[evlatests] C and X band SDM1 data

[Rick Perley]

    Michael placed various SDM1 data sets out for review. 

    I grabbed the three X and C band 'baseband' ones.   These are short 
observations of 3C273 (a heavily resolved double, unfortunately), with 
0.1 seconds averaging.   Six antennas were included for all the tests. 

    A)  X-Band.   There are four subbands, with dual polarization.  The 
FQ table looks correct.  The data quality looks excellent, with a single 
exception:  In sub-band 1 only, in both polarizations, there are two 
short periods where antenna 28 produced exact zeros.  Since noise can't 
do this, this is an 'absence of data' issue.  All channels are zeroed. 
       There are about a half dozen data 'dips', where the visibilities 
are near, but not at, zero. (A negligible fraction of the total data). 
       Since the spectral and subband ordering is reversed, I ran FLOPM 
to re-order the data in the way we like to see. 
       FRING aborted on this database -- it's not supposed to do this, 
and Eric will be investigating. 
       POSSM provides beautiful spectra. 

    B) C-Band 1.  A dual polarization, single-subband database.  All 
looks normal here, with only a few 'data dips' noted.  FRING and BPASS 
both ran normally, and the resulting spectral look very nice. 

    C) C-Band 2.  A dual polarization, four-subband database.  Again, a 
very few 'data dips' are seen, and easily removed via CLIP.  FRING and 
BPASS ran normally.  I ran CALIB, and reviewed the spectra.  The 
resulting spectra do not look right in that the various subbands are 
giving rather different (by 10s of percent in some cases) amplitudes.  
However, as the source is a partially resolved double, some oddities are 
to be expected.  I don't have a good source model handy, and don't think 
it worth the trouble to chase one down.  It's better to take some data 
on a real point source...