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

Ross Null musings

  • Please log in to reply
17 replies to this topic

#1 howlforstate

howlforstate

    Lift Off

  • *****
  • topic starter
  • Posts: 18
  • Joined: 18 Jan 2017

Posted 16 April 2018 - 10:22 AM

Good morning, 

 

I am still in the thought process of designing a testing regiment I feel accurate enough for me, a novice to master relatively quickly.  This has brought me to null testing, specifically Ross null testing.  Of the seemingly infinite wisdom afforded by these forums I was hoping that one of you might be able to answer whether or not the lens I have found would be applicable to the task.  Granted the lens I have found is small (25mm) but seeing as I have yet to receive a response from BMV Optical the company Ceravalo now recomends this might be my only hope.  The lens I am considering is linked below.  It satisfies all the requirements in the Ross Null Finder software with a working lens diameter of 19mm and has a 1/20th wave surface accuracy on fused silica. 

 

 https://www.edmundop...pcx-lens/#specs

 

I am just curious to see if maybe I missed something or am not considering something really important in ross null testing.  Thank you all for your help in advance. 

 

Ryan



#2 TOMDEY

TOMDEY

    Apollo

  • -----
  • Posts: 1406
  • Joined: 10 Feb 2014
  • Loc: Springwater, NY

Posted 16 April 2018 - 10:42 AM

If you ran the nominal test configuration thru software that confirms a good null, then this is promising. We did almost entirely null-testing at work, and it does make interpretation of the data vastly easier. A friend built his 16-inch using a Ross singlet. I tested the finished mirror in autocollimation and it was wonderful, confirming his test config as well as his skill walking around the polishing barrel. With nulling optics, proper alignment of the test is critically-important, else you inadvertently polish those errors into your mirror. The Hubble Flight PM is the textbook example of that little error resulting in a ~perfectly~ atrocious mirror!  You can prevent that possibility by executing one "sanity-check" test both with and without the null, looking for same answer.  Tom


  • starcanoe and howlforstate like this

#3 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 6964
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 16 April 2018 - 11:54 AM

Ryan, that lens could easily be workable depending on the mirror you plan to test with it. What's the mirror diameter and radius of curvature?


  • howlforstate likes this

#4 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 20728
  • Joined: 22 Feb 2010
  • Loc: Montclair, NJ

Posted 16 April 2018 - 12:33 PM

 

I have yet to receive a response from BMV Optical the company Ceravalo

$600

 

 

I am just curious to see if maybe I missed something or am not considering something really important in ross null testing.

yep. the image circle for the lens, for the mirror your testing. you should be using up to 80% of its diameter.

 

also the longer the focal length the more sensitive the test. so for fast mirrors [f3's] a 600mm fl isn't bad

 

ceravolo's is a 80x350 1/10~ surface.

 

you could go to optosigma and ask them to hand pic a stock 1/4~ pcx. or make a 1/10~ but that's not stock. $$$$


Edited by Pinbout, 16 April 2018 - 12:38 PM.


#5 howlforstate

howlforstate

    Lift Off

  • *****
  • topic starter
  • Posts: 18
  • Joined: 18 Jan 2017

Posted 16 April 2018 - 12:37 PM

MKV,  The mirror I plan to test is 14.7 in f/4.5-.7 ish I haven't fully committed to the f/ratio yet.  The only thing kind of odd that I noticed was at those specs the object to lens distance was just over an inch or so.  Lastly if anyone else could explain to me how the maximum optical path difference in waves relates to the accuracy of the test that would be great.  I'm currently coming up with 3.142 waves which seems like it might be a problem? Actually posted below are my readings if anyone would be willing to interpret.  Im assuming we want to minimize the maximum optical path difference no?

 

Results:
Object to Lens Distance           = 27.316 mm
Lens to Mirror Distance           = 1664.155 mm
Working Diameter of Lens         = 20.668 mm
70% Zone Spherical Aberration    = -0.204 mm
Marginal Spherical Aberration    = 0.001 mm
Quarter-wave Tolerence           = 0.016 mm
Maximum Optical Path Difference  = 3.142 Waves

 

Danny,  thats what I was worried about I know the price is probably worth it but this lens is half the cost and if I could make it work I'd be happy as a clam.  This is still an experimental hobby for me I'm not sure I'm comfortable with the price tags involved yet haha


Edited by howlforstate, 16 April 2018 - 12:39 PM.


#6 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 20728
  • Joined: 22 Feb 2010
  • Loc: Montclair, NJ

Posted 16 April 2018 - 12:47 PM

 

This is still an experimental hobby for me I'm not sure I'm comfortable with the price tags involved yet haha

then get this if your just experimenting

 

https://america.opto.../bk7-25061.html

 

 

this will get you use to using it and get you real close, and check with other methods also...since your just learning it.

