[evla-sw-discuss] Simple subarrays: a modest proposal

[Michael Rupen]

> On Tue, 1 Apr 2014, Michael Rupen wrote:
>
>>           For now I believe it's safe to allow any continuum experiments
>>           (3-bit or 8-bit), and I think we could write down some simple
>>           (and
>>           not overly onerous) restrictions on spectral line setups which
>>           would
>>           avoid these problems.  To my knowledge there have yet to be any
>>           approved spectral line subarray experiments, so simply not
>>           allowing
>>           those is probably fine for quite a long time.  [Basically this
>>           translates to disallowing recirculation and BlB stacking when
>>           using
>>           subarrays.  If we use the "subarray" tag in the RCT, as outline
>>           above, the RCT could enforce those restrictions on all subarray
>>           resources.]
>
> Except for extravagant use of dumptrigs if many different recirculations
> are used, why should there be a restriction on line observations?  If
> I have this right, each subarray can use its own independent set of
> dumptrigs because no antenna (station board) exists in more than one
> subarray.  What am I missing?
>

For "real" (dynamic) recirculation I think DUMPTRIG is the only issue.[*]
I am mildly worried that we haven't tried complex subarray setups, as that
might task the CBE if e.g. five subarrays are doing different levels
of stacking & recirculation, and (because we currently do BlB-by-BlB NIC
assignment) those wind up at the same NIC and hence use the same CBE node.
If we allow faster dumps than normal because of the smaller number of
baselines I can imagine this becoming problematic.  Hence the idea that
we might have to write down some simple restrictions, at least until we've
extensively tested the options we wish to offer.

             Michael

* On the DUMPTRIG end, apart from the StB-based restrictions, there are some
   limitations on how DUMPTRIG is routed from StBs to BlBs.  We don't run
   into this much for "regular" observing because we can basically use
   just one antenna's DUMPTRIGs; with subarrays we start having to
   exercise more sophisticated routing.  Sonja and I discussed this (I
   think you were involved??) in the context of how many DUMPTRIGs we
   could have total, if we started wanting the limiting case of one
   DUMPTRIG per subband...there were magic factors of two involved (above
   the obvious 16), and various things she could in principle fix up in
   the CM.  [The implication is that these are really CM rather than
   hardware limits.]  I don't think she got very far with that, as it was
   not a high priority at the time, but I don't recall the details right
   now.  Being me I probably have some of this in various e-mails, but it
   would take a while to find them.  Let me know if you're interested.