[evlatests] Running VLA antennas

[Barry Clark]

I am still doing something wrong with getting the waveguide cycles 
syschronized.  The command I send to the CMP is 
"reset at <time>"
This morning the time I was sending that successfully synchronized the
Walsh functions was "xx:xx:x9.647" for a time in UTC, corresponding to 
"xx:xx:x2.647" in TAI.  This was stable and reproducible.  This afternoon, 
that doesn't work. I am now sending "xx:xx:x9.699" (one waveguide cycle
different).  This again works stablely and reproduciblely.  I suspect that 
the modest drift in NTP time on the network has taken me across some critical 
time in the waveguide cycle.  Does anybody happen to know the phase of 
the waveguide interupt in the CMP relative to the site 10s tick?  Anyway, 
I shall assume it is offset by a large fraction of a cycle, and next time 
I work on this I will try "xx:xx:x9.673" and see if it works all day.  
Maybe.

There are clearly also problems with antenna setups at bands other than
C.  Some are fatal, some merely curious.