[evlatests] Dropouts-be-gone test

[Walter Brisken]

Today just before noon Pete W. updated L302 modules in ant 14 with a more
modern version of the software, including a task priority rearrangement 
that allowed the Walsh switching to get its deserved priority.  In the lab 
this software change made the problem go away.

Indeed, ant 14 showed no dropouts (or any other odd behavior on any 10 
second tick).  Antenna 16 continued to show dropouts (amplitude 
reductions by 25% or 50%).

This project spent alternate 5 minute scans on sources 1331+305 and 
1347+122.  At each source change there was a phase jump of -90 or 180 
degrees on the ante 14-22 baseline (the -90 deg jump being far more 
common).  Similar behavior (masked to some degree by tropospheric phase 
wander) was seen on the ant 16-22 baseline.  In fact, ante 14-16 baseline 
did not show any phase jumps -- just weather.

No phase winds were seen at either antenna, in either polarization or IF. 
Several people have suggested that Rick should schedule the next test 
since he has a certain ability to summon such behavior.

Plots on my door for the curious.

For those wishing to look more closely at the data, the experiment code 
was TBRISK on day 20051104.  X-band continuum, 6.125 MHz BW, 0.4s dumps.

-Walter