[evlatests] evla baseline results

[Ken Sowinski]

                        EVLA Baseline Results
                           Oct 11, 2005

Barry was able to fix the executor to remove the phase jumps we
had been seeing at scan beginnings and we ventured a baseline
run over the night of Oct 10-11.  The weather was calm with
good API pahse stability.  Except for antenna 4 the VLA antennas
behaved quite well.  Following is a short summary of the EVLA
results.

The fundamental bit of news is that the phase jumps in EVLA antennas
are no longer there.  This allows a more careful look at the baseline
solutions.

1.  There is a secular drift of about 60 degrees in hte residual phases
after fitting for antenna position seen in both antennas over the three 
hours of the observation.  RT phase data was available from mib -52,
but not mib -3d.  It shows about .01 turns at 512 MHz which scales to
60 degrees at 8.4 GHz.  This agreement is deceptive because 8.4/0.512
is not the correct scaling factor; we must allow for the implicit
conversion done in the transition module using a local reference frequency.

2.  There are fluctuations in both antennas, occurring at the same time 
in  both antennas which are probably elevation dependent effects.  For 
antenna 14 this  is about 40 degrees peak to peak; and about 100 degrees
in antenna 16.  Expressed in time these are about 14 ps and 33 ps 
respectively.  The "k" terms for these two antennas are 2 ps and 16 ps
and are as yet unaccounted for in the phase calculation at observe time.

3. The magnitude of the correctinos to the antenna positions is about
5 ps.  This could be applied to the antenna location parameters to 
see if we can get the transofrmation and the signs right.

4.  The RT phase data showed fast variation of about .001 turns that
could be related to antenna motion.  It would be interesting to compare
this with the pointing angles as well as with the phase residuals.