[evlatests] Basic Gain Curve Results

[Rick Perley]

    As noted earlier, we got an excellent dataset to establish the VLA 
antenna gain curves as a function of elevation.  Four frequencies in 
each of X, Ku, K, Ka, and Q bands were used, following a 10 Jy object 
from transit to setting, from 11PM through to dawn.  The time was chosen 
to minimize any temperature-depending effects.  Self-referenced pointing 
was done for every band to minimize pointing offsets. 

    Because we are changing bands, significant gain variations, 
presumably arising from the switches, commonly occurs.  These can only 
be removed through application of the PDif switched power.  For most of 
the antennas, this works very well.  But for some, there are 'issues', 
particularly troublesome at K and Ka bands.  I have no idea why these 
bands have less reliable PDif values. 

    For most antennas, we have obtained excellent gain curves.  One of 
the simplest parameters to deriver, and one which has close connection 
to the efficacy of the 'subreflector rotation trick' is the elevation of 
maximum gain.  The general goal is to arrange this to be about 50 
degrees -- the 'trick' gives us some latitude in adjusting this.

    Below is a table, showing the maximum gain elevation angle for Q, 
Ka, K, and Ku bands.  There is nice uniformity for some bands on some 
antennas, and puzzling differences between them for others.  I argue 
here that those antennas/bands for which maximum is 60 degrees or higher 
(and some are *much* higher) need to be addressed. 

    Ant.       Q         Ka         K          Ku
---------------------------------------------------
01            60          60         80         75
02            70          55         80          80
03            50          50         50          X
04            55           X         45          45
05                      out of array
06            55          50         65          75
07            40          45         40          50
08            X           55         65           X
09                      out of array
10            40          50         40          40
11            65          60         40          65
12            45          65         30          35
13            50          40         60          60
14            60          65         60          50
15            75          60         75           X
16            40          35         30          45
17            50          35         55          60
18            70          55         80         >80
19            45          35         55          55
20            X            X         X          ~70?
21            65          60         55          75
22            60          60         60          65
23            70          55         X           70
24            55          65         ??           80
25            45          45          X          45
26            55          55         55          60
27            30          30         35          45
28            50          35?       30?          X
--------------------------------------------------------------

    Some notes:

    1) The PDif values for ea20 are nonsensical at Q, Ka and K bands.  
There is no connection between these values and reality.  The situation 
is hardly better and Ku and X bands.  Something is truly broken in the 
switched power for this antenna.
    2) A similar comment can be made for ea28, although the disconnect 
from reality is not so great. 
    3) An 'X' in the table indicates either 'no receiver' or 'no 
functional switched power'.  Values with a '?' mean that the variation 
between the eight data streams (four frequencies and 2 polarizations) 
was so great that no convincing maximum gain angle could be found.