[evlatests] L-band switched power results from the band-switch test

[Rick Perley]

     35 observations of J0359+5057 were made, each one minute long, in 7 
blocks of 5 each.  In each block, the frequency was cycled back and 
forth through the other 7 cassegrain bands. System flags only were applied.

     1) For the SPWs without RFI, the switched power was extremely 
stable, with a single exception (see below).  No irregularities or signs 
of multiple levels, or instabilities were found.

     2) For one antenna-IF only, non-proper behavior was seen: ea22R has 
an ~12% modulation, of period a couple of minutes.  This is seen in 
multiple spectral windows (I have not checked them all).

     I calibrated these data with, and without the switched power applied.

     1) For the 'raw' data (no switched power applied):

     - slow gain variations, with timescales of tens of minutes, and 
amplitudes of up to 8% (in power) were common.  ea12 and ea16R were the 
worst.  For most of the others, variations were in the 2 -- 4% range.

     2) For the data with switched power applied:

     - All gain variations noted above are removed, such that residual 
variations are at the 2% level, or less, with a single exception:  ea22R 
has a ~12% (power) modulation, clearly caused by the variations in the 
switched power (PDif) values noted above. This proves that the 
fluctuations seen in the antenna-IF do not reflect gain changes.