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

Burch Null Test

  • Please log in to reply
40 replies to this topic

#1 Tucker512

Tucker512

    Mariner 2

  • *****
  • topic starter
  • Posts: 228
  • Joined: 30 Oct 2010
  • Loc: Arizona

Posted 25 February 2015 - 06:40 PM

I came across a null test that I had not seen before and thought it looked interesting: the Burch null test. It uses a spherical mirror instead of a lens to null the test mirror. It seems it would have the distinct advantage of being achromatic, which would eliminate the need for narrowband light sources as needed with the Ross or similar nulls. A quality spherical mirror would also be easier for an amateur to make than a null lens, and easier to verify the quality.

 

On the disadvantage side, the spherical mirror needs to be about 45% the focal ratio of the mirror being tested, which makes it fairly fast.  An f/2 sphere would be needed for an f/4.5 primary. But the diameter of the null mirror can be quite a bit smaller. For example, a 12.5" f/4.5 primary could be nulled with a 3.5" sphere with a 14" radius. Once the radius of the null optic in known, only the spacing between the light source and null mirror, and the null mirror and the primary, need to be adjusted. Similar to any other null test.  The light path can be folded with a flat if needed to clear the test equipment.

 

Just curious if anyone had tried such a thing. Not sure if it's any more or less practical than any other method, I just hadn't seen it before.

 

Regards,
Scott

Attached Thumbnails

  • Burch Null Test.jpg


#2 DAVIDG

DAVIDG

    Fly Me to the Moon

  • *****
  • Posts: 6679
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 26 February 2015 - 09:42 AM

The test your describing is the Waineo Null test http://www.bbastrode...com/waineo.html It was also written up in July 1992 Sky and Tel, starting on page 85.

 

                   - Dave



#3 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 27 February 2015 - 05:57 AM

Dave, C. R. Burch preceded Waineo by a few years.

 

http://articles.adsa...000456.000.html

 

 When this article was written (1936), Tom Waineo was born in that year. ;)  I wouldn't call this the Burch null test, but rather a Burch null test, as he authored literally dozens of different compensator tests for aspherics, the offshoots of which are, in addition to Waineo, also the Dahl and the Ross null tests.

 

Waineo's test is a variation of this Burch Null, with some notable and, to me personally, undesirable characteristics compared to the Burch Null: (1) The light source obstructs the view, and (2) the test also requires a perforated reference mirror. On the other hand, the Burch method  is an off-axis test, which carries potentially its own share of problems with astigmatism and coma in faster mirrors. The Burch test is rather long, so air currents and what not may be an issue as well. 

 

regards,

Mladen

Attached Thumbnails

  • Burch_Null_CN.jpg

Edited by MKV, 27 February 2015 - 08:19 AM.

  • steveastrouk and Gleb1964 like this

#4 DAVIDG

DAVIDG

    Fly Me to the Moon

  • *****
  • Posts: 6679
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 27 February 2015 - 12:01 PM

Waineo also discribed the test without the perforated spherical primary. He displayed his test setup at Stellafane in the late 60's. The '92 Sky and Tel article by our own Ed Jones discribes both methods. Ed used non perforated spherical mirror, which was illuminated by a piece of fiber optics attached to a thin metal vane. That assembly was attached to a threaded rod so one could set the light source to spherical mirror distance precisely.  Sky and Tel provided a BASIC program to calculate the needed spacings.

 

                - Dave



#5 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 27 February 2015 - 02:54 PM

 

Waineo also discribed the test without the perforated spherical primary. He displayed his test setup at Stellafane in the late 60's. The '92 Sky and Tel article by our own Ed Jones discribes both methods

 

Dave, how does Waino's test differ from Burch's? If it's the same test then it should be called the Burch Null test, because Burch published his 30 years before Waineo did his. I believe it is called the Waineo test for a reason -- probably  because it differs sufficiently from that described by Burch (1936), but i'm just guessing.

 

Regards,

Mladen 



#6 Tucker512

Tucker512

    Mariner 2

  • *****
  • topic starter
  • Posts: 228
  • Joined: 30 Oct 2010
  • Loc: Arizona

Posted 28 February 2015 - 04:48 PM

My impression was that in the Burch test the null mirror is located outside the COC of the test mirror, while in a Waineo test, the null mirror is located inside the COC, which then allows for the perforated mirror (or not, depending on spacing between the mirrors).  Otherwise I would think they are similar.  The Waineo would take up less overall length.  Not sure if the null mirror tends to be larger or smaller with one test vs the other.

