Tracking sessionJochen Markert, IKF FrankfurtJochen Markert, IKF Frankfurt

TopicsTopics ActivitiesActivities Lepton efficiency estimationLepton efficiency estimation

Implementation of efficiency in digitizerImplementation of efficiency in digitizer Dependency of efficiency on the Dependency of efficiency on the

ionization of the particle trackionization of the particle track Number of wires in clusterNumber of wires in cluster Estimation of layer efficiencyEstimation of layer efficiency Comparison of different tracking code Comparison of different tracking code

versionsversions Reconstruction of opening angle of Reconstruction of opening angle of

lepton pairslepton pairs Dependency of resolution on the Dependency of resolution on the

ionization of the particle trackionization of the particle track PID with MDC: energy lossPID with MDC: energy loss

CAL1CAL1 JochenJochenKhaledKhaledYvonneYvonne

Pulser method for offset calibrationPulser method for offset calibration Tuning of offsets and second iteration of calibrationTuning of offsets and second iteration of calibration

weeksweeksDone for ppDone for pp

CAL2CAL2 JochenJochen New GARFIELD parametersNew GARFIELD parameters Cathode planes instead of cathode wires short termCathode planes instead of cathode wires short term

Analytical description of xt-correlation and errorsAnalytical description of xt-correlation and errors

DELAYED!DELAYED!Short termShort termMedium termMedium term

TRACK TRACK

Segment Segment fitterfitter

VladimirVladimir Restructuring of the tracking code Restructuring of the tracking code Improvement of minimizationImprovement of minimization Improvement of performanceImprovement of performance

Tuning of parametersTuning of parameters

DoneDoneDoneDoneDoneDoneMedium termMedium term

ThierryThierryEmilieEmilieJean-LoisJean-Lois

Investigation on MDCIVInvestigation on MDCIVCheck of geometryCheck of geometry

Medium termMedium term

AlignmentAlignment AlexanderAlexander Alignment for inner and outer modulesAlignment for inner and outer modules Alignment with photo modeler (MDCs + Magnet)Alignment with photo modeler (MDCs + Magnet) Alignment with cosmics Alignment with cosmics Alignment of METAAlignment of META

DoneDoneDoneDoneDoneDoneDoneDone

GeydarGeydar Wire layer offsetsWire layer offsets Layer thickness (MDCIII+IV)Layer thickness (MDCIII+IV)

Done but no Done but no clear resultsclear results

Activities in MDC analysisActivities in MDC analysis

Future tasksFuture tasks First priority:First priority: Efficiency correction Efficiency correction

Tracking for high multiplicities + CPs, Tracking for high multiplicities + CPs, needed for final DSTs needed for final DSTs of SEP05!!!of SEP05!!!

CODE STABILITY!!!!! (we lost weeks for CODE STABILITY!!!!! (we lost weeks for debugging!)debugging!)

Second priority: (several weeks)Second priority: (several weeks) Time offsets from pulser method (Time offsets from pulser method (KhaledKhaled)) Development of “ideal tracking” Development of “ideal tracking”

MDC partMDC part ( done by Vladimir ( done by Vladimir)) Other detectors ? Other detectors ?

Development of embedding of simulated tracks in real events Development of embedding of simulated tracks in real events MDC part already existing MDC part already existing Other detectors ? Other detectors ?

Investigation of events with very large unphysical multiplicityInvestigation of events with very large unphysical multiplicity How many ? Definition of reasonable numbers of tracks (SIM/DATA)How many ? Definition of reasonable numbers of tracks (SIM/DATA)

Fixing of geometry of outer MDCs Fixing of geometry of outer MDCs Wire angles , layer thickness (Wire angles , layer thickness (GeydarGeydar)) Wire layer offsets (Wire layer offsets (Geydar + Thierry + EmilieGeydar + Thierry + Emilie)) Measurements on MDCIV (Measurements on MDCIV (Thierry + Jean-LoisThierry + Jean-Lois))

Optimization of cal2 parametersOptimization of cal2 parameters Smoother values + analytical description (Smoother values + analytical description (JochenJochen)) Retrieving parameters for out MDCs from DATA (Retrieving parameters for out MDCs from DATA (Thierry + EmilieThierry + Emilie))

Influence on the tracking Influence on the tracking efficiencyefficiency

MDC efficiency (cell efficiency: gas, MDC efficiency (cell efficiency: gas, thresholds, noise).thresholds, noise).

MDC hardware problems (MDC hardware problems (missing MBo, missing MBo, …).…).

