First performance results from Phobos at RHIC

1
First performance results from Phobos at RHIC

• Heinz Pernegger for the PHOBOS collaboration
• Vertex 2000

2
PHOBOS Collaboration

• ARGONNE NATIONAL LABORATORY
• Birger Back, Nigel George, Alan Wuosmaa
• BROOKHAVEN NATIONAL LABORATORY
• Mark Baker, Donald Barton, Mathew Ceglia, Alan
  Carroll, Stephen Gushue, George Heintzelman,
  Hobie Kraner ,Robert Pak,Louis Remsberg, Joseph
  Scaduto, Peter Steinberg, Andrei Sukhanov
• INSTITUTE OF NUCLEAR PHYSICS, KRAKOW
• Wojciech Bogucki, Andrzej Budzanowski, Tomir
  Coghen, Bojdan Dabrowski, Marian Despet,
  Kazimierz Galuszka, Jan Godlewski , Jerzy Halik,
  Roman Holynski, W. Kita, Jerzy Kotula, Marian
  Lemler, Jozef Ligocki, Jerzy Michalowski, Andrzej
  Olszewski?, Pawel Sawicki , Andrzej Straczek,
  Marek Stodulski, Mieczylsaw Strek, Z. Stopa, Adam
  Trzupek, Barbara Wosiek, Krzysztof Wozniak, Pawel
  Zychowski
• JAGELLONIAN UNIVERSITY, KRAKOW
• Andrzej Bialas, Wieslaw Czyz, Kacper Zalewski
• MASSACHUSETTS INSTITUTE OF TECHNOLOGY
• Wit Busza, Patrick Decowski, Piotr Fita, J.
  Fitch, C. Gomes, Kristjan Gulbrandsen, P.
  Haridas, Conor Henderson, Jay Kane , Judith Katzy
  , Piotr Kulinich, Clyde Law, Johannes
  Muelmenstaedt, Marjory Neal, P. Patel, Heinz
  Pernegger, Miro Plesko, Corey Reed, Christof
  Roland, Gunther Roland, Dale Ross, Leslie
  Rosenberg, John Ryan, Pradeep Sarin, Stephen
  Steadman, George Stephans, Katarzyna Surowiecka,
  Gerrit van Nieuwenhuizen, Carla Vale, Robin
  Verdier, Bernard Wadsworth, Bolek Wyslouch
• NATIONAL CENTRAL UNIVERSITY, TAIWAN
• Yuan-Hann Chang, Augustine Chen, Willis Lin,
  JawLuen Tang
• UNIVERSITY OF ROCHESTER
• A. Hayes, Erik Johnson, Steven Manly, Robert Pak,
  Inkyu Park, Wojtech Skulski, Teng, Frank Wolfs
• UNIVERSITY OF ILLINOIS AT CHICAGO
• Russell Betts, Christopher Conner, Clive
  Halliwell, Rudi Ganz, Dave Hofman, Richard
  Hollis, Burt Holzman,, Wojtek Kucewicz, Don
  McLeod, Rachid Nouicer, Michael Reuter
• UNIVERSITY OF MARYLAND
• Richard Baum, Richard Bindel, Jing Shea, Edmundo
  Garcia-Solis, Alice Mignerey

3
Relativistic Heavy Ion Collider

• RHIC environment
• Highest energy density ever produced in lab
• Au-Au collisions with total ?s 25TeV
• About 4000 charged particle per central collision
• 12 June 1st Collisions _at_ ?s 56 AGeV
• 24 June 1st Collisions _at_ ?s 130 AGeV
• 5 Sept end of first Au-Au physics run

4
PHOBOS Detector
5
What does Phobos measure ?

• Phobos searches for signs of Quark-Gluon Plasma
  at RHIC
• Measures multiplicity of charged particles over
  full solid angle
• Reconstruct tracks in mid-rapidity range with low
  Pt threshold and identifies them
• Measures particle ratio/spectra, particle
  correlation
• Phobos lives on analog signals of our silicon
  detectors
• Multiplicity measurement use dE/dx as
  multiplicity estimator
• Spectrometer uses dE/dx method for particle
  identification
• Analog information used to reject background
• Analog signals partially used in pattern
  recognition

