[evlatests] test data from 7/23

[Juan Uson]

Dear All,

I have reduced 3 hours of data taken on 3C147 IAT: 13:45:00 through  
16:40:10

I got all antennas and used spectral mode PA, with 15 ch, with  
Hanning at C-band default (4.881975 GHz).

I encountered the following problems:

EVLA antennas were down at the start until Barry fixed the EVLA  
script (according to operator's comments)

Later:  Antenna 13 (AC) dropped out between 15:48:30 and 16:22:20,  
i.e. the amplitude dropped by a factor of ~440, the phases became  
seemingly random.  This occurred again from 16:36:00 until the end  
(16:40:10).

Antenna 6 also dropped out (entirely) between 15:25:10 and 15:46:20.

There were no operator comments concerning these dropouts so I don't  
have any further info on these dropouts.

*** In addition, there was a "last record" problem (which I have seen  
before as Ken Sowinski knows) in that the phase of the last record is  
off (easy to see on these strong calibrators).  This is easily  
spotted with TVFLG in AIPS using "phase-difference."  In the past,  
the explanation has been that "perhaps" the array stops tracking the  
phase when the stoptime (I assume that in this case it is converted  
appropriately from the durations that are in the file) is reached  
although data are still collected until the appropriate IAT stop time  
is reached, so the phase is not tracked for the full integration.   
ken has told me in the past that this problem "comes and goes" but I  
have seen it in every spectral data observation that I have taken  
since ~2000 (not too many of them).

Cheers,

Juan