Test Reports and Airy Discs

by Roland Christen

At the risk of inviting lots of questions, I am posting some test
reports of the present production run of 130F6 lenses. I am also
posting actual images taken last night in my observatory of Vega
using the two 130F6 lenses in the report.

These two lenses represent the lower edge and middle of the distribution
curve. A third lens, which I did not have time to set up in the
observatory will be the top of the distribution curve with a P-V of
.039 P-V. This represents the extreme edge. Once the data gets below
1/10 wave, there is less confidence that it is an accurate number.

Lower Edge Lens Report and Airy Patterns
Middle Lens Report and Airy Patterns
Top Lens Report

I use this test data to do the final figuring on the lenses. After
the lenses are polished and all surfaces matched to the test plates,
they are assembled and placed on the interferometer. From the
interferogram data, I can determine the correction needed to bring
the lens to the spec level, and can figure one of the surfaces to
bring it into spec. For each lens, this process can take anywhere
from 2 hours to all day, depending on how well my pitch laps are
working.

The software has many more ways to present the data, but I use only
the two graphs, one showing the fringes, and the other showing the
surface contour. In the past, I used just a few fringes in the
interferogram to get good sensitivity of the test. The new software
from Diffraction Limited requires 6 to 7 fringes to operate properly,
and automatically places 150 to 250 data points on the fringe centers
to generate the wavefront map. While these fringes may look like
those generated by a Ronchi grating, there is no similarity at all
between them. Interference fringes are the result of wavefront
interference between the returning beam from a lens and a reference
surface. The sensitivity of this method is many times that of a
Ronchi pattern. Where a Ronchi might show smooth, equidistant lines,
the interference fringes show every tiny bump and hollow. They also
show the tiniest dust specs on all the optical surface as noisy dark
and light specs with their own interference patterns surrounding
them. The software is able to reject most of this noise and produce
an accurate wavefront profile. The repeatibilty of the test is within
10% for wavefront errors of 1/12 wave, which is plenty accurate. In
my test setup, the probing laser light is green at 543.5nm, and
passes through the lens twice (double pass), in order to increase the
sensitivity of the test.

The Airy disc patterns were imaged last night with my Olympus C-3040
digital camera. I used a Pentax 5.2 XL eyepiece and an AP 2x barlow
inserted into a Maxbright diagonal for an approximate power of 270x.
I could see a very nice Airy disc with one faint diffraction ring and
a hint of a second one on Vega. The seeing last night was extremely
steady for this aperture, so it was an ideal night for this kind of
imaging. The camera was put on 3x zoom and the resultant power at the
chip was 810x. At this magnification, the camera took images with 1
second exposures (with me holding the lens up to the eyepiece) so the
ring pattern is not as sharp and contrasty as what was shown in the
eyepiece. I averaged about 10 individual images for each pattern to
reduce the noise and to produce an image that very closely resembles
what is seen in the eyepiece.

Two things are immediately evident. The inside (left) and outside
(right) Fresnel patterns are not the same. The in-focus Airy discs of
the two lenses are virtually identical, and I doubt that anyone would
be able to tell them apart from even the best lens in this run. While
the lenses are nulled in the green, the inner and outer Fresnel
patterns show the effect of sphero-chromatism at the ends of the
spectrum (blue is overcorrected as shown in the inside pattern, red
is undercorrected as shown in the outside pattern). Also shown is the
effect of a slight amount of secondary chromatic aberration which
causes the outer pattern to have fainter Fresnel rings than the inner
pattern - the color correction is so close that there is constructive
and destructive interference of the pattern by the various colors.
However, this lens is made to be used in-focus, and the Airy pattern
shows a nice central peak of ~ 84% of the energy (burned in
unfortunately) and a first diffraction ring with approximately 7%.
The second diffraction ring is too faint to really show, but can be
seen in the original Tiff image if I stretch the contrast. What's
missing is a lot of purple blue glow, so common in fast achromats. I
measured the diameter of the first dark ring on the original Tiff
image, and it comes out almost exactly 8.4 microns, which is the
theoretical diameter of an F6.2 beam ( the lenses are a bit long at
31.8" focal length).

I hope this has addressed all the major issues with lens testing and
performance, and that I don't get a lot of Email from individuals who
want me to elaborate. There are several good books out there for
anyone wanting to dig deep into optical theory and testing. One is
Suiters Star testing volume - unfortunately he does not really
address refractor optics. The other one is Telescope Optics
Evaluation and Design by Rutten and van Venrooij - a really useful
book indeed.

Roland Christen, ASTRO-PHYSICS

Return to Roland Christen Images and Essays.