[evlatests] EVLA tests, Septermber 13 2005

[Ken Sowinski]

                           EVLA Tests
                         Sept 12, 2005

The hardware people put a concerted effort into understanding 
antenna 16 today.  They had it most of the day and it was not until 
after 1700 that both antennas were available.

Downconverters were loaded with images that leave the watchdog 
mechanism disabled.  It seemed to have been the case over the last 
week or so that only MIBs with the watchdog enabled were dieing.

When 14 was pointed all four IFS were there at all the usual bands.
14C is still noisier that the others.  I tried shifting attenuaion
between the RF and output attenuators and could discern no change 
in the correlated signal for RF attenuator settings of 0, 8 or 16 dB.
Correlated signal in 14B was erratic, varying by 15 percent or more 
from record to record.  Power measurements in the downconverter were
erratic also at a level of 1-2 dB.  Backend SD was unaware of all this.

Some time was spent trying to get pointing to not work.  I ran a file
which continually looks at 3c273 and does pointing at all bands in
turn.  Every pointing scan worked.  As a result there is a collection
of pointing offsets that can be used to infer interband collimation.
L band was unreliable because of interference.  I record them here for
comparison the next time we do this.
    Az   El    Az-X  El-X
X   0.0 -0.9   
C   2.0 -0.5   2.0   0.4
K   0.9 -1.0   0.9  -0.1

The file Rick used for which pointing did not work was modified to 
run in the time at hand and sort of worked.  The test was confused
by being started before the time of the first scan and it was not 
obvious to me that it would have worked without intervention that this
necessitated.

Antenna 16 was seen to be repaired.  What seems to have done the
trick is swpping the LVDS pair that carries 19.2 Hz to the L302s.
The little remaining time was used to verify that delays were still
good for both antennas.  A few changes were needed at L band.  K band
worked at both antennas but time ran out before delays could be well
set for either antenna.  During this time both systems switched to 3c345
at 1630 LST.  The EVLA antennas seemed to go the long way around while
the VLA antennas took the short way.