[evlatests] P-Band system report ...

[Rick Perley]

            Frank asked me to check up on the P-band performance.

     I took some data on 3C286, and a nearby blank field, early in the 
afternoon.

     Antennas 15 and 23 were out of the array.

     Problems found:

     1) ea14 has *very* low sensitivity -- by a factor of 2 to 4 in the 
center part of the band to very much worse at the lower frequencies.  
The bandpass is also odd, particularly in phase.
     All other antennas have nominal sensitivity.

     2) ea28X has a much larger 3 MHz ripple than any other antenna -- 2 
dB in power (pk -- pk).  All others are less than 1 dB.  (The frequency 
structure indicates a reflection of equivalent free-space length of 50 
meters).

     3) Cross polarization is very low on a set of antennas, and very 
high on others:

     a) Very low: (< 5%)  ea01, 3, 7, 14, 16, 22, 25 and 26.
     b) Very high: (> 40%) ea02, 06, 17, 19, and 27.

     The oddity is:  the list of low - polarization antennas have very 
little correlation with the list of antennas whose P-band dipoles have 
been stabilized.  These are:  ea01, 8, 10, 14, 20, 22, and 24.  (ea 
01,14 and 22 are the only ones common to both).

     When observing an unpolarized source (which all our standard 
calibrators are at P-band), the major source of cross polarization must 
be misaligned dipoles.  (The only alternative is very poor isolation 
between the two electronics channels).

     Thus:
     a) the dipoles in the list given in 3(b) must all be *seriously* 
out of alignment with those listed in 3(a).
     b) the dipoles in list 3(a) should all be close to parallel.
     c) The dipoles in the stabilized antennas:  ea08, 10, 20, and 24 
are not properly aligned with those in ea01, 14, and 22.

     Reminder:  Keeping the cross-polarization below 5% (the same limit 
imposed at the high frequency bands) requires the dipoles to be aligned 
to ~ 2.5 degrees.