of Lathrop California
Comments on Figuring Large Fast Mirrors
My world is figuring mirrors for Large Aperture Fast Focal Ratio Telescopes [LAFFRTs]. I just finished a
32" f/2.8 mirror, and it was quite an adventure, being how I was going to test mirrors this large and fast with a high order of
precision. On other less-demanding projects I have used the Ronchi, Caustic, Foucault, Wire and Bath Interferometer tests.
When using a Foucault test, say with only four zones, normally a person looks to see at what point on the longitudinal axis
the opposing sides' zone-holes blank out simultaneously. But with LAFFRTs, they don't blank out simultaneously, the shadow
moves across the zone hole from the side of the razor to the other side, at different rates, and one ends up looking to
see when the "terminator" of that shadow hits the center of that zone simultaneously, a very subjective act. A sight error
makes a huge optical mistake. A Bath interferometer has a difficult time gulping in that fast light cone, it requires a very
small prism and short focal length lens, and perfect laser quality, and then you put up with tons of unwrap errors, frequently
with high lips at the edges that don't actually exist. A Fizeau interferometer can gulp in the light cone, but both the Bath and
the Fizeau now have very numerous and very non-straight fringe patterns for the software to evaluate, and aliasing will possibly
not yield a satisfactory level of precision. The fix for this, at least for Fizeaus, is a Ross null lens, which has to be placed
to an unreasonably precise positioning over a large span of distance.
Perhaps the most precise of the tests for LAFFRTs is the caustic. However, the measuring accuracies required are daunting, being the
transverse measurements needs to be accurate to 0.00015" and the longitudinal set to 0.001". We open zone-pair flaps two at a time and
use an eyepiece diffraction pattern to locate the x and y positions of the zone foci, all of which can take a good four hours. If the
room temperature is not constant, the distance between the mirror and the tester is changing, and the results which can't be reproduced.
Multiple tests are required, and if they vary, you really don't have a test better than with Foucault.
Recently, I began using a new test, the TSIT invented by Bill Thomas of Grass Valley, California, (http://www.yubagold.com/tests/) and
that is the test I used to figure the 32" f/2.8, as well as every other mirror from my shop in the last three years. It uses a
caustic type mask with each hole returning an image of the slit light source which is digitally photographed with 20 to 60 second
exposures (not affected by room temperature changes). The zone readings are not done visually which reduces human subjectivity.
Instead, the zone-images are processed digitally to a tiny fraction of a pixel and with remarkable repeatability. Our experiments
have shown that the TSIT done by two different persons will yield nearly identical results.
Since the TSIT can be done quickly, I was able to immediately work errors and retest, keeping in mind that nothing can be done to
speed up the acclimatization time for large mirrors. But during figuring when not near the end of the work, TSIT lets you perform
many iterations in a day. Before the TSIT, I had to use the Foucault and Wire tests being the caustic test or interferometry takes
too long. But being that the TSIT is faster, more accurate, repeatable, and much easier than the Foucault, it brings it down to
single test while figuring.
Early on in polishing I do an interferogram to assure there is no [or negligible] astigmatism and a Ronchi to confirm there is not
a TDE. From then on it's the TSIT right up to final figuring, finalizing with a star test to confirm the absence of astigmatism,
TDE and spherical aberrations. Throughout the figuring process, besides the TSIT, I will perform Ronchi to detect TDE, and knife-edge
examination for smoothness. Star testing a mirror done with TSIT results in virtually zero spherical aberration. The TSIT has
revolutionized my shop practices more than any other single change in shop methodology
Photo of Dr. Mike Lavieri, owner of Cosmos ABAT, 32" f/2.8. Photo and mirror by Jeff Baldwin
Baldwin Astronomical Optics