Calibration qualityCalibration quality AlignmentAlignment Track finder efficiency.Track finder efficiency. Momentum reconstruction efficiency.Momentum reconstruction efficiency. Matching efficiency.Matching efficiency. Cuts efficiency (Cuts efficiency (chi2 cut etc.chi2 cut etc.).). Particle identification efficiency.Particle identification efficiency. ……

Properties of wire Properties of wire clustersclusters

CPR by properties of cluster size and number of wires in cluster

Tuned to get good agreement between simulation and experiment

Cell efficiency in digitizerCell efficiency in digitizer

Cell efficiency not depending on energy loss of particle in digitizer

Mean number of wires in Mean number of wires in clustercluster

NOV01

MDCI

MDCII

MDCIII

MDCIV

Detection efficiency of Detection efficiency of MDCMDC

Efficiency of Layer: A particle track has to be detected at least once per

efficiency of wire layer better than als 89% (MIPS)

Segment theoretical better than

98,4% !?????

Good agreement with laboratory measurement

Particle

Layer 1Layer 2Layer 3Layer 4Layer 5Layer 6

Method:

MDCI

MDCII

NOV01

Layer efficiency including Layer efficiency including the wires which have been the wires which have been removed by tukey weightsremoved by tukey weights

NOV02

Layer efficiency from fit Layer efficiency from fit accepted wiresaccepted wires

MDCI MDCII

LayLay MDMDCC

II

MDCMDC

IIII

11 0.750.75 0.900.90

22 0.680.68 0.910.91

33 0.690.69 0.900.90

44 0.700.70 0.900.90

55 0.700.70 0.900.90

66 0.660.66 0.850.85

NOV02

MDC I p-blue, --red MDC II

SEP05

Ratio fitted segments/all Ratio fitted segments/all segments of lepton pairssegments of lepton pairs

Comparison for

Exp

URQMD

PLUTOFull pair analysis and background rejection applied!

Comparison of Comparison of fitting track fitter fitting track fitter

for different for different HYDRA versionsHYDRA versionsSubtitle: long story about Subtitle: long story about

nothingnothing

Problem descriptionProblem description

Efficiency of track reconstruction of Efficiency of track reconstruction of lepton pairslepton pairs

Rumors about change in Rumors about change in reconstruction efficiency of pairs reconstruction efficiency of pairs (10%) observed by Laura between old (10%) observed by Laura between old calculation with HYDRA v7_05b and calculation with HYDRA v7_05b and new v7_07/v7_08new v7_07/v7_08

MethodMethod Tracking + ideal tracking parallel Tracking + ideal tracking parallel

(HMdcTaskSet/HMdcIdealTracking)(HMdcTaskSet/HMdcIdealTracking) Filling of ntuple with HMdcTrackingEffFilling of ntuple with HMdcTrackingEff Efficiency calculation: Efficiency calculation:

Input Pluto Sim Nov02Input Pluto Sim Nov02 Reference sample ideal segments (both inner and Reference sample ideal segments (both inner and

outer segments + Meta hits found in GEANT)outer segments + Meta hits found in GEANT) Pairs definition : inner segments cluster/fitted, no Pairs definition : inner segments cluster/fitted, no

condition on outer segments, opening angle cut of 9 condition on outer segments, opening angle cut of 9 degreedegree

Efficiency: found pairs / ideal pairsEfficiency: found pairs / ideal pairs

opening angle distribution opening angle distribution of lepton pairsof lepton pairs

Comparison of different code versions of tracking

Efficiency of lepton pairs as Efficiency of lepton pairs as function of opening anglefunction of opening angle

1ook events in simulation

No significant efficiency between the different versions

Opening angle Opening angle reconstructionreconstruction

Cut on opening angle 9 degree :

Difference between GEANT angle accepted reconstructed angle accepted gives 5% more accepted pairs.

Reconstruction of Reconstruction of invariant Massinvariant Mass

Resolution of the drift cells Drift time residuals spatial resolution :

Dependence on the primary ionization clearly visual

Drift cell resolution better than 150 m

Position resolution of the Position resolution of the track reconstrutiontrack reconstrution

distance time measured trackfit drift*t t v

Position resolution of the reconstructionMeets requirements

NOV01 Data

design value

MDCII

Energy loss measurement with Energy loss measurement with MDCs ?MDCs ?

