[evlatests] antenna positions

[Ken Sowinski]

As the antennas have been placed further out from the center
it has been apparant that there are large phase winds which
we have attributed to baseline errors.  Vivek has bullied a
baseline run into submission and found large errors for all
EVLA antennas; the largest about 10 cm for antenna 13 at AW5.

I think the problem comes about because the antenna positions
have been rotated to Greenwich using a longitude for the VLA
that is different than that used by the VLA to calculate LST.

There are several longitudes running around:
VLA:   107.617722 degrees
RCW:   107.617998
EVLA1: 107.618333
EVLA2: 107.618647

VLA is the longitude used to calculate LST to run the VLA.
RCW is the longitude that Craig found that best fits VLA
data into the official VLBI reference frame.
EVLA1 is a number coded into Array.java that is not used anywhere.
EVLA2 is the longitude used to rotate VLA antenna positions to
Greenwhich.  This was determined by comparing the EVLA antenna
positions to the VLA antenna positions.

The difference between VLA and EVLA2 is about 2 arc-sec and this
will create a position error of about 8 cm at the distance of AW5.

If we expect to correlate EVLA antennas with VLA antennas while the
Modcomps are still in use the two systems must express antenna
postions consistently.  I have created a file derived from the VLA
BASELINE file which has corrected antenna positions as required by
the EVLA system.  The VLA coordinants are rotated by VLA degrees and
converted to meters using the defined value of c as Craig would like.
The file can be found at
~ksowinsk/evla/tests/newbase

If there are no objections I will install these postions in the
EVLA database for the pads we are currently using.

There is another possible source of error in the location of the
center of the VLA in the VLBI geocentric coordinant system used
by calc.  This only enters as a high order effect because we
difference the delay returned by calc for the subject antenna
and the mythical antenna at the center of the array.