Notes from Commissioning Coordination Meeting
13 September 2002

1.  Observing Program Checkout

Carl reported that last week, two programs were moved within the status
queues, one from "active checkout" to "scheduled," and the other from
"scheduled" to back to "active checkout," because it needed more checkout.

Richard and Ron reported on spectrometer standard mode checkout.  Working
with Ron, Dave Hogg did checkouts on 5 spectrometer modes.  All of these
have problems and need further work.  Most of these use the low speed
samplers, and exhibit changes from scan to scan.  Richard has put the
results of some of the spectrometer checkouts on the web at

http://www.gb.nrao.edu/gbt/gbtstatus/checkouts/index.html

Ron did program checks on 2B5 and 2B6, both for Yusef-Zadeh.  Program 2B6
can be scheduled following one more session to check multi-IF modes.  In
the tests for this program, bad baselines were found about 30% of the
time, so a little more observing time should be scheduled than requested
to compensate for this.  Program 2B5 involves looking at a strong
continuum source.  This one is exhibiting bad baselines and needs a more
sophisticated way of spectrometer balancing (which is in the works by the
software group).  Dana sent a separate email describing the tests he has
been doing on programs he has been assigned.

Joe McMullin has the first cut at multi-IF support in AIPS++ (multiple
frequency windows in a given MS), and is ready for initial testing.

Ron listed the spectrometer mode strings that were required by programs in
the active checkout queue, which are

IN4 12-9		4W2 200
IN8 12-9		2W4 200

IN4 50-9		4W2 800
IN8 50-9		1W2 800
			2W1 800

Based on the work of the past week, three areas need addressing, in this
order of priority:

1.  Initial astronomical checkout of the spectrometer modes listed above.
2.  Engineering followup on these and the other modes with known problems.
3.  Other system issues such as bad baselines that are affecting the
programs being checked out.

Richard and Ron have the action to organize this effort early next week:

Action 13sep02-1: RMP & RJM -- Organize checkouts of priority spectrometer
modes and followup on identified problems with program checkouts.

2.  Spectrometer engineering status

Rich reported that he spent much of the past week chasing gremlins.  He
sees two types of pulses in the ACFs, a big one that appears
intermittently, and small ones that are always there.  The small ones
probably are contaminating the spectra in a small way, and the big ones
may significantly contaminate them.  There are several possibilities for
the source of these, which include digital hardware errors, data transfer
errors, or software/firmware errors.  It will take a while to track this
down.

One bug was fixed last week, a coding error that caused the integration
time in the FITS file to be wrong.

Ray Escoffier will be over next week to work with Dave Kaplan on the
pulsar spigot card.

3.  Software report

Nicole handed out a software status report.  Most projects are progressing
well.  The observers interface is now behind schedule owing to problems in
moving the code into a production environment with version control.  The
group hopes to recover some time next week.  The antenna improvements
project is underway, and the PF2 manager is ahead of schedule.

4.  Baseline investigations.

Dana reported that Roger will be working with the high pass filters in the
L-band receiver that block aviation RFI signals below 1200 MHz, that may
be distorting the spectrum.

5.  RFI work

Rich reported that the filtering of the prime focus servo system is
proceeding.  At some point in the next month or two, two nights of
downtime will be needed to allow this work to be finished.  Carl and John
will coordinate this.

6.  AIPS++ news

Joe McMullin reported that a new Stable would be out this coming week.  It
will have all the recent updates to the Filler and the multi-IF support.
The group has also been working on a convenient method to provide an
export of a current snapshot of the system to users who need an update.

As reported above, the first cut at spectrometer multi-IF support is in
place.  Joe noted that he pushed this through pretty quickly and did not
have as much time to test it as he would have liked.  Commissioning
scientists should look at this functionality carefully and report any
problems to Joe.  Joe noted that he has also added a new convolution
function that will be of use to imaging reduction.

Bob has implemented a new feature called gbtpicker that allows one to pick
individual scans out of a Measurement Set.

In other comments, Joe noted that he is thinking about coming to GB for a
visit in mid to late October.   Richard emphasized that among priorities,
multi-IF support is still highest.  Ron noted that he was seeing
inconsistencies between IARDS displays and Dish reduction displays.  Joe
said this was probably because the gbt version of AIPS++ was being used by
IARDS.  This will be updated soon.


Action summary:

Action 13sep02-1 -- RMP & RJM: Organize checkouts of priority spectrometer
modes and followup on identified problems with program checkouts.

							PRJ // 14 Sep 2002