VLA Test/Observation Coordination Meeting
                           B.G. Clark
                         March 20, 2003


1.  New K band receivers

This installation is approaching complete.  Y. Pihlstrom briefly
summarized.  The system temperature has decreased with serial
number, from about 35 K for the first few to about 24 K for the
latest ones.  The latest version as the amplifiers developed 
for MAP, which are lower noise and more uniform across the band.
The earlier versions are pretty good across the current VLA band,
but have problems at lower frequency.  One of the problems with
the earlier units is that the feed was constructed with the loaded
ring section of the throat serving as a structural element, which
makes that segment rather unstable and finicky to adjust.

2.  eVLA

Antenna 13 will come into the assembly building to start the eVLA
mechanical retrofit in less than a month.  The eVLA is upon us.
B. Clark commented on the expected timeline.  Mechanical retrofit
should be finished in July, and installation of eVLA electronics
will start at that time.  We should have actual signals from the
eVLA electronics in August.  The software to point the antenna may
be available as early as the end of August.  We may hope for first
fringes perhaps a month later.  First use of the antenna for routine
science (though still with the VLA correlator of course) is expected
in the first quarter of 2004.

P. Napier says that the antenna will be equipped with the current 
selection of bands, of which the K band and Q band receivers will be
the new style receiver; the rest are old receivers connected to the
new electronics backend.  It is hoped that the new L band feed will
be installed in August.  The new, octave bandwidth, L band receiver
will not be installed until fall, maybe in October.

3.  New controller for VLA correlator

K. Sowinski says that programming for the new correlator controller is
progressing, and he is currently finishing the design for the spectral
line part of the software.  He proposes to eliminate several features
in the current software that he feels are more confusing than they are
worth.  He is interested to receive arguments that some are necessary
if that is the case.  The features are:  1) bandpass normalization by
the total power bandpasses, 2) calculation of "channel 0", the summed
central part of the spectrum, 3) channel selection, and 4) hanning
smoothing.  He argues that the last two may be adequately done at 
data reduction time.

4.  Odds and ends

M. Goss had an example of a two subarray observation that apparently
just died (data were just noise).  Array apparently resumed working
properly with the next observing program.

C. Chandler complains that data is not flagged bad when antennas are
dumping snow.

E. Greisen is working on real-time FILLM, trying to make it more 
reliable, and requests failure reports, and especially would like to
catch it in the act of failing.