[evlatests] Another wide-band test

[Rick Perley]

    Ken set up another experiment with our full-bandwidth antennas.  
This one is at Ku-band, using 3C84. 

    A quick perusal reveals a few issues:

    1) ea23, in RCP has a sinusoidal ripple in both amplitude and phase, 
in baseband #1 (subbands 1 through 16).  The period of the ripple is 
about 11 MHz. 
    2) ea19, in LCP, as a very similar problem, in baseband #2 (subbands 
17 through 32). 
    3) Some of the 'subband 0' DC offsets appear to have gone away. 
    4) Some of the basebands are showing large steps in phase between 
adjacent spectral windows.  The phases, after removing the delay, should 
be continuous within a baseband, but this is commonly not the case.  
ea13, in baseband 4 (subbands 49 through 64), in LCP, is a particularly 
bad example. 

    Because our 8-GHz span is larger than the BW of the Ku-band 
receivers, Ken arranged the tuning to significantly overlap the nominal 
ends (12 and 18 GHz).  Remarkably, we are still getting good signals 
down to 11.6 GHz, and up to 18.4 GHz. 

    The plan is to run this test again tonight, with a proper flux 
density source.