MICE PID & trigger Detectors

Transcription

1 MICE PID & trigger Detectors V.Palladino MICE CM10, RAL Oct 29, 2004 for the PID team Bonesini, Gregoire, Kahn, Roberts, Rochford, Summers, Tilley, Tonazzo, Torun, Tortora, Wilson and more

2 Work reported at phonemeetings Wed 5.30 GMT ***********Wed Sep 1 *********** Wed Sep 15 canceled, used Sep 22 VC instead *********** Wed Sep 29 *********** Wed Oct Oct here at RAL

3 Conclusions Baseline design of devices are ready. (almost) completely General layout appear also stable G4BEAM stable tool Serious re-evaluation of µ/e performance pending G4MICE needs love&care CKOV-I, CKOV-II, MUCAL incomplete or still in debug phase

4 PID: status & goals for RAL CM-10 TOF-0 find a final site with moderate rates. < 3.8 MHz validate PMT choice (Ham 4498) CKOV-I (almost) no news, evaluate performance TOF-I find a final site, with moderate B-field, < K-gauss validate PMT choice (Ham 7791 fine mesh) Refine evaluation of performance If modified layout and/or thickness, re-evaluate TOF-II establish consensus on B-shielding and active area validate PMT choice (Ham 7791 fine mesh) CKOV-II establish consensus on active area implement simulation MUCAL establish consensus on active area debug simulation define size, Try to progress on digitization and controls evaluate performance

5 PID: status & goals for RAL CM-10 done in progress dormant done in progress done in progress TOF-0 find a final site with moderate rates. < 3.8 MHz validate PMT choice (Ham 4498) CKOV-I (almost) no news, evaluate performance TOF-I find a final site, with moderate B-field, < K-gauss validate PMT choice (Ham 7791 fine mesh) not necessary, so far Refine evaluation of performance If modified layout and/or thickness, re-evaluate TOF-II establish consensus on B-shielding and active area validate PMT choice (Ham 7791 fine mesh) CKOV-II establish consensus on active area implement simulation MUCAL establish consensus on active area debug simulation define size, in progress Try to progress on digitization and controls in progress progress too slow in progress progress too slow evaluate performance progress not fast enough resuming, new DAQ czar needed

6 Upstream detectors strongly coupled to beam design Recent Design Revisions JAN04 JAN04A MAR04 JUNE04 JUNE04A Current work SEP04

7 Isis beam SEP Μev/c, otherwise stable

8 MICE Beamline Analysis SEPT04 Tom Roberts Muons, Inc. September 22, 2004

9

10 TOF0 Distributions (Every blue dot lies right on top of a red dot.)

11 TOF0 : the issue is rate, thanks K. &T. 24 cm x 24 cm active area total SMALL!!!! must be bigger Each counter 24*4 cm*1, plexiglass light guide to Hamamatsu R4998 PMT+ boosted voltage divider

12 MICE TOF0, TOFI, TOFII status M. Bonesini INFN Milano

13 Considerations on scintillator thickness Shown time resolution is FWHM vs scintillator thickness L Green/red lines from BC408; blue line is BC404 (faster) Data from MEG tests at BTF Proposed thin solution: σ 100 ps if all goes right (perfect detector calibration,...) Actual choice : σ 60 ps

14 Considerations for PMT choice 1. Rate capability (up to some MHz) 2. Good timing properties (TTS) 3. Sustain magnetic field ( about gauss for TOF0, about.2 T for TOF2) Tests at Lasa magnet test facility (end July 04, for 15 days) with Pavia MEG group to optimize choice (M.Bonesini, F.Strati INFN Milano, G.Baccaglioni,F.Broggi, G. Volpini INFN Milano LASA, G. Cecchet, A. DeBari, R. Nardo, R. Rossella INFN Pavia).

15 Problems for high resolution scintillator based TOF (σ t < 100 ps) σ pl dominated by geometrical dimensions (L/N pe ) σ scint ps (mainly connected with produced number of γ s fast and scintillator characteristics, such as risetime) σ PMT dominated by PMT TTS σ scint+ σ PMT+ σ pl σ = + σ t N Additional problems in harsh environments: 1. B field (-> fine mesh PMTs) 2. High particle rates (tuning of operation HV for PMTs) pe 2 ele

16 Tests done at LASA Laser source to simulate MIP signal (about 300 p.e.) : fast PLP-10 laser on loan from Hamamatsu Italia Laser sync out triggers VME based acquisition (TDC + QADC) 5000 events for each data point : different PMTs (fine-mesh vs mod R4998), different B-field, different inclination vs B field axis (θ), diff laser rate to simulate incoming particle rates

