Construction and Commissioning of the CALICE Analog Hadron Calorimeter Prototype

Abstract

An analog hadron calorimeter (AHCAL) prototype of 5.3 nuclear interaction lengths thickness has been constructed by members of the CALICE Collaboration. The AHCAL prototype consists of a 38-layer sandwich structure of steel plates and highly-segmented scintillator tiles that are read out by wavelength-shifting fibers coupled to SiPMs. The signal is amplified and shaped with a custom-designed ASIC. A calibration/monitoring system based on LED light was developed to monitor the SiPM gain and to measure the full SiPM response curve in order to correct for non-linearity. Ultimately, the physics goals are the study of hadron shower shapes and testing the concept of particle flow. The technical goal consists of measuring the performance and reliability of 7608 SiPMs. The AHCAL was commissioned in test beams at DESY and CERN. The entire prototype was completed in 2007 and recorded hadron showers, electron showers and muons at different energies and incident angles in test beams at CERN and Fermilab. Comment: 36 pages, 32 figures

Full text

Page 1

Preprint typeset in JINST style - HYPER VERSION
DESY 10-032
March 2010
Construction and Commissioning of the
CALICE Analog Hadron Calorimeter
Prototype
The CALICE Collaboration
C.Adloff, Y.Karyotakis
Laboratoire d’Annecy-le-Vieux de Physique des Particules, Université de Savoie,
CNRS/IN2P3, 9 Chemin du Bellevue BP 110, F-74941 Annecy-le-Vieux Cedex, France
J.Repond
Argonne National Laboratory, 9700 S. Cass Avenue, Argonne, IL 60439-4815, USA
A.Brandt, H.Brown, K.De, C.Medina, J.Smith, J.Li, M.Sosebee, A.White,
J.Yu
Department of Physics, SH108, University of Texas, Arlington, TX 76019, USA
T.Buanes, G.Eigen
University of Bergen, Inst. of Physics, Allegaten 55, N-5007 Bergen, Norway
Y.Mikami, O.Miller, N.K.Watson, J.A.Wilson
University of Birmingham, School of Physics and Astronomy, Edgbaston, Birmingham
B15 2TT, UK
T.Goto, G.Mavromanolakis∗, M.A.Thomson, D.R.Ward, W.Yan†
University of Cambridge, Cavendish Laboratory, J J Thomson Avenue, CB3 0HE, UK
D.Benchekroun, A.Hoummada, Y.Khoulaki
Université Hassan II Aïn Chock, Faculté des sciences. B.P. 5366 Maarif, Casablanca,
Morocco
M.Oreglia
University of Chicago, Dept. of Physics, 5720 So. Ellis Ave., KPTC 201 Chicago, IL
60637-1434, USA
M.Benyamna, C.Cârloganu, P.Gay
Laboratoire de Physique Corpusculaire de Clermont-Ferrand (LPC), 24 avenue des
Landais, 63177 Aubière CEDEX, France
J.Ha
Korea Atomic Energy Research Institute, Taejon 305-600, South Korea
– 1 –
arXiv:1003.2662v1 [physics.ins-det] 13 Mar 2010

