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

Making a Fizeau Reference

ATM
  • Please log in to reply
35 replies to this topic

#1 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 28 February 2020 - 06:05 PM

Hi,

 

I am making a fizeau reference based on the ceravolo fizeau interferometer. So far I have made a 3" (76mm) test plate and the reference. Now I am in the testing phase and ran into the chicken or the egg problem. since I don´t have  access to a interferometer to check the test plate. Together I get resonable fringes, but its hard to deduce if the errors I see are in the test plate or on the reference. Has any one any tips they can give me so that I can get a better surface. I am attaching a interferogram of the reference on the test plate.  The reference convex is a steep F1 curve!

 

Thanks,

 

Alfredo

Attached Thumbnails

  • Test2.jpg


#2 Benach

Benach

    Vanguard

  • *****
  • Posts: 2,250
  • Joined: 24 Jan 2008

Posted 28 February 2020 - 06:16 PM

Ritchey common test or Foucault via flat measurement.
  • Eikonal likes this

#3 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 28 February 2020 - 06:18 PM

Hi Benach,

 

This is not a flat! It´s a F1 convex reference and concave test plate! Foucalt does not work at F1. Or at least my Focault tester.

 

Alfredo

Attached Thumbnails

  • fizeau.jpg

Edited by Alfredo, 28 February 2020 - 06:29 PM.


#4 Ed Jones

Ed Jones

    Soyuz

  • *****
  • Posts: 3,923
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 28 February 2020 - 06:34 PM

If you decenter one to the other you can differentiate zones pretty easily which are concentric to each part.  Rotating one part to the other may tell you which part has the astigmatism or maybe just which is greater.  


  • Eikonal likes this

#5 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,817
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 28 February 2020 - 06:40 PM

Unfortunately there is no absolute method of calibrating the convex reference sphere of a Ceravolo interferometer.

Convex reference spheres are best tested against a matching concentric concave spherical surface.

There are absolute methods that can be used to calibrate a concave sphere.

For such a fast sphere the random ball test is effective, however it requires an interferometer with an acessible external focus and a working distance great enough to accommodate a silicon nitride ball with a diameter of 25mm or greater in an autostigmatic setup.

 

Another option is a dual pinhole point diffraction interferometer setup like that used by Difrotec:

 http://www.difrotec.com/product

 

However that uses a pair of closely spaced submicron pinholes in a 45 degree plane mirror plus a internal single mode fibers and polarisation optics etc.

 

Its also possible to add a low working distance finite conjugate microscope objective or its equivalent to a split path PDI to measure an F/1 concave sphere.

The long working distance finite conjugate microscope objective facilitates calibration of interferometer errors via the random ball test. Suitable silicon nitride balls are readily available.



#6 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 28 February 2020 - 07:01 PM

Ed, decentering and rotating is what I have been using and it does help. I used focault and ronchi (200 lpi) to get the test plate as spherical as possible and then matched the convex reference to it. But its not that easy to figure out where the defect is, and at the moment I am only working on the reference, I am afraid of getting into a mess if a retouch both the reference and the sphere.

 

I like the PDI interferometer and will look into your suggestion, but I want to concentrate first on getting the fizeau to work before investing in making a PDI. I did try making a PDI plate with sooth and compresed air to create the pinholes, but I rather wait until I can afford to get a PDI plate from Michael in Germany.



#7 TOMDEY

TOMDEY

    Fly Me to the Moon

  • *****
  • Posts: 5,970
  • Joined: 10 Feb 2014
  • Loc: Springwater, NY

Posted 28 February 2020 - 07:49 PM

Hi, Alfredo; Nice project! Jim Wyant describes absolute cal of spheres here >>>

 

https://wp.optics.ar...calSurfaces.pdf

 

He references (one of) my Boss's earlier papers (Art Jensen). Prior to that, some of the literature had logically-incorrect approaches presented. Something to do with the sign of terms and odd aberrations that would not separate out properly, I believe. The odd Zernikes are axial coma etc. etc. Anyway, Jim's paper is a good read, useful and interesting. I do vaguely remember Art sitting at his desk with this matrix of orientations and translations that would parse out all Zernikes to odd order N... and sure looked cumbersome to me. But was analytically complex, with none of the underlying assumptions regarding absoluteness. In that sense, it was mathematically rigorous... but painfully difficult to execute properly. So we used it to calibrate only a few interferometers that way, which  we would then use to calibrate the production-floor metrology equipment (Kodak).

