[evlatests] modcomp-free pointing

[Ken Sowinski]

A Modcompless C band pointing run was carried out Thursday evening
with the goal of getting the worst pointing antennas under control.
It was windy and sunset was included in the observation so the data
is of poor quality, but is good enough to correct the antennas so
that referenced pointing, and a future X band pointing run is 
likely to be successful for all antennas.  Time has been found for
an X band pointing run Wednesday night.

We learned that idcaf sometimes garbles the azimuth it reports, and
that because of the difference in calculating the pointing model
between the two systems the az encoder offset must be adjusted as
a function of pad position.  The former has now been fixed;
adjustments for the later have been made for the occupied distant
pads.  A systematic set of corrections for all pads will be
calculated and inserted in the next few days.

Tilts were updated for antennas 1, 3 and 22.
AZ offsets were updated for antennas 3, 5, 15, 20 and 22.