Page 2

G.C.Blazey, D.Chakraborty, A.Dyshkant, K.Francis, D.Hedin, G.Lima,
V.Zutshi
NICADD, Northern Illinois University, Department of Physics, DeKalb, IL 60115, USA
V.A.Babkin, S.N.Bazylev, Yu.I.Fedotov, V.M.Slepnev, I.A.Tiapkin,
S.V.Volgin
Joint Institute for Nuclear Research, Joliot-Curie 6, 141980, Dubna, Moscow Region,
Russia
J.-Y.Hostachy, L.Morin
Laboratoire de Physique Subatomique et de Cosmologie - Université Joseph Fourier
Grenoble 1 - CNRS/IN2P3 - Institut Polytechnique de Grenoble, 53, rue des Martyrs,
38026 Grenoble CEDEX, France
N.D’Ascenzo, U.Cornett, D.David, R.Fabbri, G.Falley, N.Feege, K.Gadow,
E.Garutti, P.Göttlicher, T.Jung, S.Karstensen, V.Korbel,
A.-I.Lucaci-Timoce, B.Lutz, N.Meyer, V.Morgunov, M.Reinecke,
S.Schätzel, S.Schmidt, F.Sefkow, P.Smirnov, A.Vargas-Trevino,
N.Wattimena, O.Wendt
DESY, Notkestrasse 85, D-22603 Hamburg, Germany
M.Groll, R.-D.Heuer, S.Richter‡, J.Samson
Univ. Hamburg, Physics Department, Institut für Experimentalphysik, Luruper Chaussee
149, 22761 Hamburg, Germany
A.Kaplan, H.-Ch.Schultz-Coulon, W.Shen, A.Tadday
University of Heidelberg, Fakultat fur Physik und Astronomie, Albert Uberle Str. 3-5
2.OG Ost, D-69120 Heidelberg, Germany
B.Bilki, E.Norbeck, Y.Onel
University of Iowa, Dept. of Physics and Astronomy, 203 Van Allen Hall, Iowa City, IA
52242-1479, USA
E.J.Kim
Chonbuk National University, Jeonju, 561-756, South Korea
G.Kim, D-W.Kim, K.Lee, S.C.Lee
Kangnung National University, HEP/PD, Kangnung, South Korea
K.Kawagoe, Y.Tamura
Department of Physics, Kobe University, Kobe, 657-8501, Japan
J.A.Ballin, P.D.Dauncey, A.-M.Magnan, H.Yilmaz, O.Zorba
Imperial College, Blackett Laboratory, Department of Physics, Prince Consort Road,
London SW7 2AZ, UK
V.Bartsch§, M.Postranecky, M.Warren, M.Wing
Department of Physics and Astronomy, University College London, Gower Street, London
WC1E 6BT, UK
– 2 –

Page 3

M.Faucci Giannelli, M.G.Green, F.Salvatore¶
Royal Holloway University of London, Dept. of Physics, Egham, Surrey TW20 0EX, UK
R.Kieffer, I.Laktineh
Université de Lyon, F-69622, Lyon, France ; Université de Lyon 1, Villeurbanne ;
CNRS/IN2P3, Institut de Physique Nucléaire de Lyon
M.C Fouz
CIEMAT, Centro de Investigaciones Energeticas, Medioambientales y Tecnologicas,
Madrid. Spain
D.S.Bailey, R.J.Barlow, R.J.Thompson
The University of Manchester, School of Physics and Astronomy, Schuster Lab,
Manchester M13 9PL, UK
M.Batouritski, O.Dvornikov, Yu.Shulhevich, N.Shumeiko, A.Solin,
P.Starovoitov, V.Tchekhovski, A.Terletski
National Centre of Particle and High Energy Physics of the Belarusian State University,
M.Bogdanovich str. 153, 220040 Minsk, Belarus
B.Bobchenko, M.Chadeeva, M.Danilov, O.Markin, R.Mizuk, V.Morgunov,
E.Novikov, V.Rusinov, E.Tarkovsky
Institute of Theoretical and Experimental Physics, B. Cheremushkinskaya ul. 25,
RU-117218 Moscow, Russia
V.Andreev, N.Kirikova, A.Komar, V.Kozlov, P.Smirnov, Y.Soloviev,
A.Terkulov
P.N. Lebedev Physical Institute, Russian Academy of Sciences, 117924 GSP-1 Moscow,
B-333, Russia
P.Buzhan, B.Dolgoshein, A.Ilyin, V.Kantserov, V.Kaplin, A.Karakash,
E.Popova, S.Smirnov
Moscow Physical Engineering Inst., MEPhI, Dept. of Physics, 31, Kashirskoye shosse,
115409 Moscow, Russia
N.Baranova, E.Boos, L.Gladilin, D.Karmanov, M.Korolev, M.Merkin,
A.Savin, A.Voronin
M.V.Lomonosov Moscow State University, D.V.Skobeltsyn Institute of Nuclear Physics
(SINP MSU), 1/2 Leninskiye Gory, Moscow, 119991, Russia
A.Topkar
Bhabha Atomic Research Center, Mumbai 400085, India
A.Frey?, C.Kiesling, S.Lu, K.Prothmann, K. Seidel, F.Simon, C. Soldner, L.
Weuste
Max Planck Inst. für Physik, Föhringer Ring 6, D-80805 Munich, Germany
– 3 –

