gem-tpc resolution studies ecfa/desy lc workshop prague, november 2002 dean karlen university of...
TRANSCRIPT
GEM-TPC Resolution Studies
ECFA/DESY LC WorkshopPrague, November 2002
Dean KarlenUniversity of Victoria / TRIUMF
November 2002 Dean Karlen / GEM-TPC Resolution Studies 2
GEM-TPC Resolution StudiesA TPC read out by micropattern gas avalanche detectors can provide excellent track resolution How to do this at a reasonable cost?
i.e. how to do this with “large” pads?
Outline of Presentation: Simulation studies on pad geometry (Jeju results)
rectangles vs. chevrons rectangular pad width optimization benefit of staggering rectangular pads
Analysis of new data sample with different pad sizes
November 2002 Dean Karlen / GEM-TPC Resolution Studies 3
Simulation StudiesJava based framework for TPC simulation and data analysis ionization drifted/diffused/multiplied/collected
ionization on pads pad signals
November 2002 Dean Karlen / GEM-TPC Resolution Studies 4
Track Fittingx-y track fit uses a linear Gaussian model for the ionization cloud ie. no fluctuations
three parameter fit: x0 (x at y=0) (azimuthal angle) (transverse s.d. of cloud)
maximize the likelihood of the observed charge fractions from each row
November 2002 Dean Karlen / GEM-TPC Resolution Studies 5
Simulated TPC ParametersGas mix considered: Ar CF4
fast at low fields low transverse diffusion in magnetic fields
larger diffusion at higher fields
Example: 98% Ar, 2% CF4
E (V/cm)
0 1000 2000 3000 4000 5000
velo
city
(cm
/s)
0
2
4
6
8
E (V/cm)
0 1000 2000 3000 4000 5000
tra
nsve
rse
diff
usio
n (
cm/s
qrt
(cm
))
0.00
0.01
0.02
0.03
0.04
0.05
0.06
MagboltzB = 4T
Drift field
GEM transfer field
Electronattachment?
2 x 6 mm2
November 2002 Dean Karlen / GEM-TPC Resolution Studies 6
GEM TPC
Naïve calculation for optimum resolution:
200 cm 1cm
cm/ m 27 D cm/ m 500
G
m23mm/e 9mm 30
200 m27[cm]
primary
Nx
D
defocusingregion
November 2002 Dean Karlen / GEM-TPC Resolution Studies 7
Ar CF4 (98:2): 5 rows of 2 mm x 6 mm pads
Drift length (cm)
0 50 100 150 200
x 0 re
solu
tion
(mic
rons
)
0
20
40
60
80
100
120
140
standard row layoutchevron 2chevron 4chevron 10optimum (naive)
Chevronsunnecessaryin Ar CF4
GEM TPC
November 2002 Dean Karlen / GEM-TPC Resolution Studies 8
TPC Without Defocusing
Naïve calculation for optimum resolution:
200 cm 0.1 mm
cm/ m 27 D
m23mm/e 9mm 30
200 m27[cm]
primary
Nx
D
November 2002 Dean Karlen / GEM-TPC Resolution Studies 9
Defocusingrequired formicromegas
Ar CF4 (98:2): 5 rows of 2 mm x 6 mm pads
Drift length (cm)
0 50 100 150 200
x 0 re
solu
tion
(mic
rons
)
0
20
40
60
80
100
120
140
standard row layoutchevron 2chevron 4chevron 10optimum (naive)
November 2002 Dean Karlen / GEM-TPC Resolution Studies 10
Comparison of Pad Widths
50 cm drift: corresponds to transverse cloud s.d. of 0.58 mm
Pad width (mm)1 2 3 4
x 0 re
solu
tion
(mic
rons
)
20
22
24
26
28
30
32
34
Number of (cloud std. dev.s)1 2 3 4 5 6 7
Cos
t (ar
b. c
urre
ncy)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
2 parameter fit3 parameter fit
Cost
November 2002 Dean Karlen / GEM-TPC Resolution Studies 11
Staggering Comparison (3 par. fit)
local (rad)
0.00 0.02 0.04 0.06 0.08 0.10
x 0 re
solu
tion
(mic
rons
)
15
20
25
30
35
1 mm staggered2 mm staggered4 mm staggered1 mm not staggered2mm not staggered4 mm not staggered
“Track angleeffect”
Cloud widthpoorlydetermined
November 2002 Dean Karlen / GEM-TPC Resolution Studies 12
Staggering Comparison (2 par. fit)
local (rad)
0.00 0.02 0.04 0.06 0.08 0.10
x 0 re
solu
tion
(mic
rons
)
15
20
25
30
35