Contra: MDCs measure drift times not pulse height

„Low-mass“ - concept of MDCs not optimized for dE/dx - measurement with high resolution

Measurement of energy loss through width of the drift time signal („TTime aabove TThreshold“, t2-t1) as measure of deposed charge ? 1

1 T. Akesson et al. Nucl. Inst. and Methods, A(474):172–187, 2001.

Normalization of signal Normalization of signal width width

cell

n

m

orm

ce l dcl( 2 1) (angle , distance)t ta CT CT

norm

cellnorm

sec ll

ge

TaTnTaT

Drift cell

Impact angle , distance from

wire

Drift chamber

Gas amplification (HV)

Track segment

Mean over all cells

Impact angle

Protons and pions can be separated

Electrons and pions overlay

deuterons and protons overlay

Normalized and averaged Normalized and averaged Signal widthSignal width

Resolution of signal width Resolution of signal width measurementmeasurement

resolution for protons 6-9 %

resolution for pions 10-12%

Data

Resolution comparable with dE/dx measurement through pulse height!

- 9.8 % p 7.6 %d 7.2 %

+ 10.6 %

Correlation of signal width Correlation of signal width with dE/dxwith dE/dx

Fitted with F(dE/dxBethe-Bloch)

Correlation of signal width measurement with dE/dx

property of signal shape and readout

electronic1

Good agreement for protons and pions1 L. Ratti et al., WCC 2004, Vienna, Vortrag 2004.

norm

segmentvs dE/dxCorrelation TaT

The drift cellThe drift cell

Dimension of the drift cells 5x5 - Dimension of the drift cells 5x5 - 10x14 mm10x14 mm22

Gas mixture He/i-Butan (60/40) Simulation of the drift cells with

GARFIELD - Geometry, Field, Drift

MAGBOLZ - Gas properties HEED - Primary ionization

Field wire

Cathode wires

Cathode wires

Amplification area

Sense wire

Simulation with GARFIELDSimulation with GARFIELD

x [cm]y [cm]

drift/v cm s

Simulation:

Inhomogeneous electric Field inside the drift cell

VDrift depending on electric field

Inhomogeneous distribution of VDrift inside drift cell

drift

Time distanceTime distance

For track reconstruction space points are needed, but MDCs measure drift times

Relation between drift time and minimal distance of the particle track from sense wire has to be known

x-t- correlationx-t- correlation

2-dimensional drift cell model:

Simulation of the drift signals using GARFIELD

Parameterization through impac angle and minimum distance from wire

Implementation into track reconstruction and GEANT - Simulation

Normalization of signal Normalization of signal width (t2-t1)width (t2-t1)

MDCIIData Nov01

Normalization with one curve per impact angle step (5°)

MDCI/II normalized to the same value

Deviation for higher momenta

Normalization of signal Normalization of signal widthwidth

Normalization: Impact angle (), minimal

distance from wire All chamber types

normalized to common value

Normalization point at 450 MeV/c

Inner segment (MDCI/II) : Good agreement at small

momenta Deviation at higher

momenta

MDCIII/IV show different behavior as MDCI/II (statistic/geometry/working point?)

Data Nov01

Nomalization

Comparison of dE/dx Comparison of dE/dx resolution with other resolution with other

experimentsexperiments

2 22

inv

inv1%

2 2 2 tan( / 2)e e

e e

p p

p pMM

dE/dx resolution for gas mixtures with large fraction of hydrocarbon (Quencher) better as predicted

Empiric formula for calculation of dE/dx resolution (MIPS):

A. H. Walenta et al. Nucl. Instr. Methods, 161(45), 1979

The drift time measurementThe drift time measurement

The drift time measurement started by the induced signal at the sense wire

The signal gets amplified, shaped and discriminated

The TDC measures the time between the edges of the logic signal and an external signal („common stop“ (CMS))

Calibration of drift timesCalibration of drift times

Track reconstructionTrack reconstruction

Track fitting:

Energy loss measurement Energy loss measurement with MDCswith MDCs

Energy loss calculation with GARFIELD

Protons above 1GeV nearly minimal ionizing

Protons at 100 MeV have 4 times larger dE/dx compared to ,e,

Simulation with GARFIELDSimulation with GARFIELD

Impact of a asymmetrical Impact of a asymmetrical cathode voltagecathode voltage

Cathode voltage -1000V Cathode voltage -1000V instead t -1750V (MDCI instead t -1750V (MDCI in NOV01)in NOV01)

Electric field deformed Electric field deformed near the cathodenear the cathode

y [cm]x [cm]

drift/v cm s

Relative error of the drift time measurement compared Relative error of the drift time measurement compared to normal working conditions is largeto normal working conditions is large

Affected wire layers should nor be used in analysisAffected wire layers should nor be used in analysis

Impact of a asymmetrical Impact of a asymmetrical cathode voltagecathode voltage

Analysis of the GARFIELD Analysis of the GARFIELD SignalsSignals

„Leading“- and „trailing edge“ –times are calculated at a give threshold

Distribution of drift times of 100 tracks for a given parameter set (minimal distance, angle) are accumulated and the mean and sigma of the time measurement calculated