Page 4

B.Bouquet, S.Callier, P.Cornebise, F.Dulucq, J.Fleury, H.Li,
G.Martin-Chassard, F.Richard, Ch.de la Taille, R.Poeschl, L.Raux, M.Ruan,
N.Seguin-Moreau, F.Wicek
Laboratoire de L’accélerateur Linéaire, Centre d’Orsay, Université de Paris-Sud XI, BP
34, Bâtiment 200, F-91898 Orsay CEDEX, France
M.Anduze, V.Boudry, J-C.Brient, G.Gaycken, R.Cornat, D.Jeans, P.Mora
de Freitas, G.Musat, M.Reinhard, A.Rougé, J-Ch.Vanel, H.Videau
École Polytechnique, Laboratoire Leprince-Ringuet (LLR), Route de Saclay, F-91128
Palaiseau, CEDEX France
K-H.Park
Pohang Accelerator Laboratory, Pohang 790-784, South Korea
J.Zacek
Charles University, Institute of Particle & Nuclear Physics, V Holesovickach 2,
CZ-18000 Prague 8, Czech Republic
J.Cvach, P.Gallus, M.Havranek, M.Janata, J.Kvasnicka, M.Marcisovsky,
I.Polak, J.Popule, L.Tomasek, M.Tomasek, P.Ruzicka, P.Sicho, J. Smolik,
V.Vrba, J.Zalesak
Institute of Physics, Academy of Sciences of the Czech Republic, Na Slovance 2,
CZ-18221 Prague 8, Czech Republic
Yu.Arestov, V.Ammosov, B.Chuiko, V.Gapienko, Y.Gilitski,V.Koreshev,
A.Semak, Yu.Sviridov, V.Zaets
Institute of High Energy Physics, Moscow Region, RU-142284 Protvino, Russia
B.Belhorma, M. Belmir
Centre National de l’Energie, des Sciences et des Techniques Nucléaires, B.P. 1382, R.P.
10001, Rabat, Morocco
A.Baird, R.N.Halsall
Rutherford Appleton Laboratory, Chilton, Didcot, Oxon OX110QX, UK
S.W.Nam, I.H.Park, J.Yang
Ewha Womans University, Dept. of Physics, Seoul 120, South Korea
Jong-Seo Chai, Jong-Tae Kim, Geun-Bum Kim
Sungkyunkwan University, 300 Cheoncheon-dong, Jangan-gu, Suwon, Gyeonggi-do
440-746, South Korea
Y.Kim
Korea Institute of Radiological and Medical Sciences, 215-4 Gangeung-dong, Nowon-gu,
Seoul 139-706, SOUTH KOREA
J.Kang, Y.-J.Kwon
Yonsei University, Dept. of Physics, 134 Sinchon-dong, Sudaemoon-gu, Seoul 120-749,
South Korea
– 4 –

Page 5

Ilgoo Kim, Taeyun Lee, Jaehong Park, Jinho Sung
School of Electric Engineering and Computing Science, Seoul National University, Seoul
151-742, South Korea
S. Itoh, K.Kotera, M. Nishiyama,T.Takeshita
Shinshu Univ., Dept. of Physics, 3-1-1 Asaki, Matsumoto-shi, Nagano 390-861, Japan
S.Weber, C.Zeitnitz
Bergische Universität Wuppertal Fachbereich 8 Physik, Gausstrasse 20, D-42097
Wuppertal, GERMANY
ABSTRACT: An analog hadron calorimeter (AHCAL) prototype of 5.3 nuclear interaction
lengths thickness has been constructed by members of the CALICE Collaboration. The
AHCAL prototype consists of a 38-layer sandwich structure of steel plates and highly-
segmented scintillator tiles that are read out by wavelength-shifting fibers coupled to
SiPMs. The signal is amplified and shaped with a custom-designed ASIC. A calibra-
tion/monitoring system based on LED light was developed to monitor the SiPM gain and
to measure the full SiPM response curve in order to correct for non-linearity. Ultimately,
the physics goals are the study of hadron shower shapes and testing the concept of par-
ticle flow. The technical goal consists of measuring the performance and reliability of
7608 SiPMs. The AHCAL was commissioned in test beams at DESY and CERN. The
entire prototype was completed in 2007 and recorded hadron showers, electron showers
and muons at different energies and incident angles in test beams at CERN and Fermilab.
∗Now at CERN, Geneva
†Now at University of Science and Technology of China, Hefei
‡now at University of Heidelberg
§Now at University of Sussex, Physics and Astronomy Department, Brighton, Sussex, BN1 9QH, UK
¶Now at University of Sussex, Physics and Astronomy Department, Brighton, Sussex, BN1 9QH, UK
?now at University Göttingen

