[evlatests] Frequency-dependent sensitivity at C and L bands

[Rick Perley]

    My previous report (Sunday) noted the significantly better L-band
sensitivity on all EVLA antennas at 1385 MHz than at 1485 MHz --
by about a factor of 1.3 to 1.4.  This difference is independent of the
IF pair utilized. 

    These results were based on estimated sensitivites via 'AIPS weights'. 
I have confirmed these results directly by plotting the noise histograms of
the real and imaginary parts of the signal, when looking at cold sky. 

    I have also reviewed the C-band sensitivity results from last Thursday,
and note that for this band, the BD side gives much better sensitivity than
the AC side -- by the same ratio as seen at L-band's default 
frequencies.  However, this ratio
difference is *not* seen at X-band, where all four IFs give similar 
sensitivity values. 

    Might we be seeing something related to the LSC converter?  But -- the
L301 which drives this module doesn't change when we interchange
AC and BD frequencies.  So how could the LSC converter know when we
do this? 

    I will run a short test tonight, at C-band, interchanging the AC and BD
default frequencies (4885 and 4835 MHz) in a manner similar to that done at
L-band last Friday.