1 STAR TOF Calibration. 2 Detectors TPC(TPX) - tracking MRPC TOF (TOFr) – stop time measurement pVPD/upVPD - start time measurement Particle momentum;

1

STAR TOF Calibration

2

DetectorsTPC(TPX)- tracking

MRPC TOF (TOFr) – stop time measurement

pVPD/upVPD - start time measurement

Particle momentum; dE/dx

~8% resolution

/K separation to ~0.6GeV/c

k/p separation to ~1.0GeV/c

Aim at 100ps time resolution TOF system

/K separation to ~1.6GeV/c & k/p separation to ~3.0GeV/c

3

TOF system

pVPD/VPD consists of plastic scintillators + fast PMTs, 5.4~5.7m from TPC geometry center along the beam line

3 channels each side (runV and before) to 19 channels each side (after runV)

Full barrel TOF consists of 120 trays, 32 MRPC modules per tray, 6 channels per module

Only one tray in run V, 5 trays in run VIII, 94 trays in run IX and all 120 trays in runX

4

TOF Electronics

Start Detector

TOF Tray

TDIG TDIG

TCPU

DIFFERENTIAL DATA AND CLOCK

TINO TINO

MRPC MRPC

COMMANDS

48 CHAN

48 CHAN

48 CHAN

TRAY CAN BUS

MULTIPLICITY

THUB

COPPER: DATA, SAMPLE CLOCK, RESET TRIGGER STROBE & DATA

TOP LEVEL CAN BUS

COPPER LINKS TO 29 TRAYS

RHIC CLOCK

SIU FIBER

CAN BUS TO 29 TRAYS

TCD

DAQ

48 CHAN

4

L0 Trigger

4

PMTs TPMT TDIS TCPU

Diff. Data & Clk

Local CANbus

Co

pp

er:

Da

ta,

Sa

mp

le C

lk,

Re

set

Trg

Str

ob

e &

Da

ta

Reset & ClockFrom other THUBs

Reset & ClockTo other THUBs

2 THUB per side, 4 total 1 TCPU per tray/VPD, 122 total

1 TPMT/TDIS per VPD, 2 total8 TINO/TDIG per tray, 960 total

24 channel per TINO/TDIG, 23040 total

5

HPTDC INL CorrectionIntegral Non-Lineraity

Diffrential Non-Lineraity

6

update

Calibration results

Related software

7

TOF Calibration – runV (and later on)

Leading-edge timing signal, trailing-edge signal width (TOT)

time

Amp.

T1l T2l T2t T1t

Thre.

Time stamp

Leading-edge trigger, signal charge (ADC) is measured for correction

time

Amp.

T1 T2

Thre.

Q1

Q2

8

VPD

Collision Point

VPD PMTs •All hits in the same event arrived at the same time•As reference for each other

9

VPD calibration procedure

10

TDC-TOT correlation

Use spline fit (3rd polynomials TSplineFit based on ROOT, developed by François-Xavier Gentit)

Run VIII, 200 GeV ppRun V, 62.4 GeV CuCu

TD

C1 –

(T

DC

2+T

DC

3)/2

(n

s)

TOT (ns)TOT (ns)

11

Resolution before/after calibration

~6 ns ~0.16 ns

Bef

ore

cali

brat

ion

After

calibration

TDC1 – (TDC2+TDC3)/2 (ns)

TDC1 – (TDC2+TDC3)/2 (ns)

Run VIII200 GeV pp

12

Global offset tune• Slewing effect correction was done on east/west side separately• Absolute time from east/west side are floating• Tune the global offset between east/west side to match Vz(VPD) with Vz(TPC)

Beam line

Vertex Z

zT0+(L/2-Vz)/cT0+(L/2+Vz)/cWest East

upVPDupVPD

CP

1 1

( ) ( / / ) / 2e wi N j N

ei e wj wi j

Vz VPD t N t N c

1 1

( ) /e wi N j N

ei wj e wi j

starte w

t t N N Vz c

TN N

13

Choose sample by limiting dE/dx and momentum range (or pre-calibrated TOF)

T0 correction, different cable length and signal transition time

TOFr TOT and Z position calibration, using charged pion sample.