Page 6

Contents
1.Introduction2
2.Goals and Design Considerations3
3. Detector Layout
3.1Absorber Plates
3.2Active Layers
3.3The Scintillator SiPM System
3.3.1SiPM and Their Performance
3.3.2Manufacturing of the Scintillator Tiles
3.3.3Tile Assembly and Quality Control
3.4Cassette Design
3.5 Cassette Assembly
4
5
5
5
8
10
11
12
13
4.The Readout System
4.1The Very-Front-End ASICs
4.1.1Slow and Fast Shaping, Bias Adjustment and Input Coupling of
the SiPM
4.1.2Linearity, Gain and Noise Performance
4.2The Very-Front-End Readout Boards
14
16
16
17
18
5.Off-Detector Readout Electronics 20
6.Online Software and Data Processing21
7. The Calibration and Monitoring System
7.1The Light Distribution System
7.2Calibration and Monitoring Electronics
22
22
23
8. Calibration Procedure24
9.Commissioning and Initial Performance
9.1Test of Single Modules
9.2Test Beam Setup at CERN
9.3MIP Calibration
9.4Noise and Occupancy
9.5Performance of the SiPMs in the Test Beam
9.6Event Displays
26
26
27
28
30
32
33
– 1 –

Page 7

10. Conclusion33
11. Acknowledgments34
1. Introduction
The physics of the International Linear Collider (ILC) demands jet energy resolution of
σE/E = 30%/√E in order to efficiently separate Z0, W±and Higgs bosons via recon-
struction of their dijet invariant mass [1]. Boson separation depends on the success of the
particle flow concept [2]. The basic idea consists of reconstructing jets by separating the
contributions of charged particles, photon showers and neutral hadron showers and mea-
suring each of these components using the best suited detector subsystems. A typical jet
consists of 60% charged particles, 30% photons and 10% neutral hadrons. For example,
charged hadrons and muons are reconstructed in the tracking system. Since their mo-
menta are reconstructed with excellent resolution even for high-momentum particles, their
contribution to the jet-energy resolution is negligible. Photons are reconstructed in the
electromagnetic calorimeter with an energy resolution of typically 15%/√E, while neu-
tral hadrons are reconstructed in the hadron calorimeter with a typical energy resolution of
> 50%/√E.
In order to reconstruct accurately neutral hadron showers, energy deposits in the
calorimeter from charged hadrons, photons, electrons and muons must be reconstructed
and their energy removed from consideration. Due to overlapping showers the assignment
of a cell in the calorimeter to a particular shower is ambiguous. This effect is accounted
for by a ‘confusion’ term in the jet-energy resolution, which becomes the dominant term
as particle separation decreases. Thus, the success of this technique relies on the accuracy
with which individual cells in the hadron calorimeter are assigned to the correct shower.
For successful associations both the electromagnetic and hadronic calorimeters must have
high granularity in the longitudinal and transverse directions.
We present herein a description of the design, construction, and commissioning of the
1 m3CALICE1analog hadron calorimeter (AHCAL) prototype that consists of a steel-
scintillator sandwich structure. The scintillator planes are segmented into square tiles that
are individually read out by multipixel Geiger-mode-operated avalanche photodiodes, here
called SiPMs [4–6]. The main technical goal is to test the performance and reliability of
the SiPM readout on a large scale, since this detector uses thousands of SiPMs (7608)
for photon readout in a test beam. Ultimately, our physics goals are the study of hadron
shower shapes, reproduction of observed shower shapes in simulations, and first studies of
particle flow (PFLOW) algorithms.
1CALICE stands for Calorimeter for the Linear Collider Experiment [3].
– 2 –

