From: "James Lerch" 
To: "Nils Olof Carlin" ,
"ATM List" 
Cc: 
Reply-To: "James Lerch" 


----- Original Message -----
From: "Nils Olof Carlin" 

> James,
>
>
> > Couder Mask KE travel from z1 to z7 = 0.143"
> > Robo-Foucault KE travel from z1 to z7 = 0.133"
> >
> > Then Cos(A) = 0.133/0.143   A=21.55 degrees
> >
> > If the Couder Mask result is correct, then Robo would have to be
> misaligned
> > (with respect to the optical axis) by nearly 21.5 degrees?  If this
> is
> > right, I'm pretty sure I'd have noticed that! ;)
>
> I may be confused, but if the Robo-Foucault setup was skewed, I would
> think the KE travel would have been *longer*.

Nils,

I'm not a huge math guy, but let me give this a go.

Given a coordinate system where:

X axis is along the Platform Longitudinal axis Y axis is perpedicular to X
and along the knife lateral axis

In a perfect world the Optical axis of the mirror under test would be
parallel to the X axis of the platform.  In this perfect world measuring a
fixed distance along the optical axis would have only an X component and no
Y component.

 If by chance the Optical axis of the mirror is at some angle to the X axis,
then measuring a fixed distance along the Optical axis of the test mirror
would involve both the X and Y axis of the test platform. In theory the
distance along the optical axis would be square root of the sum of the
squares of the X & Y Platform axis, Yes?

Putting the above into plain english, if the platform ain't pointed
directly at the mirror, then the measured KE travel along the mirror's
optical axis will be less than it should be. (I think.....)


I just happend to have been listening to a lecture on Vector transformation
by Dr. Feynman Yesterday afternoon.  While 99% of what he spoke of flew way
over my head.  I did come to this conclusion:

In a typical Video Foucault setup we have THREE seperate coordinate systems
    A) X,Y,Z axis of the test optic
    B) X,Y,Z axis of the Foucault platform
    C) X,Y,Z axis of the Video Camera

In theory to measure something on the test optic coordinate system with a
micormeter attached the the platform coordiante system, we'd have to do a
vector transformation.  Fortunatly in the real world it apears a lot of
variables start to cancel each other out. For instance

Given the image of the test optic is centered in the FOV of the video
camera, and the camera's optical axis is perpedicular the knife edge, and
the Knife edge is parrellel with the lateral axis of the test platform, and
the lateral axis of the test platform is perpendicular to the longitudinal
axis, THEN distance measure on the platform X axis is 'Almost' equal to the
distance along the Test optic X axis.  I say almost because there is some
deviation as a result of the optical axis of the test mirror lying half way
between the light source and the camera lens.

With all the above in mind, I thought I do a mental excercise and try and
predict how many degrees of angular difference was required to get Robo
readings to match the Couder mask readings.  The answer Nearly 22 degrees!

> But just for kicks, and if you could move the stage (semi) manually,
> what about a test run *with* a Couder mask, matching sides the
> old-fashioned way, then remove the mask and take some shots at those
> very same KE positions? At least it would clear up any possible doubts
> about 1) software funny business and 2) wave-optical confounding
> factors.

We did an improptu version of this the day I took the optic down to our ATM
lab and had the classic Eyeball Couder mask test done.

After we finsihed with the Couder mask test, and the answer was spot on
with Carl's measurments, we replaced the 'Old School' X-Y dingbat with
Robo. However we left the Couder mask on the mirror and just displayed the
video image on the Television for all to see.  Using Robo to only MOVE the
platform we did a 'Group' consensus of zone nulls using the Television
image.  The result for this test using a Couder mask and the video camera
gave an answer about halfway between the classic Couder mask test and Robo.
 (Nothing like a third data point to really confuse things!)

I will admit that not having the optic on hand to 'play' with has lead to
many mental simulation experiments.


> > Take Care,
>
> and be *very* careful when playing with knife edges!

Copy That!

> Nils Olof
>

Take Care,
James Lerch
http://lerch.no-ip.com/atm (My telescope construction,testing, and coating site)

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