March 1999

Stockton Astronomical Society

Valley Skies

Valley Skies - March 1999 Issue

The Telescope Nut
by Jeff Baldwin

The Foucault Test

Last month we used the tornado graph to figure a mirror from a sphere to a paraboloid. What was it that we were doing to the mirror that made it change?

Using sub-diameter laps we can correct the mirror so that the central regions, or zones, obtain reduced radii of curvature. We want the concavity of the center to be more than the concavity at the edge. To do this we do a full length stroke with a small lap that goes over the center on one stroke, then slightly off center on the next stroke. As we walk around the barrel, the edge of the mirror will get 'hit' by the lap only when we are aligned with that side of the mirror, yet the center gets hit each and every stroke. This wears off the center faster than the edge, and eventually the central regions will become more 'corrected' than the outer regions.

One problem with this, especially with sub-diameter laps, is that you don't know if the middle is getting too much correction compared with, say, the 70% zone; or, for that matter, if any zone is getting too much work over another zone. That is why we check the mirror with the Foucault test periodically during the work. If one zone is winning the race to correction over another zone, then the stroke needs to be changed to prefer the under-corrected area and to diminish the work with the over-corrected area. You will hopefully get all four zones (Couder mask zones) into the center of the tornado, at which time the mirror will satisfy Raleigh's 1/4 wave criteria.

What happens when you go too far on a zone's correction? The answer to this is to anti-correct. This can be done by passing the sub-diameter lap from edge to edge in a chordal stroke that has its midpoint at the zone that needs anti-correcting. Once the zone is back to the right spot the rest of the mirror will probably be goofed up a bit, but you can then test the mirror and see where it needs correcting and where it needs anti-correcting.

If you are using a full-sized lap, then correcting strokes are different. A very long W stroke with the lap on the top and the mirror on the bottom will shear the outer regions of the mirror off causing the rest of the mirror to appear corrected with respect to its new, longer radius of curvature. To anti-correct, flip them over and do the same thing. If the mirror is very fast or large, this process can fail.

Some of these correcting strokes that I've mentioned here can cause "zoning", or troughs around the center of the mirror at different zones. The illusion in the knife-edge test is that the mirror has concentric ripples on its surface. All correcting strokes should be done carefully and smoothly to prevent this.

Small diameter mirrors with slow optical focal ratios, such as an 8" f/7, have wide tornadoes. Large diameter mirrors with fast focal ratios have skinny tornadoes. The larger mirrors get skinny tornadoes in a linear rate, while fast focal ratios get a skinny tornado in an exponential rate. When they work together, like a large mirror with a fast focal ratio, the tornado gets skinny ultra fast.

For this reason, large and fast mirrors may not be fully examined with only the Foucault test. An 8" f/7 mirror that passes the Foucault test is done. A large, fast mirror that passes the Foucault test is not done. More crucial tests include the caustic test, wire test, star test, null test, Ross and Dahl null tests, and the king of them all, the interferometer test.

Clear Glass...Jeff Baldwin
For more information on Telescope Making jump to the ATM page.