[evlatests] Bizarre Gain Stability

[Rick Perley]

    100 minutes of wideband mode data were taken Thursday night at 
L-band.  Most of this time was spent tracking 3C273 (presumably with 
lots of short scans).  I have reported earlier on results from a short 
observation of 3C84 -- here I report on the behavior from the 90+ 
minutes of tracking 3C84.

    Data quality appears good in general -- all subbands are present, 
and so far as I can tell, there are no missing pieces.  The fraction of 
identical zeros at times more than 1 minute from change of source is 
miniscule. 

    What is notable in this observations -- and which I have not seen 
before -- is a weird assortment of amplitude instabilities affecting 
perhaps half of the antennas.  Since 'a picture is worth 1000 words', 
I'll spare the readers a length discussion, and show some gain plots at 
the next appropriate meeting.  But a brief summary is perhaps in order:

    A)  Some antennas/IFs showed excellent stability throughout the 90 
minute track:  2C, 4C, 7A, 8C, 9, 11A, 12, 14, 16,  19, 21, 23, 25A, 27C. 
    B) Some antennas showed excellent continuity (= no jumps), but had 
gain changes which are far beyond what is possible at the band and 
elevation (45 degrees) over so short a time.  Notable is antenna 13, 
where in both A and C the fringe amplitude rose by 6%.  This cannot be 
pointing, or weather, or shadowing.  It could be the amplifiers changing 
their gain -- but having both polarizations do this identically is 
asking a bit much.  What is especially intriguing is that *all* the 
outermost north arm antennas behaved in the same way:  besides antenna 
13, I see the save behvior in 7, 18, 22, and 28. 
    C) Much more troubling are the discontinous gain jumps seen on all 
the other antennas -- usually in great numbers -- dozens of jumps in 
some cases.  Only a few of these are of the '1 dB' variety.  The jumps 
are always independent of each other -- there is no commonality in jump 
amplitude or timing between IFs of the same antennas, or between 
antennas in the same IF. 
    D) Phase stability is generally excellent.  But a few antenna-IFs 
show phase discontinuities which match the amplitude jumps:  on 3C, 4C, 
5A, 22A and 25C.  I will show plots to illustrate these. 

    All the above jumps etc. are the same on all subbands, to the extent 
that I can tell with the RFI so abundant in some of them. 

    I have not tried any imaging, and probably will not have the 
opportunity to do so since I leave on Tuesday.