[evlatests] Amplitude and Phase Stability

[Rick Perley]

    I used 2 hours of test time Friday night to check EVLA amplitude and 
phase
stability.   The data were taken in 50 MHz continuum, with 400 msec 
averaging. 
Six northern sources were observed cyclicly, to check overall phase 
connectivity.
All 8 IFs were functioning well. 

    1) Amplitudes.

    Antennas 14 and 16 are both suffering occasional amplitude 
dropouts.  These
are always short (nearly always 400 msec, or less, very occasionally affect
two consecutive 400 msec integrations), and almost always affect both 
antennas
at the same time.  Some scans show no drop-outs at all.  The effect is 
seen only
in amplitude -- phase is unaffected.  The drop-outs within any one scan are
*always* separated by 10 seconds, but the epoch of drops is not the same in
different scans.  For example, in one scan, the dropouts will be 'on the 
5s',
while on the next one, they are 'on the 1s'.  All IFs are affected in 
the same
way. 

    2) Phase. 

    Within any scan, phase behavior seems normal.  But there is no phase
connectivity for antennas 14 and 16 between scans.  This seems to be true
not only for adjacent observations (of different sources), but even for
the separated observations of the same source!  This behavior seems 
qualitatively
different from that seen last week, when we saw what seemed to be a clear
sign of a large Bz error -- and which Ken quickly explained was due to an
incorrect value of the speed of light. 

    3) Sensitivity. 

    Antenna 14's sensitivity, as judged the the calibrated ampscalar rms on
scans when dropouts did not occur, seems nearly as good as, or equal to
that of good VLA antennas.  It is certainly *not* more than 10% worse. 
Antenna 16's sensitivity, as judged by the same means, is poorer than
VLA antennas, by a few tens of percent.   This may be due to a pointing
error -- Ken S. will shortly report on the results of the pointing run, 
taken
immediately following these tests.