 10 October 2017 - 01:42 PM

I understand.   Therefore, finding a spherical mirror that has half the focal length of the primary is the key.    Guess that is not easy to find.   I presume it also would need to be a very accurate mirror so not to introduce any error of its own.  I keep seeing these cheap mirrors on ebay with a really fast curve on them. I think they use them in solar collectors.  I presume their figure would not be accurate enough for optical testing?

 

i find this listing on ebay intersting

 

http://www.ebay.com/...jYAAOSwv0tU6S8j


Edited by starman876, 10 October 2017 - 01:46 PM.


#33 Ben H

Ben H

    Viking 1

  • -----
  • Posts: 526
  • Joined: 12 Jun 2012

Posted 10 October 2017 - 02:18 PM

Those solar cookers you see on eBay have to be absolutely horrible for use in optics. I've no idea what their actual error is, but I do know there are several labs working on casting epoxy mirrors and they've struggled to get the peak to valley error <50 microns. I can only imagine the error in a mirror manufactured without the intent of using it for optics. 

 



#34 starman876

starman876

    Hubble

  • *****
  • Posts: 18654
  • Joined: 28 Apr 2008
  • Loc: VA

Posted 10 October 2017 - 02:52 PM

that's what I figured.  I knew someone would have a good answer.  However, those Edmund mirrors sure look interesting.



#35 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8575
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 10 October 2017 - 03:53 PM

 You don't need a spherical mirror that is exactly 1/2 the focal of the primary mirror that  is in the cassegrain you want to test the secondary from. What you do is position the secondary so it gives you the back focal length you want with the spherical Hindle mirror. That gives the correct angle to the beam hitting the secondary. Depending on the focal length of the Hindle sphere what can happen thou is the beam hitting the  secondary may not fully illuminate it so you won't be testing the full aperture of the secondary.

  Also with a sphere and having the light source and the focal plane of the returning image coplanar you can only get a null test for secondary that is meant to be used with a parabolic primary. The sphere has no spherical aberration when the light source is at the radius of curvature, just like a parabola has no spherical aberration when it is used with parallel light. So the secondary see the some wavefront that is does in the telescope when you do the Hindle test. 

 To test a secondary used in  a Ritchey Chretien telescope, which uses a hyperbolic primary  , you need to calculate a different position were the light source is located away from the secondary. This is because hyperbolic primary   produces an over corrected wave  and the test needs to reproduce this. By moving the light source to a calculated position, this adds  the needed amount of spherical aberration so the sphere no acts like the hyperbolic primary. 

 

              - Dave 



#36 junkbum35

junkbum35

    Sputnik

  • *****
  • topic starter
  • Posts: 28
  • Joined: 25 Aug 2012

Posted 09 September 2018 - 01:38 PM

Some more testing results here.

https://guysmathastr...segrain-mirror/https://guysmathastr...segrain-mirror/

(I haven't tried testing the secondary yet.)



#37 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8575
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 09 September 2018 - 08:37 PM

 Two of your test show it to be over corrected, those being standard Foucault and Double pass  autocollimation and one test may show it to be a parabola, that being the matching Ronchi test. Double pass has very few sources of error. There is no way I know of that a bad parabola can show good results in this test. A good parabola could only show bad results if it wasn't aligned correctly but that is very easy to see. The Ronchi pattern would not be symmetrical, ie the lines would either bow in one direction like this )))) or show key stoning like this /|\ .  So if you saw  a symmetrical pattern like this  ((|)) on the inside of focus with Double pass automation then the odds are 99.9% sure that your mirror is over corrected by a good amount like your Foucault test also indicated.

  The matching Ronchi pattern test has a couple of sources of error. You need to know the radius of curvature to  a high degree of accuracy and you need to place the grating at an exact position inside the focal plane. If not you can make the pattern match what a good parabola will show when it is not. 

    The simplest answer is that the optics were just made wrong. It happens more then people would want to believe. So I would set the primary back up in double pass and refigure it to a good parabola, then either refigure or make a new secondary. 

 

                      - Dave 


Edited by DAVIDG, 10 September 2018 - 09:44 AM.

  • rcwolpert likes this

#38 Jeff B1

Jeff B1

    Vanguard

  • -----
  • Posts: 2356
  • Joined: 07 Mar 2014
  • Loc: South Central Florida

Posted 10 September 2018 - 12:19 PM

Some more testing results here.

https://guysmathastr...segrain-mirror/https://guysmathastr...segrain-mirror/

(I haven't tried testing the secondary yet.)

Hum, got a anti-virus alert there:  "Scam Insight: Personal Information Risk"



#39 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 621
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 10 September 2018 - 01:58 PM

If the primary is a smooth figure of revolution but overcorrected why not figure the secondary to match, as Dave mentioned above. Would it matter if the finished scope ended up somewhere between a classical Cassegrain and an R.C? In any event you'd need to do the final figuring of the secondary as part of the whole system in DPAC, so why not go straight there?

 

Could the secondary's problems be due to stresses from its cell? (Also mentioned above, I've just seen). A long shot, I guess. Why not replace the focusing assembly with a temporary fixture which will handily take the secondary during figuring?

 

David


Edited by davidc135, 10 September 2018 - 02:06 PM.


#40 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8575
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 10 September 2018 - 03:15 PM

 With a telescope that has been around for many years, that gives poor images and has been  past from one group to the next , it is safe to bet that all the simple stuff has been tried to make it work. The answer is simply, the optics were not made correctly and the fix is to refigure them.

   If you go back to a parabola, you have an excellent null test for it in DPAC. Once it shows a null you know you have a one good surface and the scope can be used as both a Cass. and Newt. Now you can either refigure or make a new secondary and you can easily calculate the figure that is required. A Hindle sphere can be used to null figure it along with DPAC. So again your using null tests which have less sources of error so the odds are much better that the results will be excellent.

   The issue I keep seeing with these large commercial Cass is that the people making the optics just didn't have a good understand of what it takes to make them right. and you had companies claiming that the optics were 1/10 wave i with nothing to back that claim.

     You have a fast primary and the Foucault test has many sources of error. The result is a primary that is not figured to the needed  accuracy. One also has a  convex hyperbola that few have the experience in figuring and you  need the right test equipment to do it right. If not the  the result is a poor image. The poor image is blamed on seeing, alignment, not enough cool down time etc etc until finally people give up and buy a new scope.   

    

                 - Dave 




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: cassegrain, classic, equipment, imaging, optics, reflector, ATM



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics