[evlatests] Non-linearity of ea24 at C-band remains ...

[Rick Perley]

    I reported a couple weeks ago on the reduction in PDif when we 
observe the strong source Cygnus A.  
    At C-band, the antenna which has by far the largest reduction is 
ea24, in RCP.  IF 'A' had a reduction of 30%, IF 'B' was reduced by 
12%.  (By contrast, IFs C and D had virtually no reduction at all).

    Because the T302 may be implicated in this behavior, Jim and Rob 
replaced antenna 24's T302.  They asked yesterday if the behavior seen 
last month has changed.  I repeated the on-off Cyg A test this morning. 

    The reduction in PDif when on Cyg A is *unchanged* on this antenna 
(and indeed remains the same on all the others also).  It would appear 
that the non-linearity seen on this antenna is independent of the T302. 

    Worth emphasizing is the spread in behavior of PDif.  For most 
antennas, there is a reduction when going onto a strong source.  But for 
some antennas, there is an enhancement!   Most notable are 4A (+10%), 
23A and B (+5%), and 28C (+12%). 

    The change seen is generally different from IF to IF (so that A and 
B are generally different).  This would seem to exonerate the T302, and 
implicate the T304.  But on the other hand, IFs that show severe 
compression at one band show little or none at another band.  So for 
example, 24A, which is easily the worst at C-band, is one of the better 
antennas at S and L bands, and shows no effect at all at X-band.  Given 
this, how can the T304 be responsible? 

    It's all very odd...