[evlatests] Initial look at switched power

[Rick Perley]

    Having been out of town for quite a while, I'm a little behind on 
this switched power stuff. 

    I loaded (via Bill/Eric's nice BDF2AIPS procedure) a 17-minute long 
RSRO-type C-band dataset, used for system testing last Thursday (before 
the power outage).  The visibility data look great, and most of the 
switched power looks great --- but not all.  Immediate things I noted are:

    1) Antenna 10 and 24 showed zero switched power on both 
polarizations and all 16 subbands, but did show total power. 
    2) Antenna 19 has a 'wobble' in its switched power on all IFs in 
RCP.   The wobble amplitude is a few percent, and the frequency changes 
notably over time.  In LCP, there is a nearly 50% ramp/oscillation in 
the switched power, acting on very fast (seconds) timescale. 
    3) Subband 16 shows a total power completely unlike any other 
subband, except for antennas 2, 11 18 and 20.  For the rest, the total 
power is changing on timescales of ~1 minute, by factors of 10 to 30%.  
The visibilities are certainly not reflecting this, nor is any other 
subband -- it's hard to believe these power variations are real. 

    I have not yet tried to apply these data to the visibilities.