[evlatests] ea 25 DTS problem

[Rick Perley]

    The Saturday night test has revealed a problem with ea25's bandpass 
spectrum.  It shows the 'multi-lobed' bandpass shapes characteristic of 
a DTS problem.   The problem exists throughout the entire run in LCP.  
In RCP, it started mid-way through -- at approximately 2:50 IAT. 

    Besides the characteristic 'multi-lobe' spectra, the problem also 
shows up as oscillating delays, with the changes between two apparently 
stable states of a few tens of nanoseconds, switching states on 
timescales of ~1 second (or less).  Interestingly, the amplitudes remain 
stable, with no obvious loss in sensitivity or amplitude.