SiStrip Simulation overview

Special meeting 13-05-2014

Claude Nuttens (claude.nuttens@cern.ch)

Table of Contents

Strip Digitizer Structure ↓

From Geant 4

Strip signal simulation steps

Sim Hit deposit simulation

‹ 0 ›

• "Divide deposit" and "Crosstalk" in previous slide step-picture
• be carefull! scales, strip density, ... are not trustfull, some effects are exaggerated for the representation/understanding (also to be check strip orientation in the local reference frame, x or y?)

SiStrip Digitizer structure

Detailled view

Module view

Time View

First Project ↓

Impact of the Temp. param.

• Study the impact of the Temperature parameter on the simulation
  Seen in previous slide :
  Temperature = cms.double(273.0),
• This parameter is used in the diffusion constant computation required to compute the drift time of the charge in the sensor (we will see a bit further)
  → As seen before, the charge drift is one of the first step
• In 2015 runs, the temperature is lowered to ~-20°C,
  • How does distributions are impacted by this change?
  • If you see impact, how a few degrees variations around this value impact the distributions
    → to see if keeping a global temperature is a fair approximation
• Note that it may not be the only temperature effect in the simulation, as other parameters used/defined are temperature dependent but not defined from this configuration temperature value. You can try to list temperature dependence in the sim.
  (not mandatory, we can see together when you have the first results)

Details of modules ↓

Depletion and applied voltage

Depletion and applied voltage

• In SiHitDigitizer.cc :

   double timeNormalisation = (moduleThickness*moduleThickness)/(2.*depletionVoltage*chargeMobility); 

  
  $ t_{Norm} = \frac{Thickness_{mod}^2}{2. V_{Depl}. chargeMobility}$ where $ chargeMobility \propto \left[ \frac{cm^2}{V.s} \right] $
• In SiLinearChargeCollectionDrifter (drift time in the sensor)

  double driftTime = -timeNormalisation *log(1.-2*depletionVoltage*thicknessFraction/(depletionVoltage+appliedVoltage)) +chargeDistributionRMS

  SignalPoint drift (EnergyDepositUnit edu, Localvector drift, moduleThickness, timeNormalisation) see next slide

• This have an effect on coupling wich use the chargeSpread

Depletion and applied voltage

SignalPoint drift (EnergyDepositUnit edu, Localvector drift, moduleThickness, timeNormalisation)

• computes the fraction of the module the charge has to drift through,
  $ depth=\frac{moduleThickness} {2}- energyDeposit.z()$
  $ thicknesFraction=\frac{depth}{moduleThickness}$
• ensure thicknessFraction $\in [0,1]$
• Compute the drift time in the sensor
  $ driftTime = −t_{Norm}∗log \left(1−\frac{2.V_{Depl}}{V_{Depl.}+V_{Appl.}}\cdot thicknessFraction \right) +chargeDistributionRMS $
• Return the SignalPoint(x,y,sigma,amplitude): pos., an energy on the surface, with a size due to diffusion.
  $SignalPoint(edu.x ()+\frac{depth∗drift.x ()}{drift.z()}, edu.y ()+\frac{depth∗drift.y()}{drift.z()},\sqrt{2.∗diffusionConstant∗driftTime} , edu.energy ())$ where
  $diffusionConstant= \frac{CBOLTZ}{e}∗chargeMobility∗Temp∗{1.0 if noDiffusion \choose 1.0e-3! noDiffusion}$

Noises, pedestal and baseline

Noises in Zero suppression

Noise in Zero suppression

• Produce noise using GaussianTailNoiseGenerator

   > generatedNoise;
  genNoise->generate(numStrips, threshold, noiseRMS, generatedNoise);

• Gaussian Noise on strips with signal with CLHEP::RandGaussQ(rndEngine):
  • Loop on channels and add noise if not null
    in[iChannel] += gaussDistribution->fire(0.,noiseRMS);
• Noise on the other strips
  • Loop on generatedNoise vector pair generated at top
    • Add to channel the corresponding noise (and check that value ==0)
      in[(*p).first] += (*p).second;

Gains

Couplings