From: "Herb Kasler" 
To: "Michael Burr" , "ATM list"

Reply-To: "Herb Kasler" 



Michael,


RE:


>>But my understanding from reading the ATMJ and TM articles is that you
have to measure the lateral motion. Is that correct? <<

I have not read the articles, but I suspect that they are referring to the
caustic test and not what is commonly known as the wire test. In the
caustic test, I think what you measure is the amount of Y offset required
to move the shadow of your wire from some zonal radius on one side of your
mirror to the same radius on the other side. If you know the X offset from
the CoC of the center zone of the mirror at the same time, you are then
sampling this trumpet-shaped surface that is the caustic of the mirror,
which is what aspheric surfaces have instead of real radii of curvature
(Thanks for setting me straight , Richard ;)). If you take enough
sufficiently accurate data points ( I presume at several different X
offsets too), you apparently get a much better picture of the surface than
you can get with a simple Focault-type setup. People bandy about 1/50th
wave as the range of accuracy you can get. But please, take this with a
grain of salt. I have never done the test, and I was kind of scratching at
my simian cranium while I was reading about it, so I could be talking
squarely out the wrong orifice.

What  I am pretty darn sure about is that you don't need to measure lateral
offsets when you do a regular wire test. In Sixtests, there is no place to
enter Y offset data if you select "wire test". In fact, there is
no difference in the output from what you get if you enter the same numbers
under "focault test". More importantly, I used the wire test on
my mirror, and it seems to work just fine. The Ronchi picture is dead
indistinguishable from what RonWin says it should look like, even if you
measure every line as accurately as possible with a pinstick at several
different offsets. It looked that way even when the wire test said it was
still around 1/2 wave undercorrected. Smooth errors in correction can be
really hard to spot with the Ronchi test. With the wire test, I got it
within at least 1/5 wave, possibly 1/9 wave. The Ronchi pictures still
looked exactly the same. Anyway, the point is that if I had been doing the
test totally wrong, that would definitely not be the case. And I wouldn't
ever see nice, sharp pointlike star images at 550X with the resulting
telescope, which I occasionally do, when the sky is exceptionally still.
AIUI, what makes the wire test better is that on big, short mirrors it is
much easier to judge where the wire shadow is on your mirror's surface than
where the boudary of the KE shadow is. I found it especially helpful to use
a painted hair because the center of the hair transmitted enough light to
project a very sharp dark-bright-dark triplet of lines that was much easier
to accurately line up against the teeth of my pinstick than the more fuzzy
boundary of a KE shadow, especially on the outer zones. In truth, what I
did was to use a KE on the innermost zones and a wire on the outer ones,
because the wire shadow gets kind of dicey near the middle of the mirror.
It is also really easy to line up that round Phi shape with multiple
pinstick teeth. I had a cross-shaped pinstick with 8 teeth for each zone.
It gave me a MUCH better idea of what I was looking at than a Couder mask.
You can see immediately if something is out of whack, so you don't waste
hours generating bogus numbers, as I did once when using a 9-zone Couder
mask that was 2 millimeters displaced to the right.


Anyway, I hope this clarifies what I'm saying about the wire test.


Later,

Herb.

--- BBBS/NT v4.01 Flag-4