[evlatests] EVLa tests, Oct 27 2005

[Ken Sowinski]

                     EVLA Tests, Oct 27 2005

Much of the time today was spent getting things in order.  Delays 
were checked and corrected at all bands.  14B had changed by 22 ns 
probably because of a module swap some time ago.  Focus was checked 
at K and Q band because of Rick's report of focus offsets while 
observing Venus.  At an elevation of 57 degrees antenna 14 was 
within a millimeter of the optimal setting at both K and Q band.

There was a power glitch during the night.  It was reported that 
X and L were warm in antenna 14 and X band in antenna 16.  From
a quick survey of G/T at all bands it can be verified that the
X band system in antenna 16 was very warm.  G/T results, in arbitrary
units,  estimated from the measured correlation coefficients.
Band 14    16     VLA
Q   1.5   ---   1.4-1.6 (with pointing corrected)
K   2.4   2.6   best antennas are 3 to 3.5; most between 2.0 and 3.0
X               unreliable because of cryo problems
C   11     9    13 to 16    (Bob and Rick report the receiver in 16
                             is not as good as it should be)
L   16    16    best antennas ~22; most between 18 and 20
I would guess that EVLA antennas look more like VLA antennas at
K and Q band because there is a larger variation of performance
in the VLA antennas for the high frequency bands.  The Q band
performance of antenna 14 hints that there is no serious LO problem.

Q band was working well enough on 14 to warrant a pointing check to
pin down the collimation relative to X band.  The errors found were
small but improved the response by about 30 percent.  The system
misbehaved frequently during the pointing check.  It never was
reliably in sync with the Modcomps for more than about half an hour.
The program had to be started twice to work at all; the first time it
seemed not to respond to Modcomp on-source signals.  Inspired by this
I spent the last hour or so doing normal X band pointing which worked 
fine.  My guess for now is that something was wrong in the translation
from .OBS to .evla for observe file used in the collimation tests.

Focus was checked again at an elevation of about 30 degrees.  K band
wanted to be about 5 mm more negative and Q band about 2 mm.  The K band
observation is a little surprising because the elevation coefficient 
of focus for VLA antennas is about 6 mm between horizon and zenith.

The downconverter in 14C was changed yesterday.  This was apparant in
the level setting at Q band.  The required RF attenuator setting is now
21 dB instead of the former 10 dB.  Other bands were not checked but
there seems not to be any large changes.  Perhaps someone can explain
why the change is not the same for all bands.

During the pointing check over lunch MIBs in all four downconverters
in 16 stopped responding.  They were reset and have worked since.
They did fail again not long afterward.  Pete and the HW people seem
to have worked out a short term course of action to investigate the
failures.  If a downconverter is swapped attenuator setting will have
to be checked again.

Except for the troposphere effects while at Q band, the phase behavior 
looked normal during all this work.