Detector Requirements

Some detector features were discussed at Bloomington Workshop July 14-16, 1997 and were summarized in A. Dzierba letter to participants. It was agreed, that detector should be designed employing state-of-art detector technology. I would suggest to consider 4$\pi$ detector with equivalent systems to be able to perform complete analysis for all particles in backward and forward directions (momentum, energy, angular resolution, particle ID ...). As it is seen from above, CP,CPT study gives good example about quality of the detector. It should include:

- good tracking in all directions with momentum measurement to reconstruct kaon mass and predict the direction and momentum of another kaon, decaying into photons.

- good segmented 4$\pi$ calorimeter to reconstruct effective kaon mass and remove background from $K_L \rightarrow\pi^0\pi^0\pi^0$ decay. It should be able to reconstruct decay point with accuracy better than one $\tau_S$.

- good particle ID. For CP study the main background decay into two charged pions is $K_L \rightarrow\pi\mu\nu_{\mu}$ decay, when neutrino is soft and muon can be selected as pion. Boost can give some advantage compare to colliding beams experiments, when pions and muons are soft and muon range system cannot be used.

- good timing

Beam line should prevent detector from direct exposition of huge initial flux of electrons and photons.

The target and recoil proton region should have about 1.5m length surrounded by cylindrical drift chamber and cylindrical calorimeter covering polar angles from about 20 to 180 degrees with no holes. This recoil region should be instrumented with all above systems. Good example with similar dementions and resolutions is CMD-2 detector. CsI crystals (pure, with good timing) can be used. (After producing about 10 tons of crystals for KEK, Novosibirsk can do another similar job for TJNAL) Recoil proton has 200-700 MeV/c momentum and also will be clearly seen in the same detector.

But for the trigger one should use fast timing (with DE/DX recognition). Full hadronic rate is expected at the level of $10^6$ and trigger system should reduce this rate to about $10^3$ or so to be able to record data. So fast (scintillators?) counters should be set before DC with amplitude analysis for DE/DX of protons. (May be after DC? Or crystals are fast enough?)

The forward going system should cover $\pm$20(?) degrees with uniform acceptance. One dipole magnet will not give this uniformity, but two in perpendicular direction could do. Torroidal magnetic spectrometer can also be considered. By detecting charged particle in forward direction, it should be posibility to reconstruct decay point in the decay region with accuracy about few cm.

Good segmented calorimeter with excelent spatial resolution is required. With four detected photons from two neutral pions one should be able to reconstruct kaon decay point with about 10 cm accuracy or better. It may be crystal calorimeter or liquid Krypton as in CERN NA48 or in KEDR detector in Novosibirsk.

For CP study it is not nesessery to select kaons, so Cerenkov counters are not needed, but muon range system should be used. But Cerenkov counters will be usefull for identifying charged kaons from other decays. About 50% of $\phi$ decays give $K^+ K^-$ pairs and rare decays and interactions of tagged charged kaons can be studied. CP-violation in asymmetry in $K^{\pm}\rightarrow\pi^{\pm}\pi^+\pi^-$ rates can be searched.

The trigger rate should be an order of 1 MHz and with sofisticated trigger processors the rate should be reduced to about 1 kHz to be able to record to tapes only interesting events. (Remember we expect $10^4 \phi$/s.)