1

The STAR Time Projection Chamber

Fabrice Retière (LBNL)

for the STAR collaboration

2

Introduction

TPC functionLarge acceptance gas detector

||<1.8

Full azimuthal coverage

Momentum reconstructionTracking with design hit position resolution ~500 m

Pid using dE/dxDesign resolution : 7%

TPC design

Tuning the TPCPosition reconstruction

Drift velocity

Drift distortion

dE/dxUnderstanding ionization

Gain calibration

3

STAR detector

0.5 Tesla magnet0.25 for year 1

Trigger CTBZDCLevel 3

Year 1 detectorsTPCRICH1 SVT ladder

4

5

TPC gas volume

Gas : P10 (Ar-CH4 90%-10%) @ 1 atm

Drift voltage : -31 kV

6

Pad readout

2×12 super-sectors

60 cm

127 cm

190 cm

Outer sector6.2 × 19.5 mm2 pad

3940 pads

Inner sector2.85 × 11.5 mm2 pad

1750 pads

7

Electronic readout

FEE, custom design IC : SAS + SCA (512 time bins)Readout 140K channels, i.e. 70M pixels

Readout boardCarry ~1000 Channels to DAQ

SCAADC

SCAADC

X 16

MUX

TPCPad

PreampShaperAmp

AnalogMemory

Fiber optictransmitterto DAQ

SAS IC SCA + ADCIC

FEEBoard

ReadoutBoard

8

TPC at workFirst RHIC events

Detector very stableGood for physics without calibration

9

TPC at work dE/dx measurement before calibration

K pd

dEdx resolution good for Pid

e

10

Tuning the TPCProcesses to control

IonizationPid using dE/dx

Electron driftDrift distortion

Drift velocity (laser)

GainGas gain

Electronic gain

Particle

11

Electron drift Drift velocity under control

Pressure (mbar)Alexei Lebedev, Bill Love, Jeff Porter (BNL)

Dri

ft v

eloc

ity

(cm

/s)

5.44

5.45

1010 1020

Laser for coarse value

Fine adjustment from tracking matching both side of the TPC

12

Electron drift Drift correction in TPC

Distortion sourcesRadial B field (<2mm)

End cap location (800 m)

E field corrections to field cage (400 m)

0.5 mrad E/B field misalignment (400 m)

Detected using residual average over many tracks

Corrections using field maps and geometry survey

No tuning on data required

Ave

rage

res

idua

l (m

m)

0.3

-0.3

0.

60. 100. 140. Radius (cm)

13

Electron drift B field map correction

Field map allows parameter free calculation

TPC active volume TPC active volume

Calculated distortion = ExBrMeasured Br/Bz

Bill Love, Al Saulys (BNL), Jim Thomas (LBNL)

20

60

100

140

Rad

ius

(cm

)

20

60

100

140

Rad

ius

(cm

)

-200 2000-100 100Distance to central membrane (cm)

-200 2000-100 100Distance to central membrane (cm)

0.4

-0.8

-0.4

0.

Br/

Bz

(%)

R/

dis

tort

ion

(mm

)

0

-1.

1.

14

Electron driftInner/outer sector boundary

Data

Calculation

No wires at the boundary between inner and outer sector

E field leakE field radial component

ExB effect on R/

Radius (cm)

Outer sector Inner sector

Gating grid = -127 VGround plane = 0 V

1.6 cm

Pad row #10 20 30

Ave

rage

res

idua

l (m

m)

Ave

rage

res

idua

l (m

m)

0.2

0.1

-0.1

0.

gap

Inner sector

Inner sector

Outer sector

Outer sector

15

Electron driftdistortions under control

TPC active volume

Huan Huang, Hui Long and Steve Trentelange (UCLA)Jim Thomas (LBNL)

20

60

100

140

Rad

ius

(cm

)

-200 2000-100 100Distance to central membrane (cm)

R/

dis

tort

ion

(mm

)

0

-1.

1.

-2.Pad row #10 20 30

Ave

rage

res

idua

l (m

m)

0.2

-0.1

0.

Inner sector Outer sector

All calculated distortionsAverage residual

16

Gain uniformityGas gain

Eugene Yamamoto (UCLA)

Gain variationOver TPC sectors With time

PressureTemperature…

Correction using average dEdx

Require a lot of events to cancel out fluctuations

Gain monitor chamber being builtPulser for electronic gain calibration

1.3

1.4

1.5

Gai

n (a

rb. u

nit.)

1000

1.6

1010 1020 1030Pressure (mbar)

essure

essure

Gain

Gain

Pr

Pr7.3

17

Gain uniformityElectronic gain

To measure uniformity of electronic gain

8% sigma variation but 20% RMS (tail)Precise channel level correction

Pulser also identifies dead channels = 0.25%

Pad pl

ane

Anode

Groun

d plan

e

Gating

grid

Pulser

TPC drift volume

Inner sector

Outer sector

Pu

lser

am

plit

ude

(arb

. uni

t.)

18

Ionization and gain uniformity dEdx resolution

Remaining issue : correlation of dE/dx between pad rows

Yuri Fisyak (BNL)

No calibration 9 %

With calibration 7.5%

Design 6.7%

Track length (cm)

dE

/dx/

(dE

/dx)

(%

)

19

Conclusion particle identification

Aihong Tang (Kent State U)

K

p d

e

dE/d

x (k

eV/c

m)

0

12

8

4

20

The TPC is an excellent tool for physics

Approaching design performance

Good particle separation using dE/dx

7.5%

-proton separation : 1.3 GeV/c

Position resolution500 m

2-Track resolution2.5 cm

Momentum resolution 2%

Future challengesAchieve turn-key operationHandle increased luminosity

Lots of physics from the year 1 data

Collective flowIdentified particle spectraParticle correlationsEvent by event physicsStrangeness…