[evlatests] AR642 2jul07 04:31-05:31 Mostly good, with a couple oddities

[mrupen at nrao.edu]

AR642:  2jul07 04:31-05:31  C band continuum, fast switching, 3.3s averaging

SUMMARY: much better than 30jun07 -- a few problems.
  Log is mis-labelled, though data are not (!).
  One sub-scan shows anomalous but consistent gains, something I've not seen
    before.

* OBS file identical to mostly-failed run on 30jun07

* Correctly labelled in archive, but log claims AK642 (i.e., K rather than R)

* Ant. 19,21 have no C-band receivers, but are not flagged after the
  first two fast-switching subscans (there is ONE fast-switching scan,
  followed by one scan on 1331+305)

* One record during fast switching has no fringes -- wrong source?
  (04:54:57)

* Last scan ends at 05:31:37
  All antennas vanish for the next record (05:31:40); then EVLA-EVLA
  baselines re-appear (with random phases) for 05:31:43

* Phases quite stable -- no sign of crazy baselines

* EVLA ant. 13, IF 2L correctly flagged
                IF 2R is very weak
  EVLA ant. 18 is very unstable in all IFs -- good phases, but amp. very
    high (factor 10-15) on many records.  Not flagged.

* One subscan (in the middle: 05:02:20-05:03:00) shows very odd amplitude
  gains, very consistent within the subscan but 10-50% different from the
  adjacent subscans.
  - No obvious pattern (VLA vs. EVLA, E/N/W arm, baseline length, ...)
  - IF 1R (=IF A):
    Low gains for antennas  5, 7, 10, 13, 14, 23
    Roughly constant gains for  4, 22
    The rest have high gains
  - Sense of change (higher/lower gains) is mostly same for all IFs, but amount
    is different from IF to IF
  - Antennas return to previous gains after this one subscan
  - Can't tell whether this extends to the adjacent on-source subscans,
    since that source is so weak and the changes are not large enough to
    show up in the rms noise as seen by TVFLG
  - May correspond to a 20-60d phase jump on some antennas, but with
    the atmosphere it's hard to tell 

================================================================================