ISSN 1063-7796, Physics of Particles and Nuclei, 2008, Vol. 39, No. 3, pp. 410–423. © Pleiades Publishing, Ltd., 2008. Original Russian Text © A.M. Artikov, O.E. Pukhov, G.A. Chlachidze, D. Chokheli, 2008, published in Fizika Elementarnykh Chastits i Atomnogo Yadra, 2008, Vol. 39, No. 3.
Scintillation Counters of the Muon System at CDF II A. M. Artikova, b, O. E. Pukhova, G. A. Chlachidzea, c, and D. Chokhelia, d
d
a Joint Institute for Nuclear Research, Dubna, Moscow region, 141980 Russia b Nuclear Physics Laboratory, Navoi State University, Samarkand, Uzbekistan c Fermi National Accelerator Laboratory, PO Box 500, Batavia, Il 60510, USA
Institute for High-Energy-Particle Investigations, Tbilisi State University, Georgia Abstract—The framework of scintillation counters for the CDF II muon system at the Tevatron collider at Fermilab is described. Information from the detectors of the muon system is essential for forming triggers of the first and second levels and for an “off-line” data analysis related to studies in the field of the heavy quark physics, Standard Model tests, search for phenomena beyond its limits, and for many other CDF II experiments with p p collisions at the energy s = 1.96 TeV. PACS numbers: 29.40.Mc DOI: 10.1134/S1063779608030040
1. INTRODUCTION The muon system at CDF II at Fermilab plays a central role in a wide program of investigating the physics of c (charm), b (bottom), and t (top) quarks, verifying the standard model (SM), searching for phenomena beyond its limits, and studying other processes of p p interactions at
s = 1.96 TeV at the Tevatron.
Scintillation counters and drift chambers of the muon system are used to form triggers of the first and second levels and in an “off-line” analysis for selecting and investigating a wide range of p p interactions where muon production is expected from experimental conditions. The scintillation counters allow the selection of the desired event among those detected by the drift chambers: the drift time (~1–1.5 μs) exceeds the period (396 ns) between intersections of Tevatron p p bunches, thus introducing uncertainty in the event timing [1].
2. THE SYSTEM OF MUON SCINTILLATION COUNTERS AT CDF II The muon system at CDF II comprises scintillation counters and drift chambers. After the muon system was upgraded (1996–2000), its acceptance increased by more than 60% (Fig. 1). The muon scintillation counters cover an area of pseudorapidity 0 < |η| < 1.5 and are grouped into the following main subsystems (Fig. 2): (i) an upgraded central muon scintillation detector (counters of the central scintillator upgrade (CSP) in the region 0 ≤ |η| ≤ 0.6); (ii) the extension of a central muon scintillation detector (counters of the central scintillator extension (CSX) in the region 0.6 ≤ |η| ≤ 1.0); and (iii) a forward muon scintillation subsystem (counters of the barrel scintillator upgrade (BSU) on toroids and counters of the toroid scintillator upgrade (TSU) inside the toroids in the region 1.0 ≤ |η| ≤ 1.5).
The framework of scintillation counters of the CDF II muon system, including high-voltage power supplies, read-out electronics, and the system of monitoring the parameters of counters and electronics, is described in the review. The necessary stability of scintillation counter parameters during long-term data taking is maintained by a special computer program integrated into the CDF II overall monitoring system of operational control and automated supervision. The penultimate section illustrates the principal or essential role played by an efficiently running muon system in achieving a number of important physical results. 410
CMX
–1
CMP
0
CMU
1
CMX
CMP
–1.5 –1.0
0
CMU
IMU
1.0 1.5
φ
η
φ
η
Fig. 1. Acceptance of the CDF muon system during (left) Run I and (right) Run II.
SCINTILLATION COUNTERS OF THE MUON SYSTEM AT CDF II North
CSP Top
CSX Internal
East
BSU Rear
CSP Wall CSX External BSU Forward BSU Rear
y
411
p
θ
φ
CSX External
z
x p
CSP Wall CSP Bottom
West TSU
South
CSX Miniskirt
Fig. 2. Layout of scintillation counters at the upgraded CDF II.
H 40 Mirror
T
Number of wavelength-shifting fibers 1 mm
L Ribbon of wavelength-shifting fibers
Adapter
150 50
Type CSP L1
L 2400
H 305
T 20
N 20
CSP L2
3100
305
20
20
CSP L3
3200
305
20
20
BSU
1638
166
15
15
Fig. 3. New generation of the CSP and BSU scintillation counters and their dimensions (in mm).
For light collection, Lucite light guides are used in so-called “old” counters, wavelength-shifting fibers in new-generation counters, and both of these types of light collection in “upgraded” counters. 2.1. A Group of CSP Counters in the Region 0 ≤ |η| ≤ 0.6 and Δϕ = 360° They are positioned (Fig. 2) directly after the drift chambers of the central muon upgrade (CMP) and at a distance of ≈1.2 m from the drift chambers of the central muon unit (CMU), comprising the muon identification system. Counters of the CSP and chambers of the CMP are installed behind a steel shield 61 cm in thickness, which almost completely suppresses the radiation background coming from the facility center. The CSP is equipped with 276 counters of two sorts. At first, Lucite light guides of the “fishtail” type were used to collect light from muon scintillation counters during RUN I. Because of aging, a portion of these counters was replaced with new-generation scintillation counters manufactured at JINR; their chief feature is that the light collection is performed by wavelength-shifting fiber ribbons (produced by Kuraray, Japan, and Pol. Hi. Tech., Italy) [2]. PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
The CSP Bottom and CSP Top, consisting of 144 such counters, are respectively positioned below and above the facility (Fig. 2). They are made from rectangular polystyrene scintillators of three typical sizes (Fig. 3); this was caused by the existence of supports in the lower part of CDF II and was aimed at covering the maximum available area. The CSP Wall consisting of 132 counters used earlier in RUN I was upgraded by adding the wavelengthshifting optical fibers to the plastic light guides used as a basic method. Owing to this, the amount of light collected by a photomultiplier tube (PMT) from the farthest part of a counter was increased more than twofold [3]. The upgraded counters are located on the north and south sides of CDF II (Fig. 2); they are made of a rectangular polyvinyltoluene scintillator (Fig. 4). 2.2. A Group of CSX Counters in the Region 0.6 ≤ |η| ≤ 1.0 and Δϕ = 360° The CSX counters are arranged as two layers around the plane of drift chambers comprising the central muon extension (CMX), and the CSX “miniskirt” (MSK) counters as one layer directly in front of the drift chambers if they are viewed from the facility cen2008
412
ARTIKOV et al.