 

Pragmatic would be to seek someone outside who has a NIST-traceable rig to qualify yours.    Tom



#8 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 28 February 2020 - 08:00 PM

Hi Tom,

 

Will be sure to read the article, thanks. Unfortunately I live in the third world, where interferometers are dificult to come by, that is why I decided to build my own. At the moment I do not even have the interferometer built. The fringe foto are from placing the convex reference on the test plate. The convex reference has a aplanatic back so I can do the test without fringe distorting due to the F1 curvature.

 

I was imagining some  way to use center 1" of the convex reference to do sub-aperture interferograms of the test plate and some how using zernikes deduce the test plate profile. But I still have not figured out how to do this.



#9 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,817
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 28 February 2020 - 09:08 PM

Stitching interferometry could be used to compare the centre of the convex reference with the entire surface of a concave testplate but that is both mechanically and computationally complex.

 

When testing an F/1 sphere by PDI the diffraction limited point test source should be close to its image formed by the test surface. This is difficult to achieve without introducing additional aberrations due to the use of beamsplitters etc.

 

However if the PDI operates at F/10 or slower together with a finite conjugate microscope objective with adequate working distance is used to transform the F/10 beam to F/1 or faster then on axis PDI is practical as the random ball test can be used to calibrate the errors of the microscope objective and interferometer including any additional optics such as beamsplitters etc.

 

Even with an aplanatic back distortion is incurred when imaging a spherical surface onto the flat surface of the image sensor unless the camera lens is designed to correct such distortion. Done correctly the linear displacement of the image on the image sensor of a point on the test surface from the image of the center of the test surface will be proportional to the sine of the field angle of the chief ray. Otherwise correction needs to be made for the effect of such mapping distortions. 

 

A ball bearing can also be used as the ball in the RBT ( random ball test) but they are easily damaged unlike silicon nitride balls.

 

The RBT is mechanically less complex than the 10 axis stage required by Jensen's method.

A 3 axis stage will suffice to place the ball (sitting in a kinematic nest formed by 3 smaller balls) so that the center of the ball is coincident with the focus of the microscope objective.

 

It may be possible to use something like an eyepiece ( with low exit pupil aberrations) instead of an expensive long working distance microscope objective but raytracing  would be needed to check the performance of a candidate eyepiece design in that application. 



#10 PEterW

PEterW

    Vanguard

  • *****
  • Posts: 2,002
  • Joined: 02 Jan 2006
  • Loc: SW London, UK

Posted 29 February 2020 - 06:01 AM

To get flats grind 3 plates together in rotation to eliminate any surface shapes other than flat being generated. You could also look to use a liquid surface that’s big enough that the meniscus doesn’t affect you... mercury is best! Small high grade flats aren’t too costly. Good luck.

Peter

#11 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 29 February 2020 - 06:24 AM

BGRE,

 

So maybe I can use software to remap the fringes generated from the spherical surface to a flat surface, thus eliminating the error before using a fringe analyses. 

 

I also like the idea of converting the F1 cone of light to F10, I will give it a try when I build the PDI.

 

Alfredo



#12 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,817
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 29 February 2020 - 07:14 AM

Alfredo

 

Yes software correction is possible.

Its relatively easy to derive the mapping correction for spheres and other conicoids.

Its also necessary to correct for any camera lens distortion when imaging a plane surface.

 

 

A Zygo paper outlines a method of correction using measurements obtained for various tilts of the test surface.

 

A 28mm RKE eyepiece is only good for light cones of about F/6 or slower.

 

An F/1 light cone requires a microscope objective with an NA of at least 0.5 or so.

 

In principle a coherently illuminated dual single mode fiber interferometer (has very small errors ~ 1/1000 wave ptv or less) could be used to measure the errors in an F/1 sphere using the stitching method.



#13 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 29 February 2020 - 09:43 AM

Can you give zygo paper title. I searched but did not find it.



#14 Ed Jones

Ed Jones

    Soyuz

  • *****
  • Posts: 3,923
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 29 February 2020 - 11:19 AM

Out of curiosity why do you need a large F/1 sphere?



#15 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 29 February 2020 - 12:37 PM

Hi Ed,

 

I wanted to be able to test up to F3 optics. It was made 3" in diameter with 6" radius and the original plan was to cut out the 1" center.

 

Alfredo


Edited by Alfredo, 29 February 2020 - 12:41 PM.


#16 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,817
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 29 February 2020 - 05:02 PM

If you are testing F/3 optics at CoC then the test beam is F/6.

F/6 test beams are easily measured with a Bath interferometer.

A 28mm RKE eyepiece can work well as a finite conjugate microscope objective for F/6 test beams and the working distance (with modified mount) is adequate for RBT calibration with a 25mm (or larger) diameter ball.

The optical design parameters can be downloaded from the Edmund optics site (NB the Zemax zmx file uses unicode and not ascii characters - conversion may be required -notepad can do this).

