While trying to find and work out a suitable design for the Cherenkov Imaging Detectors,
on comes across nice ideas, but several of them do not work out in the end. But if one
discards them quietly, they tend to crop up as helpful suggestions again and again.
Time-of-Propagation for the Endcap-DIRC
Idea: measure the time and determine the Cherenkov angle.
Caveat: light speed in medium is determined by the group velocity, not the phase velocity. One may express this with an idex of refraction n_group different from the usual n==n_phase. For fused silica from 300nm to 600nm n_phase varies 2%, n_group 6% and (projection of) light propagation normal to incident particle still 4%, see TimeOfPropagation
Result: pion-kaon-4sigma-separation is practically limited to about 3-4GeV/c.
Cs_I photocathodes (as used in HADES)
Idea: use photon detector more efficient and cheaper than standard PMT
Caveat: Cs_I wavelength range is VUV 146nm-210nm, fused silica or aerogel needs >250nm to >300nm
Result: no wavelength overlap
photon readout with wavelength shifters
Idea: concentrate the photons onto a small photon detector surface, compress the phase space
(properties of Bicron BC-484)
Caveat1: only photons 330nm-400nm are shifted
Caveat2: only 1/6 (singe-sided readout) or 1/3 of shifted photons geometrically transmitted
Caveat3: no 2-dimensional readout possible - shifted photon can only be in one "scitillator" bar,
hence most of the photons do not yield a useful angle information
Result: too little photons (on average 0.5 to 2 useful photons detected)
- 01 May 2006