[evlatests] Multiple amplitudes

[Rick Perley]

    As noted yesterday, antenna 24B has reverted to the 'multiple 
amplitude' state behavior.  I took a closer look at the phenomenon, with 
the following observations:

    * The levels are very clearly quantized. 
    * There are six different levels in yesterday's data. 
    * The levels are not evenly spaced in either amplitude or power.  
The power ratios, in dB, for the six levels are:
          0.0, -1.7, -3.8, -6.6, -9.8, -16.6
    * There is a smooth, gaussian-like distribution in the number of 
times the power is in one of the six states:  The most common is the 
third, followed by the 2nd and 4th.  The first (the highest power, and 
the one which is correct, in comparison to the other IFs) is only rarely 
seen.  The 5th is uncommon, and the 6th (the weakest) about as rare as 
the first. 
    * There is no obvious periodicity in the sequence of levels. 
    * There is no 'memory' in the levels -- an existing state has no 
influence on the next. 
    * There is no measureable phase effect. 

    These characteristics are the same as for every other IF, when this 
effect is 'on'.