1. Electron ID - Working Group Thursday, March 10, 14:00-17:00 14:00High rate beam test of gas detectorsV. Golovatyuk (JINR) 14:20High counting rate TRDM. Petris (NIPNE) 14:40ATLAS straw-tube TRT test beam resultsV. Tikhomirov (Moscow) 14:55Straw chamber read-outK. Zaremba (Warsaw) 15:10MWPCs vs straw tubes for the CBM TRDC. Garabatos (GSI) 15:30RICH resolution studies IC. Hoehne (GSI) 15:50RICH resolution studies IIB. Politchouk (IHEP) 16:10Status of ECAL for CBMI. Korolko (ITEP) 16:30Discussion next steps, TPJ. Wessels (Muenster)

2. Beam Test Goal of the experiment: detector performance in high counting rate environment Experimental Setup: - 2 Scintillators (ToF, trigger); - 2 Si - Strip Detectors (position information); - 2 MWPC - GSI (10 x 10 cm 2 ) - 1 MWPC - Bucharest (24 x 24 cm 2 ) -1 MWPC - Dubna (10 x 10 cm 2 ) - 1 GEM - Dubna - Pb - glass calorimeter (last run) - FADC readout ; DAQ (MBS ) M. Petris (NIPNE)

3. Beam Composition e/  vs. rate  not possible, too low e intensities at SIS  /p @ 1 GeV/c  too low rates;  p/d @ 2 GeV/c rate dependence studies.  p  p pd pd M. Petris (NIPNE)

4. e/  Discrimination M. Petris (NIPNE)

5. Fig.5. Beam intensity distribution during the spill. The figures represent the case when extraction time was 0.15 and 2.0 sec. 0.15 sec The information from the upstream scintillation counter which covers the beam also was used for a total beam intensity estimation. Number of counts in this counter happened in time from the previous trigger was recorded. Having in addition information from the clock about time between triggers we are able to recover the time structure of the beam passed through our detectors. 2.0 sec Beam Intensity Estimation V. Golovatyuk (JINR)

6. Average pulse shape from FADC (50 bins x 30 nsec) for different readout chambers and different spill length V. Golovatyuk (JINR)

7. Stability of the charge signal from GEM and Dubna chamber vs beams intensity V. Golovatyuk (JINR)

8. Stability of the charge signal from GSI chambers vs beam intensity V. Golovatyuk (JINR)

9. Space resolution vs beam intensity (Dubna chamber ) V. Golovatyuk (JINR)

10. High Counting Rate Effect (4.2  1.2) % (3.2  2.1) % (11.7  1.8) % (0.4  1) % M. Petris (NIPNE)

11. 2 s 1850 V 0.2 s 1850 V Gain - 2s spill length, for protons M. Petris (NIPNE)

12. Conclusions Diagnostic system not specific for the environment  limited performance The results of the first in-beam tests of such a geometry seem to recommend it as a solution for a major percentage of the CBM TRD subdetector. Participants: Bucharest, Dubna, GSI, Heidelberg and Münster Modest statistics  limiting factor for detailed analysis in terms of: - Beam profile - Spill profile Much too low statistics for negatives (e/  ) Although M. Petris (NIPNE)

13. ATLAS TRT Barrel module setup V. Tikhomirov (Moscow)

14. Rejection: threshold scan Optimal threshold is around 6-7 keV Rejection power better than 2% can be achieved ATLAS Inner Detector TDR gives 5% at central rapidity, but: 1) with smaller number of active straws due to double wire join 2) with magnetic field 3) with full material in front of TRT taken into account V. Tikhomirov (Moscow)

15. High rate prototype Goal: estimate straw performance in multiparticle environment. Single straw in the beam line Straw matrix, irradiated by Fe 55 source Mixed output signal V. Tikhomirov (Moscow)

16. High rate prototype (2) Deterioration in both tracking accuracy and efficiency with rate increasing. V. Tikhomirov (Moscow)

17. High rate prototype (3) Still good performance is demonstrated up to 20 MHz counting rate V. Tikhomirov (Moscow)

18. Radiator prototype (2) TR radiation from straw walls has to be taken into account to describe spectrum above 8 keV V. Tikhomirov (Moscow)

