From: "James Lerch" 
To: 
Reply-To: "James Lerch" 


----- Original Message -----
From: "Jim Burrows" 
>
> At 14:27 2003-07-21 -0400, James Lerch wrote:
>
> >Ok, Let the FUN begin 
> >
> >Unit = Waves
> >Wavelength = 633nm
> >
> >RMS Wavefront = 0.145  {46nm Surface}
> >PV wavefront  = 0.67       {212nm Surface}
> >Strehl Ratio    =  0.44
>
> Mike Peck gave me his guesses for the units, order, and normalization of
> the Zernike coefficients (you need all three).  With that information and
> the values up to "8th order",

Hi Jim,

I'm hoping when the mirror arrives, the10th order coefficients are
included. I've asked for them via Email, but I haven't received them, nor
have I received the order and units for the coefficients we do have...

>the easy computation is the RMS; I get 0.1484
> wavefront, 46.96 nm surface (assuming 632.8 nm wavelength (He-Ne laser)),
> close enough for government work.
>
> The sign and value of the third "2nd order" coefficient, -0.384812 are
> curious.  The sign could be due to "reflection", and I
wonder if that sign
> goes through the whole table - might solve the under/over correction
> conundrum.
>The value should be 1/2R, giving R = 1/(2*.384812*632.8E-6) =
> 2053.3 mm vs. 111.63" = 2835.4 mm (not close enough for government work
> ).  [James, what IS R for the mirror?  I hope one of those
values .]

Well, at the lab we measured R as 2829mm, with zone one nulled.  Any error
in this R measurement, should push R a little shorter (IE zone 1 did have a
radius of 29mm, so we weren't exactly at the null for the center of the
optic, and while the tape measure was taught, it might have had a little
sag)

Here's something else interesting, if the 3rd "2nd order"
coefficient is supposed to be 1/2R, this gives an R = 2053mm, and I put
this value in for an X Bias in Sixtests, this shows the mirror as
undercorrected with a surface RMS of 188nm.

NOW, if I double this value (don't know why I should, but I did), and put 4106mm
in for an X-Bias, I get an Over-Corrected result, with a surface RMS of
40nm, and a Strehl of 0.434....  Gee, that seems awfully close to the
interferometry result....... (I don't think we miss measured R by 2
METERS!....)

> If we could solve these mysteries, the tough computation of surface height
> and slope is possible, first along the reference diameter.  From that we
> can derive what Robo-Foucault readings should have been.

YEA, WHEEEEE 

As a wise man told me recently "Funny how research usually spawns more questions
than it answers."


James, signing off, Banging head against wall, repeatedly mumbling "why?"

--- BBBS/NT v4.01 Flag-5