17 Rate capabilities of PMTs To have a linear signal the mean average anode current (100 µa for R4998, 10 µa for R7761 or R5505, 100 µa for R5924 fine-mesh PMTs) must not be exceeded -> damage to dynodes... shorter PMT lifetime This gives a theoretical rate capability of: 267 KHZ with R KHZ with R7761,R5505 BUT!!! Divider can be modified only for R4998 (going up to 1.67 MHZ)

18 Test magnet at LASA (B up to 1.2T) PMT under test 1. B field up to 1.2 T 2. Free space 12 cm in height

19 Used laser light source Light source: Hamamatsu fast laser ( λ 405 nm, FWHM 60 ps, 250 mw peak power) PLP-10 Optical system: x,y,z flexure movement to inject light into a CERAM/OPTEC multimode fiber (spread 14 ps/m) PMT under test Laser light Signal ~ 300 p.e. to reproduce a MIP as measured with an OPHIR Laser powermeter

20 1. R4998 PMT rate studies R4998 with modified divider circuit: booster for last dynodes R 4998 R 5505 Structure Linear Focused Fine Mesh Stages Gain B= B=1 T Rise Time 0.7 ns 1.5 ns Transit Time 10 ns 5.6 ns Transit Time Jitter 0.16 ns 0.35 ns Nominal: up to 1.5 MHz

21 R4998 with modified divider (booster for last dynodes) + mu-metal shielding One specimen on loan in Jly 2004 for tests

22 Gain in magnetic field for R4998 Y Z x

23 Timimg properties of R4998 in B field

24 Rate effects studies for R MHz done with available R4998 with modified divider from Hamamatsu (booster on last dynodes) Light signal corresponds to ~ 300 p.e.

25 Cherenkov1 Distributions (Every blue dot lies right on top of a red dot.)

26 No news on CKOV I π µ e 20 cm 20 cm

27 TOF1 Distributions (Every blue dot lies right on top of a red dot.)

28 TOFI : B field is larger KG or so upstream shield needed 48 cm x 48 cm active area total fine-mesh KG Each counter 48 x 6 x 1, plexiglass light guide, Ham R7761 PMT

29 Fine Mesh Photomultiplier Tubes Secondary electrons accelerated parallel to the B-field. Gain with no field: 5 x With B=1.0 Tesla: 2 x x 10 5 Prompt risetime and good TTS Manufactured by Hamamatsu Photonics R5505 R7761 R5924 Tube diameter No. Of stages Q.E.at peak Gain (B=0 T) 5.0 x x x 10 7 Gain (B= 1 T) 1.8 x x x 10 5 Risetime (ns) TTS (ns)

30 Gain in B field (various orientations) G(B)/G(B=0T) G(T)/G(0) B θ PMT axis B (T) θ > critical angle B(T)

31

32 Pulse height resolution in B field 1 2

33 Rate effects (as a function of HV) rate capability is limited by maximum anode mean current (tipically 0.1mA for a 2 R5924 PMT) HV increases this is the ONLY relevant point, e.g. in B field if gain is lower by a factor F rate capability increases by 1/F with very high particle rates: try to reduce mean current

34 Rate effect as function of B field B field increases

35 Rate effects as function of B field B field increases

36 Timing studies

37 Time resolution

38 Time resolution vs Npe A / N pe c The number of p.e. is deduced from the PMT measured gain directly ( Q ADC = N pe e G )

39 Conclusions to have good TOF resolutions with high rate and high B fields one can try to optimize PMT working point (setting a proper HV + N pe ), but we have to study σ t as a function of rate in B field (this study has been done only at fixed rate) for TOF2 it is not worthwhile to reduce too much the B field (in the range.2-.5 T it is OK) inclination angle θ is a critical parameter, even F.M. PMTs do not work in very small B field if θ > critical value (B fields maps are an essential ingredient to design TOF!!!!) it may imply lightguides of quite complicate design to have θ < θ c. for moderate B fields < 200 Gauss another good choice is a R4998 with booster, but standard mu-metal shield is good only at B < 120 Gauss -> increase shielding for TOF0,TOF1 Thickness of TOF detectors can be optimized (but You loose N pe )

40 JUNE04A π/µ/e Discrimination, with Estimated Gaussian Resolutions in P perp, P z, and TOF Is JUNE04B OK? Is 0.5 Thickness OK? How much rate does that buy? Is TOF resolution dominant? Tails?

