[evlatests] Widar data over the weekend.

[Vivek Dhawan]

Summary:

1. WIC loops at C, X band: Fringes on most antennas; assorted dead IFs.
   Image dynamic range is only 1000-2000  but no obvious bugs.

2. X-band delay data, 2 files on Sunday: An oddity where RCP phase is
   steady while LCP twin has a phase rate of a turn in a few minutes,
   different on each antenna.

--------------------

Details:

1. WIC loops C and X band, Friday evening. 24 antennas in the array, not all
   fringing, not always the same ones:

   WIC loop C band starting at MJD 55233.041
13 R,L dead. 11 OK. 15 ? . 25 R, 26 L, 27 L dead.

   XWIC loop at 55233.062
11 R,L dead. 13 OK. 15 R dead. 25 R absent, L dead. 26 R,L dead. 27 L dead.

FRINGE fails on RCP (refant 2) though phases vs. time and bandpass look
normal, largeish delays but not aliased - so the reason is yet unknown.
LCP works fine.

For both the above data sets, phase transfer looks reasonable, no obvious
closure violation. But the images of the 4Jy calibrator are only ~1000:1
at C band, and ~2000:1 at X band. Images of the secondary source  ~300:1
after calibration transfer, and ~1000:1 after self-cal (~1Jy). All images
used the central 40MHz of the 3 subbands averaged.


2. X-band delay setting: 2 files from Sunday afternoon,  Xdelay 55234.873
   and 55234.896,  each had 3 subbands of 128chan, RR & LL. In both,
   RCP is fine but LCP has a phase rate; this is a polarization pair,
   the rate cannot arise at the antenna since they share the same LO,
   so its likely in the correlator. We saw this in a short file last week,
   I think Ken had some ideas but I forget what he said then - we hoped
   it would never appear again -alas.  The new factoid is that the rate
   changed when sources changed 3C345 to 3C84. It looked like a steady
   rate for ~20min on 3C84. I have not had a chance to look at the delays
   that were found from this data.

Vivek.