 

A Hastings triplet (f = 1") also works well as a monochromatic microscope objective at F/6. The working distance is larger allowing a 38.1 mm diameter ball to be used.

 

Since chromatic aberrations are unimportant with a monochromatic source simpler (and cheaper??) designs should be possible for F/6 beams.  


Edited by BGRE, 29 February 2020 - 06:02 PM.


#17 Arjan

Arjan

    Apollo

  • -----
  • Posts: 1,110
  • Joined: 21 Jan 2009
  • Loc: Netherlands

Posted 29 February 2020 - 05:28 PM

Alfredo, any reason not to build a Bath?

#18 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 29 February 2020 - 07:06 PM

I've built a bath before and it works. Just wanted a good challenge, and  I like the fizeau because its on axis and can do fast optics without having to cancel out astigmatism in the software.

 

Today I worked on the reference and got quite a good result. Just my pride is hurt from a few sleeks (maybe scratches, not sure). Well they do not interfere with its use as a reference and they are outside the center 1.5". So I am happy.


Edited by Alfredo, 29 February 2020 - 07:06 PM.


#19 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,817
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 29 February 2020 - 07:59 PM

Since the reference surface isn't in focus at the image sensor, some smoothing of small reference surface features will occur.



#20 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 29 February 2020 - 08:35 PM

HI,

 

Just took three fotos using only 1.5" (40mm) of the reference. Placed it on the center, left and right positions on the test plate. Does not look too shabby!  DFTFring shows very good values. Dont want to really beleive the values, but at least I am achieving my goal of 1/20 wave, which would alow me to test wave fronts up to 1/10 wave. Still got to do alot more work to confirm what I got. 

 

For those interrested I am using a 405nm laser diode.

 

I have to get a better camera setup. I am using a hand held camera and its very difficult to get a good image.

 

I think I will stop for now and build the interferometer. Than I can use the other tecniques suggested in the thread to validate the reference. Maybe even make a new reference from scratch!

Attached Thumbnails

  • fizeau 40mm Centered.JPG
  • Fizeau 40mm Left.JPG

Edited by Alfredo, 29 February 2020 - 08:46 PM.

  • LarsMalmgren and Pinbout like this

#21 Alfredo

Alfredo

    Lift Off

  • -----
  • topic starter
  • Posts: 15
  • Joined: 09 May 2014
  • Loc: Curitiba, Brasil

Posted 29 February 2020 - 08:35 PM

and the right side image.

Attached Thumbnails

  • Fizeau 40mm Right.JPG

  • Pinbout likes this

#22 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,817
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 29 February 2020 - 08:45 PM

If a non contact "testplate" is used in a Fizeau setup then the non radially symmetric errors can be averaged out by rotating the tesplate and averaing over a large number of rotational orientations. This just leaves the radially symmetric errors of the test setup.

 

Alternatively, measuring the errors of the concave testplate is relatively straightforward with a suitable interferometer (not a Fizeau with a convex reference surface).



#23 MKV

MKV

    Cosmos

  • *****
  • Posts: 8,467
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 29 February 2020 - 11:16 PM

I've built a bath before and it works. Just wanted a good challenge, and  I like the fizeau because its on axis and can do fast optics without having to cancel out astigmatism in the software.

 

Today I worked on the reference and got quite a good result. Just my pride is hurt from a few sleeks (maybe scratches, not sure). Well they do not interfere with its use as a reference and they are outside the center 1.5". So I am happy.

Why not build a Twyman-Green IF? It's on-axis and the reference surface is easy to procure?

 

T-G_LUPI.jpg

 

D63_R300_1 TGL_LR.JPG



#24 BGRE

BGRE

    Mercury-Atlas

  • *****
  • Posts: 2,817
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 29 February 2020 - 11:25 PM

It can also be calibrated using the RBT as long as the F# of the test beam isn't slower than about F/8 and the working distance is greater than the radius of the calibration ball.

 

One drawback is that the laser needs a long coherence length whereas a common path interferometer only needs a short coherence length source. 



#25 MKV

MKV

    Cosmos

  • *****
  • Posts: 8,467
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 01 March 2020 - 11:10 AM

I used an inexpensive DPSS green laser diode I purchased on eBay, a simple plano-negative lens and a small achromat for a beam expander, and two 25 mm surplus Hastigns triplets I bought years ago at Edmund's, the beamsplitter is a 25 mm cube from Surplus Shed, and the reference surface is a portion of as larger 1/10 wave flat.

 

Originally, I wanted to use my HeNe laser but the whole thing was too bulky.

 

twyman_Green_IF_2a.jpg

 

Mladen


  • PrestonE and Gleb1964 like this


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: ATM



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics