[evlatests] EVLA-VLA Global Phase Changes

[Rick Perley]

    There were three EVLA-VLA global phase changes during the recent 
run.  All three occurred early in the run, while we were running in the 
0.416 second integration mode.    All three took place during a scan. 

    Two of these changes were abrupt -- the duration of the change is 
less than 1.0 seconds.  One occurred while observing at X-band, the 
other while observing at C-band.  For each, a *single* 0.41 sec record 
showed a reduced amplitude -- (as seen on EVLA-VLA baselines) -- low by 
about 20%.  The phase jump -- for each band -- was 70, +/- 10 degrees.  
(My statement that the phase jump was proportional to wavelength, made 
in my last message, is wrong -- I was fooled by an atmospheric shift 
occuring at the same time at X-band).   There is no corresponding effect 
on EVLA-EVLA, or on VLA-VLA baselines. 

    Ken wonders if any of these changes accompanied a system reset by 
him or the operator (Ken corrected a minor error in my observe file 6.5 
minutes into the run).  The answer is 'no'.  His reset is easily visible 
(no data for 20 seconds), and is not coincident with any phase  jump.  
There are no other similar gaps in the record. 

    The third phase change is very different.  This one took place over 
a 30 second time duration.  I had originally thought it was a 
atmospheric effect -- but if so, there's a conspiracy out there, as ONLY 
the EVLA antennas show the effect.  The 'change' was at X-band, and is 
of the same magnitude as the jumps described above -- 70 to 80 degrees.  
There is no corresponding amplitude effect (unsurprisingly).