[evlatests] Failure to tune at x-band

[Rick Perley]

     A beam-holography script was run last night, at X-band.

     Repeated 'failures to tune correctly' occurred on four instances.  
Three of these were after x-band referenced pointing was done.  The 
referenced pointing was done with different frequency tunings (but in 
the same band).  The other instance was at the beginning of the 
observation.  For all four, the failures were only in the 'A/C' IFs.  
The B/D pair were fine.

     Details:

     1) Instance A:  First scan in the run -- antennas 5 6 18 and 19 did 
not tune correctly (no fringes), on SPWs 1 through 8, and both 
polarizations.

     2) Instance B:  Following the first referenced pointing. Antennas 
6, 9, 18, 24 and 26.  (On the following scan, all antennas except ea06 
came back to life.  Ea06 returned on the next scan).

     3) Instance C:  Following the 2nd referenced pointing.  Antennas 3 
5 11 and 18.

     4) Instance D:  following the 3rd (and last) referenced pointing.  
Antennas 9 11 13 19 24 26 and 27.

     This failure is particularly insidious since it is *not* caught by 
the flagging system.  In nearly all cases, the non-fringing antenna 
returns on the following scan.

     This failure mode is long-standing.  I believe I have seen it in 
every case that I can remember where and X-band observation follows an 
X-band referenced pointing done at a different frequency.