CHIPP June 2008Jean-Sébastien Graulich MICE Calorimeter R&D Jean-Sebastien Graulich, Univ. Genève o o Introduction to MICE o o The MICE Calorimeter o o Sensor and Electronics o o Test beam results o o Future developments o o Conclusion CHIPP Workshop on Detector R&D June - University of Geneva
CHIPP June 2008Jean-Sébastien GraulichSlide 2 Introduction MICE is Accelerator Physics more than HEP MICE is part of the R&D program towards the neutrino factory Aim at Demonstrating Ionization Cooling of a muon beam Validate new simulation tools for future design MICE should deliver results before 2011
CHIPP June 2008Jean-Sébastien GraulichSlide 3 MICE Design Focus coils and LH2 Absorbers Coupling coils and 200 MHz RF Cavities Solenoids, Matching coils and Scintillating fiber Tracker
CHIPP June 2008Jean-Sébastien GraulichSlide 4 MICE PID Design Time of Flight Stations Cherenkov Counters Electron/Muon Calorimeter Some Beam Optics here
CHIPP June 2008Jean-Sébastien GraulichSlide 5 MICE RAL, UK Pion production and capture in ISIS vault MICE hall at RAL Pion decay solenoid (PSI) First beam Protons/Pions
CHIPP June 2008Jean-Sébastien GraulichSlide 6 MICE Electron-Muon Calorimeter Aim at separating muons from decay electrons Necessary to reduce systematic errors on emittance measurement Muons and electrons have ≠ Single Particle Emittance
CHIPP June 2008Jean-Sébastien GraulichSlide 7 Electron Muon Separation Muons and electrons have Overlapping P z in the Tracker (where it is measured) Time of flight is not sufficient to separate them “Nobody said it was going to be easy, and nobody was right.“ President George Bush, quoted by Rikard Sandstroem “Nobody said it was going to be easy, and nobody was right.“ President George Bush, quoted by Rikard Sandstroem
CHIPP June 2008Jean-Sébastien GraulichSlide 8 EMCal Conceptual Design Design proposed by Geneva Reduce the need for charge measurement (ADC are expensive, aren’t they…) Backboned by detailed simulation Made of two parts: High Z active pre-shower layer: Forces electrons to shower Thick stack of plastic scintillator: Stop the muons Measure range and track topology e+e+e+e+ µ+µ+µ+µ+
CHIPP June 2008Jean-Sébastien GraulichSlide 9 Pre-shower: KLOE Like
CHIPP June 2008Jean-Sébastien GraulichSlide 10 KL Construction Produced by INFN Roma III “In the mail” to RAL Coupled to individual PMTs Readout by 100 MHz waveform digitizer (after shaper) This gives both charge and time (< 1 ns) information KL + shielding KL Naked Signal Shape well understood
CHIPP June 2008Jean-Sébastien GraulichSlide 11 Scintillator Wall (SW) ~1 m 3 bloc of scintillator Longitudinal segmentation needed to measure the range Increasing thickness optimize the relative precision on the range w.r.t the number of readout channels Transverse segmentation needed to reduce the occupancy per channel Main issue: Muons stopping in the EMCal produce background when they decay At equilibrium: ~1 muon in the detector at all time
CHIPP June 2008Jean-Sébastien GraulichSlide 12 SW R&D Use extruded scintillators + WLS fibers A la Miner a First prototype built in INFN Trieste using rectangular shape scintillator received from Fermilab 19 mm 15 mm
Full scale single Layer Prototype 1/5 scale 2x4 Layers Prototype Simple mechanical assembly Easily stackable Assembled, equipped and tested in Trieste Has been tested in beam at CERN last week
CHIPP June 2008Jean-Sébastien GraulichSlide 14 SW Light Sensor choice Multi-Anode PMTs Like in Miner a itself 64 or 256 channels per PMT Last dynode can be used to measure the total energy deposit per plane. SiPM (or Hamamatsu MPPC) Like in T2K (P0D for example) Give some d.o.f. for the mechanical assembly Two Options
CHIPP June 2008Jean-Sébastien GraulichSlide 15 SW Front-End Electronics TDC with ~1.5 ns resolution is cheap Discriminator from existing ASIC (e.g. VA64TAP2.1) A simple FPGA can give the time w.r.t trigger ~ 20 CHF/channel can be achieved ADC is more expensive In particular if dead time has to be small Commercial waveform 1 GHz: ~2000 CHF/ch Note that cost decreases rapidly for lower sampling rates 2 Options (among others) Keep the full segmentation allowed by the design (~1500 ch) but with only 1 fADC per plane Reduce the segmentation by grouping fibers and equip each channel with shaper + fADC at 100 MHz (like for KL)
CHIPP June 2008Jean-Sébastien GraulichSlide 16 CERN T9 beam line Si-detectors TRACKING Plastic Scintillator TRIGGER 1m rods MICE-SW-EMR Recent Testbeam at CERN
CHIPP June 2008Jean-Sébastien GraulichSlide 17 Participants: Gianrossano Giannini, Pietro Chimenti, Erik Vallazza, Stefano Reia (M.T.), Dario Iugovaz(M.T.), Mauro Bari(E.T.), Giulio Orzan(E.T.) Trieste INFN and Trieste University- Physics Department Michela Prest, Valerio Mascagna & students: Andrea, Daniela, Davide, Said Como Univ. and INFN-MiB Jean-Sebastien Graulich Geneva Univ.
