[evlatests] Wideband Modes (cont.)

[Rick Perley]

    I loaded the data, minus the flags, to permit discrimination between 
'real flags' and 'zero records'. 

    The data are in general very find indeed, with spectacular stability 
in both amplitude and phase.  But not all is perfect ...

    a) There are a lot of reported 'perfect zeros':  20% of the first 
test, 7% of the second, and 25% of the third.  I'm having trouble 
actually finding all the 'zero data', but there's no question that a lot 
of data is missing.  Unlike the last go-around, these missing data 
(a.k.a. zero records) do not cluster at the beginning of scans, but are 
found, more or less continuously, associated with some antennas and/or 
baselines. 

    b) Antennas 2, 4, 11, 27, reported earlier as having been flagged, 
are present in full force in this (unflagged) database -- I can see 
nothing wrong with the data attached to them.  Clearly they were flagged 
by the system for a reason -- but whatever it was, the data quality was 
not seriously affected. 

    All other issues reported in my last posting remain as before.

    I did a more or less normal calibration only of the third test -- 1 
GHz of bandwidth, comprised of 8 adjacent subband of 128 MHz each, full 
polarization.  The data (except for the missing antennas and IFs) are 
simply gorgeous.  There were some 'burps ' in amplitude at the 
beginning, after which all is stable to a small fraction of 1% in 
amplitude.  The bandpasses are excellent.  Delays are tiny (except for 
those antennas already reported).