Page 8

First tests of the AHCAL prototype were accomplished with test beams at CERN in
2006 and 2007. With a Si-W electromagnetic calorimeter (ECAL) prototype [7] in front
and a tail catcher [8] behind the AHCAL, we have measured energy deposits of muons,
electrons and pions. In addition, some data were taken with the AHCAL alone. Electron
energies were varied between 10 GeV and 60 GeV, and pion energies between 2 GeV
and 120 GeV. We also varied the incident beam angle from normal incidence (90◦) to
60◦in steps of 10◦. Further studies with a focus on lower energy data points continued at
Fermilab in 2008 and 2009. Here, we also tested an electromagnetic calorimeter with W-
scintillator-strips.
This article is organized as follows. In Section Two we discuss goals and design
considerations. In Section Three we present the detector layout and in Section Four we
give details on the readout system. In Sections Five and Six we respectively give a short
summary of the detector electronics and online software and data processing. In Sections
Seven and Eight we discuss our calibration/monitoring system and our calibration method.
We present commissioning and initial performance in Section Nine before concluding in
Section Ten.
2. Goals and Design Considerations
The design of the AHCAL prototype described here is inspired by the calorimeter layout
of the Large Detector concept (LDC) [9], which has evolved to the International Large
Detector Concept (ILD) [10]. The AHCAL is constructed as a sampling calorimeter using
a material of low magnetic permeability (µ < 1.01) as absorbers and scintillator plates,
subdivided into tiles, as the active medium. The millimeter-size SiPM devices that are
mounted on each tile allow operation in high magnetic fields [11]. The scintillation light is
collected from the tile via wavelength-shifting (WLS) fibers embedded in a groove. This
concept is different from existing tile calorimeters that have long fiber readout. Ultimately,
this allows us to integrate both the photosensors and the front-end electronics into the
detector volume. The high granularity imposed by particle flow methods is realizable with
scintillators in a very compact way at reasonable cost. The baseline absorber material
is stainless steel for reasons of cost and mechanical rigidity. However, since we do not
anticipate operating the AHCAL prototype in magnetic fields in test beams, we have used
absorber plates manufactured from standard steel (see Section 3.1).
We have two main goals for the physics prototype. First, we want to test the novel
SiPM readout technology on a large scale, identify critical operational issues, develop
quality control procedures and establish reliable calibration concepts that include test
bench data. We operate here nearly two orders of magnitude more SiPMs than in our
first test calorimeter [12]. Second, we want to accumulate very large data samples of
hadronic showers in several test beams. These samples are needed to investigate hadronic
shower shapes and to test simulation models, since it is not possible to extract this infor-
– 3 –

Page 9

mation from existing calorimeter data. The test beam data samples are also very useful for
studying and tuning particle flow reconstruction algorithms with real events.
Transverse dimensions of 1 m×1 m and a depth of 1 m represent an adequate choice
considering performance and cost. This guarantees that the core of hadron showers with
energies up to several tens of GeV (the range being most relevant for the ILC) is laterally
contained. The depth is comparable to that of a realistic ILC AHCAL which has to fit
inside the diameter of the solenoid. The longitudinal segmentation should be of the order
of one X0and the transverse dimension of the order of a Molière radius to resolve the
electromagnetic substructure in the shower. Detailed simulation studies of overlapping
hadronic showers in ILC events have shown that a transverse tile dimension of 3 cm×3 cm
provides optimal two-particle separation [13] and the anticipated jet energy resolution [2].
The total thickness of the AHCAL prototype is 5.3 nuclear interaction lengths (λn) or 4.3
pion interaction lengths (λπ).
The design is largely based on established technologies and contains a high degree of
redundancy to ensure stable and reliable operation of the AHCAL prototype over several
years while testing the novel SiPM readout. Since this layout is not a section of a full
detector for the ILC, it is not scalable and many of its external components still need to be
integrated into the ILC detector volume.
In test beam operation, the hadron calorimeter prototype is augmented with a tail
catcher and muon tracker (TCMT) system [8] to record leakage out of the rear of the
AHCAL prototype which becomes important particularly for high-energy hadrons. While
for a 10 GeV pion the energy leakage is 3%, it reaches 8% for a 80 GeV pion [15]. The
TCMT is a sandwich structure of steel absorber plates and scintillator strips read out by
SiPMs connected to the same electronics as the AHCAL prototype. The first eight layers
of the TCMT (1.1λn) have the same longitudinal sampling as the AHCAL prototype, while
thenexteightlayershaveacoarsesamplingcorrespondingto4λn. Inaddition, theendplate
of the AHCAL (0.12λn) contributes. The ECAL prototype [7] in front of the AHCAL
prototypehasadepthof0.9λn. Theoverallcalorimeterdepthis0.88λn+5.26λn+0.1λn+
5.76λn= 12.0λn.
Since particles in a colliding-beam detector are produced over a large range of po-
lar angles and charged-particle tracks are curved in the strong magnetic field of the ILC
detector, the typical angle of incidence at the calorimeter front face will differ from 90◦.
Thus, we have built a mechanical support structure that accommodates incidence angles
up to 55◦without turning the support structure itself. The detector layout is modular such
that active layers are exchangeable in order to test different readout technologies within
the same absorber structure.
3. Detector Layout
The AHCAL prototype consists of a sandwich structure of 38 absorber plates, 38 active
layers containing 7608 scintillator cells and an endplate. A schematic layout is displayed
– 4 –

