[evlatests] Strange goings-on for AW703

[Mark Claussen]

I took a look  at AW703, observed on July 29, 2007 as part of
checking on modcomp-free science observations, and found, that,
unlike most of the data sets I have looked at, this project had
some pretty strange goings-on.  I post this to evlatests in order
that others might comment or try to figure out what happened.

This project was standard X-band continuum.  The start IAT was 19:12.  
There were a couple of user errors that made this project iffy to
begin with: 1) the source names had embedded blanks, and most targets 
were 3C or 4C sources so the names were like "4C 51.2".  Thus AIPS
interpreted all these sources as the same object with different 
qualifiers.  Will be pretty tricky to sort out; 2) the project
code was ADA000 rather than AW703, so the data showed up in the
archive under ADA000, and marked *public*.

The only log info about this project rather than the usual messages
were that 1) Antenna 18 was stowed because of a foc/rot problem and
that 2) for about 1/2 hour (2035 - 2104) VISPIPE gave no data.

The curiosities with this data were:

1) For most scans, there is some period of time at the beginning
(can last up to minutes) where there are no fringes for any combination,
then at some point the fringes pop up and are there for the rest of the
scan, as if the antennas are off source or something.  But the data is
not flagged by the online system.   A specific example of this is:

     There is a 10 minute scan on 1411+215 in the middle of the
     run (215620 - 220600, scan listing).  For all antennas, there 
     are no fringes on this source until 220430.  Nothing in the
     log, and the data is not flagged.  After 220430, fringes on all
     baselines look good.

2) There is data coming from Antenna 18.  No fringes, as would be expected
since the antenna was apparently stowed, but the data wasn't flagged.

3) Antenna 11 had data on EVLA baselines.  There were no fringes, and data
11-VLA baselines must have been flagged.

4) For at least one scan and possibly others, there were data only for
EVLA-EVLA baselines.



It seems like something was fairly messed up for this run, as opposed to
other data sets that I have looked at recently, where the data looks very
good.

Mark