19. Angular effects in case of an anode readout x d  track  =  (1/cos  ) xd max. resol.worsening  15 % K.Zaremba, J.Marzec, CBM Meeting, 10.03.2005

20. Angular effects in case of a pad readout K.Zaremba, J.Marzec, CBM Meeting, 10.03.2005

21. ASD-8BIRIMIO TechnologybipolarBiCMOS Input range100 fC Input impedance120 Ω10 Ω Input noise1000 e ShapingGauss 3 x int t P = 12 ns Gauss 4 x int t P = 7 ns Tail cancellationFor t 0 = 2 ns Base line restorerno0.017 fC/ns (+) 0.17 fC/ns (-) Threshold0-150 fC nonlinear 0-20 fC linear Power supply+3 v, -3V+3V, -2V Power consumption41 mW per ch.35 mW per ch. K.Zaremba, J.Marzec, CBM Meeting, 10.03.2005 ASD8-B versus IRIMIO parameters

22. Occupancy, granularity MWPC: Pad size chosen to match occupancy and resolution in the bend direction. Straws: Straw length chosen to match occupancy. TSR layerschannelsoccupancy MWPC9550 00010 % Straws18300 00018 %  At the end, the number of channels should be equal for equal occupancy C. Garabatos (GSI)

23. Resolution, rate capability MWPC First estimation from test beam data C. Garabatos (GSI)

24. Resolution, rate capability Straws: CBM rates no problem for the straw tubes Resolution worse than quoted 500 kHz/cm 2 C. Garabatos (GSI)

25. Segmentation MWPC III III Increasing segmentation, determined by chamber size Small cracks (frames, services) C. Garabatos (GSI)

26. Segmentation Straws Many dead areas Cracks with material (end-plugs, electronics, services) Non-projective geometry  Need a careful estimation of coverage and radiation length III III C. Garabatos (GSI)

27. Leak rate  cost MWPC 10% vol/year (1 mbar l/h) 1.5 k€/yr Straws 0.2 m 3 /day 50 k€/yr C. Garabatos (GSI)

28. Conclusions Comparable performance (to 1 st order) Multiple scattering will probably drive the tracking performance: need good estimates of material budget Stability and gas for the straws to be clarified Both detectors would need substantial R&D to be ready for CBM C. Garabatos (GSI)  discussion on roadmap for common TRD tomorrow

29. particle identification with RICH ring finding ring finder: Hough Transform, Elastic Net to be implemented in framework → efficiencies... determination of center and radius of ring/ Cherenkov angle matching of rings with tracks → tracking (momentum and position resolution), information from other detectors pid by combining ring radius and momentum information  detailed knowledge of resolution necessary! C. Hoehne (GSI)

30. emission point rings(  ) -  polar angle,  azimuth angle no diffusion at reflection no magnetic field, no multiple scattering to do: quantify and correct for distortions at large  improve focussing/ position of focal plane correct for remaining distortions  = 80 o 60 o 40 o 20 o  = 5 o 10 o 15 o 20 o 25 o 30 o 35 o one quarter of mirror/ photodetector: → restrict investigation of resolution to "good" area in central region and wait for optimized setup C. Hoehne (GSI)

31. chromatic dispersion [nm]  [mrad] 60024.42 20026.15 15028 10036.75 strong increase of n( ) in UV region however, dN/d also increases in UV region and N2:N2: Boris Polichtchouk [Landolt Boernstein Series, 6 th Edition, volume II/8 Ph.D. thesis of Annick Bideau-Mehu (1982)] N2N2 4mrad   ~2mrad (~0.4cm) C. Hoehne (GSI)

32. total resolution (I) multiple scattering   ~ 1 mrad (p=1 GeV) magnetic stray field   < 1 mrad (p=1 GeV) emission point   small because of corrections, optimization angular deviation of mirror   < 0.1 mrad chromatic dispersion   > 1 mrad (strongly dependent on min ) pixel size   ~ 1-2 mrad  couple of mrad contributions, independent errors  c =24.4 mrad    ~2-3% of  c C. Hoehne (GSI)

33. total resolution (II) gaussian distributed Cherenkov angles/ radii → calculate separation power for e and  in terms of   for different   1% 2% 3% 4% 5% C. Hoehne (GSI)

