[evlatests] PSum, PDif tests

[Rick Perley]

    We are testing the noise calibration system, chasing down the cause 
of the 'PDif' compression which has been previous reported.

    A short test today, at 6 cm -- while not useful for the purpose 
stated due to other problems -- has revealed:

    Antenna 24, LCP:  The noise diode power is extremely weak -- about 
1% of its proper value. 

    Antenna 8, LCP:  The values of PDif are very noisy -- the scatter 
from point to point is at least an order of magnitude higher than on the 
RCP side.  This scatter is not seen in the PSum values, which appear 
normal in all ways. 

    All other switched powers from the working antennas looked 
reasonable.  (Antennas 2, 3, 18, 21, and 26 were out of the array).