41 Downstream PID Distributions Same simulation, except: The downstream IronShield is physically removed Its effect on the magnetic field remains TOFX is added to the trigger; it is 1.5 by 1.5 meters, 1 mm thick, made of Air, located 1mm ahead of where the downstream IronShield would have been. TOF2 is extended to be 1.5 by 1.5 meters Cherenkov2 is extended to be 1.5 meters in diameter Calorimeter is extended to be 2.0 by 2.0 meters Relaxed Trigger: Good mu+ requires only 14 hits out of 18 counters (i.e. nothing after TOFX is required, no cut on particle type)

42 TOFX Distributions 1 (Every blue dot lies right on top of a red dot.)

43 TOFX Distributions 2

44 TOF2 Distributions (Every blue dot lies right on top of a red dot.)

45 TOF II should not present any problem 48 cm x 48 cm active area total fine-mesh KG Each counter 48 x 6 x 1, plexiglass light guide, Ham R7761 PMT

46 Cherenkov2 Distributions (Every blue dot lies right on top of a red dot.)

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65 Calorimeter Distributions (Every blue dot lies right on top of a red dot.)

66 Calorimeter layout Scintillating fibers embedded in grooved lead Side view: 2 blocks of 72x36x16 cm 3 72 cm Beam fibers Readout: 18 PMTs per layer at both ends Y Minimum cell size 4x4cm 2 due to PMT support X Z 16 cm 72 cm (no limitation)

67 Fine Grained Calorimeter Option for Muon Identifier Fine Grained Calorimeter Option for Muon Identifier The construction technique consists in embedding 1 mm diameter polystyrene based blue scintillating fibers between thin grooved lead plates, obtained by plastic deformation of 0.3 mm thick lead foil. Fibers are glued to the lead plates and run parallel to each other with a pitch of 1.35 mm and are mostly orthogonal to the entering particles. Same Construction Technique as KLOE EmCal fiber Density 3,7 g/cm 3 Radiation length 2,1 cm Moliere radius 3,4 cm ( estimate without glue)

68 Fiber-Lead composite 0,3 mm Lead + 1 mm Fiber 0,5 mm Lead + 1 mm Fiber

69 e/µ separation algorithm Muon efficiciency vs momentum and Comparable efficiency with all the sampling options muons P(GeV/c) at cal Fraction of electrons surviving the cuts is ~10% for P e >120 MeV (lager with very thin Pb layers) - but there might be some problems in the simulation we are checking Our favourite choice at the moment: Pb layer thickness 0.3mm

70 Basic elements for calorimeter construction Lead spool Big lead shaping machine The grooving rollers

71 Big swanging machine at RomaIII

72

73

74

75

76

77

78 TOFX Distributions 1 G4BEAM G4MICE

79

80 Preliminary Evaluation of PID Performance Steve Kahn 28 October 2004

81 Conclusions These results are quite preliminary. There are enough inconsistencies present, that one should expect these numbers to change. The EmCal selections plots are somewhat different from those previously shown. These differences need to be understood. The Ckov selections are not based on PMT photoelectrons captured. They are based on simple kinematics. This needs to be done better. The Tof time differences are not effectively used in the analysis. There is much that can be done to improve this in the future. But we have gotten started.

82

83

84 limitation to MICE DAQ rates (Emilio) expected to the largest extent from TDCs running in deadtimeless mode. Buffers get full and subsequent hits go unrecorded..rates up to 1 should be feasible but tests are necessary in order to establish that and consequently the MICE baseline DAQ rate.

85

86 PID: status & goals for RAL CM-10 done in progress dormant done in progress done in progress TOF-0 find a final site with moderate rates. < 3.8 MHz validate PMT choice (Ham 4498) CKOV-I (almost) no news, evaluate performance TOF-I find a final site, with moderate B-field, < K-gauss validate PMT choice (Ham 7791 fine mesh) not necessary, so far Refine evaluation of performance If modified layout and/or thickness, re-evaluate TOF-II establish consensus on B-shielding and active area validate PMT choice (Ham 7791 fine mesh) CKOV-II establish consensus on active area implement simulation MUCAL establish consensus on active area debug simulation define size, in progress Try to progress on digitization and controls in progress progress too slow in progress progress too slow evaluate performance progress not fast enough resuming, new DAQ czar needed

87 Baseline design of devices are ready. (almost) completely Conclusions General layout appear also stable G4BEAM stable tool Serious re-evaluation of µ/e performance pending G4MICE needs fresh young manpower CKOV-I, CKOV-II, MUCAL incomplete or still in debug phase

88 The end

89 TOFX Distributions 1 (Every blue dot lies right on top of a red dot.)

90 TOFX Distributions 1