[daip] [!6915]: AIPS - GBT data (Lband) - phaseing/scheduling issue?

[Michael Bietenholz]

Michael Bietenholz updated #6915
--------------------------------

GBT data (Lband) - phaseing/scheduling issue?
---------------------------------------------

           Ticket ID: 6915
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/6915



I have a phase-stability, or possibly scheduling issue with the GBT data from  GB076 (28-OCT-2014; 2048 Mbits/s; 1.6 GHz), where there are phase jumps between contiguous scans where both scans were pointed at the same source.

The schedule for this was a bit complicated.  It was a phase-referencing run, so the scans nominally alternated between shorter (~1min) scans on our calibrator (M81) and long (~3min scans) on our program source (SN1993J).  However, except for the phase VLA and WB, both sources are in fact in the beam, so our strategy was to actually use in-beam phase-referencing.  So, for all antennas (except VLA and WB), all scans were actually pointed at the midpoint between M81 and SN1993J (MIDPOINT).  We then obtained two correlations, one at the M81 position and a second at the SN1993J position.

In the M81 correlation (I haven't looked at the second correlation yet), for GBT (as for most antennas), all the scans should have been pointed at MIDOINT, and correlated at the M81 position, and there should be no difference between the short and long scans.

However, I noticed systematic phase jumps between the two sets of scans for GBT (but not for other dishes).  I attach a VPLOT output, GB076_visphases_GBT.png, showing a short example section of the uncalibrated visibility phase on a selection of GBT baselines.  The grey lines mark the scan boundaries.  Note that phase discontinuity between the long and short scans.  I also attach a corresponding plot (GB076_visphases_EF.png) showing the visibility phases on EF baselines - where the phases (except for EF-GBT) are continuous.

Below is a chunk of gb076sch.gb correspondng to the start of the plotted segment, retrieved from
http://www.vlba.nrao.edu/astro/VOBS/astronomy/oct14/gb076.  According to this, anyway, GBT should not have switched sources.

So where do the phase-jumps come from??  For GBT, nothing should have changed between the short and long scans, so vis-phase should be nice and smooth (which is how it is for all the non-GBT baselines).  Did GBT actually switch sources, even though the sch file says not to?  Did the correlator do something funny?

...from gb076sch.gb...

01 39 30  MIDPT        22 49 24  17.8   4.9-11.1     -10.8   -20     1585   01 39 50
01 42 10  ---          22 52 05  17.9   5.2-11.1     -11.4   140     1603   01 39 50

01 42 10  MIDPT        22 52 05  17.9   5.2-11.1     -11.4   -20     1603   01 42 30
01 43 30  ---          22 53 25  17.9   5.3-11.1     -11.6    60     1611   01 42 30


01 44 00  MIDPT        22 53 55  17.9   5.4-11.0     -11.8    10     1611   01 44 00
01 46 40  ---          22 56 36  18.0   5.6-11.0     -12.3   160     1631   01 44 00

01 46 40  MIDPT        22 56 36  18.0   5.6-11.0     -12.3   -20     1631   01 47 00
01 48 00  ---          22 57 56  18.0   5.7-11.0     -12.6    60     1639   01 47 00


            cheers,           michael



------------------------------------------------------
Staff CP:  https://help.nrao.edu/staff
-------------- next part --------------
A non-text attachment was scrubbed...
Name: GB076_visphases_GBT.png
Type: image/png
Size: 16872 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/daip/attachments/20150720/edf58a64/attachment-0002.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: GB076_visphases_EF.png
Type: image/png
Size: 34034 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/daip/attachments/20150720/edf58a64/attachment-0003.png>