[mmaimcal] Re: subarrays

[Michael Rupen]

Solar flares.  They don't last long enough to wait around while changing freqs.

VLBI whilst using the array for any other purposes (I assume maintanence+regular
usage+VLBI).

Simultaneous SD & array observations to maintain identical atmospheric
conditions & calibration. 

But basically I totally disagree with the attitude that convenience or
efficiency is not an argument when put up against inconvenience or cost of
design.  Making a telescope more efficient is equivalent to enlarging the
collecting area, even more so when working at wavelengths where the atmosphere
changes rapidly & frequently prohibits work at the highest observing 
frequencies.  A $400M (or is it $600M now?) telescope should not be provided
with only the barest minimum of capabilities.  My car does have air
conditioning, and I suspect yours does as well.  It can even carry more than
one passenger, although I could always drive back to pick #2-4 up later. 
We should not have to scrape by on a day-to-day basis, for normal observations,
at the very limit of what we can do.  If half the antennas are not outfitted 
with receiver X, and only 3 antennas have single-dish capability, and the 
engineers/software types want to futz with 4 more, we should be able to use 
*all the antennas* despite this complexity: that means
  subarray 1  freq. X (equipped antennas)
  subarray 2  freq. Y
  subarray 3  single dish (equipped antennas)
  subarray 4  maintenance/improvement work
and I haven't even invoked VLBI yet.   I admit that given infinite observing
time we can just use one or two subarrays, but if we had infinite time
we could just bang away with PdB or SMA.  More importantly, you haven't 
mentioned any strong, *quantitative* driver for the truly minimal number
of subarrays you suggest.   How painful would it really be to allow 6 or so
subarrays?  How much would it cost, in $$$, in design time, and in maintenance?
Productivity and efficiency are important, as well as uniqueness.

-- Michael