[evlatests] Longer-term phases

[Rick Perley]

    Barry took some data this afternoon on 3C84 at X-band, to determine 
if the 'ramps' on VLA antennas (when under modcomp-free control) were 
gone.  They are.  The detailed results have been sent to those concerned.

    Here I report on the phase behavior over longer terms -- about 12 
minutes. 
    Phase behavior in general was not great -- fluctuations of timescale 
~1 minute, and amplitude ~1 radian are seen on all antennas.  They are 
correlated between adjacent antennas (especially notable for antennas 
near the center of the array), so I can confidently ascribe these to 
atmospheric variations. 
    Superposed on these fluctuations is a longer-term drift, which is 
quite large for some antennas.  The worst ones are:

       22 (at W72) -- +500 degrees over 12 minutes. 
       3  (at E64)       - 350 degrees over 12 min.
       15 (E40)          - 300
        20 (E48)         -220
       23 (E72)          -200

    Although these large negative slopes are all on the east arm, there 
is little correlation in size or sign for the other antennas.  The 
slopes for the EVLA antennas in all cases except one (23, at the end of 
the E arm) are less than those of nearby VLA antennas. 
    The facile explanation is that these are due to baseline errors.  
But we'll have to wait for a much calmer atmosphere to make progress.