Polarisation / spin-echo tuning scans

[Tom-Willemsen]

• \\ndxlarmor\c$\Instrument\Settings\config\NDXLARMOR\Python\LSS\SESANSroutines.py
• echoscan_axis
• Collect data into pairs of periods via e.g. a PolarisingDae class
• Need to be able to have a stage/unstage in PolarisingDae that sets up different wiring tables/tcbs etc and puts them back post-scan
• Chop up the data into N (perhaps overlapping) wavelength bands
• Calculate polarisation for each for each scan point (scanning over a current)
  • Might need a custom device for the magnet current as block SP is an integer in mA and rbv is a float in A. Alternatively make the IOC have a more sensible set of PVs to expose as a block...
• Guesses for fits need to be careful as data will be noisy etc:
  • Fit widest (full) wavelength band first
  • Use center from that fit as the guess for the center of the next fit

Discussed planning 02/04/2025 00:10:12

[Tom-Willemsen]

Subtasks

Write a PolarisingDae:

• This may not necessarily be a SimpleDae, it might be best to subclass Dae directly for this one
• N configurable wavelength bands (each independent from one another, but may be overlapping). Chopping into wavelength bands was previously implemented in Sum detector/monitor spectra with tof/wavelength bounds #42 .
• Each count() needs to use two DAE periods under the hood. Call these two runs A and B (which will really correspond to "up" and "down" polarisation). So the total number of periods needed for a scan measuring at 20 points is 40.
• Fiddle with a flipper in between the two runs (Likely need to use ophyd-async Reference type here)
• Will need to use the logic implemented in Polarization/asymmetry #58 to calculate polarisation
• The outputs of this PolarisingDae object, for each scan point (i.e. the signals available to be read) are:
  • For each wavelength band
    • det_counts_a & det_counts_b
    • mon_counts_a & mon_counts_b
    • intensity_a & intensity_b where intensity = det / mon
    • polarisation = (intensity_a - intensity_b) / (intensity_a + intensity_b)
    • polarisation_ratio = intensity_a / intensity_b
• For all of the above we need to be propagating uncertainties correctly

Write a device for magnet current on LARMOR

• Probably would live in LARMOR's inst area long term, not ibex_bluesky_core.
• Or fix the IOC instead

Write a spin-echo tuning callback

• This probably isn't "just" a normal LiveFit callback - but might be able to inherit from it (?) and re-use our normal model functions (?)
• Need to do a fit on wavelength band N first, and then use that as the initial guess for wavelength band N+1
• Return all N of the fit results (the fit for each wavelength band is independent, so for N wavelength bands, N fit results are returned

Write a spin-echo tuning plan

• Will eventually live in e.g. LARMOR's inst area, not ibex_bluesky_core. But if helpers make sense to put in ibc, that's fine.
• Needs to change wiring tables (Dae can already do this) and have a finalize wrapper to put them back at the end
• Ties together the device for magnet current ("motor"), PolarisingDae ("detector"), and the spin-echo tuning callback into one top-level plan.

[jackbdoughty]

• polarisation & spin echo ibc PR. look at this for vast amount of detail
• larmor_spin_echo_plan larmor plans, not to be merged to be put in inst scripts
• magnet ioc issue issue created to patch larmor magnet ioc
• As part of the top level inst script/plan, we need to be able to make backups/restore backups/change dae settings which is covered by this issue