[evlatests] progress

[Ken Sowinski]

A few things got cleared up today; perhaps they will remain
working until next week.  Rob took care of the problems in 16
which I reported Tuesday.  Both L302's are working and the
problem with 16D was traced to a bad downconverter and a bad
fiber connection between downconverter and DTS.  The problem
reported with 14C was finally found to be that it was still
filtering out subband 5 rather than the required subband 6.
That no correlated signal was seen in 10 second integrations 
allows us to deduce the not surprising result that there is a
very good lower bound on the correlated signal that spills over 
from one subband to its neighbor.  

I spent most of the time measuring SD and correlation coefficient
for a large range of sampler input power and deformatter gain
parameter.  This will be summarized later.