Iteration several times (if needed)

TOFr Calibration procedure

• Try channel-by-channel first• Not enough statistics? Then

try module-by-module, or board-by-board

14

T0 correction

TOFmeasured – TOFexpected (ns)

Com

ts

Channel-by-ChannelSimple Gaussian function FitShift the mean of TOFmeasured – TOFexpected to 0

Run VIII200 GeV pp

15

TOFr TOT slewing correction

Run V 200 GeV CuCuSignal ringing effect

Run VIII 200 GeV ppBetter signal termination achieved

TOT (ns)

TOT (ns)

TO

Fm

easu

red

– T

OF

exp

ecte

d (n

s)

16

Hit position (Z) corrction

TO

Fm

easu

red –

TO

Fex

pec

ted

(ns)

Hit Position Z (cm)

Run V 200 GeV CuCu

Run VIII 200 GeV pp

MRPC stripMRPC stripReadoutReadout

Finite signal propagation speed vs. hit position

17

TOF Time Resolution Run V CuCu 62.4GeV

Run V CuCu 200GeV

Run VIII pp 200GeV

18

TOF resolution summary

Operation conditionTime Resolution (ps)

pVPDTOFr

(overall)TOFr

(stop time)

Run III200GeV d+Au ~85 ~120 ~85

200GeV p+p ~140 ~160 ~80

Run IV

62GeV (Au+Au) ~55 ~105 ~89

200GeV (Au+Au)

FF/RFF ~27 ~74 ~70

HF ~20 ~74 ~71

Run V200GeV Cu+Cu (TOT) ~ 50 ~92 ~7562GeV Cu+Cu (TOT) ~ 82 ~125 ~94

Run VIII200 GeV d+Au NA NA NA

200 GeV p+p (TOT) ~83 ~112 ~75

19

Expectation (ideal)

TOF calibration w/o start detector

nTTtofZcorTOTcorTDCTn

i iiiii

10__

Fast simulation

TOF mean time resolution: 90ps

Time-TOT/Z correlation taken from run V CuCu data

T0 obtained by direct average of Ti

Several iteration of calibration

20

T0 w/o start detector – from data

expected) (as ps74~0

2ps74~0

T

nTTi

60ps) :(expected ps63~0

3ps89~0

T

nTTi

n

i i

n

i iiii tofZcorTOTcorTDCT1

2

1

2 __0

Run V 62.4 GeV CuCu

After 6-round calibration iteration

21

Summary VPD Calibration

INL correction Slewing correction

TOF tray Calibration INL correction T0 correction Slewing correction Hit position correction

Calibration database INL table T0 table T-TOT table T-Z table

Calibration w/o VPD Fast simulation Data results

22

Backup Slides

23

Output calibration parameter

• Inconvinient to retrieve and use the parameters of all of the Splines • Upload each TOT bin boundaries and the corresponding corrections• Interpolate in each TOT bin

Bin boundarySame statistics in each bin

24

#Events needed for calibration

Collisions <dNchraw/d>a

rXiv:

0808.2041

×1/4(pure pion)

×80% (match)

×2 ()

Useable hits per channel

#Events for

channel-by-

channel

#Events for

module-by-

module

#Events for

board-by-

board

#Events for T0

(500/channel

)

p+p (MB) 2.4 0.96 4.2e-5 240 M 40 M 10 M 12 M

d+Au(MB) 10.2 4.1 1.8e-4 56 M 9.3 M 2.3 M 2.8 M

Au+Au(MB) 200 80 3.5e-3 2.9 M 0.5 M 0.12 M 0.15 M

Au+Au

(0-10%)

515 206 8.9e-3 1.2 M 0.2 M 0.05 M 0.06 M

10 K for slewing correction

Note in pp collisions, the upVPD efficiency is only ~40% if requiring >=2 hits on both sides

25

VPD fired channels

• ~35% events have >=3 hits on either side (useable for calibration)• ~40% events have >=2 hits on both sides (better start time or • ~15% events have >=3 hits on both sides Vz resolution)

Run VIII200 GeV pp

Zebo Tang

26

Track projection and Vz correlation

2D DCA

Beam line uncalibrated, fix to (0.5,0) cmShould be different between fillsUnnecessary to parallel to z direction

Vertex from track projection (No vertex from TPX ): project tracks matched to TOF hits, get the closest position to beam line.

27

Calibration Procedure – before run V

Leading-edge trigger, signal charge (ADC) is measured for correction

time

Amp.

T1 T2

Thre.

Q1

Q2

28

pVPD Calibration – start time

xdx

c

x

baxf

29

TOFr Calibration – MRPC TA & TZ

TA slewing correction Hit position (TZ) correction

2x

e

xx

d

x

c

x

baxf

7th-order polynomial

30

Time resolution determination

332~

22~

0

321

21

nTTT

nTT

TTi 35.1~

22~

0

nT

nT

TT

i

i

i

Increase with n (assume equal time resolution for each channel)

320~3~

20~2~

0

nTTT

nTTT

T

ii

ii

31

Calibrated time resolution

32

Hadron PID

33

Electron PID

34

Muon PID

200 GeV CuCu

35

Run-VIII PID

1.2<p<1.4 GeV/c