[evlatests] Crazy phase winding at Ku-band

[Rick Perley]

    A 30 minute test file was run last evening, with the goal of 
checking for discrepant Tcal values.  As usual, the most curious issues 
found are the unanticipated ones...  One in particular deserves a 
special note.

    At Ku band, I tuned the AC and BD IFs to 13.2 and 16.8 GHz (center), 
with the usual 128 MHz subband width (using the old 'OSRO' mode to keep 
the data volume to within manageable levels).  The BD IFs give lovely 
stable fringes (like all other bands), but the AC IFs gave highly 
different visibility ampliltudes amongst various baselines.  FRING 
failed to find stable delay solutions on this IF, so I took a closer look:

    The phases are spinning at rates at or above 1 Hz!  Only a few 
baselines show trackable phases, and these bear no spatial relation to 
one other.  For example, antennas 4, 14, and 23 give trackable winding 
phases amongst themselves.  These antennas are each located near the 
middle of the arms.   It would thus appear that these rapid phase 
windings are not the result of bad geometry, or bad frequency. 
    However, the baselines with slowly winding phases are not those with 
the highest amplitudes!   There does appear to be a preference for the 
baselines with high amplitudes to be adjacent. 

    My best guess is that we have, for this IF only, a phase rate of a 
few Hz, which may have a relation on antenna location.  Why this 
happened for this band, and this IF only, while leaving all others 
unaffected, is beyond me.