[evlatests] Wideband L band: quick look

[Rick Perley]

    I've taken a quick look, with the following top-level results. 

    Phase/amplitude connectivity will have to wait a bit. 

    1) FITLD reported 6.1% integer zeros.  Most of these are associated 
with particular antenna-IFs (see below).

    2) Four antennas have no receivers, and were flagged (via UVFLG):  
6, 10, 17, 20. 

    3) Although antenna 6 has no receiver, the visibilities are still 
there -- pure noise (as they should be), except for 6B, which was simply 
blank -- meaning that IF was all integer zero. 

    4) Antenna 1, in subband 7, LCP only, was all integer zero.

    5) Delays were large for antenna 15 in RCP, for IF A only.  (IF B 
was fine).

    6) Bandpasses all look fine.  These reveal that subband 1 is *not* 
the infamous 'subband 0' in WIDAR-ese.  The spectrum looks peculiar 
because of the general roll-off at the low edge of the band.  But there 
is no 'channel 1 of subband 1' effect which reveals the useage of the 
bad first WIDAR subband. 

    A brief scan of the data show that all the data look quite good.  A 
more detailed look will be taken later. 

   

>>
>> Basic questions are as before -- pure zeroes, partial or total
>> incoherence, frequency labels, subband ordering, gain stability,
>> ability to transfer gains from calibrator to target.  I believe
>> the subbands should be in a reasonable order now.
>>
>>
>>