Mirror Evaluation

I have been working on a 10" f5.5 mirror that is my first homemade mirror, and will be my second telescope overall. The foucault tests have it coming in consistently at between 1/8 and 1/12 wave (although there are outlier data sets that have it at 1/3 wave.) These tests show the mirror as being excellent over the outer 3 inches with slightly low center starting about 2 inches out. In other words it is slighly overcorrected. Last night we star-tested the mirror, and the instructor of the class I'm taking proclaimed it done. Granted this was done as the SoCal marine layer was rolling in and the star we were using was always at least partially obscured. Tonight I took the mirror out again, and the weather was much more cooperative. Despite my limited experience star testing, I was able to easily spot two flaws, astigmatism and spherical aberation. I know that the star test is sensitive enough to reveal trivial errors, but I am concerned that if my untrained eye can see an error it must be nontrivial. When I got back home, I checked Suiter's book and I would estimate that the astigmatism is about 1/4 wave and the SA is perhaps 1/8 wave (interestingly, the star test reveals my mirror to be undercorrected, the opposite of what the foucault data predicts). For comparison, I also star tested the nearest newt I could get my hands on, a homemade 6" f5 made with an orion mirror. I have a scope made with the same mirror and can attest that these mirrors are surprisingly good. In the star test, I was unable to detect any error in the 6". My gut impression is that if my handmade f5.5 is unable to beat a mass-produced f5, I still have more work to go. However, most other people are telling me that the observed errors are insignificant. After all, I was able to split the double-double tonight. I would appreaciate any and all comments here.

Jeremy

Hi, Jeremy, I’m currently at work so I’ll have to keep this short. A few comments:

1) The must puzzling thing is that the star test is indicating undercorrected and the Foucault test is indicating overcorrected. This cannot be – one of them is lying to you. You may be misinterpreting the Foucault test results. When you say the center is “low” do you mean that the measurements (i.e. radius of curvature) near the center are too long relative to the parabola? If so, that really means that the center is “high”, because the radius is too long. Still, if the outer regions are good, then that is more important because there’s much less area in the center 2 inches than in the outer 3.

2) Make sure the “overcorrection” you’re seeing in the star test isn’t really a turned down edge. Double check that section of Suiter’s book.

3) The astigmatism you’re seeing may be the result of miscollimation or a not flat secondary. First ensure that everything is collimated – an f/5.5 will be quite susceptible to misalignment and it may be the culprit. If you eliminate that, then it could be the secondary. First make a note or drawing of the direction of the astigmatism you’re seeing. Then rotate the primary in its cell 90 degrees, re-collimate carefully, and then try again. If the astigmatism is still in the same direction, then it’s being caused by the secondary. If it has rotated 90 degrees, then it’s in the primary.

4) You need to take your 1/8 – 1/12 wave numbers with a grain of salt, especially since this is your first mirror, no offence please. It takes testing quite a few mirrors to really get good at doing the Foucault test right, and even then it’s not a “perfect” test. One thing to look out for with a fast mirror like yours is when to judge that the shading in the zone openings are equally “gray”. Make sure you judge them when they are just starting to darken, not after they have gotten really noticeably dark. You want the knife to be just encroaching into the converging beam, not well into it. If you compare when the mask openings are too dark then you’ll get erroneous readings, most likely leading to an undercorrected mirror (I think). That being said, if the mirror truly is 1/8 – 1/12 wave at the wavefront, then that’s really good. Are your numbers peak to valley or wavefront? Plus I don’t know what you mean when you say “although there are outlier data sets that have it at 1/3 wave”. That's a big discrepancy.

5) Almost no mirrors will pass the star test without some defects being seen. It’s that sensitive. There really is no such thing as a perfect mirror in practice. The real test is what does the diffraction pattern look like in focus. Try the double double again at really high magnification, say 400x or higher, seeing permitting. If you can see clean Airy disks with diffraction rings then your mirror is probably good, especially considering it’s your first mirror. Also the double double isn’t that good of a test – you might want to try something closer – Zeta Bootis at 0.7” in your 10” might be good, but you may not have good enough seeing to split something that close.

The bottom line is that the star test will nearly always reveal something wrong with any mirror. But if in-focus images are good, then the mirror is indeed good.

Hope this helps,
Doug