[evlatests] Last record problem, and more.

[Rick Perley]

    In the rapid band change experiment, a couple of other issues, not 
related to switches, were noted:

    1) For many bands, the last record of a scan was incorrect -- low 
amplitude, and not flagged.  But this only occurred in LCP, and only at 
L, C, X, K, and Q bands. 

    2) Antenna 5, IF 'C', at Ku-band had a very large delay:  -20 nsec.  
All other antennas showed small delays at all IFs and bands ('small' ~ 2 
nsec or less). 

    3) Bandpass stability was generally quite good (although high 
accuracy is not possible with this object (3 Jy), and short integration 
(30 seconds).   The most unstable band by far appears to be X-band, 
where a number of antennas show remarkable changes in the bandpasses 
over the two  hours:  notably:  ea01 in LCP, ea02 in RCP,  ea14 in LCP, 
ea17 in IF 'D', ea24 in LCP (this one is particularly bad!), ea28 in 
RCP.  All the above are unstable over the full two hours.  Many other 
antennas have individual scans with deviant bandpass solutions. 
    In other bands, only two antennas stand out:  ea25 in RCP at L-band, 
and ea18 in all IFs, at L-band.  In both cases, large changes in 
visibility gain were also noted.