[mmaimcal] Implementation of ALMA correlator multi-resolution modes

[Al Wootten]

Hi Todd, Brian Darrel and Stuartt

Thanks for the input on modes.  The meeting was very instructive; here 
are a few comments on items touching the use cases you provided.

With one quadrant, such as we have now, there is no way to observe with 
one window >1 GHz and then up to 16 carefully placed narrower windows of 
different resolution.  But you can do that with 1 GHz, and you can have 
up to 16 windows and 4 resolutions--eventually.  Of course that is not 
implemented in software yet (R7.1).  The third quadrant will go to Chile 
soon and the second quadrant wired for use (not sure of that exact 
schedule).  The final phase has 4 quadrants working as designed; there 
is no 3 quadrant phase.  Beginning with two quadrants we could have the 
ability to simultaneously observe one 2 GHz window encompassing several 
narrower windows depending upon the implementation plan for those modes.

Eventually one can devote up to 4 quadrants to observe at the reduced 
resolution--all on top of one another if one wished, or arrayed within 
the IF in some more inventive way.  One quadrant or more could then be 
devoted to the multi-window (<=16) multi-res (<=4) observations but you 
would lose that quadrant for the 2 GHz wider view.  There is a 
gotcha--different digitizers would be used for different quadrants and 
calibration could be tricky.  All data nominally has the same dump duration.

The limit on four resolutions comes from an executive decision in the 
correlator software and could be relaxed.  It seems like a generous 
place to begin.

The limit to 16 windows comes from the availability of connectors; there 
are not enough connectors to allow the 32 theoretically possible 
windows.  16 filters are left without power owing to this hardware feature.

Software will implement modes as they can.  I think we could provide 
input on that--you might discuss it with Jeff Kern if you get a chance. 
   Most likely Richard and Alison have done that.  I think that the 
deliverables for R9 are being discussed now.

One possible confusion which already bit me is that the OT allows one to 
set up just about any possible mode, even if it is e should discuss 
whether there should be a more limited version of the OT which just 
allows modes actually implemented e.g. for early science.  Currently one 
can have one window, one resolution.  Later one may have multiple 
windows.  Later still, multiple windows and resolutions.  I heard on the 
grapevine that some multiple windowing could be available in R8 I think.

Clear skies,
Al