6
The Multiplicity detector
Vertex
octagon

• 1 layer of large silicon pad detectors
  everywhere
• Count single hits or sum of analog signals in a
  detector area as measure of particle multiplicity
• Has to deal with high occupancy (gt80)

7
The silicon spectrometer
1x1mm to 0.7x19mm

• 16 layer of smaller silicon pad detectors near
  mid rapdity
• Tracks and Identifies particles (dE/dx) in 2T
  magnetic field
• All silicon readout with Viking VAHDR1 chips
• Very high dynamic range (gt100MIPs), peaking time
  1.1ms

8
Our silicon detectors
Double Metal, Single sided, AC coupled,
polysilicon biased detectors produced by ERSO,
Taiwan
AC coupled Pad (p-implant metal 1
pad) polisilicon bias resistor metal 2 readout
line contact hole metal 1- metal 2
9
Before installation

• The full silicon detector in numbers
• 500 wafers, 1600 Viking VAHDR1 readout chips
• 9 different wafer layouts produced by
  Miracle/Erso, Taiwan
• Assembled to 240 modules with 140 000 channels
• Commissioning setup (15 of full) March-July
• Study environment and measure first collisions
• Full installation for physics run on July 13
• 200/200 modules functional
• 1082/1084 chips functional 99.8
• In channels 98.8 channels fully functional
• Peak Signal/Noise 131 to 201 depending on
  sensor layout
• Original requirements S/Ngt10 and full
  functional channels gt95

10
RHIC beams in Phobos
RHIC Integrated Luminosity 6565 GeV
Integrated Luminosity ?B-1
PR00 Start6 Bunches
Start 55 Bunches
Physics Run 2000
Date
Luminosity estimated using coincidence of signals
in the Zero Degree Calorimeters. ?10.7barn used
to convert counts to luminosity.
11
Run 5332 Event 35225 08/31/00 065924PHOBOS
Online Event Display
Trigger Scintillators P
Spectrometer Arm P
Octagon Multiplicity detector
Au-Au Beam Momentum 65.12 GeV/c
Spectrometer Arm N
Trigger Scintillators N
Not to scale
Not all sub-detectors shown
12
Performance of the Multiplicity Detector
Opening to Spec
phi
Opening to Vtx
Opening to Spec
Opening to Vtx
Z (beam)

• One high multiplicity event in the octagon
• occupancy up to 80
• Color encodes pulse height

13
Dealing with high occupancy
Base line before and after correction

• Problems associated with high occupancy
• Few channels left to determine common-mode-noise
  correction
• Event-by-event baseline shift dependent on input
  signal

14
Signal dependence on occupancy

• Problems associated with high occupancy
• Gain dependence on occupancy can distort the
  multiplicity measurement
• Multiplicity measured
• dE(meas)/ltdE(part)gt
• Gain loss at highest occupancy
• 20 NO baseline corr.
• 6 WITH baseline corr.

15
Multiplicity sensor uniformity
3.6 x 8.4 cm
Smp 93 keV
/- 3
8.3 cm x 6.5 cm
Smp 85keV
/- 1
No substantial signal variation due to different
layout (double metal line routing/ varying pad
size)
16
Performance of tracking detectors
Hits in SPEC
Tracks in SPEC
Hits in VTX
130 AGeV
130 AGeV
56 AGeV
17
Signal uniformity in Spec/Vertex
T3 Smp 85 T4 Smp85 T5 Smp85
T1 Smp 90 T2 Smp85

• Signal distrbutions for different layouts
• All signal distribution after calibration (20
  effect!)
• Small pads (type 1 2 , 1mm2)
• Larger pads (type 3,4,5 10 mm2)
• strips (vertex 0.4x20 mm2)
• Very uniform in shape and peak