 

Scott



#7 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 14 December 2015 - 11:44 PM

Waineo also discribed the test without the perforated spherical primary. He displayed his test setup at Stellafane in the late 60's. The '92 Sky and Tel article by our own Ed Jones discribes both methods. 

 

Actually, Paul Valleli, a retired ITEK optical engieneer, and a frequent contributor on the Yahoo Interferometry Group, and someone who personally worked with Tom Waineo wrote to me a little first-hand historical perspective:

 

"Mladen,

 

Despite Tom Waineo being my close associate for many years since 1961, this test should NOT bear his name. It was described by D.D. Maksutov when Tom was an infant, but was not published during WWII. 

 

We learned of the test when ITEK engineers and library researchers described it to their opticians and optical technicians about 1962. Tom made a small demo unit to show the layout mounted on a camera tripod at Stellafane.

 

Tom then spent a lot of time promoting the test to atm's as a reflective version of the Ross Null Test that does not need a  narrow band filter or laser light source. The majority of professionals call it "The Maksutov Null Test for Concave Paraboloids". It is an ideal alternative to autocollimation testing for fast mirrors.

 

Amateurs should not fool around with alternate nomenclature caused by incomplete information or lack of knowledge."

 

I think we can call it the Maksutov (or Just Mak) Null Test and not do it injustice. The reason so few of us know about this is, of course, because Maksutov's work remained largely unknown in the West because of USSR's isolation between the two world wars. However, verifiable contemporary Soviet puclicaitons prove beyond any doubt that D. D. Maksutvo pionneered  many of these tests.

 

Mladen


Edited by MKV, 14 December 2015 - 11:51 PM.


#8 Mark Harry

Mark Harry

    Fly Me to the Moon

  • *****
  • Posts: 6720
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 15 December 2015 - 06:08 AM

"Amateurs should not fool around with alternate nomenclature caused by incomplete information or lack of knowledge."
*******
I agree 100%.
----------------------
That being said, how about the null test layout with concave reference inside the ROC of the mirror being tested? Parameters, such as diameters, spacings etc? How does sensitivity compare to a test with an AC flat?
M.



#9 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 15 December 2015 - 09:50 AM

"Amateurs should not fool around with alternate nomenclature caused by incomplete information or lack of knowledge."
*******
I agree 100%.
----------------------
That being said, how about the null test layout with concave reference inside the ROC of the mirror being tested? Parameters, such as diameters, spacings etc? How does sensitivity compare to a test with an AC flat?
M.

 

I agree, but it is also true that even those quite familiar with optics are not always familiar with the works of D. D. Maksutov, and it is also more a habit to call it (incorrectly) The Waineo Null Test (WNT).

 

Any conjugate test, including the double-pass autocollimation test (ACT), will be sensitive to accurate spacing.. Likewise, accurate alignment of all elements is just as important. The tolerances vary according to the focal ratio of the mirror being tested, but very few conjugate null test tolerances will be as generous as those of the ACT.

 

The real advantage of the Mak Null Test (MNT) is that it is a better test than the Ross Null Test (RNT), because the quality of the nulling Maksutov sphere is known, whereas the quality of the Ross null lens is not -- because for most ATMs its' difficult to accurately test the concave surface of the Ross lens. Also, using a reflection surface avoids any substrate inhomogeneity problems present in refractive optics, especially bargain basement lenses. But the quality of a reflective surface has to be better than that of the reractive. 

 

It is easy to show that a  for moderate apertures the MNT is adequate for mirrors as fast as f/3. For example, the Mak Null Test can correct an 8-inch f/3 paraboloid to about 0.16λ ptv, but a 32-inch f/3 only to about 0.7λ. 

 

By far the best nulling compensator for all diameters, including very large (6 meter) and fast mirrors (down to f/1.25) is the Offner test, and the tolerances are tight but not too tight. People often mention the The Hubble Space telescope fiasco as an example, but that unfortunate error was caused by gross negligence. If anyone is interested in the story of this event here is a link

 

http://people.tamu.e.../658/Hubble.pdf

 

As can be seen in this article, a technician not only used a wrong reflection as reference but to "fix" the problem inserted an ordinary hardware washer as a spacer in the nulling compensator! This is at least 1.2 to 1.5 mm when tolerances range in the microns! It's very unlikely for a meticulous worker to do that under ordinary circumstances or outside ATMs' capabilities.

