[mmaimcal] forwarded message on subarrays from Jacob Baars

[hliszt]

Perhaps this has been covered before in Leiden or Garching or Tucson 
or Timbuktoo...

I don't see any purpose whatsoever in putting a limitation on the #
of subarrays which the software can deal with until and unless forced
to do so by some external circumstance. I offer this with several 
prejudices in mind:

   There wouldn't be any reason for limiting the number if one were
   just writing a piece of code "in free space" and of one's own 
   "free will" , but putting such a limit in the spec might embed 
   limitations which would be hard to overcome later.
   
   Such limitations are much more likely to be forced (and earlier) 
   on the hardware; 

   Such limitations are much more likely to be forced (and earlier) 
   on the software which is embedded inside the correlator than the 
   software which exists outside it.

regards, Harvey
-------------------------------
Al Wootten wrote:
> 
> Imcalers:  Comments on this note from Jaap?  I think it looks workable.
> 
> Al
> 
> I have finally found the time to work on a possible wording of the
> "sub array" issue.
> We, or at least some of us, discussed that a short note might be
> appended to the ALMA System Preliminary Design Recommendations, as
> well as to the ASAC minutes. Using mainly a proposed version by
> Stephane, here is a draft proposal of such a note.
> 
> Any comments and improvements, as well as alternatives to its
> dissemination are are welcome.
> 
> regards,
> 
> Jaap
> =====================================
> DRAFT
> 
> There exists confusion, and possibly differences of opinion, among
> the scientists and engineers of the ALMA Project as to the final
> decision regarding the number of sub-arrays, which would be supported
> in ALMA. After discussions in the ASAC and among the Project
> Scientists and System Engineers, we have agreed that the following
> definition is appropriate.
> 
> We distinguish three types of sub-array levels:
> 
>      1) Number of different frequencies the LO system can drive simultaneously
>          Specification: up to 4.
>      2) Number of antenna sets the correlator can treat independently:
>          Specification: up to 6
>      3) Number of antenna sets the software can handle independently:
>          Specification: up to 8
>      4) Independent control of any antenna will be possible (for
> maintenance and repair).