320
Mirror
Number of wavelength-shifting fibers 1 mm 3200
Ribbon of wavelength-shifting fibers
20
Fig. 4. Upgraded scintillation counters of the CSP Wall and their dimensions (in mm).
RUN I
RUN II Additional shielding Toroid is pulled nearer to the center
Forward calorimeter
Old forward calorimeter New Plug calorimeter decreases the gap
Fig. 5. Improvement in the muon-detector protection from secondary particles in the pseudorapidity range 0.6 < |η| < 1.0.
CSX External CMX chambers
400 mm
Counters of the CSX External
300 mm
55.0° 1800 mm
300 mm
Counters of the CSX Internal 48.25° (design angle 48.542°) CSX Internal 41.25°
400 mm
1800 mm
Fig. 6. Scintillation counters of the CSX.
ter (Fig. 2). They all look like a truncated cone widening to the facility center. The counters are installed on the east and west sides of CDF relative to the plane η = 0. For RUN II, the passive protection of the muon-system detectors in this region was considerably improved by additional steel shields put on the toroid of an old CDF magnet (Fig. 5). A “gap” through which secondary particles emerging from the CDF II region where p p beams collide was also decreased by substituting modern lead-scintillation sandwich calorimeters for outdated gaseous calorimeters in the forward region. The CSX counters have not been upgraded since RUN I and, as shown in investigations [4, 5], can successfully be used during RUN II, covering the region 0.6 ≤ |η| ≤ 1.0 and Δϕ = 270° (the range of the azimuthal angle from ϕ1 = –45° to ϕ2 = 225°); they are
made from polyvinyltoluene scintillation plates of trapezoidal shape 25 mm in thickness (Fig. 6). So-called “internal” (Internal CSX) counters are positioned on the truncated cone, on the side surface nearer to the center of CDF. Light is collected from the wide face of the plate (the long base of the trapezium). The total number of counters in the Internal CSX is equal to 134. So-called “external” (External CSX) counters are located on the truncated cone, on its side surface which is more distant from the center of the CDF. Light is collected from the narrow face of the plate (the short base of the trapezium). The total number of counters in the External CSX is 136. The MSK counters also have a trapezoidal shape (Fig. 7) and cover the region 0.6 ≤ |η| < 1.0 and Δϕ = 90° (from ϕ1 = 225° to ϕ2 = 315°) [6].
PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
2008
SCINTILLATION COUNTERS OF THE MUON SYSTEM AT CDF II
413
MSK # 2–4
MSK # 2–4 MSK # 5–16
MSK # 5–16
Fig. 7. Layout of the MSK counters of the CDF II. Ribbon of wavelength-shifting fibers
c PMT
B
A Aluminum coating msk 5 msk 6 msk 7 msk 8 msk 9 msk 10 msk 11 msk 12 msk 13 msk 14 msk 15 msk 16
A
B
C
171.0 169.0 167.0 165.5 164.0 162.5 161.0 160.0 159.0 158.5 158.0 157.5
243.5 239.5 236.0 233.0 230.0 227.5 225.0 223.5 222.0 220.5 219.5 219.0
1783.3 1757.3 1733.9 1713.0 1694.4 1678.0 1663.8 1651.5 1641.2 1632.7 1626.0 1621.1
Mylar
152
C
H5783
B
A
A msk 2 msk 3 msk 4
B
C
179.0 275.5 1879.2 176.0 251.5 1844.0 173.5 248.0 1812.2
Fig. 8. Scintillation counters of MSK and their dimensions (in mm).
Their linear dimensions vary with the azimuthal angle ϕ, because the space for their installation at CDF II is limited. For the same reason, the MSK counters are arranged only in one layer on the internal (near to the center of CDF) side surface of the truncated cone. In the MSK counters, unlike the counters of the CSX, light is collected from the both ends of a polyvinyltoluene scintillation plate 15 mm in thickness (Fig. 8). The total number of the MSK counters of CDF II is 48. Additionally twelve counters (the first three counters on each end shown in Fig. 7 and in the right panel of Fig. 8) equipped with the combined light collection will be installed at CDF II after the preparation of corresponding drift chambers. 2.3. Counters in the Range 1.0 ≤ |η| ≤ 1.5 Counters of the so-called Intermediate Muon Upgrade (IMU) system were added when the CDF was upgraded for RUN II. They are located (symmetrically about the plane η = 0, Fig. 2) on the toroids of the CDF “warm” magnet currently not in use; the system consists of two types of muon scintillation counters, BSU and TSU, and covers the area 1.0 ≤ |η| ≤ 1.5. The former counters are located on drift chambers of the BMU, and the latter within the toroid (Fig. 2). The thickness of the PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
toroid wall amounts to 60 cm, and it almost entirely suppresses the radiation background coming from the region of colliding p p beams of CDF II. The total number of scintillation counters of both types in the IMU group is 551. The BSU counters were manufactured at JINR and are located in two rings on the outer side of the toroids, in the direction parallel to p p beams (Fig. 2); they cover the range Δϕ = 270° (from ϕ1 = –45° to ϕ2 = 225°) and, respectively, 1.0 < |η| < 1.25 (“forward,” BSUF) and 1.25 < |η| < 1.5 (“rear,” BSUR). The counters of BSU are entirely analogous to the CSP Top and Bottom counters of the new generation but have a polystyrene scintillator of smaller size (Fig. 3) [2]. The total number of BSU counters of the CDF is 407. TSU counters are annularly arranged in the toroids, in the direction perpendicular to the p p beams (Fig. 2), cover the area Δϕ = 360° and 1.25 < |η| < 1.5, and are made of trapezoidal polyvinyltoluene scintillation plates 10 mm in thickness (Fig. 9) [7]. Long and short counters are used in TSU because of the intricate profile of the inner toroid surface where the counters are mounted. The TSU counters were designed at Michigan State University (USA), are the scintillation detectors of the 2008
414
ARTIKOV et al. c Ribbon of wavelength-shifting fibersPMT
B
C
A PAD
T
30
A B C c T
Long
Short
285 315 550 605 10
285 330 650 705 10
Aluminum coating
Fig. 9. TSU scintillation counters and their dimensions (in mm).
