The NA48 detector


NA48 is a typical fixed target experiment, the beam line and the decay volume extending over ~200m. The decay volume is contained in a ~100 m long evacuated tank, being followed by the main detector, which is about 30 m long. The main components are a magnetic spectrometer to measure charged particles and an electromagnetic calorimeter based on liquid krypton for neutral decay products.

The detector components

The magnetic spectrometer consists of four drift chambers located in a tank filled with helium. Each drift chamber is made of four double layers of sense wires oriented along four directions, each one rotated by 45 degrees with respect to the previous one to avoid ambiguities in the reconstruction of tracks. A dipole magnet between chamber two and three; the resulting deflection of the charged particles allows to reconstruct their momentum.

The electromagnetic spectrometer (LKr) is used to detect neutral particles and to identify electrons and positrons. It is a quasi homogeneous liquid krypton ionization chamber with more than 13000 readout cells of 2x2cm2 cross section, which are formed by electrodes extending longitudinally from the front to the back. The operation principle is similar to that of a drift chamber: the charged particles originating from the electromagnetic shower ionize the krypton, and the resulting ions are collected at the electrodes. The collected charge is a measure of the total energy of the shower in each cell. The energy resolution of the calorimeter is 3.2%/sqrt(E) (with energy E in GeV), and the fine segmentation provides an excellent spatial resolution.

The liquid krypton calorimeter is followed by the hadronic calorimeter to measure the energy of hadrons which did not deposit all energy in the electromagnetic calorimeter before. As the energy resolution is significantly worse compared to that of the LKr, its information is used mainly for the trigger. In order to detect muons, three planes of plastic scintillators are located after hadronic calorimeter, each shielded by an 80cm thick iron wall to stop all particles except muons. For the time measurement, there are two hodoscopes: one for charged particles, located right before the LKr , the second one integrated inside of the LKr for neutral particles. Both hodoscopes detect the particles in scintillators with photomultiplier readout; the time resolution achieved is 300-400 ps. To detect particles escaping the decay volume before reaching the detector, scintillators arranged in rings are placed along the vacuum tank. Finally, the beam counter, a simple detector at the end of the beam line composed of strips of scintillator, measures the instantaneous intensity of the beam.

During the first phase of the experiment, which was dedicated to measure direct CP violation, an additional detector was used to decide whether particles seen in the main detector resulted from a KS or a KL decay. For that, the tagging detector needed to registrate (or tag) every proton before reaching the second target to produce the KS beam. In case of coincidence between the event time and the proton time, the decay was assigned to a KS. The tagging detector consisted of small and very thin pieces of scintillator with photomultiplier readout, arranged in a ladder-like structure, thus cutting the proton into slices to distribute the high proton rate on several counters. The inefficiency of the small sophisticated detector was only a few 10-4 at rates up to 35 MHz.

Trigger and Data Acquisition System

The data acquisition of the experiment is built as a three-level trigger system. The first level is fully realized in hardware, processing fast information from the drift chambers, the calorimeters and the scintillator detectors.

The second level is partly implemented in software. For decays into charged particles, dedicated fast PCs compute the track momenta and the invariant masses. For neutral decays, the momenta of the energy distribution in the electromagnetic calorimeter are computed. Based on these informations, the level 2 trigger decides whether to accept an event or not, the rate of positive decisions reaching up to 10kHz.  If an event is accepted, the corresponding informations from all detectors are collected on a special PC farm and built to a global event. Finally, the data are sent to the CERN computing centre.

The third level is a pure software trigger running offline on a big PC farm at the CERN computing centre. It uses all informations, including calibration constants etc. to reconstruct and categorize the events, compute physical values and transform the data into a suitable format. Originally designed to reduce the data volume by rejecting events which are of no interest for the coming analyses, this feature was used only during a few run periods.

Contributions by the Mainz working group

The Mainz working group was responsible for developing, setting up and maintaining the following detector components:
  • Hadronic calorimeter and beam counter
  • Tagging detector
  • Online PC farm for event building
  • Third trigger level (Level 3)