[evlatests] Channel 14B

[Rick Perley]

    It is generally known that the amplitudes from 14B, at all bands,
fluctuate by over a factor of two when observing a strong source. 
I found that when running in our fastest continuum mode (averaging
time ~400 ms), the distribution of the correlated amplitudes separated
into five distinct amplitude bands, spaced roughly by ~15% --
corresponding to power steps of about 1 dB.  The suspicion is that
the T304 is somewhat randomly inserting attenuation on a timescale
as fast, or faster, than 400 ms. 
    The distribution of points is not equal in each state -- perhaps half
of them are in the central state, 20% in each of the states immediately
above and below, and the rest in the +2dB and -2dB states. 
    I check to see if there was any memory, or latency between
succesive points.  There is none that can't be explained by the
probabilities quoted above. 
    Finally, I checked to see if there is any phase effect corresponding
to this amplitude effect.  There is none at all.  This is purely an 
amplitude
(or power) effect. 

    Rick