[evlatests] Antenna 1, LCP, 16 nsec delay changes

[Rick Perley]

    In yesterday's cursory review of the new WIDAR data, I had noted 
that antenna 1, in LCP, did not give good fringes on two of the 42 
separate observations.  I dutifully flagged these out, and proceeded as 
normal.  But there was clearly something odd about this antenna-IF, so 
this morning I looked a little more closely....

    It turns out that the delays for this antenna-IF changed during 
those two 'dead' scans, so that the subsequent observations each had a 
different delay.  The change in delay is a notable number:  *exactly* - 
16 nsec.  So, the data from 23:14:10 through 23:34:45 have a delay -16 
nsec less than that at the beginning of the run, while the data from 
23:34:45 through to the end have a delay exactly -32 nsec less than that 
at the beginning. 

    To emphasize:  This only occurred on antenna 1, in LCP.  RCP is rock 
solid. 

    And also to emphasize:  All other antennas, on both polarizations, 
have rock-steady delays.  Changes are less than 1 nsec throughout the 90 
minutes of observation.