new generation, and are equipped with wavelengthshifting fibers for light collection and a H5783 photosensor (produced by Hamamatsu Photonics, Japan) on the basis of a miniature PMT, R5600. The wavelengthshifting fibers in the TSU counters, unlike the JINR counters (CSP and BSU), are closely attached to the face of a scintillation plate without gluing; i.e., there is the possibility of a further increase in light collection by gluing the fibers to the plate. The total number of the TSU counters of the CDF is 144. 3. SYSTEM OF DATA ACQUISITION AND CONTROL The total number of the muon scintillation counters of the CDF II exceeds 1140; the operation of this system is maintained and supervised by a system of control and monitoring [8] (see Section 3.3). Two sorts of PMTs employed in the CDF II divide the muon scintillation counters into two types and chiefly determine the methods used in the control system. 3.1. Counters with a Classical Method of Readout, and Large PMTs These are the CSX, MSK, and CSP Wall counters implementing a light collection equipped with a Lucite light guide and an EMI 9814B PMT having a photocathode diameter of 51 mm (produced by Electron Tubes Inc., UK). Power is supplied by high-voltage power supplies1 via distribution units called “Pisa Boxes”.2 1A
Gamma Power Supply high-voltage source provides an output voltage up to 3000 V at a strength of current of 75 mA (produced by Gamma High Voltage Research Inc., USA). 2 Pisa Box (produced by Costruzioni Apparecchiature Elettroniche Nucleari (CAEN), Italy) can supply as many as 40 PMTs with high voltage and maximum current up to 2 mA per channel. The maximum range of varying the output voltage is in the limits of 15–20% of the input voltage and the accuracy of fixing the value is better than 0.1%. The unit is controlled by means of a serial port. The distribution units can be combined into a unified control network using its serial port.
During RUN I, high voltage at a PMT was adjusted and controlled with the help of a Pisa Box portable unit, but only one channel could be tuned at one time. For RUN II, the earlier (RUN I) scheme of the highvoltage supply and information readout was retained, and the system of adjusting and controlling the high voltage of the Pisa Box distribution unit was introduced instead of the former one. A special interface commutator, the so-called Pisa Driver, produced by CAEN in accordance with specifications for computer-aided measurements and control (CAMAC), provided the communication between the computer and the distribution unit by means of a small-computer system interface (SCSI) controller of the Jorway Model 73A type (Figs. 10 and 11). The new technique [8] reduced the time of adjusting the voltage to several seconds per channel, thus reducing the time of adjusting and controlling the high voltage for a group of old scintillation counters by some ten times when compared with the manual control. Currently, it takes less than one hour to adjust high voltage at all 450 counters of this type. A schematic diagram of the high-voltage power supply and reading information from a group of old counters is shown in Fig. 10. A signal from a counter is routed to a discriminator of the LeCroy 4413 or LeCroy 4416 type which has 16 separate channels with a threshold voltage of 15 mV. An output signal formed by the discriminator has a standard 30-ns form for devices with emitter-coupled logic (ECL) and is routed via a twisted-pair to the input of a time-to-digit converter (TDC [9]) and further to the input of the data acquisition system. Signals from overlapping pairs of CSX counters, as well as from both ends of MSK counters, after discriminators, are routed to so-called “meantimers,” the output signals of which are also routed to TDCs (Fig. 11). The meantimer circuit [4] is a special coincidence circuit for signals of two PMTs and triggered at the instant when p p bunches enter the CDF II. It produces a signal having a constant delay equal to the time needed for light to pass through the full length of the counter plate
PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
2008
Experimental hall
PISA BOX
GAMMA box
H. V.
CSX, CSW, or MSK counters Scintillator
EMI 9814 PMT
H. V.
LeCroy 4413 discriminator
Output signal
Counter
415
CAMMA PISA-driver
SCINTILLATION COUNTERS OF THE MUON SYSTEM AT CDF II
TDC
Fig. 10. Block diagram of the high voltage power supply and the data readout CSP Wall scintillation counters.
Experimental hall Pickup Coils
CSX Internal counter Scintillator
Delay unit
PMT
CSX External counter PMT
Meantimer
TDC
LeCroy 4413 discriminator
TDC
TTL-to-NIM converter Logic FAN-IN FAN-OUT
GAMMA Box
Scintillator PISA BOX
GAMMA PISA-driver
Fig. 11. Block diagram of the high voltage supply and the data readout for CSX counters and meantimers.
