Difference between revisions of "Pol He-3 d2n analysis plan"

From HallCWiki
Jump to: navigation, search
(Complete/Update run database)
Line 29: Line 29:
= Analysis stages =
= Analysis stages =
== Documentation what we did ==
== Documentation what we did ==
* run list, run database, etc
* Make a high-level ''Calendar'' page that summarizes our run period
* go through the RC reports and call out anything 'important'
* Run list, run database, etc
** [[https://hallcweb.jlab.org/wiki/index.php/Pol_He-3_d2n_analysis_plan#Run Lists]]
** [[https://hallcweb.jlab.org/wiki/index.php/Pol_He-3_d2n_analysis_plan#Complete/Update run database]]
* Go through the RC reports and call out anything 'important'
* Document survey information for SHMS/HMS angle determination
* Document survey information for SHMS/HMS angle determination
** verify it is correctly used in the analyzer (part of the run database)
** Verify it is correctly used in the analyzer (part of the run database)
== Low level calibrations/checks ==
== Low level calibrations/checks ==

Revision as of 18:36, 17 October 2020


  • Move analysis to ifarm/farm as soon as we can
  • Do /not/ work on your laptop, personal computer, etc
    • You *must* work on the farm or other JLab system with valid backups!

Run Lists

  • production runs for each kinematic setting
  • calibration runs (optics, etc)
  • 1-pass runs
  • link to 'A1n' run database/logs (for our reference)
  • issues/comments should be flagged
  • identify any configuration changes
    • target changes, significant beam changes
    • angle changes, of course
  • target polarization (rough)
  • beam polarization (rough)
  • Flag runs that may need special attention
    • beam problem, hardware problem, etc
  • 'Notable events' list

Complete/Update run database

  • Should have all the information above in it
    • later iteration can fill in a model-corrected target and beam polarization
  • prescale settings
  • deadtime corrections
    • NOTE: There was an issue with the HMS trigger scaler for part of the d2n run and likely all of the A1n run. See HClog
    • As part of this, it would be good to establish a way to create a database containing scaler rates vs. timestamp for the 'redundant' scaler channels between the two arms

Analysis stages

Documentation what we did

  • Make a high-level Calendar page that summarizes our run period
  • Run list, run database, etc
  • Go through the RC reports and call out anything 'important'
  • Document survey information for SHMS/HMS angle determination
    • Verify it is correctly used in the analyzer (part of the run database)

Low level calibrations/checks

  • need to verify and/or complete for all detectors
  • check basic stuff like
    • gains, thresholds for anything in the trigger (hodoscopes for 3/4; hodo+cer for other)
    • check reference time cuts for all detectors
    • check 'good hit' selection cuts
    • check walk-correction
    • check beta plots
  • wire chamber drift-time calibrations
    • t0 calibrations
    • drift time 'flatness' calibration
  • shower calibrations
    • relies on good optics to match P0
    • relies on good cluster identification
    • I gather there has been history of repeating this shower calibration frequently
      • This is manifestly wrong -- nothing in the detector system changed. The fact this is done indicates that the calibration is /not/ done correctly in the first place. The 'recalibration' is literally hiding errors in the prior calibration and introducing hidden errors in the new calibiration each time. We should look into this ourselves and decide if there is a better way.
    • Watch for low statistics blocks that skew the fits

Mid-level calibrations/checks

  • Hodoscope start time check
    • These involve path-length and PMT/cabling level corrections used to work out a best 'start time' for a particle passing through the detector stack.
      • ie. you get a TDC time for your particle as it passed through each of the hodoscope planes (2 PMTs/bar, and hits on at least 1 bar in each of at least 3 of 4 hodoscope planes). The TDC times for those hits depend on the flight path, and where in each bar the particle hit.
      • Correct and fit those data to interpolate the time at which the particle passed through an ideal "focal plane".
    • Note that changing this 'start-time' will impact some of the 'low-level' calibrations like the drift-time t0's, etc. Need to go back and check/iterate until things are stable.
  • Shower/Preshower pion contamination/rejection
    • determine electron efficiency: (electrons detected)/(total electrons) vs.
      • 'loose/tight' cuts
      • spectrometer momentum
    • estimate pion rejection efficiency (pions rejected by cuts)/(pions that are mis-identified as electrons) vs.
      • 'loose/tight' cuts
      • spectrometer momentum,
  • Cerenkov pion contamination/rejection
    • same process as above

Develop some automated 'quality/stability' checks and histograms for each run

  • Beam trips/hour over the run
  • E/p peak value
  • Wire chamber 't0' peak positions
  • Cherenkov ADC peak and TDC peak
  • Beta plots
  • 'hit time' peaks from all (most?) detectors
  • Wire chamber tracking residual plots
  • Wire chamber single-track tracking efficiency
  • Dead times
  • Scaler data checks
    • some will only be useful after we skim our data and remove trips, etc
    • trigger_rate/beam_current ('luminosity' check)

Target analysis

  • polarization calibration constants
    • cross checks between EPR, water NMR signal
    • polarizing gradient between target and pumping chamber
  • N2 dilution analysis
  • 3He pressure curve analysis
  • 1 pass elastic 3He polarization cross checks
  • Target density and temperature analysis
  • Target Holding field direction (compass measurements)

Beam calibrations

  • Beam polarization (Moller)
    • table/calendar with Moller results, half-wave plate changes
  • charge asym analysis
    • compute from each run, compare with 'parity daq'
      • also good per-run sanity check plot
  • BCM calibration
  • Raster calibration/correction
    • Verify raster calibrations from bulls-eye scan data still look reasonable for d2n 'carbon-hole' checks (for example)

Moving from - Acceptance corrections - Deadtime corrections -