[evlatests] evla decay

[Ken Sowinski]

Several things were done with test time today.

Rick made a series of observations to capture more
information about the 8797.5 MHz birdie.  That is
now begin digested.


There have been various reports over the last few days
of failures or poor performance.  I have been able to
track most of these down and provide a summary.

1.  Bob Hayward and I responding to Michael's report of
this morning, found that many EVLA receivers on antennas
14 and 16 were warm.  It might be worth speculating about
whether there is some common cause of the refrigerators
tripping and some of the other problems described below.

2.  Three L301s are not functioning.
ea14-l301-1 goes to standby and does not work better after
being rebooted.
Neither l301 in antenna 13 responds, but the BD L301 seems
to still be working at the frequency for C band.
The result of this is that 13 and 14 fully work only at
X band.

3.  14C looked weak at the bands for which it worked at all;
the autocorrelation spectrum looked unnatural.  Resetting
the DTS module formatter board fixed the problem.

4.  16D was driving the sampler to hard at, at least, K and
L bands.  This was enough to reduce the reponse by a factor
of two.  I left the attenuators unchanged in the hope that
once the receivers were cool again things would look better.

4.  At L band the main loop in the L302 for antenna 18AC was
going in and out of lock.

5.  16BD is dead at Q band.  No obvious reason.

6.  24 still works at X band, but IF C is about a factor of
ten too weak.  Again there are no obvious reasons.  My guess
is a spurious signal in the passband presented to the sampler
or in the output of the transition module.


The remainder of the time was devoted to a long observation
of 3c273 with short integration time.  The intent was to
provide more information about the ten second dropout, but
Walter's announcement may make that moot.  I note that,
in contradiction to Walter's explanation, that 24AC was glitching
when this observation began.  In any case this may provide
useful data with five EVLA antennas for Rick to examine EVLA
closure properties.