January PREx Test Run: Compton Photon Analysis

1
January PREx Test RunCompton Photon Analysis

• Diana Parno
• Carnegie Mellon University
• HAPPEX Collaboration Meeting

2
Outline

• Compton photon DAQ
• Integrating method
• FADC design
• Test run results
• Problem areas

3
Photon DAQ Integrating Method

• 3 ways to measure a Compton-scattering asymmetry
• Differential scattered photon count as a
  function of energy
• Integrated scattered photon count without energy
  information
• Energy-weighted total energy deposited (no
  photon counts)
• New DAQ allows us to use energy-weighted method
• At low energies, detector response function
  becomes complicated and thus harder to know with
  precision
• Energy-weighted integrated method is less
  sensitive to the precision of the detector
  response function

4
FADC Design

• As specified, the FADC (from Struck DE) samples
  the data at 200 or 250 MHz (programmable)
• Six accumulators sum over a time interval (30
  ms).
• Each signal sample contributes to at least two
  accumulators, depending on which criteria it
  meets.
• Data can be read out in sampling mode (first
  50000 sample words are included along with
  accumulator words) or integrating mode
  (accumulator values only)

5
FADC Design Accumulators

• Thresholds, and degree of stretching, are
  programmable
• Ideally, Compton events fall within the window
  so we can compute asymmetries using
  Accumulators 0, 2, and 4

6
Test Run Goals DAQ

• See Compton events with new FADC
• Test signal splitting (to use original DAQ and
  new DAQ simultaneously)
• Compute asymmetries and beam polarization using
  only energy-weighted integrated data
• Study software, hardware behavior and systematics

7
Outline

• Compton photon DAQ
• Results
• FADC functionality
• PMT performance
• Accumulator signals
• Asymmetries and polarization measurement
• Problem areas

8
Studying Pulses Two Modes

• Integration mode Read accumulators only no
  deadtime
• Several pulses in each 30-ms event
• Sampling mode Read individual 5-ns samples.
  Detect peaks with a software threshold
• Detailed study of individual pulses (e.g.
  snapshots)
• Precise location of pedestal

9
Saturation

• Plotting pulse area vs. pulse amplitude shows
  clear saturation during the test run

• PMT (12-stage, -2500 V) major contributor
• Compton events are in relatively linear region

10
Accumulator Physics Signals

• The raw accumulator values can be used to extract
  the total physics signal
• We can apply a deadtime correction to the physics
  signals from the window accumulators (2 and 4)

Number of samples
Average signal
Average pedestal value
11
Sample Run Accumulator Signals
12
Dilution Factors

• S includes Compton signal C and background signal
  B
• The measured asymmetry in S differs slightly from
  the asymmetry in C
• We can correct for this by dividing the asymmetry
  in S by a dilution factor D

13
Dilution Factors

• Computed dilution factors for four production runs

Run 60325
Run 60326

• Small D means low SN

Run 60327
Run 60328
14
Asymmetries!
Sign flip IHWP change

• 3 accumulators, 4 production runs

Sign flip Laser pol. change
15
Accumulator Combinations

• Suppose we combine two accumulators to compute
  asymmetries

Left circularly polarized laser
Laser off background only
16
Sensitivity to Pedestal

• A small mistake in finding the pedestal has a
  large effect on the computed asymmetries

Correct pedestal value (2017.2)
Incorrect pedestal value (2016)
17
Electron Beam Polarization

• Megan Friends Geant simulation of analyzing
  power 0.02316 for test run
• Still to be included PMT, PMT nonlinearity
• We have everything we need to compute beam
  polarization

18
Outline

• Compton photon DAQ
• Results
• Problem areas
• Shape of energy spectrum
• Signal size
• Discrepancy between left/right polarization
  states

19
Problem Shape of Energy Spectrum

• Compton cross section s(?) is a parabola. In the
  past, this shape has been echoed in the detected
  photon energy spectrum.

• In January, the measured energy spectrum did not
  look remotely like a parabola. Why not?

January test run (central Saclay crystal)
20
Signal Size Mystery

• During the test run, signal from the new detector
  seemed smaller than expected
• Afterward, cosmic ray measurements at CMU (same
  detector) showed a much bigger response

Photon source Approx. deposited energy Pulse area response Pulse amplitude response
Compton 130.9 MeV 44 RAU-S/MeV 7.3 RAU/MeV
Cosmics 20.4 MeV 138 RAU-S/MeV 39 RAU/MeV
21
Discrepancy Left and Right

• Large discrepancy between asymmetries (and thus
  polarizations) measured when laser
  left-circularly polarized vs. right-circularly
  polarized
• Always in the same direction even when IHWP
  flips electron beam helicity

22
Future Work

• With GEANT, investigate energy spectrum shape
• Incorporate PMT into analyzing power model
• Investigate L/R discrepancy
• Improve analysis code incorporate coincidence
  data
• Improve detectors and mounts in beamline

23

• Thank you!