[evlatests] ea06 and ea16 S-band PDif variations

[Rick Perley]

    I have previous reported on the apparently ever-present oscillations 
in the switched power from ea06 and ea16 at S-band. 

    I've trolled through all 24 recently-taken databases from a 
recently-completed project.  Each is at S-band, each is from 1 to 3 
hours long, gathering basic statistics on this curious phenomenon.  
Below are the basic results, which show great consistency.

    Fourteen of the databases have both ea06 and ea16 present.  (ea06 
was out of the array for the others).    For *all* of these, the varying 
PDif was present in:

    a) All four IFs in ea06

    b) In the 'A' and 'B' IFs only (i.e., RCP only) in ea16.   
           -- The remaining 10 databases, with ea16 only, show the same 
result -- oscillations only in RCP. 

    c) The variations are identical for all subbands for a given 
polarization.  In other words, the variations are the same in IF 'A' and 
IF 'B', and in IF 'C' with IF 'D'. 

    d) The mirror asymmetry described before is actually between 
opposite polarizations -- not between ea06 and ea16!   In other words, 
the variations in RCP are mirror anti-symmetric in LCP (this for ea06 
only -- ea16 has no variations in LCP). 

    e) No other antenna shows these variations (to about 1% in amplitude). 

    f) the amplitude of the phenomenon varies between 2 and 10 percent, 
with a higher typical value, and greater variation in amplitude, in 
ea06.  ea16 has a very constant amplitude -- typically 2.5% in both IF's 
A and B.  The variations in ea06 in IFs 'A' and 'C' are usually larger 
than in 'B' or 'D'. 

    g) The time-scale of the phenomenon is remarkably variable, with 
rather vast variations of a few seconds period transitioning, within a 
few minutes, to slow 'wobbles', of timescales of minutes.  The character 
of the opposite polarizations are not identical. 

    h) There can be no dependence on 'u' or 'w' (baseline coordinates) 
-- the changing of the nature of the variations is just too rapid. 

    i) These variations cannot be due to system gain variations -- there 
measureable effect in the total power. 

    There has been talk of these being somehow due to the RFI in the 
band -- but this is hard to find evidence in support.  The variations 
are there whether the total power is 5 or 15 counts.  They are never in 
any other antenna.  They are always in ea06 and ea16, no matter where 
they are, or where they point.   These variations are never seen (so far 
as I know) at any other band. 

    Perhaps some re-wiring of ea06 (swapping modules, for example) might 
help clarify where this is coming from.  Given the inverted response 
between R and L, this would seem to exonerate the front end, at least.