Page 10

in Figure 1. All scintillator tiles in a layer are housed inside a rigid cassette, i.e. a closed
box with steel sheet top and bottom covers. A cassette with the calibration/monitoring
board on one side and the readout electronics on the other side is called a module. Fig-
ure 2 displays the schematic layout of a module and a photograph of the scintillator tiles
in a cassette for layers 1–30. The segmentation is 3 cm×3 cm in the core and coarser
elsewhere (see Section 3.3). Figure 3 shows a schematic cross section of a cassette and Ta-
ble 1summarizes dimensions, X0, λπand λnof the individual components. A VME-based
system for digitization and data acquisition is placed in a separate crate. The AHCAL
prototype is placed on a movable stage that allows us to move the prototype up/down and
left/right as well as to rotate it. In a rotated configuration, the individual layers have to be
realigned to ensure that the beam still traverses through the center of each layer.
3.1 Absorber Plates
The steel plates are 1 m×1 m wide and on average 17.4 mm thick. We use standard S235
steel that is a composite of iron, carbon, manganese, phosphorus and sulphur. Table 1
summarizes the characteristic parameters of each layer2while Table 2 shows the number
of λπ, λnand X0for the entire AHCAL. The magnetic properties are irrelevant, because we
do not plan any measurements inside a magnetic field. The absorber plates are mounted
on support bars with four bolts. The gap width between plates is adjustable. For a perpen-
dicular beam direction, the gaps are 1.4 cm wide including a 2 mm tolerance to account
for the aplanarity of the steel plates and to allow a smooth insertion of the cassettes. The
cover sheets, each 2mm thick, together with the absorber plates yield on average a total
absorber thickness of 21.4 mm per layer.
3.2 Active Layers
Active layers one through 30 contain 216 scintillator tiles, while the last eight layers con-
tain 141 scintillator tiles. Table 3 summarizes the details of the layout of the AHCAL
layers. Housing all scintillator tiles of a layer in cassettes allows us to independently test
modules in electron beams, where typically four of them were stacked together without
additional absorber plates. These tests are important for obtaining a first set of calibration
constants. The modular design allows us to exchange individual modules easily in case of
problems.
3.3 The Scintillator SiPM System
Figure 4 shows the fiber-SiPM readout of the three different size tiles. A rectangular-
shaped groove with a cross section of 1/25” is milled into each scintillator tile using
2The steel plates are covered with a thin layer of Zn to prevent rusting. The thickness is typically 100 µm
on each side which occasionally may increase to 250µm. In the calculation of the properties we have
assumed steel for the entire thickness. Including 100 µm thick Zn coating on each side decreases λπby
0.0047 pion interaction lengths and increases X0by 0.0037 radiation lengths.
– 5 –

Page 11

Figure 1. Schematic layout of the AHCAL prototype placed on a moving stage. The steel plates
mounted on rods are shown rotated with respect to the beam that enters from the right hand side.
The rack in the front houses the supply voltages, trigger electronics, and data acquisition system.
HV
HV
data
data
VFE electronicsCassette
temperature sensors
UV LED
PIN diode
CMB
CAN?BUS
ASIC chip
AHCAL MODULE
Figure 2. Schematic tile layout of a scintillator module for layers 1–30 (left) and a photograph of
tiles in a module (right). The red dots indicate the position of the thermosensors.
a computer-controlled milling machine. A Kuraray Y11 WLS fiber is inserted into the
groove that collects the scintillation light. Using double cladding and coupling the fiber
via an air gap to the tile maintains the total-reflection properties of the fiber. One fiber end
is pressed against a 3M reflector foil, while the other end is coupled via an air gap to the
– 6 –