 

I think its a nice test, easy to setup, and easy to look at roc then move everything up a bunch and test with the lens.



#7 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 20728
  • Joined: 22 Feb 2010
  • Loc: Montclair, NJ

Posted 16 April 2018 - 12:48 PM

 

Lens to Mirror Distance           = 1664.155 mm

that's all you need to concern yourself with when measuring.



#8 howlforstate

howlforstate

    Lift Off

  • *****
  • topic starter
  • Posts: 18
  • Joined: 18 Jan 2017

Posted 16 April 2018 - 01:18 PM

Im so full of questions.... First when I inputted the lens you referenced it gave me a working lens diameter larger than that of the lens itself?  Lastly would it be easier and cheaper to just build a bath IF and would that give better accuracy? I just want the best readings so Im not figuring into a hole or end up with a hubble.



#9 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 6964
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 16 April 2018 - 01:20 PM

then get this if your just experimenting

 

https://america.opto.../bk7-25061.html

 

 

this will get you use to using it and get you real close, and check with other methods also...since your just learning it.

 

I think its a nice test, easy to setup, and easy to look at roc then move everything up a bunch and test with the lens.

The referenced lens doesn't specify surface accuracy. The Ceravolo lens is 1/10 wave which is like 1/20 wave mirror, the OP's 1 inch lens is 1/20 wave, which is like 1/40 wave mirror. My concern is that 1/4 wave tolerance is ±0.016 mm.  Can you set the separation distance to that accuracy?  What's the point of using a lens with unknown surface accuracy and such tight tolerance? BTW, have you ever had your mirrors that you nulled with your Ross test verified by an interferometer?

 

OP, a 600 mm f.l Cervaolo's lens has a 1/4 wave tolerance of 1/8 a mm (0.126 mm), which is 8 times larger. Usually, if you can measure the distance from the lens to the mirror within 1 mm you're doing good. 


Edited by MKV, 16 April 2018 - 01:24 PM.


#10 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 6964
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 16 April 2018 - 01:23 PM

Lastly would it be easier and cheaper to just build a bath IF and would that give better accuracy?

Yes! The Bath will cost you about $60. The software is free. Join Dale's interferometry group at 

 

https://groups.io/g/Interferometry

 

and you'll get plenty of support and advice form experienced ATMs and real pros.



#11 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 20728
  • Joined: 22 Feb 2010
  • Loc: Montclair, NJ

Posted 16 April 2018 - 01:45 PM

Im so full of questions.... First when I inputted the lens you referenced it gave me a working lens diameter larger than that of the lens itself?  Lastly would it be easier and cheaper to just build a bath IF and would that give better accuracy? I just want the best readings so Im not figuring into a hole or end up with a hubble.

not the same to work the mirror with as a instant test.



#12 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 20728
  • Joined: 22 Feb 2010
  • Loc: Montclair, NJ

Posted 16 April 2018 - 01:47 PM

 

The referenced lens doesn't specify surface accuracy. The Ceravolo lens is 1/10 wave which is like 1/20 wave mirror, the OP's 1 inch lens is 1/20 wave, which is like 1/40 wave mirror. My concern is that 1/4 wave tolerance is ±0.016 mm.  Can you set the separation distance to that accuracy?

I hear how you feel about it. but seams like your too caught up in numbers and the lens will help test the shape, the rest is just correction. and he can verify with other tests.

 

using if everytime you go around just 1x test 2x around test...becomes not very fluid with IF.


Edited by Pinbout, 16 April 2018 - 02:47 PM.


#13 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 20728
  • Joined: 22 Feb 2010
  • Loc: Montclair, NJ

Posted 16 April 2018 - 02:47 PM

Yes! The Bath will cost you about $60. The software is free. Join Dale's interferometry group at 

 

https://groups.io/g/Interferometry

 

and you'll get plenty of support and advice form experienced ATMs and real pros.

why not both...test and verify.



#14 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 6964
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 16 April 2018 - 02:51 PM

@OP -- personally I would go with the Ceravolo's lens. If you decide not to use it, you can sell and it will snatched. Such lenses are hard to come by. It can be used as an Offner nulling compenstaor with a Fizeau or Twyman-Green interfeormeter for testing very large and very fast mirrors. The little lens is not likely to generate much interest, imho. 

 

With a Bath you can take several igrams quickly and run them through the software as a batch, so using an IF to keep track of progress in figuring is quick and easy. One of the IF group member (Arjan) is doing it, and so are others. Look him up on CN to get an idea.

