[evlatests] Switched Power at S-band

[Rick Perley]

    The results of the on/off Cygnus with switched power on and off for 
S-band, where the system temperature when observing Cygnus is 
approximately triple that of cold sky:

    A) Most antennas show a decrement in the PDif values -- typically 5 
to 10% -- when observing Cygnus A, compared to nearby cold sky.  As at 
L-band, some antennas show no, or very little decrement.  Well-behaved 
antennas are:  ea09, ea14, ea22, ea24, ea26.  (Note that these are 
generally NOT the same ones as at L-band). 

    B) Many antennas show a small but significant zero offset in PDif 
when the switched power is turned off.  ('Small' means 0.1% or less of 
the total power).  For many of these, the decrement is larger when on 
Cygnus A than on cold sky.  Antennas which do NOT show a negative offset 
on any IF include ea09, ea22, ea24, ea26 and ea27.  Note that these are 
generally the same antennas as those which show no decrement in PDif 
when on Cygnus A.

    Minor issues include:

    A) ea06 has a variable (timescales of seconds) PDif signal in LCP.
    B) ea18's switched power did not turn off when commanded to do so!
    C) ea24 had no switched power data in IF#1 (same problem as at L-band).
    D) As at L-band, there is a wide range of power level, indicating 
that the 'set and remember' algorithm did not work correctly.  
Especially notable is ea28 in LCP -- the power level was a factor of 10 
higher than other antennas. 
    E)  In general, the S-band cold-sky powers were a factor of 3 lower 
than expected -- perhaps because the set-and-remember did its thing 
while sitting on Cyg A?