and independent of the place where a muon hits the plate, thus specifying the time instant after which all background events are suppressed. Meantimers were proposed in RUN I to cut off secondary background particles delayed by 12–15 ns with respect to the particles escaping from the muon production region (Fig. 5). In RUN II, meantimers will be used in a trigger with an increase in Tevatron luminosity to a value larger than 4 × 1032 cm–2 s–1. In total, 8 high voltage power supplies, 16 Pisa Box distribution units, 38 discriminators (LeCroy 4413 and LeCroy 4416), and 24 meantimers will be used. 3.2. Counters Using Wavelength-Shifting Fibers and Miniature-PMT Photosensors for Light Readout These are new-generation scintillation counters using, for light readout, wavelength-shifting fibers and an H5783 photosensor on the basis of the R5600 PMT; this category includes all the BSU, TSU, and CSP Top and Bottom counters. Both power supply and data readout for these counters essentially differ from those described in the previous section. A PMT amplifier and discriminator (PAD) is used for the control of an H5783 photosensor in the counter (Fig. 12) [10]. It can vary high voltage at PMTs, PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
amplify PMT signals, and discriminate among them in accordance with a regulated threshold. Each PAD is connected to a custom-design control and concentrator unit (CCU) [11].3 Each channel is connected to a CCU by means of an RJ-45 connector in a so-called registered-jack standard and cables of so-called category 5 (CAT5). One cable with its four twisted pairs can simultaneously be used not only to control and adjust voltage but also to read information from the channel, as well as to monitor performance of the entire line by sending signals to a blue light-emitting diode (LED) located on the far end of the counter. Information from a CCU, with the help of two flat 50-cables consisting of 25 twisted pairs, is routed to a TDC and, then, to the input of the data acquisition system. The total number of employed CCU receiver/distributors is 20; they are connected to as many as 700 scintillation counters. Eight independent communication (COM) ports are used for the connection of a computer to CCUs arranged in groups positioned in different regions of the CDF. 3A
CCU can supply voltage to 48 PADs and simultaneously receive information in the ECL standard from them for transmission of it to a TDC. 2008
416
ARTIKOV et al. Experimental hall
Scintillator
CCU
TDC
H5783 PAD
LED Counter
Fig. 12. Block diagram of high voltage power supply and data readout for new-generation scintillation counters.
3.3. System to Control Scintillation Counters and Its Embedding in the CDF II Overall Control System MuonMonitor, a computer program, was written using Microsoft Visual C and the Visual Basic program kit and installed for controlling and monitoring effectively the entire system of the muon scintillation counters. The MuonMonitor program can control the operation of all the CCU’s and Pisa Box units. Using a database describing the chief parameters of the scintillation counters (plateau voltage, discrimination-threshold voltage, position, particular number, etc.), a user of the MuonMonitor sets needed high voltages at PMTs and threshold levels for PADs. A further logical evolution of the system of controlling the muon scintillation counters was its integration with an overall control system of the CDF II. The iFIX software, a licensed program package produced by Intellution, USA, is used for monitoring continuously the parameters of the scintillation counters. A MuonMain program, along with its embedded subroutines and utilities, based on this pack and incorporated into the overall system of control, allows real-time control over the pre-assigned parameters of the muon scintillation counters (Fig. 13). High voltages of ~1200 PMTs and thresholds of about 700 PADs are checked each 15 min for a duration of 30 s. Besides these program indications, to warn of possible malfunctions, an operation-test system built into a CDF overall control system is used to check reference voltages of some equipment: CCUs, Gamma Boxes, Pisa Boxes, all CAMAC crates, and a base computer of monitoring and control. Information on the status of the muon scintillation system is displayed on the central control monitors of the CDF II. A warning or alert signal accompanied by a corresponding sound is produced in the case of system failures, which can be caused either by a significant drift of the preset parameters outside the specified range or by emergency shutdown of some equipment or a channel. In emergency, a shift operator should fix the problem or call experts. For on-line monitoring of stable operation of the scintillation counters, a shift operator also routinely
plots two-dimensional histograms showing, on the screen, the number of counters responding to globaltrigger signals.4 A change in the reference (stable) form of the histogram generally means a fault in a counter or its circuit. The long-term stability of the counters is judged from efficiency plots produced by using data collected over 2–3 months. Figure 14 illustrates the efficiency of the CSP Wall counters; it was obtained by comparing the number of counter hits with the number of tracks produced in corresponding CMP/CMU chambers. The program for control and monitoring of the muon scintillation counters is used in the cases described above (at the first stage) to find the causes of low efficiency or failure of a particular counter or a group of counters. It is needed to check the presence of high voltage, current through PMTs and their signals, and the performance of specific channels of the Pisa Boxes and electronics. In the second stage, when the Tevatron is shut down, it is possible to search for bad cables, PMTs, and the PAD or a failure in light isolation. In extraordinary cases, a counter is removed from the CDF for its parameters to be checked at a special test facility [12]. The above measures maintain the stable effective operation of the scintillation counters of the CDF II muon system during long-term sessions of data taking. 4. SCINTILLATION COUNTERS IN A MUON “ON-LINE” TRIGGER AND “OFF-LINE” ANALYSIS The scintillation counters take part in the formation of the muon “on-line” triggers. A small part of “online” triggers from a complete trigger set, which was mandatory at the CDF II until January 2006 for selecting the candidates for muon events, is illustrated in the table. As seen, scintillation counters of the CSX, BSU, and TSU are widely used in triggers of the first and sec4 The
global trigger [1] of the CDF generates a readout enabled signal to record events if coincidence signals of triggers of the first, second, and third levels form some tabulated codes. In this case, signals of all detectors are registered, irrespective of whether they have formed a global-trigger signal.
PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
2008
SCINTILLATION COUNTERS OF THE MUON SYSTEM AT CDF II CCU control chain
417
PisaBox control chain
Fig. 13. Chain of screen images of the MuonMain program [9].
ond levels with the requirement of the necessary presence of their signal proving the selection of an event with a muon. A group of the CSP scintillation counters does not take part in on-line triggers. They are situated in the area relatively protected against radiation, and their inclusion into an on-line trigger decreases the background by 10%. Since the effect is small, it was decided not to strain on-line triggers with the addition of the CSP counters. However, the use of the CSP counters is necessary in many cases when the accumulated data is analyzed, for example, in measuring the top quark mass. Detailed guidelines of forming the triggers of the first and second levels for all subsystems of the muon PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
scintillation counters are described below. Then, some experiments are briefly considered to illustrate the role of the scintillation counters in on-line triggers and offline analysis in the muon selection. 4.1. Muon Trigger for the IMU (1.0 ≤ |η| ≤ 1.5) The IMU trigger of the first level (L1) “geometrically” covers the forward (1.0 ≤ |η| ≤ 1.25) and rear (1.25 ≤ |η| ≤ 1.5) muon detectors comprising the IMU subsystem (Fig. 15). In the range 1.0 ≤ |η| ≤ 1.25, the muon trigger is formed by the coincidence of signals from the BSUF counters, the BMU drift chambers, the track-extrapola2008
418
ARTIKOV et al.
