All three spectrometers of the A1 collaboration are equipped with similar detector systems.
The Cerenkov counters are part of the trigger system and are necessary to identify electrons and charged pions. In the momentum range of interest these particles are not distinguishable by the scintillation detectors of the trigger system, since they are both minimum ionizing particles. The Cerenkov counters are designed to give a signal if an electron passes the detector; a charged pion gives no signal. Therefore we can suppress electrons and identify charged pions by using the Cerenkov detector as a veto counter. On the other hand it is sometimes useful to have a good electron identification if there is a lot of background.
Geometry:
- Spectrometer A, C:
2 × 6 main mirrors in one plane
Winston-funnel mirrors in front of each photomultiplier (picture above) - Spectrometer B:
5 main mirrors overlapping each other in a stage arrangement
no Winston-funnel mirrors
Main mirrors:
- spherical geometry with 800 mm radius
- size: 400 mm × 500 mm
- substrate: lucite 2 mm thick
- front aluminisation with MgF2-coating
reflectivity at least 70% in the range of 250-650 nm
Winston-funnel mirrors:
- diameter at outlet (photomultiplier): 110 mm
- diameter at inlet: 92 mm
- length: 150 mm
- angle of acceptance: 36°
- front aluminisation
reflectivity at least 90% in the range of 250-650 nm
Radiator gas:
- Decafluorobutane, R 3-1-10, C4F10 (until 2002: R 114, C2F2Cl4)
- index of refraction: 1.0013 at 400 nm
- threshold momentum for pion: 2700 MeV/c; for electrons: 10 MeV
- radiator gas is inside a closed cycle; the detectors are filled by exchanging the volume at least five times; the air in the fluorocarbon/air mixture coming from the detectors is then separated by liquifying out the freon at -40 °C and dismissing the air via an overpressure valve
Photomultiplier:
- type: XP4500B (Philips)
- cathode window: UV-borumsilicon
transmission: 50% at 240 nm; >92% at 280 nm - useable diameter of cathode: 110 mm
- number of dynodes: 10
- gain: 107
The efficiency of the detectors is tuned to have at least eight photoelectrons reaching the first dynode of one of the photomultipliers per event. This gives an efficiency of at least 99.966%. We checked the efficiency with a three detector setup, using large scintillator paddles behind the Cerenkov counters. For spectrometer A, C we have an integral efficiency of 99.98%. For spectrometer B this value counts not for the full acceptance, since 2% of the momentum acceptance at the lower end are not "seen" by the Cerenkov detector.