[evlatests] AR642 30jun07: ok for 17mins, dead for remaining 40mins

[Michael Rupen]

Hi Ken,

   I successfully loaded all the data using on-line FILLM.  The problem
is that the data loaded are mostly useless.  The project code problem would
show up as no data whatsoever.  I've run into that one as well ;)

   I did note the lack of a calibrator code and the difference in naming
conventions for fast-switching scans, both of which are annoying.  I didn't 
bother mentioning those since they are known problems.

   I realize fast switching stresses the system somewhat, but this is a very 
important mode for A configuration, and I'd rather stress the system now, 
injuring data for "us locals", than have this hit more widespread and less 
wary observers later...

   Cheers,

           Michael

On Sat, 30 Jun 2007, Ken Sowinski wrote:

> On Sat, 30 Jun 2007, Michael Rupen wrote:
>
>> AR642  30jun07  02:42-03:40
>>  dynamic run, fast-switching, continuum, C-band, 3.3s averaging
>> 
>> Summary: looks ok for 1st 17 mins (a couple bad flags), then no fringes
>>   for the remaining 40 mins, including one long non-fast-switching scan on
>>   3C286.  2/3 USELESS FOR SCIENCE.
>
> Michael,
>
> There has been an extended non-public discussion about how
> project codes are getting lost or distorted on the way to
> the archive record.  Fast switching is the easiest way
> to trigger this.  I am sure your data is in the archive but
> the project code is not what you expect.  Also, if you do
> find the data be warned that the calibrator will not be marked
> with a non-blank cal code.
>
> Ken
>
>