 

Good luck. 



#15 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 6964
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 16 April 2018 - 02:57 PM

I hear how you feel about it. but seams like your too caught up in numbers and the lens will help test the shape, the rest is just correction. and he can verify with other tests.

 

using if everytime you go around just 1x test 2x around test...becomes not very fluid with IF.

He can use a Foucault test for free to get an idea of the shape, but using an IF to track progress is a lot easier and shows you the whole surface. The problem with a lens of unknown surface accuracy and dubious distance measurements is that you're adding those errors into the mirror. That's how the Hubble ended up being "myopic" even though it tested as a null.


  • kfrederick likes this

#16 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 20728
  • Joined: 22 Feb 2010
  • Loc: Montclair, NJ

Posted 16 April 2018 - 05:10 PM

Not as fluid either, let alone judging shadows.

 

the Ross null requires judging as well as dpac. The best way to judge with the nulls is to flip inside outside RoC / focus quickly to see any difference in slope. 

 

And the ke your still inputting measurements into a computer. Actually nulling zones in steps so you don’t get a overall image of the mirrors shape in one view.



#17 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 20728
  • Joined: 22 Feb 2010
  • Loc: Montclair, NJ

Posted 16 April 2018 - 05:23 PM

 

dubious distance measurements is that you're adding those errors into the mirror. That's how the Hubble ended up being "myopic" even though it tested as a null.

Tests never add errors to a mirror 

 

the laser on the Hubble reflected off the wrong surface (a nulling if test not ronchi), can’t have that on a Ross setup. Ceravolo’s anecdote would be a better ref for setting the distance incorrectly.

 

the Ross setup for a 14 f4.5 has a tol in setup that’s not hard.

 

but your missing the point. Using the null is a very fluid, quick and instant view to help judge how to shape the surface. Get the shape correct the rest is just correction, and you can take a longer time with the ke or if or star test to verify.

 

I have no more to say about the benifits of the Ross null test. Either you can appreciate the process while making a mirror or...

 

when was the last time you made a mirror using these tests? Seriously.


Edited by Pinbout, 16 April 2018 - 05:26 PM.

  • DAVIDG likes this

#18 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 6964
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 18 April 2018 - 11:12 AM

Tests never add errors to a mirror 

 

the laser on the Hubble reflected off the wrong surface (a nulling if test not ronchi), can’t have that on a Ross setup. Ceravolo’s anecdote would be a better ref for setting the distance incorrectly.

 

the Ross setup for a 14 f4.5 has a tol in setup that’s not hard.

 

but your missing the point. Using the null is a very fluid, quick and instant view to help judge how to shape the surface. Get the shape correct the rest is just correction, and you can take a longer time with the ke or if or star test to verify.

 

I have no more to say about the benifits of the Ross null test. Either you can appreciate the process while making a mirror or...

 

when was the last time you made a mirror using these tests? Seriously.

Errors of the test device add to the wavefront. If you're nulling a mirror/lens against a wavefornt which also has errors of the test device/setup you'll null it but the mirror/lens true figure will not be of the figure yo're shoot for. 

 

If you use a flat in a DPSC but it actually unbeknownst to you contributes 1/5 wave error on the wavefront (due to either the flat being faulty or the setup or both), and you null your mirror, that mirror will in reality be off by 1/4 wave off on the surface -- hardly a "perfect paraboloid".

 

Sure, using the null is a quick and instant picture of the wavefornt, but unless you can discount the errors of the nulling element from that "null" that null is misleading.

 

You seem to neglect the fact that the Ross null test is not a double pass (double sensitve) test for the test optic, so a Ronchi Ross test is hardly a good "judge" as you say. But since the Ross lens gets traversed twice it has to be of exceptional optical quality -- because any errors from the lens will be doubled!

 

Thus, a null is valid only if you absolutely know with certainty that your test device does not contribute greater error than what you're shooting for. If your "Ross" lens is of unknown surface and figure quality, or your setup error is unknown, you can't make an assumption that the null you're observing reflects the correction on the test optic.

 

As for the last time I made optics using the Ross null test the answer is never. I would never make any optic using a "Ross" lens of unknown quality, and since I don't have a certified test lens, it's pointless. And so is a ":quick look" with a Ross using a lens of unknown quality.

 

That's why I asked you when was the last time you a mirror using your Ross test and had that mirror tested with an inteferofmeter to see just how good it is? Seriously.  To me it makes more sense to monitor the progress using a knife-edge for a "quick look" and/or an IF.




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






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics