[evlatests] Phase jumps on EA14

Wed Apr 24 11:25:09 EDT 2013

I've been contemplating pretty much what Barry suggests, with some
additional attention paid to what the R-L phase might do.  The
practical point of a reference 'antenna' is two-fold: enforce
interpolatable continuity among solutions as a function of time
(i.e., the presumption made when choosing the calibrator cadence),
_and_ (for polarimetry) enforce some sense of R-L phase stability
(a presumption we are fond of, and which is probably reasonable,
but which we have not really quantified from a commissioning point
of view, to my knowledge).

Relying on any single antenna (for all subbands for whole
observations) will often be problematic (flagging), and a single
antenna probably isn't the formally best R-L reference, in general.
The upshot is probably that a traditional referencing with a specific
antenna is applied at a specific time (to establish a definite frame),
and then minimum median deviation (over time) and median R-L stability
(over fields) over the _ensemble_ of available antennas is enforced.  This
is
a stronger and less vulnerable realization of the presumptions
we want to make.   Use of  medians in this enforcement will render
any minority of misbehaving antennas (including phase jumps, and
also other more-subtle pathologies like differential phase drifts)
(a) irrelevant to the referencing operation itself, and  (b) more
plainly obvious  for flagging or exclusion purposes.

For the R-L part of all of this, external factors like the ionosphere
have to be accounted for separately, of course.

-George

On Wed, Apr 24, 2013 at 8:57 AM, Barry Clark <bclark at nrao.edu> wrote:

> Perhaps the better long term solution is to get rid of the concept of
> reference antenna for pipeline purposes.  Phase solutions are
> intrinsically floating (that is, the matrix is singular).  Assuming
> one antenna has phase zero is one way of supplying the extra bit of
> information to make the solution possible; more particularly, assuming
> one antenna doesn't have phase jumps is one way to tie two phase
> solutions together.  For lots of antennas, a better way is to say
> that the mean change for all antennas is zero.  (Probably one would
> want to use one of the robust means.)
>
> On 04/24/2013 07:31 AM, Claire Chandler wrote:
> > We are working on this from the pipeline side, but at the same time, such
> > things need to be fixed from the technical sided as well.  The pipeline
> is
> > part of end-to-end operations now.  It really does try to do something
> > sensible in picking a refant -- it is based on location in the array and
> > the amount of unflagged data (to avoid picking an antenna that is
> shadowed
> > or missing a receiver).  But it cannot test for phase jumps.  I don't
> > think using the same on as Telcal is always going to work either.  We are
> > working on providing a list of antennas that it should NOT use.
> >
> > Claire
> >
> > On Wed, 24 Apr 2013, Michael Rupen wrote:
> >
> >> In general it would be nice if the pipeline were smarter about choosing
> >> reference antennas.  ea14 should of course be fixed, but one bad antenna
> >> should not produce such a "ripple effect" downstream.
> >>
> >> The intelligent choice of a ref.ant. is not entirely trivial.
> >> An interim solution might be to use the same antenna we use as a
> reference
> >> for TelCal, which we tend to keep a careful eye on operationally.
> >> This has the benefit of matching the reference used for phased array
> >> work.  That reference antenna at the moment is ea12, but might more
> >> usefully be described as the antenna on pad E8 -- one of the three
> >> antennas which is never moved when we change configurations, and central
> >> enough when it matters (i.e., for the more extended configurations).
> >>
> >> There is no obvious place in the SDM proper at the moment to store the
> >> suggested reference antenna.  We might consider adding an element to the
> >> ExecBlock Table to handle this:
> >>
> >>    Optional:
> >>    numRefAntenna  int
> >>    refAntenna[Nrefant], Tag[Nrefant]
> >>    description: "default reference antennas for calibration, in order
> >>      of preference"
> >>
> >> We actually do send such suggestions to TelCal at the moment, so
> recording
> >> that information (esp. useful for phased-array data) would seem
> reasonable.
> >>
> >>             Michael
> >>
> >>> Hi All,
> >>>
> >>> We are seeing phase jumps on EA14 and right now the pipeline is
> choosing
> >>> EA14 as a reference antenna so, when there is a phase jump, the
> pipeline
> >>> output is not usable.
> >>>
> >>> Is there anyway someone can go out to EA14 on Maintenance day tomorrow
> to
> >>> see if this can be fixed?
> >>>
> >>> Thanks,
> >>> Deb
> >>>
> >>>
> >>>
> >> _______________________________________________
> >> evlatests mailing list
> >> evlatests at listmgr.cv.nrao.edu
> >> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
> >>
> > _______________________________________________
> > evlatests mailing list
> > evlatests at listmgr.cv.nrao.edu
> > http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20130424/b1e7957d/attachment.html>