[mmaimcal] Correlator reconfiguration time

[Mark Holdaway]

> Does anyone have a spec for how long the final MMA correlator can take to do
> a reconfiguration? It came up at a discussion in Socorro today where Chuck
> Broadwell and a few of us discussed some correlator interface issues, some
> of which impact how long it might take the correlator to reconfigure.
> 
> For example, I would guess that with fast switching one might want to
> observe the calibrator with a different spectral resolution etc, which would
> imply a ~1s reconfiguration time. Are there observations that will drive it
> faster than this?

There will be times when fast switching will be really fast (or COULD be
really fast).  Using my Gaussian velocity profiles, I could usually get
from traget source to calibrator source in about a second, which means
often you have calibrators which just happen to be much closer, and you
can get there in UNDER a second -- though not WAY under a second... in
order to NOT excite the antenna, you've got to start up the move very
slowly and end up the move very gradually, so there is a fair bit of
overhead; in other words, time required is not proportional to distance
traveled.

So, I would aim for somewhat better than a second -- like half a second,
assuming that the correlator set up can occur while you are slewing the
antennas -- though 1 s would not be a disaster.

	-Mark