[evlatests] Planet Tracking works!

[Rick Perley]

    Barry tried tracking Venus yesterday afternoon, as part of his 
'Modcomp-Free Zone' tests.  He took about 30 minutes of data on Venus, 
at C-Band. 

    I've reviewed the data.  First the good news:

    The phases, for all antennas which were working, are stable over the 
entire duration.    Nice! 

    Now the not-so-good news (most of which I hope have easy explanations):

    1)  Only 17 antennas gave useable data.  There are three classes of 
problem:
             A) No data at all:  19 and 24.  Easy to explain: 19 is in 
the barn, and I presume 24 was out of the array.
             B) Data present, but dead:  21, 23, 25, 26, 27, and 28.  We 
know 21 and 23 have no receivers, but 25 through 28?  Did somebody 
forget about adding these in?
             C) Antennas 17 and 22 were very weak on all IFs.  But -- 
their cross-correlation was close to full strength.  This translates to 
a huge 'closure' error. 
             D) Antenna 14 B had the phase oscillation problem (this has 
nothing to do with the Modcomp conversion). 
             D) Antenna 10D and 14D were very very weak, for no clear 
reason. 


    2)  There were plenty of short-term amplitude drop-outs, during 
which the antennas were essentially dead.  These always affected all 
four IFs at the same time.  In total, only perhaps 10% of the data were 
affected. 

    Otherwise, it all looks very good, and very normal. 

    I attempted a point-source self-calibration -- despite the obvious 
heavy resolution of the disk of the planet.  Phases came out fine.  
Amplitude gains resulted in the expected calibrated visibility, where 
short spacings formed from adjacent antennas at the ends of the arms are 
way overcalibrated.