My latest project is a 12.5" Dall-Kirkham mirror. I decided to prepare the convex secondary as a cut-out of a larger mirror, and to check it using fringe testing. This is a poor-man's interferometry method, and requires a concave reference mirror of precisely the same focal length as the secondary. I used the Pyrex tool and polished it to a good sphere. The sphere was evaluated by Foucault and Ronchi tests. When the secondary was polished out I was ready to test it.

Fringe testing requires a reasonably monochromatic light source. A number of light sources have been suggested. The best ones are usually discharge lamps, because they have very sharp bands. I started using a simple but not very monochromatic source: a Fluorescent "energy saver" bulb that fits into a normal light socket. Mine was a 15 W model that I picked up at the hardware store. It works and does give visible fringes, but the multiple wavelengths of light lead to a rainbow of fringes. With lots of fringes visible it looks fine, but when you try to observe a few fringes it gets to be very hard to find the edge of the rainbow fringes.

My new and improved light source is a laser pointer! Laser pointers are very useful when checking for pits on polishing, and I had one lying around. I used a small achromate lens to spread out the beam, but almost any lens should work as the beam is monochromatic. The direct beam did not give good fringes, so I added a diffuser: a piece of tissue paper over the lens. The entire ensemble was held together with model builder's putty. It took less than a minute to set up, and everything can be stripped off the laser pointer with no problem. A picture of the monochromatic light source is shown below.

Fringes show up very well using this light source. It works best near the ROC of the mirror set. I took pictures of the fringes and they came out surprisingly well. The fringes were photographed with a hand-held 35 mm camera on auto-exposure using 400 ASA film. The laser light source was also hand-held, so you can see that this is a very quick and dirty setup. A mounted camera should give a better picture. One picture of the fringes is shown below.

The convex mirror is on top, and I only need the inner section. The concave reference surface has a TDE, which is obvious in the fringe pattern around the edge. The ROCs for the two mirrors are not perfectly matched (the concave mirror is too deep) and this produces the differential curvature from top to bottom. This is not a perfect test, but it shows that the center of the secondary is basically spherical with a good edge. I will use it in the telescope and see how it performs.

I had not appreciated how much of a pain it is to match the curvature of a convex and concave mirror. The rest of the process is simple, but matching curvatures is just tedious. The laser pointer light source works very well.

FYI: A standard red laser pointer has a wavelength of 650 nm. The fringes are separated by lambda/2 or 325 nm.