CHIPP June 2008Jean-Sébastien GraulichSlide 18 GLUED NOT GLUED multianode PMT (16 channels) XY Silicon detectors (resol. ~ 40 m) 2.0 GeV CERN T9 beam (e-/ -) 2 FIBERS SINGLE FIBER Light yield study
CHIPP June 2008Jean-Sébastien GraulichSlide 19 Pulse height comparison
CHIPP June 2008Jean-Sébastien GraulichSlide 20 Electron Muon Ranger Efficiency in SW
CHIPP June 2008Jean-Sébastien GraulichSlide 21 Efficiency versus beam particle position in scintillator bars (1.9 cm height) in all 4 vertical layers superimposed Wide and deep efficiency drop between individual bars Being analyzed more carefully..
CHIPP June 2008Jean-Sébastien GraulichSlide 22 The first two layers are equipped with WLS fibers on both sides
CHIPP June 2008Jean-Sébastien GraulichSlide 23 Obninsk/CPTA SiPM (Russian flavor)
CHIPP June 2008Jean-Sébastien GraulichSlide 24 IRST SiPM (Italian flavor) After cut on the opposite scintillator side (readout with MA-PMt) Easy calibration ! Energy spectrum 2 GeV/c)
CHIPP June 2008Jean-Sébastien GraulichSlide 25 Further developments Considering triangular shape for better occupancy Minerva die set gives a triangle with a base of 33mm and ht. of 17 mm nominal Allow grouping channels with marginal loss in resolution => savings in FEE Miner a test bench
Each layer 59 triangular bars: 30 triangular bars x 33 mm =990 mm + 29 matching triangular bars Each layer Active Scintillator region : 990 mm x 990 mm x 17 mm Thickness with fiberglass cover~18 mm Total=18 mm x 40 layers = 720 mm Half triangular edge mechanical elements
CHIPP June 2008Jean-Sébastien GraulichSlide 27 Conclusion MICE will demonstrate Ionization Cooling for muons by 2010/2011 It requires electron/muon separation for p between 140 and 240 MeV/c at level, with 99.8% efficiency A PID detector system has been designed in Geneva and is constructed by FNAL/Geneva/INFN collaboration It includes a fully active extruded scintillator detector (~1m 3 ) with WLS fiber and (Si)PM readout We are developing cheap and efficient construction techniques and Front End Electronics Economical FEE able to do charge measurement at MICE rate (1MHz) is something to be…
28 Spare Slides
CHIPP June 2008Jean-Sébastien GraulichSlide 29
CHIPP June 2008Jean-Sébastien GraulichSlide 30
How efficiency is defined: 6) 6)To overcome this problem an “or” condition has been added so that the efficiency is computed not only for the expected rod but taking into account also the previous and the next one. In other words, if the particle is expected to pass through rod number 3, the pulse heights of rod number 2, 3 and 4 are considered, if anyone of them is over-threshold, the particle in this event is considered as “detected” (that is a 1 is put into the profile histograms described before). The dead zone width is reduced. Anyway a (smaller) region seems to have poorer efficiency... maybe rods profile is not rectangular!
CHIPP June 2008Jean-Sébastien Graulich Curved edges Dead zone up to ~ mm 1m long rods of the same type of the EMR ones
CHIPP June 2008Jean-Sébastien Graulich ENERGY SCAN At 15 GeV dead region seems to reduce at less than 0.5 mm (as expected from rods shape)