[evlatests] X-band status, and flagging issues

[Rick Perley]

    The modcomp-free flagging is quite curious.  From the X-band test 
done at noon today, I had two observations (in 'fast continuum), 
separated by 8 minutes in time. 

    The first X-band scan was the first in the Modcomp-free test.  It 
was followed by an L-band observation.

    The second X-band scan was preceded by another X-band (pointing) 
observation, and was followed by a U-band observation. 

    First Scan:

    Unflagged data appeared at 18:00:00.  But no antennas were on 
source, so the data are noise.

    At 18:00:38, good data from the VLA antennas began.  Very curiously, 
all data were flagged for the four seconds preceding this.  It's acting 
as if the flagger 'woke up' just as the antennas neared their destination. 

    The EVLA antennas reached the source at 18:01:22, (44 seconds after 
the VLA antennas reached it).   Curiously, some EVLA antennas were 
flagged for a few seconds prior to their arrival. 

    At 18:01:23 through the end of the scan at 18:01:30, all the VLA 
antennas were flagged as bad -- but the data are in fact perfectly 
good.  EVLA antennas were not flagged, and are good. 

    Second Scan:

    Unflagged data began at 18:08:30.  These are good data.  This scan 
followed an X-band pointing scan, so there are no band transition issues. 

    At the end of the scan, some truly curious behavior was noted.  The 
EVLA and VLA became incoherent with each other for the last 10 
seconds.   The VLA antennas were flagged out for the last 9 of these 10 
seconds.  However, the data are perfectly good through the entire 
interval.  The EVLA antennas were not flagged, and produced 
semi-sensible correlations *with each other only*.  I say 
'semi-sensible', because all the phases jumped to new values, and the 
amplitudes dropped notably at the beginning, recovering partially after 
about 5 seconds. 

    I suspect this odd behavior has something to do with the next scan 
being at a different frequency -- U-band. 

    Other than these 'transition' issues, the data are very good, with 
the exception of a few antennas which I'll report on separately.