The CMS Central Tracker
The tracking detector for the Compact Muon Solenoid (CMS) is one of the most ambitious LHC projects. The design must be robust and versatile to cope with the full range of accessible physics, while the search for new particles and the study of quark and gluon interactions at an unprecedented energy scale require exceptional track and vertex reconstruction.
The detector reconstructs isolated muons and electrons with efficiencies higher than 98% and 95% respectively. The high track densities inside jets of particles can be studied by measuring the non-isolated hadrons with an average efficiency higher than 90% and down to very low track momentum.
Long-lived particles produced in heavy quark reactions can be identified by precision measurements of their distance of approach to the interaction point and by reconstructing the decays. Typical identification efficiency of 60% can be achieved at the price of about 1% contamination. Finally, excellent momentum resolution will allow for detection of narrow resonances and effective background suppression.
Design constraints
The detector will be operated in a very complex environment. At the LHC design luminosity, 1000 soft tracks due to secondary charged particles will illuminate the detector for each bunch crossing (every 25 ns). In addition, a mass of signals will be generated by low-momentum particles forced into helical trajectories by the 4 T CMS solenoidal magnetic field. A "typical" event will generate more than 40 000 hits in the detector.
The need for efficient two-track resolution and high-precision momentum measurement has pointed the selection of detectors to high granularity and precision and fast response time.
Performance must also be unaffected by the severe radiation due to the flux of secondary particles detector and electronics will have to survive megarad radiation doses. Special radiation damage R&D studies have shown that detectors based on silicon crystal technology will have to be operated continuously at a temperature as low as 10 °C. This, for sensors dissipating many kilowatts of power over about 7 m3, is a serious technical challenge.
In addition, heavily ionizing nuclear fragments (created when high-energy particles hit the detector infrastructure) will pose serious constraints on the operation of gas detectors at the highest luminosity, and will require a careful choice of operating point. Extensive tests have been carried out. An additional constraint is to keep the detector material budget as low as possible, to optimize the combined performance of the tracker and the high-resolution CMS electromagnetic calorimeter.
The CMS tracker has an overall length of 6 m and a diameter of 2.5 m (figure 1). All detector elements are arranged in cylindrical coaxial layers in the barrel region, while the endcap region uses discs normal to the beam direction. The total sensor area amounts to about 300 m2 with a granularity of about 45 million readout channels.
Detector technologies
Three detector technologies are used, each best matched to the stringent resolution and robustness requirements in the regions handling high, medium and low particle fluxes Silicon Pixels, Silicon Microstrip and Micro Strip Gas Chambers (MSGC) respectively.
To achieve reliable track identification, the detector segmentation is such that typical channel occupancies are between 1 and 2% in the whole tracker. All three detectors are fast, thus limiting event pile-up to a single bunch crossing in the solid-state sensors and to little more than two in the gas detectors.
The Pixel system occupies the radial region between 7 cm and 20 cm from the interaction point. The 150 x 150 micron cells are connected via an indium bump to the readout chip (figure 2). Sensors are arranged tangentially to cylindrical surfaces so that the strong magnetic field drifts the charge cluster over neighbouring cells, yielding a position resolution of better than 15 microns.
Silicon-strip detectors occupy the intermediate region extending up to 60 cm radius. Thanks to an active sensor thickness of only 300 microns, the fine readout pitch (about 100 microns) and their radiation hardness, silicon-strip detectors are especially suited to the medium occupancy region. Position resolutions ranging between 15 and 35 microns will be obtained in the r-phi direction. The orthogonal coordinate will be measured in about 60% of the hits by sensors with strips at a small angle to the beam direction.
A prototype detector module is shown in figure 3. In the outer region, extending up to 120 cm, the lower expected count permits lower granularity. The MSGCs are a very performant answer to the specifications. Due to gas diffusion, hit resolutions as good as 40 microns can be achieved with a readout pitch of 200 microns. The strip length will vary between 12.5 cm and 25 cm to match the decrease of occupancy with radius. These chambers will be operated at room temperature, thus simplifying the system. For about 50% of the hits the orthogonal coordinate too will be measured using stereo detectors. A prototype detector module is shown in figure 4.
The electronics will provide analogue data readout in radiation-hard technology. Care will be taken to minimize power consumption and material. Electronic noise immunity and robustness have driven the system design. High-speed signal processing and identification of individual bunch crossings needs to be guaranteed. Main components are the front-end circuit, an analogue optical link, the front-end driver and the control system. The front-end chip amplifies detector signals, stores them in a pipeline and multiplexes, after some analogue signal processing, to the counting room using the optical link. The link technology relies on edge-emitting lasers operating at 1300 nm illuminating single-mode fibres.
The control system will distribute clock, trigger and commands via the digital communication links and a control chip serving each group of modules. Pulse heights are received by a photodiode amplifier on the front-end driver which digitizes and processes the signals and stores results in a local memory for higher level data acquisition.
With the
recent formal approval of the
technical design report by the
research board,
the
CMS tracker collaboration has passed one of its major milestones.
The collaboration represents a multinational team with members belonging to more than 40 institutes from 12 countries.
They are looking forward to the
construction of the
final prototypes and
the
testing of a pre-production series of detectors by the
end of this year.
Mass production of detectors is planned for next year.
Horst Breuker and Alessandra Caner,
CERN.