 

The Offner can be both refractive and reflective. High precision and high-quality melts are required for the refractive model, and the reflective version looks very much like a miniature MNT! :)

 

Regards,

Mladen


Edited by MKV, 15 December 2015 - 09:58 AM.


#10 Ed Jones

Ed Jones

    Gemini

  • *****
  • Posts: 3351
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 15 December 2015 - 10:18 AM

I've used the test and was unaware of how Waineo was using it.  I used it both inside of focus like the Dall which works better with longer F ratios and also like Burch.  I just called it the reflective null in my Sky & Tel article and wrote a short basic program for calculating spacings.  A single mode fiber worked great and the null mirror is a lot easier to make and test than a null lens.   



#11 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 15 December 2015 - 11:33 AM

Ed, Tom Waineo was using the Maksutov spherical null mirror with a hole in it

 

http://www.bbastrode...aineotest1.jpg 

 

The configuration is also described by Vla on his site

 

http://www.telescope...o_null_test.htm

 

A single mode fiber is used in many papers published on the subject, even a dual fiber method which results in Young's interference pattern (the only problem with that method is the fact that the mirror diameter cannot be accurately assessed without some "backpropagating" the wavefront to the test mirror, which is something most of us ATMs will not bother with).

 

Regards,

Mladen



#12 df_2112

df_2112

    Sputnik

  • *****
  • Posts: 49
  • Joined: 28 Nov 2007

Posted 15 December 2015 - 11:38 AM

What would the layout of the reflective Offner test look like and what spec/size would the elements be? Would it be better for larger mirrors?

#13 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 15 December 2015 - 12:53 PM

The all-reflective Offner compensator looks just like a miniature Maksutov/Waineo two-mirror setup.

 

http://www.pointsdev...es/fig_6_13.png

 

Note: this is not necessarily drawn to scale. 

 

Both Offner types are very effective at nulling large and fast mirrors. A refractive type Offner was used with a Shack-cube Fizeau compact interfeormeter to null a 6.5 meter(256-inch)  f/1.25 mirror.  

 

http://www.loft.opti...IE 2199 658.pdf

 

Regards,

Mladen


Edited by MKV, 15 December 2015 - 12:54 PM.


#14 saemark30

saemark30

    Apollo

  • -----
  • Posts: 1468
  • Joined: 21 Feb 2012

Posted 15 December 2015 - 06:42 PM

Can a laser diode be used as the light source in place of the single mode fibre?



#15 Ed Jones

Ed Jones

    Gemini

  • *****
  • Posts: 3351
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 15 December 2015 - 08:24 PM

A laser diode in the light path will create some amount of heat that may cause heat  currents. 

 

The Burch layout can also add a field lens to make it an Offner, it doesn't need to have 2 perforated mirrors.  It does end up being a pretty fast F ratio.

 

Ed



#16 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 15 December 2015 - 11:04 PM

 

The Burch layout can also add a field lens to make it an Offner, it doesn't need to have 2 perforated mirrors.

The Burch layout is an off-axis configuration, resulting in astigmatism. This would be particularly offensive for faster mirrors, and a field lens would do very little to remedy this problem. The field lens is primarily responsible for removing 3rd and 5th order OPD, but nto astigmatism. As a general rule, for fast optics an on-axis test is needed to avoid predominant astigmatism. This is why the Bath is limited to about f/3 mirrors of moderate aperture.

 

Mladen



#17 Ed Jones

Ed Jones

    Gemini

  • *****
  • Posts: 3351
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 16 December 2015 - 10:21 AM

Mladen,

  Yes there is astigmatism from being offset just as there is in the KE test  You can see in Dave's raytrace it returns as a line image which nulls nicely as it does the KE test.  You would want to keep the separation as small as possible but you evaluate astigmatism by other tests.