1 mm staggered2 mm staggered4 mm staggered1 mm not staggered2 mm not staggered4 mm not staggered
cloud width fixed to 0.58 mm
GEM-TPC Data Analysis
November 2002 Dean Karlen / GEM-TPC Resolution Studies 14
GEM-TPC Data AnalysisCanadian LC - TPC Group: R. Carnegie, M. Dixit, S. Kennedy, H. Mes, E. Neuheimer,
A. Rankin, K. Sachs, V. Strickland – Carleton University J.-P. Martin – University
of Montreal P. Birney, D. Karlen,
G. Rosenbaum – University of Victoria / TRIUMF
TPC 1B: 15 cm drift (no B field) double GEM ALEPH preamp + custom
FADC readout
November 2002 Dean Karlen / GEM-TPC Resolution Studies 15
Pad LayoutWe have recorded a large dataset with a new pad layout: 3x multiplexed Ar CO2 (90:10) and P10 ~ 400 K events ~ 75 GB
Test set 2 - (16-n) channels (49+n+1 to 64)
48 45
46
46
47 45
45
4525
39 40 4543 44 37 3844 37 41 4240 41 42 4346 37 38 39 40 41 42 43 44
35 36 4531 32 33 3434 35 36 3030 31 32 3333 34 35 3646 30 31 32
26 27 2828 29 23 2424 25 26 2746 23 24 25 26 27 28 29 23
Test set 1 - n channels (49 to 49+n)
19 20 2122 16 17 1818 19 20 2121 22 16 1746 16 17 18 19 20
14 1513 14 15 99 10 4510 11 12 1311 12
46
46 9 10 11 12 13 14 15
3 4 457 8 1 28 1 5 64 5 6 75 6 7 81 2 3 4
old
new
November 2002 Dean Karlen / GEM-TPC Resolution Studies 16
Example EventsOuter 6 rows are used to define track parameters inner two rows: resolution studies (fit for x0 alone) 2 mm x 6 mm / 3 mm x 5 mm
November 2002 Dean Karlen / GEM-TPC Resolution Studies 17
Transverse DiffusionTransverse cloud size: 22
02 ]cm[dD
drift time (5 ns bins)
sigm
a2 (m
m2 )
uncorrected drift time (5 ns bins)
sigm
a2 (m
m2 )
Ar CO2 P10
0 = 450-500 mD = 190 m / cm
0 = 450-500 mD = 500 m / cm
November 2002 Dean Karlen / GEM-TPC Resolution Studies 18
x0 Resolution
Example: for Ar CO2 (90:10) d < 2 cm || < 0.1 rad
pad width: 2 mm <> = 0.5 mm w/<> = 4
resolution:140 m
residual (mm)
November 2002 Dean Karlen / GEM-TPC Resolution Studies 19
Resolution vs. Drift Distance
Drift distance (cm)
0 2 4 6 8 10 12 14
Res
olut
ion
(mic
rons
)
0
50
100
150
200
250
300
350
3 mm x 5 mm pads
2 mm x 6 mm pads
Naïve optimum: - all primaries collected - upper: 0 before amp. - lower: 0 after amp.
MC simulations: - GEM efficiencies: collection 1.0 extraction 0.7
Ar CO2|| < 0.1
November 2002 Dean Karlen / GEM-TPC Resolution Studies 20
Drift distance (cm)
0 2 4 6 8 10 12 14
Res
olut
ion
(mic
rons
)
0
100
200
300
400
500
600
Resolution vs. Drift Distance
3 mm x 5 mm pads
2 mm x 6 mm pads
Naïve optimum: - all primaries collected - upper: 0 before amp. - lower: 0 after amp.
P 10
It appears that a smallfraction of primariescontribute to track fit.
|| < 0.1
November 2002 Dean Karlen / GEM-TPC Resolution Studies 21
Distance to edge of pad (mm)
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Res
olut
ion
(mic
rons
)
0
50
100
150
200
250
300
350
Resolution vs. Distance to Pad Edge
3 mm x 5 mm pads
2 mm x 6 mm pads
|| < 0.05d < 4 cm<> = 0.52 mm
November 2002 Dean Karlen / GEM-TPC Resolution Studies 22
Fixing the Cloud WidthSmall improvement in resolution for 2 par. fit Ar CO2 (90:10) d < 2 cm || < 0.05 rad
pad width: 2 mm <> = 0.5 mm w/<> = 4
resolution:130 m
residual (mm)
November 2002 Dean Karlen / GEM-TPC Resolution Studies 23
SummarySimulation studies Rectangular pads favoured Pad widths should be kept below 3-4 cloud s.d.
pad width may be determined by 2 track resolution
Staggering helps determine cloud width
TPC data analysis Large new data set: analysis only just begun Have attained 130 m resolution with 2 6 mm2 pads Should be able to do much better
need improved GEM operating point (?)