Beamformer observation requirements

[rubyvanrooyen]

Issue carried from retired repository https://github.com/rubyvanrooyen/astrokat/issues/1

Information supplied by Maciej This is placeholder for beamformer-related observation aspects:

• noise diode (modes of operations, switching cycles),
• calibrators,
• multiple targets?
• beam shaping,
• quantization settings,
• gain settings,
• how do we specify those (array, single value etc.)?
• drift scans, spider scans, spiral scans, raster scans,
• use cases (how do we observe, observing strategy),

[rubyvanrooyen]

Implementation of noise diode into observation framework, follow issue #5

[rubyvanrooyen]

Discussion inherited from related to max duration parameter (obs_duration) from Anton https://github.com/rubyvanrooyen/astrokat/issues/6

I'm wondering about max duration for PTUSE work as well. Some scripts have a fixed chunk of work (e.g. visit N targets for M minutes each). Other just want to run as long as possible, and keep looping through their sources. Think it would help if we at least include a place to define max duration in the astrokat config. Maybe we allow an empty value, or -1 to indicate no limit.

[rubyvanrooyen]

Comments by @mserylak from https://github.com/rubyvanrooyen/astrokat/issues/3 to take forward

About the pol-/flux-cal, especially in the beamformer (bf) case: currently we have only adopted single dish approach to pol/flux calibration. For fluxcal:

• slew the telescope to on-source (flux calibrator) and record it with noise diode firing,
• slew the telescope to an offset position of +/-X deg. in declination (that is to be decided) of flux calibrator and record it with noise diode firing (so 2 obs in total),

The flux calibration observations should happen at least once per observing session, in general the observations of flux cal targets are done at least once per week if not more often. For polcal:

• slew the telescope to an offset position of X deg. in declination (that is to be decided) of target of interest (pulsar) and record it with noise diode firing,
• this is done before or immediately after target of interest of observation,

This strategy has been working fine: my rev. 1 report on SD calibration will demonstrate it, hopefully will be released before the "BRP-odyssey" ;-). Given improved calibration pipeline (removing cross-hand leakages, as per BenH/Ludwig latest feedback) we will AFAIK apply the same approach for large subarray for now.

So the catalogue2config.py should have (for bf case):

• offset in deg from target for both flux and pol cal,
• a strategy of pointing,

OK, the above is not very descriptive (and there are other issues lurking around the bushes which I think, we should winkle out). Shall we make a brief meeting about that? Tomorrow, before lunch looks fine for me...