Shape of the signalsShape of the signals

parallel tracks

Broad arrival time distribution near sense wire

slow electrons from the edge of the drift cell

Anzahl der Cluster pro cm Anzahl der Cluster pro cm als Funktion der als Funktion der TeilchenenergieTeilchenenergie

Nimmt mit steigender Energie ab

Unterschiede zwischen Teilchenspezies ver-schwinden bei hohen Energien

Anzahl der Cluster pro cm Anzahl der Cluster pro cm als Funktion der als Funktion der

GasmischungGasmischung Ändert sich mit

der Zusammensetzung des Zählgases

Nimmt mit steigendem i-Butan Anteil zu

Signalbreite versus Signalbreite versus TeilchenimpulsTeilchenimpuls

Messung einzelner Driftzellen (oberer Reihe)

Normalisierte Signalbreite für ein Segment (unten)

Data Nov01

Segment

Single cellSingle cell

Korrelation der Signalbreite Korrelation der Signalbreite mit dE/dxmit dE/dx

Freie Anpassung mit

Korrelation der Signalbreitenmessung gegenüber dE/dx

Eigenschaft der Ausleseelektronik1

Gute Übereinstimmung für Protonen und Pionen

par0* log( / ) par1dE dx

1 L. Ratti et al., WCC 2004, Vienna, Vortrag 2004.

Data

Zeitauflösung als Funktion Zeitauflösung als Funktion des Schwellenwertesdes Schwellenwertes

Zeitauflösung verschlechtert sich mit steigender Schwelle

Der Effekt ist nahe am Auslesedraht und in den Randbereichen der Driftzelle stärker ausgeprägt

DUBNA

Zeitauflösung als Funktion Zeitauflösung als Funktion der Teilchenenergieder Teilchenenergie

Zeitauflösung verschlechtert sich mit steigender Energie

Der Effekt ist nahe dem Auslesdraht und in den Randbereichen der Driftzelle stärker ausgeprägt

DUBNA

Zeitauflösung als Funktion Zeitauflösung als Funktion der Teilchenenergieder Teilchenenergie

Zeitauflösung verschlechtert sich mit steigender Energie

Data Nov01impact 90°

Zeitauflösung als Funktion Zeitauflösung als Funktion der Schwelleder Schwelle

Änderungen in der Zeitauflösung führen zu einer Verschiebung der Driftzeitmessung mit steigender Energie

resolution in the middle of the cell

DUBNA

Verschiebung in der Verschiebung in der DriftzeitmessungDriftzeitmessung

xt – Relation für 100/1000 MeV Protonen

Effekt ist nahe dem Auslesedraht und in den Randbereichen der Driftzelle stärker ausgeprägt

DUBNA

Verschiebung in der Verschiebung in der DriftzeitmessungDriftzeitmessung

Änderungen in der Zeitauflösung (verursacht durch Änderungen der Ionisation) führt zu einer Verschiebung der Driftzeitmessung mit zunehmender Energie

Timing shift in the middle of the cellDUBNA

VVD D als Funktion der als Funktion der GasmischungGasmischung

Driftgeschwindigkeit in der Mitte der Driftzelle

i-Butan verringert die Driftgeschwindigkeit

Relativer Fehler der DriftzeitmessungRelativer Fehler der Driftzeitmessung

VVD D als Funktion des als Funktion des GasdruckesGasdruckes

Driftgeschwindigkeit in der Mitte der Driftzelle

Driftgeschwindigkeit verringert sich mit steigendem Druck

Relativer Fehler der DriftzeitmessungRelativer Fehler der Driftzeitmessung

VVD D als Funktion der als Funktion der GastemperaturGastemperatur

Driftgeschwindigkeit in der Mitte der Driftzelle

Driftgeschwindigkeit steigt mit steigender Temperatur

Relativer Fehler der DriftzeitmessungRelativer Fehler der Driftzeitmessung

VVD D als Funktion der Oals Funktion der O22 und und NN22 Konzentration Konzentration

Driftgeschwindigkeit in der Mitte der Driftzelle

Effekt vernachlässigbar

VVD D als Funktion der Hals Funktion der H22O O KonzentrationKonzentration

Driftgeschwindigkeit in der Mitte der Driftzelle

Driftgeschwindigkeit nimmt mit steigender H2O-Kozentration ab

Relativer Fehler der DriftzeitmessungRelativer Fehler der Driftzeitmessung

Townsend KoeffizientTownsend Koeffizient

Nimmt mit steigendem i-Butananteil zu

Attachment KoeffizientAttachment Koeffizient

DiffusionskoeffizientenDiffusionskoeffizienten