Page 12

material
ρλπ
[cm]
20.4
21.5
52.0
70.9
64.0
20.5
107.2
59.1
106.8
52.0
1134
56.6
71.4
107.2
98.9
107.2
93.7
101k
λπ/ρ
[g/cm2]
160.8
160.2
117.8
141.7
140.7
160.8
113.7
137.7
121.9
117.8
80.3
175.5
121.4
113.7
128.5
113.7
126.5
122
λn
[cm]
16.8
17.7
37.9
56.2
50.6
16.8
77.1
46.5
79.0
37.9
734.6
47.5
52.6
77.1
74.6
77.1
70.2
74.8k
λn/ρ
[g/cm2]
132.1
131.4
85.8
112.4
111.4
132.1
81.7
108.4
90.2
85.8
52.0
147.2
89.45
81.7
97.0
81.7
94.8
90.1
X0
[cm]
1.76
1.97
18.9
9.75
9.64
1.76
41.3
9.37
30.01
18.9
890.4
3.68
17.5
41.3
19.6
41.3
19.9
30.4k
X0/ρ
[g/cm2]
13.8
14.6
42.7
19.5
21.2
13.9
43.8
40.2
34.2
42.7
63.0
11.4
29.8
43.8
25.5
43.8
26.9
36.6
RM
[cm]
1.72
1.85
4.89
5.77
5.39
1.72
9.41
4.94
9.52
4.89
67.92
4.52
6.06
9.41
8.34
9.41
7.95
7.3k
f
[g/cm3]
7.87
7.44
2.27
2.0
2.2
7.86
1.06
2.33
[%]
98.34
1.4
0.17
0.045
0.045
100
100
18.1
40.6
27.8
6.8
6.7
100
100
87.2
11.9
100
0.9
Fe
Mn
C
S
P
steel
tile
Si
O
C
H
Br
FR4
3M foil
PVC
polystyr.
Cable
air
2.27
3.1
1.7
1.06
1.3
1.06
1.35
Table 1. Composition and properties of the absorber (cassette) plates and materials used in a
cassette of one AHCAL layer [14], where ρ, RMand f respectively denote the density, Molière
radius and fraction of components in composite materials (steel, PCB boards and cables) while
other quantities are defined in the text.
PCB
Scintillator
Steel cassette
Steel cassette
Air
Air
(steel)
Absorber
Cable?fibre mix
3M foil
z
y
Figure 3. Schematic cross section of a cassette (not to scale).
– 7 –

Page 13

material
Steel plate
Cassette plates
Scintillator tile
FR4
3M foil
Air gaps
Cable mix
AHCAL
#λπ
3.237
0.743
0.177
0.053
0.008
#λn
3.941
0.905
0.247
0.072
0.011
#X0
37.555
8.624
0.460
0.217
0.021
t [cm]
66.19
15.2
19.0
3.8
0.9
9.5
5.7
120.26
tlayer[cm]
1.74
2×0.2
0.5
0.1
0.023
2×0.125
0.15
3.163
0.061
4.28
0.081
5.26
0.286
47.16
Table 2. Number of pion interaction lengths, nuclear interaction lengths, radiation lengths, and
thickness in the 38 layers of the AHCAL. The last column shows the average thickness of an
individual layer. The exact thickness per layer varies from 3.093 cm to 3.183 cm due to variable
sizes of the steel plates.
granularity# layers # tiles
3 cm×3 cm
100
# tiles
6 cm×6 cm
96
121
# tilestiles/
layer
216
141
total
# tiles
6480
1128
7608
12 cm×12 cm
20
20
fine
coarse
30
8
38
Table 3. AHCAL total amount of readout channels.
SiPM. In the 30 cm×30 cm core region the tile sizes are 3 cm×3 cm and the groove has
a quarter-circle shape. A full circle is not achievable, since the bending radius becomes
too small. The quarter-circle shape yields a higher light collection efficiency than a simple
diagonal readout [16]. The layer core is surrounded by three rings containing 6 cm×6 cm
tiles that have circular grooves, while the outer ring consists of 12 cm×12 cm tiles that
also have circular grooves into which the Y11 fibers are inserted. The varying tile sizes
represent a balance between shower sampling and cost. All scintillator tiles have a thick-
ness of 5 mm, which is thick enough to ensure a sufficient signal-to-noise separation for
MIPs.
3.3.1 SiPM and Their Performance
A SiPM is a multipixel silicon photodiode operated in the Geiger mode [4–6]. The photo-
sensitive area is 1.1 mm×1.1 mm containing 1156 pixels, each 32 µm×32 µm in size.
SiPMs are operated with a reverse bias voltage of ∼ 50 V, which lies a few volts above
the breakdown voltage, resulting in a gain of ∼ 106. Once a pixel is fired it produces a
Geiger discharge. The analog information is obtained by summing the signals from all
pixels. Thus, the dynamic range is limited by the total number of pixels. Each pixel has
– 8 –

