Pixel Detector Simulation with Magnetic Field

1
Pixel Detector Simulationwith Magnetic Field
JC and Marina 12/18/00

• Effects with magnetic Field
• Deflection
• Effective mobility
• Non-constant Hall mobility
• Comparison with magnet run

2
Magnetic Field
3
Deflection in Magnetic Field
Effect of magnetic field ? Effect of
incident angle
4
Deflection in Magnetic Field
Assumption mH constant meff m
The charge distributions are the same with B or q
5
Effective Mobility

• Very small effect
• B field ? Cloud drifts slower ? Larger Spread

6
E-dependent Mobility
Larger E ? Smaller mHall?? ? Smaller deflection
7
Effect of Magnetic Field

• With non-constant mHall(E), and meff

The charge distributions are slightly different
8
Conclusion
We Dont expect the magnetic field will reduce
the ratio between two-pixel clusters and
one-pixel clusters
9
Magnet Run

• Track at 1.3? angle ( 0.242 Tesla )
• Measurement Rmin 0.193
• Simulation ( Qth 2.5Ke ) Rmin 0.250

What is wrong ?
10
Magnet Run
Just to Play around

• Track at 1.32? ( 23 mrad ) ( 0.245 Tesla )
• Measurement Rmin 0.193
• Simulation ( Qth 4.0Ke ) Rmin 0.197

But there is no other evidence that threshold was
wrong
11
Magnetic Field

• Measured with Gauss meter 45A ? 0.585 Tesla
• Assume overall scale of the magnetic field
  uncertain
• Use MC simulation to fit ? 0.517 Tesla (error?)
• Error of Hall factor (mHall/m 1.15) can make up
  the difference

12
Summary

• MC simulation accurate both for drift and B
  field simulation
• Magnet Run data is not fully understood yet

13
Lorentz Angle

• Lorentz angle is not a constant along z,
  shown are overall effects
• Larger bias voltage
• ? Larger E field
• ? Smaller mobility
• ? Smaller Lorentz angle

14
Lorentz Angle Comparison
Magnetic field 1.4 Tesla Udepletion is
calculated with effective thickness of the sensor
15
Diffusion Constant
Einstein equation D k T m / q

• Reduce Teff from 390K to 300K
• ?
• Reduce fraction of doublet by 1.4

16
Angular Fit with ? Distribution
FPIX0-Pstop Normal Incidence Data

• MC 0.5? interval, interpolation between
  points
• Normal incident data ? (1.77 ? 0.09)?

17
Angular Fit with ? Distribution
FPIX0-Pstop
18
Detector Inclination Angle

• In Minuit fit, the beam is assumed to normal
  incident on the first SSD, the overall rotation
  of the detector can not be fitted
• In ? fit, interpolation between MC points (0.5
  interval) is used
• The error from the fit is about 0.1, which
  should be smaller than error from MC model