[evlatests] WIDAR: IF C this morning

[Michael Rupen]

A somewhat longer but otherwise similar run this morning recording dual pol'n
(RR+LL, IFs A and C) using WIDAR-0 (11 antennas) showed:

1- 40 minutes of 20 second scans with fshift=20 kHz shows NO phase jumps,
   apart from one glitch on all baselines.

2- Delays for IF C changed considerably from last night's values.  Ken
   and I have the numbers for those who care.  IF A by contrast gave nice
   consistent values.

3- ea03, IF C shows a steeply falling cross-corr'n response in subband 3
   (also evident in subband 4), reaching a 10dB-ish "suckout" in amplitude
   around 4990 MHz.  There is a corresponding phase jump of 80-90 degrees.
   The amp suckout does NOT show up in the auto-correlations, the StB wideband
   correlations, or Walter/Mike's autocorrelation plots.  IF A looks normal
   in both amp and phase.  The suckout and phase jump in IF C are seen on all
   baselines, and  (in spot checks) throughout the observing run.

More work to do, obviously...

         Michael