[evlatests] Last 6-second phase problem

[Rick Perley]

    It was asked this morning if the 'last 6 second' problem affected 
all bands.

    It does.   The ~3 degree random distribution of non-closing phase 
seen only in the final two records (3 second averaging) is seen equally 
in my Q, Ka, K, and X band data.  (The order of observing was:

    X -> XBand pointing -> Q -> Ka -> K