34. pid versus  R 1% momentum resolution 1.3 mrad resolution in azimuth angle 0.8 mrad resolution in deep angle 200  m position resolution in mirror ideal tracking: with of  R –distribution due to method for ring center determination finite tracking: distribution widened cut on  R important for efficiency and purity! C. Hoehne (GSI)

35. Resolution vs UV cutoff Two concurrent mechanisms: 1)n(lambda)-> const => better resolution; 2) Nhits decreases (see next slide)=> less number of points to fit => worse radius resolution ➔ 130-150 nm – optimal transparency cut for PMT window B. Politchouk (IHEP)

36. Number of fired tubes per ring vs UV cutoff B. Politchouk (IHEP)

37. Resolution vs pad size of photodetector B. Politchouk (IHEP)

38. Conclusions ➔ 130-150 nm is an optimal transparency cut for PMT window in it's current design (R=0.3cm) ➔ Resolution drops sharply if UV cutoff > 200nm (for Rpmt = 0.3cm). ➔ For larger PMT diameters, the transparency requirements in UV region are harder. ➔ High multiplicity environment with overlapped rings and realistic ring recognition program are expected to lead to additional deterioration of ring radius resolution... B. Politchouk (IHEP)

39. summary/ outlook particle identification with the RICH detector aim: momentum dependent pid efficiency and purity efficiency: ring finders to come purity: started with detailed analysis of ring radius resolution  for  =3% of  c we have 3  separation between e and  at 13.5 GeV/c  impact on detector layout: granularity of photodetector maximum wavelength range for photodetection purity: extend tracking algorithms for extrapolation of tracks to photodetector plane combine with information from other detectors C. Hoehne (GSI)

40. SROccupancy (24 K channels) Occupancies are VERY high ! 1. SR thresholds – 50, 60, 80 MeV 2. 10 MeV thresholds – 40-50% occ. Do we need very high energy resolution ? 1. HEP experience – NO 2. ALICE example – YES And it is much more interesting to build such a good calorimeter… Usual “shashlik” technology exists already I. Korolko (ITEP)

41. SRe-π separation 1. Compare track momentum with CALO energy Use 2x2 matrix to minimize pile-up from neighbors Most simple approach – improvements possible 2. Shower shape (2x2 / 3x3 ratio) Gives additional factor 2 Rather simple approach – improvements possible 3. Preshower Gives additional factor 2-3 (with 95% electron efficiency) For rough sampling ECAL – could perform better I. Korolko (ITEP)

42. Small exercises Repeat SR simulations with rough sampling almost the same results for e-pi separation Repeat SR simulations with less CALO cells worse results for e-pi separation (still in progress) I. Korolko (ITEP)

43. “smaller” Calorimeter 1. Reduce CALO acceptance a)Outer acceptance b)Inner acceptance 2. Make it rougher (number of plates) 3. Look for cheap photo-detectors (lower gain) 4. Reduce number of channels a)lateral granularity b)longitudinal granularity (Preshower) I. Korolko (ITEP)

44. electrons from J/  decays P t > 1 GeV/c Strong dependence- Small number of cells- Mostly high P t particles- Worse e-pi separation+ Saving material, not money Outer acceptance I. Korolko (ITEP)

45. Electron ID - Working Group Discussion: next steps in detector research –dedicated high rate test environment (DAQ, trigger…) –comparison of different gas mixtures and gains –quantitative evaluation of position resolution –tests of thicker detectors MWPC, GEM and radiators –characterization of X-ray response –faster PASA development –high rate and high multiplicity tests –straw prototype test and electronics development –evaluation and choice of PMT –choice of radiator gas –readout of RICHes based on gaseous detectors to be studied –evaluation of beam data from ECAL measurement

46. Electron ID - Working Group Discussion: towards a technical proposal –improve e/pi evaluation of individual detectors –combined e/pi of ALL detectors with realistic materials –evaluation of straw/MWPC options of TRD - test beams –explore straws at ‘low’ momentum –combined tracking for optimization of detector numbers, position resolution requirement, subdivision, placement –include noise and background in pattern recognition –ring distortions at the edges –acceptance/performance studies for ‘light’ ECAL –evaluate possible trigger/data compression options