Date: Fri, 31 Oct 2008 20:48:31 +0000
From: Richard Hills <firstname.lastname@example.org>
To: Todd R. Hunter <email@example.com>, firstname.lastname@example.org
Here is a version with the beam squint calculated and also the defocus.
It assumes that you have two data sets taken with the source in the same
position - assuming (?) cross-pol effects can be neglected, this is best
done by setting it to 45 degrees polarisation. The data needs to be posted
into the two sheets. You then need to solve for the XYZ position of the
phase centre on each sheet. (I seem to have broken Josh's buttons, but
Solve still works.) Note that it is what I describe as the position with
respect to the nominal beam direction that is important here. If you use
the X and Y positions with respect to the scanner frame there will appear to
be a spurious squint if there is any difference in the z-positions.
The beam squint calculation is in pink. You need
1) the "plate scale", p, which converts movement in the focal plane to
angle: p = 1/cass focal length, which is 96m.
2) the differences between the X and Y phase centres - these are in J8 and
K8 and are then made into a radial distance in L8.
3) the FWHM is taken to be 1.15 * lambda / dish diameter as an angle. This
is in R10 and converted into a distance in R11.
4) The ratio of 2 and 3 gives us what we need.
The Defocus is in green. Basically it just takes the average of the phase
centre Z distances from the two sheets. You then have to put those in B10
"Nom Z_off" in order for the calculation in Q6 to be correct.
Since this is band 6 where the polarization splitting takes place in an OMT
behind the horn, we expect both the squint and the defocus to be small, so
the fact that they are is reassuring and evidence that we are doing OK on
- 03 Nov 2008