Inner Vtx Smp87 Outer Vtx Smp85
18
Uniformity within sensors
/- 2
1 x 1 mm2
0.4 x 6 mm2
0.7 x 7.5 mm2
Relative signal variation
0.7 x 15 mm2
0.7 x 19 mm2
0.3 x 23 mm2

• Typical variation lt/-1 within sensor over large
  range of pad size and readout line length

0.3 x 46 mm2
Pad row (along readout lines accros sensor)
19
Signal/Noise vs sensor layout
Closest to beam
Signal peak e-
24000e-
Large pads Longs readout lines (high capacitance)
Noise e-
Chip dominated base offset (ENC 900 e-5e-/pF _at_
1.1ms)
20
Focus on Si signal simulation

• Critical test of detector understanding
• Both distributions contain the same number of
  central events
• Points are for VTX data
• No correction for detector thickness
• Histogram is for simulated VTX signals
• GEANT
• Response from test-beam
• Electronics noise
• Shulek correction

(CR setup)
21
Optimizing our signal simulation

• Measured dE/dx in silicon in a testbeam and
  verified simulation
• Measure dE/dx and distribution shape, test PID
• Cover large momentum range (130MeV 8GeV),
  measure p K

? Data

• p
• K

? Geant
22
Measuring charged multiplicity

• VTX Tracklets
• Two hit combinations that point to the vertex
• dh h2 h1
• Good tracklets have dhlt.1

• SPEC Tracklets
• Two hit combinations that point to the vertex
• dR ? (dh2 df2)
• Good tracklets have dRlt.015

23
Results PHOBOS Measurement of Charged Particle
Multiplicitynear Mid-rapidity
dNch/d? (?lt1) at ? sNN 56 GeV 4081230
dNch/d? (?lt1) at ? sNN130 GeV 5551235
hep-ex/0007036 Accepted for publication in PRL
Oct 02 2000
24
Summary

• The good performance allowed a very fast physics
  analysis
• Submitted within 5 week after first recorded
  collision
• The first publication of all RHIC experiment
• Phobos successfully completed its first physics
  run
• 3.5 million Au-Au collisions on tape (collected
  mainly in 2 weeks)
• Phobos silicon detector operated flawlessly
• 98 off al channels fully functional
• Not a single module failure during installation
  and all running
• Operates at S/N gt15
• Phobos is well equipped for future analysis
• Very uniform and well calibrated signal response
• Can operate at high occupancies
• Detector showed to be reliable and stable

25
Next transparencies are backup additions
26
Readout Calibration system

• Readout with Viking VAHDR1 chips
• Very high dynamic range (gt100MIPs), peaking time
  1.1ms
• Phobos lives on analog signals
• Multiplicity measurement use dE/dx as
  multiplicity estimator
• Spectrometer uses dE/dx method for particle
  identification
• Analog information used to reject background
• Analog signals partially used in pattern
  recognition
• Dedicated calibration system
• Measures full gain curve for each channel
  (1-2/day)
• Verifies functionality and normalizes gain of
  different detector modules and sensors

27
Derivation of dN/dh

• Extract a(Z) from correlation of
• Primaries in 1 lt h lt 1
• Measured number of tracklets

5ltzlt10
Number of Tracklets
VTX
SPEC
dN/dh
28
Measuring dN/dh with tracklets

• Number of reconstructed tracklets is proportional
  to dN/dh hlt1
• To reconstruct tracklets
• Reconstruct vertex
• Define tracklets based on the vertex and hits in
  the front planes of SPEC and VTX
• Redundancy essentially eliminates feed-down,
  secondaries, random noise hits
• To determine a
• Run the same algorithm through the MC
• Folds in detector response and acceptance

29
Measuring Vertex

• Spectrometer sits very close to vertex
• High resolution tracking in 6 planes gives
  excellent vertex resolution

• Pointing accuracy describes how extrapolated
  tracks deviate from calculated vertex.
• Compares well with HIJING simulation