[evlatests] More on Cyg A Test Results, at X-band

[Rick Perley]

    Continuing on from the report this morning: 

    I SPLIT out the central 2 MHZ-wide channel from the data in order to 
apply the standard methods for self-calibrating Cygnus A. 

    Some further oddities in amplitude and phase behavior were found 
from this:

    Antenna 5 has a large phase offset between Cygnus A and the nearby 
calibrator -- this looks to me like a Bz error -- and it has to be BIG!  
The offset is about 30 degrees over 10 arcminuts in declination!!

    Numerous antennas have semi-random phase jumps, in the sense that 
there is no phase continuity at all between scans.  (Note that other 
frequency bands separate these observations).  The badly behaved 
antennas are:

    8, 10, and 14 on the AC side
    11 and 14 on the BD side. 

    A number of other antennas have a small number of large phase 
offsets (1 to three) for individual scans.  Curiously, the phase offset 
seems to be about the same -- 150 degrees -- for all of these

   ea03 on all IFs
   ea10 on BD only
    ea19 on all IFs
    ea23 on AC only
    ea27 on all IFs

    As noted earlier, amplitude stability is rather poor, at the typical 
level of up to 10 percent in power.  Three antennas (5 in all IFs, 17 in 
AC, 12C) have clear '1-db' steps in power. 
    Antenna 20 has an oscillatory gain change of a few percent, with 
period of 20 to 30 minutes, visible on all IFs.  Looks like a large 
pointing wobble ...

    Self-cal was able to remove most of the above without trouble.  
However, the resulting image, although quite nice, is far short of what 
I had hoped to see -- about 2000:1 in dynamic range.  This is about what 
we were able to get with the VLA's correlator, and is 'rather short' of 
what I had hoped to see...