Energy Calibration with Compton Data

1
Energy Calibration with Compton Data

• June 23, 2006

2
Outline

• Revisit Z using DA Compton calibration
• SA Compton calibration
• Using production on carbon, lead
• SA gains vs LMS
• How to improve LMS performance
• Need a consistent normalization scheme for the
  LMS gains during the production period.

3
Target-Hycal Distancefrom Double Arm Calibration
Calibration procedure Gain correction factors
are found through cluster coordinates and
Z (E?(el)Eb/12Eb/mesin2(??(el)/2)) Incorrect
value of Z will result in elasticity
distribution not centered exactly at 1 ! try to
vary z, recalibrate, check elasticity
after
before
Mean1.001
Mean1.017
(e1e2)/Eb
(e1e2)/Eb
4
Target-Hycal Distance from Double Arm
Calibration (cont)
Mean of the elasticity distribution vs
hycal-target distance (Z) used in
calibration for 5 groups of runs
731.4 lt Z lt 732.1 cm
Resolution does not change with z
5
Target-Hycal Distancefrom Compton kinematics

• Could obtain Z from Compton kinematics, setting
  z1z2
• z(Ebr1r2/me/2)0.5
• Not so sensitive to the gains
• Sensitive to beam misalignment changing x by 1
  mm changes z by 6-7 cm

Z (cm)
Run 4871, Be target
6
Target-Hycal Distancerun by run
Difference between the first group of carbon
runs and the rest of the runs should be 7.62 cm
We see 6.52

9Be Z732.08 cm 12C (ave) Z 733 cm
7
Single Arm Compton Calibration

• Objectives
• Monitor gain change during the production data
  taking
• Compare with the LMS behavior
• Use LMS gains in the part of HyCal where no
  Compton gains are available
• Procedure
• 320 production runs with carbon target and 70
  runs with the lead target
• Look for events for one neutral cluster and fill
  the following distributionecl/eC,
• eCEb/(1Eb/mesin2(?/2))
• Fit ! gain corr1/mean

carbon
lead
8
SA Calibration PSCarbon
30 eff.
PS cut
Require 1 hit in the PS within the physics time
window
lt5 eff
9
SA Calibration PS (cont)Lead
8 eff.
Very low PS efficiency for lead runs! lose most
of events
10
Life and times of module 1494
SA Lead
SA carbon
LMS
Reference pmt
Gain correction factors relative to the snake
calibrations vs run
11
2 parts to life of module 1494
Relative to snake 1
Relative to snake 2
LMS data is not properly normalized for this
time period
Looks good
12
Look closer
SA gains stat. errors are small (for
carbon) Systematic uncertainty comes mostly from
possible beam misalignment shift of 1 mm in X
results in 1 change of the gain value
LMS and SA gains are consistent but LMS is more
scattered
13
SA gains vs LMS
carbon
lead
lead
carbon
carbon
carbon
Could find correlation between the LMS and SA
gains
14
Tri-modal behavior of LMS
What are we going to do about this?
15
Summary

• 120 modules can be calibrated with the SA gains
• Good agreement with the Double Arm gains
• Reasonable agreement is observed between the SA
  and LMS gains for both the carbon and lead
  targets
• Can/Will find the correlation between LMS and SA
  gains for every module
• LMS gains look reasonable but are scattered and
  sometime tri-modal! what to do about it?
• Parameterize LMS with SA gains?
• How to correct for tri-modality?
• Need to use a consistent scheme for LMS
  normalization throughout the run ! currently it
  is wrong for runs lt4838