Events 779
0 5 10 15 20 25 30
North_E efficiency
1.0 0.8 0.6 0.4 0.2 0
0 5 10 15 20 25 30
Events 780 1.0 0.8 0.6 0.4 0.2 0
South_E efficiency
North_W efficiency South_W efficiency
Events 779 1.0 0.8 0.6 0.4 0.2 0
1.0 0.8 0.6 0.4 0.2 0
CSP, CSP wall Events 276 Average 0.9375 RMS 0.09452
CSP, CSP wall 100 90 80 70 60 50 40 30 20 10
0 5 10 15 20 25 30 Events 809
0
0.2
0.4
0.6
0.8
1.0
0 5 10 15 20 25 30
Fig. 14. Efficiencies of the CSP Wall counters (left) and the distribution of efficiencies of the CSP and CSP Wall counters (right).
Chambers of CMX Counters of CSX Internal
6
Upper notch steel η = 0.6
η = 1.28 Counters Counters of BSU rear of BSU front Chambers of BMU η = 1.5
Chambers of CMU
5 η = 1.0
Calorimeter of CHA
4
Counters SSU
3
Calorimeter of WHA
2
m
Counters of TSU
Counters of CSP
Chambers of CSP
Counters of CSX External
COT 1
Calorimeter of PHA
0 9
8
7
6
5
4
3
2
1
0
m Fig. 15. Cutaway view of the CDF II set-up.
tion system (XTRP), and the system of timing a hadron (Fig. 16).
for synchronization with muons originating in the collision area.6
After a track in the central outer tracker (COT) is identified by the extremely fast tracker (XFT), it is extrapolated to the BMU chambers by the XTRP (Fig. 15).5
In the region 1.25 ≤ |η| ≤ 1.5, the muon trigger is formed by the coincidence of signals from counters of BSUR and TSU, the BMU drift chambers, and the hadron timing system. Signals (the information from towers of the WHA and PHA hadron calorimeters with corresponding η, Fig. 15) of the hadron timing system
Signals (the information from towers of the WHA hadron calorimeter with corresponding η, Fig. 15) of the hadron timing system (HTS) are used in the trigger 5 Until January 2006, the XFT trigger for the area 1.0 ≤ |η| ≤ 1.2 con-
sisted of no less than three necessary coincidences in superlayers of the COT, and more than three coincidences for 0 ≤ |η| ≤ 1 [1].
6 The
hadron timing system gives information on the time interval between the instant when p p bunches enter the CDF II and the instant when a minimum-ionizing particle (MIP) traverses the hadron calorimeter (CHA, WHA, and PHA) [1].
PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
2008
SCINTILLATION COUNTERS OF THE MUON SYSTEM AT CDF II
419
Examples of “on-line” triggers (their full formulas are shown in the second column) when the scintillation detectors enumerated in the third column were used in the formation of the triggers of the first and second levels. The references in the fourth column correspond to experiments where the scintillation counters were used in the on-line and off-line event selection and which are described in Section 5 Used scintillation counters
Examples of experiments
1 L1_BMU10_ BSUR_TSUO_&_CLC L2_CJET15_L1_BMU10_ BSUR_TSUO L3_CENTRAL_JET_20 MUON_CENTRAL_JET20_L1_BMU10_BSUR MUON_BMU_1
BSUR and TSU 1.25 < |η| < 1.5
[20]
2 L1_BMU10_ BSU _PT11 L2_CJET15_L1_BMU10_ BSU _PT11 L3_CENTRAL_JET_20 MUON_CENTRAL_JET20_L1_BMU10_PT11 MUON_BMU_1 L2_RL2HZ_L1_BMU10_ BSU _PT11 L3_BMU9 MUON_BMU9_L1_BMU10_BSU_PT11 MUON_BMU_1
BSUF 1.0 < |η| < 1.25
[18], [20]
3 L1_CMU1.5_PT1.5_&_CMX1.5_PT2_ CSX L2_CMU1.5_PT1.5_&_CMX1.5_PT2_DPHI120_OPPQ L3_JPSI_CMUCMX JPSI_CMU1.5_CMX2 JPSIMUMU_1 L3_LOWMASS_CMUCMX_SUMPT RAREB_CMUCMX_SUMPT B_RARE_1 L3_RAREB_CMUCMX_LXY RAREB_CMUCMX_LXY B_RARE_1 L2_CMU6_PT4_&_CMX1.5_PT4_ CSX L3_DIMUON_CMU4CMX4 DIMUON_CMU4_CMX4 SUSY_DILEPTON_2
CSX and MSK 0.6 < |η| < 1.0
[21]
4 L1_CMX6_PT8_ CSX L2_AUTO_L1_CMX6_PT8_ CSX L3_CMX8_TRACK5_ISO TAU_CMX8_TRACK5_ISO TAU_LEPTON_1 L2_CMX6_PT15_JET10 L3_MUON_CMX18 MUON_CMX18 HIGH_PT_MUON_1
CSX and MSK 0.6 < |η| < 1.0
[16–20]
N
Full trigger formula
5 Being used in the “off-line” analysis, being prepared for use in the “on-line” CSP and CSP Wall trigger 0 < |η| < 0.6 PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
2008
[16–21]
420
ARTIKOV et al. Data flow for IMU Trigger
COT signals
XFT
XTRP Front IMU min
BMU Muon hits
|t2 – t0|, |t3 – t1|
Hadron Ca1 signals
Time window
BSU/TSU scintillator signals
Muon “2.5°” MatchBox
Rear IMU
Time window
Fig. 16. Block diagram of forming the IMU trigger of the L1 to select muons in the region 1.0 ≤ |η| ≤ 1.5 (necessary explanations are given in the text) [1, 13].
