[evlatests] Peculiar switched power behavior ...

[Rick Perley]

    The 'Demo Science' Herc A data from two nights ago are now on disk, 
and I'm beginning the process of reduction. 

    The run lasted a bit less than 7 hours, starting at 2:00 AM, local 
time, and ending near 9AM. 

    In reviewing the X-band data, I noted some remarkable behavior in 
the switched power from antenna 1.

    A) On both the RCP and LCP sides, there is a smooth decline in the 
PDif measurements over a duration of 2.5 hours (2:30 -- 5 AM), 
indicating a drop in gain of about 7% (if the calculated PDif values are 
truly proportional to power and not voltage).  Although most other 
antennas show smooth variations, none are anywhere near as large as 
antenna 1!   The variation is certainly real (meaning, it is telling us 
something about gain changes), as it affects both PSum and PDif equally 
-- the values of Tsys do not change throughout.  Or, it is telling us 
something about a change in the value of the power produced by the noise 
diode?  An inspection of the visibility amplitudes can reveal which (if 
either) of these explanations is correct.  This remains to be done.

    B) Perhaps more disturbing is what is happening on the LCP side 
only.  There are a number of small 'steps' in the reported value of 
PDif, probably initiated by band changes (I observed cyclically through 
X, C, S and L bands).  What is particularly odd is that the number of, 
and size of  these steps ***depends on the subband!***.  For example, in 
IF 'C', subbands 3, 5, and 7 show 5 periods where PDif is low by 7% 
compared.  However, subbands 1 and 2 show 7 such periods, one of which 
is low by 7%, the others are low by only 2%.  subbands 4 and 6 show only 
a single drop, of 7%.  Meanwhile, in IF 'D', subband 1 shows 4 periods 
of PDif low by 2%, while subband 2 shows the same 4 periods, but they 
are now *high* by 2%.  The remaining 6 subbands in that IF show now 
visible steps at all! 
    I can't see how this behavior can be explained by a bad band 
switch.  Can anybody offer a plausible explanation? 

    Each antenna has its own interesting behavior.  I'll try make some 
sense of it all, and give a summary later today ...