[evlatests] L-band switched power characteristics

[rperley]

The 16-hour holography run on 13 March allowed a careful look at the 
switched power in the two spectral windows.

A)Good Behavior

In about 20 of the 26 antennas in the array, the switched power was 
stable and reflected the observed gains -- virtually no change in gain 
during the run.  An example of such an antenna is attached for ea20.
Plotted are (from top to bottom) are the switched power ('PDif'), total 
power ('PSum'), and Tsys.  Red points are observations on the calibrator 
3C147, blue points are for the holography scan.  The slow changes in 
PSum and Tsys reflect the changing spillover contribution as the 
antennas change elevation.  The asymmetry between red and blue points is 
due to the horizontal scanning starting 12 degrees above the calibrator 
(left side) and 12 degrees below the calibrator (right side).  The 
single blue point seen immediately next to the red points at 1/01h 24m 
is when the horizontal scan went through the main beam.

Note that the PDif (reflecting the changes in gain) is very stable.

B) Odd Behavior

For  five other antennas, the same overall pattern was found, but some 
very odd additional features occurred.  Examples are shown for ea03 and 
ea08, attached.  For ea03, three 'events', of up to two hours, changed 
the observed power sum and system temperature, but with little effect on 
the gain.  This behavior was seen on both spectral windows.

For ea05, eight ~30minute long events changed all three quantities -- 
these are not just a gain change, as the system temperature was modified 
in synchronism with the PDif and PSum.  (A pure gain change will affect 
both PSum and PDif identically, so the ratio -- Tsys -- will not be 
affected).  The effect is not large -- about 8% in PDif.  This behavior 
cannot be blamed on proximity to the others -- this antennas was at E72, 
about as far from another antennas as one can get.

Three other antennas behaved like this -- ea09, ea14, and ea16.  For all 
of these, the 'bumps' are larger in the lower frequency spectral window.

C) Weird Behavior.

One antennas did something unlike any of the others -- ea08.  In 
addition to the 'bumps' that were seen in the 5 antennas listed above 
(seen in the PSum), we had two extraordinary and extended 'gain 
changes', as seen in the PDif.  These are definitely no due to a gain 
changes -- they are not seen at all in the PSum.  Adjacent antennas show 
none of this.  And the apparent gain changes are not seen in the gain 
solutions.  (This antenna was a reference antenna, so the differences 
between 3C147 and the holography data are not seen).

Suggestions of the origin of these weirdnesses are welcome ...

Rick
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ea20Plots.png
Type: image/png
Size: 79439 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20210408/17ca0d8d/attachment-0004.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ea03Plots.png
Type: image/png
Size: 78537 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20210408/17ca0d8d/attachment-0005.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ea05Plots.png
Type: image/png
Size: 76964 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20210408/17ca0d8d/attachment-0006.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ea08Plots.png
Type: image/png
Size: 56009 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20210408/17ca0d8d/attachment-0007.png>