(HTS) are used in the trigger for synchronization with muons produced in the collision area. The logic of the IMU trigger L1 is built as follows [13]: (i) L1F = BSUF & BMU & XTPR & WHA (row 2 in the table) for the forward part, and (ii) L1R = BSUR & TSU & BMU &(WHA + PHA) (row 1 in the table) for the rear part. At an initial instant luminosity of 1.2 × 1032 cm–2 s–1, the level of the L1 trigger did not exceed 200 and 300 Hz, respectively, for the forward and rear parts of the IMU system (October 2005, [14]). Depending on the investigation task, additional requirements are imposed upon the trigger of the second level. For example, when processes under investigation are related to t quarks, the IMU trigger of the L2 is formed from the L1 trigger by the additional requirement for a hadron calorimeter to record jets with the energy ET > 15 GeV in the region |η| < 1.1 (condition CJET15 in rows 1 and 2 of the table). 4.2. Muon Trigger for the CMX (0.6 ≤ |η| ≤ 1.0) The L1 trigger for the CMX is formed by the coincidence of signals from the CMX chambers, the XTRP module, and a pair of the CSX counters or one of the MSK counters. After a track in the COT is identified by the XFT fast trigger, it is extrapolated to the CMX chambers by the XTRP (Fig. 15). The requirement for the coincidence of four COT superlayers (see Footnote 6 in the previous subsection) is sufficient to maintain the necessary trig-
ger level without coincidence with a signal from the hadron timing system. The L1 trigger of the CMX is expressed as follows: L1 = (CSXInt + CSXExt) & CMX & XTPR (row 4 in table). The level of the L1 trigger for the CMX did not exceed 100 Hz for the initial luminosity of 1.2 × 1032 cm–2 s–1 (October 2005, [14]). Different L2 triggers are used for the CMX subsystem in the CDF II. For example, for the investigation of many processes where it is required to select both muons with high transverse momentum and a jet in the central region of the CDF (conditions PT15 and JET10 in row 4 of the table), the L2 trigger for the CMX has the following logic: L2 = (CSXInt + CSXExt) & CMX & XTPR & (ET > 10 GeV at |ηseed | < 1.1). 4.3. Muon Trigger for CMP/CMU (0 ≤ |η| ≤ 0.6) A muon trigger confirming the detection of a track in the central region of the CDF is generated by a coincidence of signals from drift chambers of the CMP subsystem and signals from the CMU subsystem. Also required is coincidence with a signal from the XTRP module extrapolating a track found in the COT by the XFT trigger to the CMU/CMP chambers. For triggering in this region, coincidence with a signal from the hadron timing system is not also required (see the preceding subsection). The L1 trigger logic for the CMP/CMU system is expressed as follows: L1 = CMP & CMU & XTRP. Depending on the problem under investigation, additional requirements are included in a trigger of the second level. For example, for investigating processes with t quarks, the L2 trigger for CMP/CMU is formed from the L1 trigger by adding a requirement for jets with an energy of ET > 15 GeV in the range |η| < 1.1 to be detected in the hadron calorimeter: L2 = CMP & CMU & XTRP & (ET > 15 GeV at |ηseed | < 1.1). When the initial instant luminosity of Tevatron was 1.2 × 1032 cm–2 s–1 (October 2005, [14]), a desired level (<300 Hz) of the L1 muon trigger in the range 0 ≤ |η| ≤ 0.6 was attained owing to the use of two spaced drift chambers. The addition of CSP counters to the on-line trigger reduces the frequency by 10%. Therefore, as long as the Tevatron luminosity is not increased, CSP counters are not used in forming the on-line trigger. Counters of CSP are used in an off-line data analysis to find a muon in events selected by the L1 trigger of CMU/CMP. Presently (May–September 2007), CSP counters are being prepared for use in the on-line trigger if it will not reduce the allowed level of the L1 trigger (≤1 kHz)
PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
2008
SCINTILLATION COUNTERS OF THE MUON SYSTEM AT CDF II
Events/(10 GeV/c 2)
when the Tevatron luminosity reaches a value of 3 × 1032 cm–2 s–1 or higher. 5. EXAMPLES OF EXPERIMENTS IN WHICH SCINTILLATION COUNTERS OF THE CDF MUON TRIGGER HAVE BEEN USED IN EVENT SELECTION The references in the right column of the table point out experiments when the corresponding trigger was used. 5.1. Measurement of Top Quark Mass The increase in Tevatron luminosity, in CDF II acceptance, and in s raised yield of t quark events in the RUN II experimental session started in 2002, thus allowing investigation of the statistically sound physics of the t quark rather than a limited number of events. For p p collisions with the energy p p = 1.96 TeV, t quarks are created in pairs of tt , chiefly through qq annihilation (≈85%) or gluon–gluon fusion (≈15%) [1]. In the Standard Model, a dominant mode of t-quark decay is t Wb. The topologies of final states are determined by W decays: (i) hadronic decays for two kinds of the W boson: +
tt
–
W bW b
qqqqbb (jets, 44%);
(1)
(ii) a hadronic decay for one W boson and a leptonic decay for another: tt
+
–
W bW b
qqlνl bb (lepton + jets, 30%); (2)
(iii) leptonic decays of both W bosons: tt
+
–
W bW b
lν l lν l bb (dilepton + jets, 5%);(3)
where l = e, μ, or τ. The W τντ mode is not investigated because it is difficult to identify a τ lepton. A purely hadronic mode is most probable but characterized by a large background in a form of QCD jets. The top quark mass is most accurately measured in processes (2) and (3). All methods of measuring the mass of a top quark are based on a statistical comparison of experimental distributions with a simulated sample of tt and background events; the latter are determined from the standard model; any discrepancy between the masses measured in the different modes of the t quark decay may indicate the existence of new physical processes. In CDF before 2006, masses of a t quark were measured in modes (2) and (3). 5.1.1. Topology of “lepton + jets.” An electron or muon “candidate” with pT > 20 GeV/c in the region |η| < 1, which includes the CSP and CSX scintillation counters, and a candidate (from missing energy) for a neutrino with energy loss exceeding 20 GeV are PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
18 16 14 12 10 8 6 4 2 0
Total MC Prediction Signal Background Data: 63 events
30 25 20 15 10 5 0
421 Joint likelihood
166 170 174 178 182 168 172 176 180 184 M, GeV/c 2
80
120 160 200 240 280 100 140 180 220 260 300 Per-event mass at maximum likelihood, GeV/c2
Fig. 17. Top quark mass and (inset) the corresponding likelihood function. The experimental data (points) are compared with the expected distribution (histogram) obtained by simulating tt (Mtop = 172.5 GeV/c2) decays and background events [16].
