Re: [ATM] Foucault Test Analysis Question

[Mike Lockwood <melockwo@uiuc.edu>]

Doug,

Douglas Arion wrote:
> I'm finishing up a 6" f/4.5, which Figure XP rates at 1/9
> right now, and a Strehl of .95. Now, what it shows for the
> residual errors includes a high center and a high zone
> around 70%, both of which I could reduce with suitable
> polishing. But, the question is, to what extent can one
> trust the specific profile that Figure XP shows? Realizing
> that it is using only four points (the four window
> positions in the Couder screen) to infer the full shape, do
> I dare polish precisely where it shows the high areas?

If you have doubts as to whether the high area exists, then 
congratulations!  You're questioning you test results because you 
realize that they're not perfect.

 From what I've seen, FigureXP is quite reliable (I use it regularly), 
but more testing is required.

If you want to see if the high zone is really there, I recommend 
testing with a 5- or 6-zone mask (if you can read the narrow zones) or 
with a mask that has the center of the zones where the edges of your 
current zones are.  This will give you a different "sample" of the 
surface slopes in different locations.

Then compare the surface error profiles that you get from the data 
obtained with the different masks.  If the error profiles are quite 
similar and show the high spot in roughly the same place, then you can 
have much more confidence that it actually exists.

If the profiles do not match, then you may well have run into the 
limit of your testing ability or the test setup.  (This typically 
happens to me at about 1/10 to 1/20th wave of wavefront error.)

By the way, the surface you describe sounds like slight 
overcorrection, which may likely be reduced with a SHORT spell of 
polishing with the normal stroke.

	Mike Lockwood



_______________________________________________
ATM mailing list http://www.atmlist.net/