[evlatests] Some results from full polarimetry

[Rick Perley]

    Some things noted ...

    1) The R-L delay error is actually 135 nsec. 
    2) Antenna 28, LCP on both subbands is bizarrely bad.  Michael 
anticipated this. 
    3) Antenna 25, RCP, in both subbands has variable gains, as reported 
on  yesterday.  This does not look like an attenuator problem, per se. 
    4) Antenna 4, LCP in both subbands has two different amplitude 
states, separated by exactly 1 dB in power.  It transitioned three times 
between these two states over the one hour. 
    5) The cross-polarization amplitude is about as expected, about 
10%.  There is considerable frequency structure -- more on this later. 
    6) Some cross-polarization spectra have large 'RFI-type' spikes, 
which are not seen in the parallel-hand spectra.  Something odd here.  
More investigation is warranted.
    7) There are a large number of zero-visibilities (identically equal 
to zero).  The fraction appears to be about 0.01%, which may seem small, 
but is far larger than it should be. 
    8) The parallel hand data seem littered with sudden jumps (duration 
1 second) in gain -- these look just like the zero visibilities, but are 
not zero.  I'm not sure yet if these are antenna based or correlator 
based. 
    
    The phase and amplitude stability (other than the antennas noted 
above) appear good overall.  The phase wind seen yesterday on antennas 3 
and 23 is absent. 

    More to come.