[evlatests] Phase Jumps Fixed -- (maybe)

[Rick Perley]

    Data were taken a short time ago, to see if the fixes that Bryan and 
Barry
coded up were effective in preventing the scan-change phase-change. 
    The test consisted of three 15-minute files, the first with the default
system, the 2nd with Bryan's fix, and the 3rd with Barry's fix.
    Each file comprised 7 observations, the first four were of different
sources, the last three of a single source (not the same as the first 
four). 
Data were taken in continuum, with 3 second averaging. 

    A)  Default. 

    Phase jumps occured for both antennas, of different magnitudes, but
at the same times.  The first source was missed (slow startup of EVLA
antennas).  The phase jumps occured between sources 2 and 3 and
between 3 and 4.  No jump occured between the next pair.  No jump
occurred between the first repeating pair, but did occur bewteen
the 2nd and 3rd repeating pair. 

    B)  Bryan's Fix. 

    Behavior was much better, but not perfect.  A single jump occured
between observation #1, and #2.  (Perhaps the change was slow to
take effect?) 

    C) Barry's Fix.

    Even better, but still not perfect.  No jumps were seen.  But to ensure
that we don't get too confident, the phases spun around by a couple of turns
(at the remarkable rate of about 20 degrees per second) during part, or all,
of the 4th source.  (Different IFs spun for different durations!)  The 
amplitudes
were o.k. throughout. 

    Amplitude drops were seen, at 10-second intervals, during one of the
sources with Barry's fix in.  However, because the 3.3 second averaging
dilutes the effect, we need to run to fast continuum mode to get a better
judgement.