selected. The trigger efficiency was 96% for electrons and 90% for muons, and had a weak dependence on pT . The selection requirement also included no fewer than four jets with transverse energy ET > 15 GeV (the fourth jet could have ET > 8 GeV). For an accumulated luminosity of 318 pb–1, 165 tt events were selected by this method; the procedures of evaluating the background are described in [15, 16]. For an accumulated luminosity of 318 pb–1, the mass of a top quark in mode (2) of “lepton + jets” was measured by two techniques. The first technique [16] uses the maximum likelihood method to find the top quark mass as a function of a main matrix element of products in tt decays of each event. The “joint” likelihood function determining the top quark mass is obtained by multiplying the likelihood functions for each event. The second technique, the so-called method of “templates,” uses the reconstructed effective mass reco reco m t and a subsequent comparison of the m t distribution with template distributions, which are obtained by simulating a sample of the top quark masses Mtop = (140–220) GeV/c2, corrected for background [17]. A distribution of reconstructed effective masses at the maximum of the likelihood function obtained by the first method is shown in Fig. 17. A sample of 63 events was used in measuring the top quark mass Mtop = +2.6
173.2 –2.4 (statistical) ± 3.2 (systematic) GeV/c2 [16]. The results of determining the top quark mass by the template method for a different number of b-tagged (associated with a b quark) jets are illustrated in Fig. 18. The histograms are seen to agree with the fitting func+3.7 tion in all cases. The obtained result is Mtop = 173.5 –3.6 (statistical) ± 1.3 (systematic) GeV/c2 [17]. 2008
422
ARTIKOV et al.
Events/(15 GeV/c 2) 8 6 4 2 0 5 4 3 2 1 0
2-tag: 16 events Data Signal + Bkgd Bkgd only
100
200
300
400
1-tag Loose: 25 events
100
200
300
400
16 12 8 4 0 8 6 4 2 0
1-tag Tight: 57 events
with the value of the top quark mass obtained in CDF II from the channel of “a lepton + jets.” 5.2. Other Experiments
100
200
300
400
0-tag: 40 events
100
200
300 400 2 mreco t , GeV/c
reco
Fig. 18. Distribution of masses m t from the template method; the curves show results of the combined fit with the use of normalized signals and a background [17].
5.2.1. Search for anomalous production of multilepton events in p p collisions at s = 1.96 TeV [19]. The events were recorded if a trigger detected one muon with pT > 18 GeV/c in the region |η| < 1 (table). Then, an off-line analysis selected muons with pT > 20 GeV/c, pT > 8 GeV/c, and pT > 5 GeV/c; dimuon events were rejected as background. Analyzed data corresponded to an accumulated luminosity of 346 pb–1. A search was carried out separately for such events as “3 leptons” and “≥4 leptons.” In both cases, a number of the events agree with expected backgrounds predicted by the SM. In the framework of the “supergravity with R-parity violation” model, the bounds of the 0 masses of the lightest neutralino M( χ˜ 1 ) > 110 GeV/c2 ±
5.1.2. Topology of “dilepton + jets.” This topology is characterized by two jets from b quarks and two leptons having large transverse momenta and large missing energy (two neutrinos) from decays of W. Dilepton data were selected with the help of inclusive triggers on electron with the transverse energy ET > 18 GeV in the central calorimeter, or on muon with the transverse momentum pT > 18 GeV/c in the region |η| < 1.1, which includes the scintillation counters of CSP, CSX, and BSUF. Electrons in the forward region of the calorimeter are required to have ET > 20 GeV. The events should be characterized by the missing transverse energy ET > 15 GeV. The number of candidates for the tt events selected with the help of the inclusive lepton trigger was 33. The procedure of reconstructing the top quark mass from the dilepton mode with the help of the template method [18] consists of the following: (i) a reconstruction of the top quark mass in each event with the use of additional assumptions about azimuthal angles of the neutrino to perform a kinematic reconstruction of the event; (ii) a calculation of templates for simulated signal and background events, and a parametrization of these templates to find the probability density function for the distribution of the top quark masses; and (iii) a maximum likelihood fit of a mass sample of the experimental events involving the probability density function to obtain a final value for the top quark mass. In the CDF II, the mass of the top quark determined from the dilepton channel amounted to 170.1 ± 6.0 (statistical) ± 4.1 (systematic) GeV/c2. This result agrees
and chargino M( χ˜ 1 ) > 203 GeV/c2 were obtained at a 95% C.L. 5.2.2. Measurements of the tt production crosssection in p p collisions at s = 1.96 TeV in the allhadronic decay mode [20]. The cross-section for the +3.3 formation of a tt pair, σ tt = 7.5 ± 2.1 (statistical ) –2.2 +0.5
(systematic ) –0.4 (luminosity) pb, measured from data of an accumulated luminosity of 311 pb–1, is in accordance with the predictions of the SM. The events were selected with the requirement of “≥6 jets” (the hadronic mode of tt -pair decay). In this case, all scintillation counters are used in an off-line analysis to reject the events with a detected muon (the leptonic mode of a tt -pair decay). 0
0
5.2.3. Measurement of the Λ b lifetime in Λ b J/yL0 in collisions at
s = 1.96 TeV [21]. The life-
0 Λb
0
time of a hadron in the inclusive decay Λ b 0 J/ψΛ was measured from data corresponding to an 0 accumulated luminosity of 1 fb–1. The value τ( Λ b ) = +0.083
1.593 –0.078 (statistical) ± 0.033 (systematic) ps was obtained with the use of fully reconstructed decays, having better accuracy than the published current world average. Muons were selected in the region |η| ≤ 0.6 using CSP counters in an off-line analysis with the requirement that pT > 1.4 GeV/c (table, row 5), and in the region 0.6 ≤ |η| ≤ 1 by a trigger (with the use of CSX counters) on two muons from J/ψ μμ with the requirement that pT > 2 GeV/c (table, row 3).
PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
2008
SCINTILLATION COUNTERS OF THE MUON SYSTEM AT CDF II
6. CONCLUSIONS The muon system of the upgraded CDF II numbers consists of over 1140 scintillation counters, more than half of which are new-generation counters with a ribbon of wavelength-shifted fibers for light collection. Data taking, control over parameters, operational control, and monitoring of the scintillation counters are performed by a complex of apparatuses and programs which maintain the necessary efficiency of the CDF II muon system in accumulating physical information during Run 2 at Tevatron at FNAL. Quite accurate measurements of the t quark and W 0
0
B s oscillations, boson masses, observation of B s etc. are among the results achieved using a muon trigger. ACKNOWLEDGMENTS We are grateful to D. Bellittini, Yu.A. Budagov, A.N. Sissakian, and G. Pauletta for support and G. Velev, V.V. Glagolev, A.A. Semenov, and I.E. Chirikov-Zorin for discussions and valuable advice in writing this article. REFERENCES 1. The CDF II Collaboration, “The CDF-II Detector Technical Design Report,” Fermilab-Pub-96/390-E (1996). 2. A. Artikov et al., “Design and Construction of New Central and Forward Muon Counters for CDF II,” Nucl. Instrum. Methods Phys. Res. A 538, 358–371 (2005). 3. S. Cabrera et al., “Making the Most of Aging Scintillator,” Nucl. Instrum. Methods Phys. Res. A 453, 245–248 (2000). 4. P. Giromini et al., “The Central Muon Extension Scintillators (CSX),” CDF Note 3989. 5. J. Fernandez et al., “Test of the Central Muon Extension Scintillators (CSX),” CDF Note 5006. 6. A. Artikov et al., “The "Miniskirt” Counter Array at CDF II, Part. Nucl. Lett. 114, 25–39 (2002). 7. J. N. Bellinger et al., “Intermediate Angle Muon Detectors for CDF II,” Nucl. Instrum. Methods Phys. Res. A (in press). 8. O. Pukhov et al., “Automatization of the Monitoring and Control of the Muon Scintillation Counters at CDF II,” Part. Nucl. Lett. 5, 72–81 (2002).
PHYSICS OF PARTICLES AND NUCLEI
Vol. 39
No. 3
423
9. E. James and M. Soderberg, “Operation of Michigan TDC boards at CDF,” CDF/DOC/ONLINE/CDFR/7164. 10. C. Bromberg, “Gain and Threshold Control of Scintillation Counters in the CDF Muon Upgrade for Run II,” Intern. J. Mod. Phys. A 16, 1143–1146 (2001). 11. C. Bromberg et al., “A System to Control the Hamamatsu H5783 PMT Module and Condition Signals for TDC Readout,” CDF Note 4990. 12. A. Artikov, G. Bellettini, J. Budagov, et al., “On the Aging of the CSP and CSX Counters,” CDF/PUB/MUON/PUBLIC/7033. 13. C. M. Ginsburg et al., “CDF Intermediate Muon Trigger,” CDF Note 7694 (2005). 14. CDF online webpage, “RUN Summary for run 205991,” http://www-cdfonline.fnal.gov/java/cdfdb/servlet/RUN_ NUMBER=205991. 15. D. Acosta et al. (CDF Collaboration), Phys. Rev. D: Part. Fields 71, 052003 (2005). 16. A. Abulencia et al. (CDF Collaboration), “Measurement of the Top Quark Mass with the Dynamical Likelihood Method Using Lepton Plus Jets Events with b-Tags in p p Collisions at s1/2 = 1.96 TeV,” Phys. Rev. D: Part. Fields 73, 092002 (2006). 17. A. Abulencia et al. (CDF Collaboration), “Top Quark Mass Measurement Using the Template Method in the Lepton + Jets Channel at CDF II,” Phys. Rev. D: Part. Fields 73, 032003 (2006). 18. A. Abulencia et al. (CDF Collaboration), “Measurement of the Top Quark Mass Using Template Method on Dilepton Events in Proton-Antiproton Collisions at s1/2 = 1.96 TeV,” Phys. Rev. D: Part. Fields 73, 112 006 (2006). 19. A. Abulencia et al. (CDF collaboration), “Search for Anomalous Production of Multi-Lepton Events in p p Collision at s = 1.96 TeV,” FERMILAB-PUB-06-482-E (Jan. 2007). 20. A. Abulencia et al. (CDF–Run II Collaboration), “Measurement of the tt Production Cross-Section in p p Collisions at s = 1.96 TeV in the All-Hadronic Decay Mode,” Phys. Rev. D: Part Fields 74, 072005 (2006). 21. A. Abulencia et al. (CDF collaboration), “Measurement 0
0
of the Λ b Lifetime in Λ b
J/ψΛ0 in Collisions at
s =1.96 TeV,” Phys. Rev. Lett. 98, 122 001 (2007).
2008