 

Ed



#18 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 16 December 2015 - 01:10 PM

Ed, first the raytrace (spot diagram) is not Dave's, but mine. Second, the mirror in question is a 7.3 -inch f/8.7, so the effects (mostly astigmatism) of the off-axis configuration are present but "below the radar". Null tests for paraboolids have their greatest value when the mirror is large and fast because at that point the conventional tools (even the Bath interferometer) become less accurate, and a large astigmatism has a lot to do with that. Try the same Burch off-axis configuration for, say a 12.5-inch f/3 paraboloid and see what happens. 

 

regards,

Mladen


Edited by MKV, 16 December 2015 - 01:13 PM.


#19 Mark Harry

Mark Harry

    Fly Me to the Moon

  • *****
  • Posts: 6720
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 16 December 2015 - 04:02 PM

 A 7" F/8.7 is a slam-dunk.
*******

Use another shorter focus mirror to do any evaluation as to how too proceed. You're loading this discussion as to what's acceptable.
M.
(PS: this is pre-school!)



#20 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 16 December 2015 - 06:55 PM

Mark, can you be less cryptic? 

 

Thanks!

Mladen


Edited by MKV, 16 December 2015 - 06:57 PM.


#21 Ed Jones

Ed Jones

    Gemini

  • *****
  • Posts: 3351
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 16 December 2015 - 07:09 PM

Mladen,

   My mistake I thought Dave posted it.  Anyway here is a 12.3 inch F/2.8 using a one mirror reflective Offner null with about 1/2 inch separation which ought to be enough.  There is astigmatism but the KE will only see a null of less than 1/100 wave.  This is the same situation as in the Foucault KE with a sphere.   One issue I see is that the numerical aperture of a fiber might not be high enough to illuminate the test mirror, you might need to use something like a laser diode without a lens.

Attached Thumbnails

  • Burch2.JPG


#22 Mark Harry

Mark Harry

    Fly Me to the Moon

  • *****
  • Posts: 6720
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 17 December 2015 - 07:15 AM

Is there a way for the test you post to work without the field lens? Could the reference sphere be modified to work without it?
At this point, I assume to eliminate the lens, the reference sphere would have to have a conic, correct? (confocal paraboloid?)
M.



#23 Ed Jones

Ed Jones

    Gemini

  • *****
  • Posts: 3351
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 17 December 2015 - 08:56 AM

Mark,

  With no field lens the null mirror could be an ellipse but the N.A. of the light source gets even bigger.  A field lens is much much easier.  You could also make on axis using a good beam splitter/ lens combination.

 

Ed



#24 MKV

MKV

    Fly Me to the Moon

  • *****
  • Posts: 5179
  • Joined: 20 Jan 2011

Posted 17 December 2015 - 09:41 AM

Ed, illuminating a 4-inch mirror from a distance of about 6.5 inches is a challenge (in this case it's like testing a 4 inch f/1.6 at its ROC), and I'm not sure a lensless laser diode would be suitable. I don't have one handy (I'm remodeling my house, so all the "goodies" are in storage), but it's easy to test it if you have a diode with a removable lens.

 

A beam splitter will result in spherical aberration (unless the beam passes through it as a plane wave). For converging/diverging cones of light, a Weinstein beam splitter is needed. Fortunately,  It's fairly easy to make (add two PCX lenses to the splitter cube) to correct the SA. But then you're adding more refractive surfaces and increased the possibility of error due to substrate inhomogeneity. 

 

However, a suitable high quality small flat mirror positioned at 45 degrees relative to the optical axis could divert the nulling mirror away from the return beam and make the focus more accessible.

 

Unfortunately, I don't get your spot results unless I use a very small tilt. In order for the offset to be 1/2 inch as you said, the tilt angle is 0.411 degrees which results in image degradation in my OSLO version. perhaps you entered the tangent of the angle (0.5/69.595 = 0.00718) vs the angle itself, i.e. arctan(0.00718). Your printout doesn't seem to show the angle or the wavelength used. 

 

Regards,

Mladen

 

EdJones_reflector_Offner w fld lns.JPG



#25 Ed Jones

Ed Jones

    Gemini

  • *****
  • Posts: 3351
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 17 December 2015 - 09:54 AM

Mladen,

The mirror was tilted about .19 degrees giving .47 inch separation between the beams @ .55 nm although the field lens introduces very little color so white light would probably work.  I optimized in X only.  You can account for a beam splitter in the optimization if you wanted to use one.  The biggest challenge is illuminating the fast mirror.

 

Ed




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