[evlatests] Tests, Jan 6 2006, and pointing

[Ken Sowinski]

When we began 14C was dead, there was not TP at the VLA T5.
Rob and Jerry found that there was not power at the deformatter
output spigot either and restored it by cycling power to the
deformatter.  C band is not available at antenna 16 until the
smoke is put back in the C band card cage.

After we got things to work we tried an experiment to learn more
about the "dead scans" that Rick has reported.  A file was created
which changes among all bands frequently and ran it using two 
executors: antenna 14 was controlled by the executor the operator
invoked by the EVLA control web page, and antenna 16 was controlled
by an executor which I ran from a shell.  If there are any failures
it may be instructive to see if the both fail or only one at a time.
This sequence began about 1822 IAT.  Vivek will report what is seen
when he has digested the data.  I think this marks the first time
that we intentionally had three executors running: one for the
operator controlling antenna 14, one for me controlling antenna 16,
and one for Barry controlling a handful of VLA antennas in a second
subarray.

Barry reports that there are still problems with the Walsh sequences
generated by the CMP.


Antennas 14 and 16 were included in a pointing run Tuesday night.
They are still in reasonably good shape; 14 is a little better than
16.  a surprise is that when fitting for refraction the answer was 
zero.  Apparantly the weather model used by calc is a good match to
this particular January night.  The pointing fit is not quite as
good as from the last pointing run and I intend to look at this
more carefully to see if anything systematic can be discerned.

The EVLA antennas did not make it into the following collimation run
because the wrong script was used.