[mmaimcal] Re: subarrays

[Larry D'Addario]

Simon,

If subarrays were free, or if they could be created with "only
software," perhaps the I&C group's idea would be feasible.  This is
not the case.  Lots of expensive hardware must be multiplied by K for
K subarrays.

Like all other proposals, this one does not deserve to "get into the
specs" without thorough justification.  To be seriously considered, we
need a document showing why 10 subarrays are *needed*, and considering
the cost tradeoffs.  If, as I suspect, there is no such need but the
number is just the largest one for which anyone could imagine a use,
then we need not consider it at all.

I believe we should support two subarrays.  This provides hardware
redundancy when only one is needed, and it supports testing and
calibration on some antennas simultaneously with science on the
others.	 I would not be surprised if justification is found for
extending this to three subarrays, but I think it will be hard to make
a case for going beyond that.  Nevertheless, I'd be happy to hear any
such arguments.

--Larry