[evlatests] Does Switched Power Monitoring Help?

[Rick Perley]

    ... by which I mean, does the application of the detected PDif to 
the visibility amplitudes remove gain fluctuations introduced by the 
electronics? 

    The short answer is 'YES!', but there are caveats. 

    A) There is no question that in most cases, short-term variations in 
raw observed visibility amplitudes are effectively removed by 
application of switched power.   There are spectacular examples of large 
gain changes which are completely removed. 

    B) There are a few clear counter-examples to this -- and it will be 
important to understand why.  In principle, these will be due to effects 
outside the measurement loop:  either due to atmosphere or antenna 
pointing, or something between the station board and the baseline board 
(unlikely, I hope!). 

    C) There remain, following application of switched power 
calibration, unexplained long-term 'drifts' in visibility amplitude, 
which may be an elevation effect, and/or a temporal drift', and/or a 
miscalculation in atmospheric opacity/antenna gain in post-processing 
(unlikely, I think), or something else.  If this effect is 
elevation-dependent, and repeatable, it is not too worrisome..., but we 
would like to understand its origin. 

    D) The switched power measurement system is high susceptible to 
RFI!  (This is no surprise, given the wide 128 MHz of bandwidth 
utilized).  Both L and S band observations show 'false' power readings 
which are obviously induced by RFI within the bands.  This observations 
emphasizes the need to use 'safe' subbands when monitoring system gain 
for observations in RFI-full subbands. 

    The detailed results are sufficiently diverse that conclusions 
beyond the above are not warranted.  I plan to organize a meeting for 
interested parties, at which I can show some of the plots, with the goal 
of better understanding what is happening.