[mmaimcal]Re: [Almasci] Antenna Acceleration

[Mark Holdaway]

>>>>> Jeff wrote: 
> I think that the discussion regarding antenna performance should focus
> on the requirements rather than how those requirements are met.
> Asking questions about how a specification might be met is
> interesting, but "getting the cart before the horse".
> 
> So, a more useful exercise would be to do a better job of specifying
> the fast motion requirements (fast switching, total power OTF, and
> interferometric OTF).  What is missing in the current prototype
> antenna specs are things like:
> 
> -- Duty cycle for OTF.
> 
> -- Specific turnaround time for OTF.

These will depend upon 1) how large the source is,
2) how fast the antenna is moving, 3) exactly what sort of
antenna motion profile you are using.

So, it isn't a simple number or a simple requirement.  And there isn't
anything special about this that the science requires -- faster is better
(faster means more efficiency, more observing time);  but if its slower,
we'll live with it.  If it is REALLY REALLY slow, then we'd need more
than 4 antennas measuring total power... but that number "4" isn't really
well justified from science anyway. 


Tracking tolerance has some well-defined numbers we can attach,
one for OTF mosaicing (ie, the standard pointing error requirement)
and one for total power OTF (ie, a fraction of a beam such that we
are still Nyquist sampled -- but if we can't make that spec, we
just oversample more between scans).

> -- Tracking tolerance.
> 
> For example, saying that the antenna should turn around during an OTF
> raster "as fast as possible" is not a specification.  This requires a
> number, which should derive from a duty cycle requirement.
> 
> Once a requirement which derives from the science is defined, then the
> feasibility of these requirements can be determined.
> 
> Cheers,
> 
> Jeff
>