Page 14

Figure 4. Readout of 3 cm×3 cm (left),6 cm×6 cm (middle), and 12 cm×12 cm tiles (right)
with WLS fibers and SiPMs.
a quenching resistor of the order of a few MΩ built in, which is necessary to break off
the Geiger discharge. Photons from a Geiger discharge in one pixel may fire neighboring
pixels yielding inter-pixel cross talk. For stable operations we selected detectors with an
inter-pixel cross talk of less than 35% and with moderate dark current caused by pile-up
from thermal noise-induced signals. The pixel recovery time is of the order of 100 ns.
However, recovery times as low as 20 ns are achievable by reducing the resistance. For
short recovery times, the light pulse of a tile is sufficiently long that a pixel may fire a
second time. This is a disadvantage as the SiPM saturation depends on the signal shape.
Thus, we use relatively high quenching resistors. Furthermore, the sensors are unaffected
by magnetic fields as tests in magnetic fields up to 4 T confirm [11].
More than 10,000 SiPMs have been produced by the MEPhI/PULSAR group and
have been tested at ITEP. The tests are performed in an automated setup, where 15 SiPMs
are simultaneously illuminated with calibrated light from a bundle of Kuraray Y11 WLS
fibers excited by a UV LED. During the first 48 hours, the SiPMs are operated at a bias
voltage that is about 2 V above the normal operation voltage. This procedure allows us to
reject detectors with unstable currents caused by long discharge. Next, the gain, noise and
relative efficiency with respect to a reference photomultiplier are measured as a function
of the reverse-bias voltage. The reverse-bias voltage working point is chosen such that
a signal from minimum-ionizing particle (MIP), provided by the calibrated LED light,
yields 15 pixels in order to ascertain a large dynamic range and to have the MIP signal
well separated from the pedestal.
At the working point, we measure several SiPM characteristics. With low-light inten-
sities of the LED, we record pulse height spectra that are used for the gain calibration. A
typicalpulseheightspectrumisshowninFigure5(leftplot), inwhichuptonineindividual
peaks corresponding to different numbers of fired pixels are clearly visible.
This excellent resolution is extremely important for calorimetric applications, since it
provides self-calibration and monitoring of each channel. We record the response function
– 9 –

Page 15

?
?????????????????????
???
???
?
? ?????? ???
Light intensity [MIP]
# pixel
0
500
1000
1500
050100150 200250
Figure 5. A typical SiPM spectrum for low-intensity light (left) showing the pedestal in the first
peak and up to eight fired pixels in the successive peaks. The response function for SiPMs of fired
pixels versus input light (right). The curves are taken as a set of twenty measurements at increasing
light intensities and can be fit with a sum of two exponential functions.
of each SiPM over the entire dynamic range (zero to saturation). Figure 5 (right plot)
shows the number of pixels fired versus the light intensity in units of MIPs for different
SiPMs. TheshapeoftheresponsefunctionofallSiPMsissimilarandindividualcurvesare
generally within 15% of one another. In addition, we measure the noise rate at a threshold
of 0.5 MIPs, the inter-pixel cross talk, and the SiPM current as shown in Figure 6. Arrows
in the figures indicate the requirements for our detector selection.
The relative variation of SiPM parameters for a 0.1 V change of the bias voltage is
also measured and is shown in Figure 7. Thus, a voltage change of 0.1 V modifies most
SiPM parameters by 2–3%. The largest effects of about 5% are found for the cross talk
and the SiPM response.
We use SiPMs with noise rates of less than 3 kHz at half a MIP threshold.3Additional
requirements on other parameters reduce the yield only slightly. The main noise sources
are the SiPM dark current caused by pile-up from thermal noise-induced signals and cross
talk. Both sources depend on temperature (T), fluctuations in the bias voltage (∆V) and the
readoutelectronics. Adecreaseintemperatureby2◦Cleadstoadecreaseofthebreakdown
voltageby0.1V, whichisequivalenttoanincreaseofthebiasvoltagebythesameamount.
In addition, a decrease in temperature leads to a decrease of SiPM dark rate.
3.3.2 Manufacturing of the Scintillator Tiles
The scintillator tiles have been produced by the UNIPLAST plant in Vladimir, Russia.
3Here, a threshold of 0.5 MIPs corresponds approximately to 7.5 fired pixels.
– 10 –