[evlatests] Constant Squint

[Rick Perley]

    At today's coordination meeting, Ken suggested that the values of 
the EVLA
antennas' squint (separation between RCP and LCP beams) as measured by the
pointing test last Saturday showed possibly variable values, from which 
it could
be inferred that the 'wiggle' pattern and/or solutions obtained are 
faulty. 

    I have looked at the data quite carefully, and conclude that no such 
variations
occur.  The squint values are remarkably constant, for each antenna and 
frequency,
(with a likely exception at K-band -- see below). 

    1) C-Band.  The beam separations in elevation and azimuth for 
antennas 13, 14 and 16 are -0.52 and +.10 arcminutes, respectively.  The 
peak variations in the solutions from these eyeball averages are about 
.06 arcminutes, the mean differences are much less.  The solutions are 
stable enough that small (~.02 arcmin) differences between the antennas 
can be seen easily. 

    2) X-Band  The separations are -.26 and -.12 arcminutes in elevation 
and azimuth, respectively.   The stability is ever better than at C-band. 

    3) K-Band  The separations are .10 and .05 arcminutes in el and az, 
and are as stable as the other bands.  However, here it is noted that 
the orientation of the squint is not the same for each antenna.  As the 
feed is at an angle of about 30 degrees w.r.t. the elevation axis, the  
azimuth squint should be about 1/3 of the elevation squint.  This is the 
case for antenna 14, but not for 13 (where it ratio is about 1/2) or 16 
(where the squint is entirely in elevation).  

    4) Q-Band.  Only antenna 13 returned any solutions.  They were very 
stable, but the squint is so small I didn't bother to compute it.  There 
are bigger problems here to look at first!