1. May 16 th 2006M. Bruschi – CSN1 Roma1/20 A. Bertin, M. Bruschi, D. Caforio, S. De Castro, L. Fabbri, P. Faccioli, B. Giacobbe, F. Grimaldi, I. Massa, M. Piccinini, M. Poli, C. Sbarra, ASbrizzi, N. Semprini-Cesari, R. Spighi, M. Villa, A. Vitale, A. Zoccoli Stato del Luminometro (LUCID) Outline:  The Lucid detector  Test beam results  Tests on Bench  The installation scenario  The electronics  Conclusions

2. May 16 th 2006M. Bruschi – CSN1 Roma2/20 Activities of the Group since Sep.05 Main activities of the Bologna Group in LUCID:  development of the electronics  Preparation (electronics,daq) Test Beam @ Desy (Nov. 05) and data analysis  Monte Carlo: development and optimization of the code (to reproduce the test beam results). Simulation of the backgrounds.  Test on bench @ CERN to check the detector performances and tune the MC  Discussion of the detector design and of the installation strategies

3. May 16 th 2006M. Bruschi – CSN1 Roma3/20 LUCID: relative Luminosity Monitoring LUCID : “LUminosity measurement using Cerenkov Integrating Detector A bundle of ~200 (per end) projective Al Cerenkov tubes around the beam pipe Cerenkov radiator gas isobutane Light fed into PMTs via quartz fibres

4. May 16 th 2006M. Bruschi – CSN1 Roma4/20 LUCID position Planned installation after the Big Wheels A/C.

5. May 16 th 2006M. Bruschi – CSN1 Roma5/20 Test Beam @ Desy (Nov. 05): The T22 teststand has a SiMS telescope - resolution ≈30µm (LUCID) (Roman Pot) Cerenkov tubes Test beam stand for Roman Pots Winston cones window of gas vessel under pressure test Test Beam stand for LUCID

6. May 16 th 2006M. Bruschi – CSN1 Roma6/20 Test Beam analysis Data analysis performed in Bologna together with MC simulations. Fit of a typical ADC spectrum:

7. May 16 th 2006M. Bruschi – CSN1 Roma7/20 Test Beam results Pressure scan C 4 H 10  Number of pe’s/Cerenkov-tube for C 4 H 10 at 1 atm. pressure is 5.3  C 4 F 10 pressure scan - essentially the same as that of isobutan  Fall of in signal a factor of two in 2 degrees (  nature of LUCID)  Disagreement of a factor ~7 w.r.t. MC  Number of pe’s/Cerenkov-tube for C 4 H 10 at 1 atm. pressure is 5.3  C 4 F 10 pressure scan - essentially the same as that of isobutan  Fall of in signal a factor of two in 2 degrees (  nature of LUCID)  Disagreement of a factor ~7 w.r.t. MC Angle scan C 4 H 10 1 bar

8. May 16 th 2006M. Bruschi – CSN1 Roma8/20 Comparison with Monte Carlo Stage A  Improvements in MC still ongoing (light diffusion, etc.)  Checks of the detector performances on bench ongoing Now MC predictions much closer to RD results: ~10 p.e.

9. May 16 th 2006M. Bruschi – CSN1 Roma9/20 Test on bench: quantities studied

10. May 16 th 2006M. Bruschi – CSN1 Roma10/20 Test on bench @ CERN Performed under the Bologna responsibility: Studies on all the different parts of the detector:  tube reflectivity between 75% & 95% (depending on the incident angle, typ: 90%), variation up to 10% observed along the tube  Not responsible of the poor performances  light attenuation along the fibers is about 75%  WC – Fiber light collection

11. May 16 th 2006M. Bruschi – CSN1 Roma11/20 Test bench results All quantities measured on bench. Quite good understanding of the detector performances.

12. May 16 th 2006M. Bruschi – CSN1 Roma12/20 Possible improvements & next steps Improvements on the detector; Tubes: improvements in the internal surface (~ 1.1 - 1.2) Winston cone: new design + internal surface (~ 1.2 - 1.7) Fibers: improve fiber quality. ~2 less attenuation (at 300 nm), factor of ~2 large angular acceptance plus a larger (core)/(clad) area ratio (0.78 compared to 0.68) ( ~2) PM + Fibre: enlarge the light spectrum sensitivity of PM +fibers (UV range) (~ 2)  Possible improvements in light collection: factor ~6/7 Next steps: New tubes + WC + fibers will be delivered @ CERN middle May  Test on bench Test beam @ CERN end of June to test the new detector Test beam @ CERN in November to test the final detector

13. May 16 th 2006M. Bruschi – CSN1 Roma13/20 Backup solution (under study) Place single miniature PMTs directly onto the end of each Winston cone? Advantages: –More light because of direct coupling –Lower electronics cost –Swap quartz fibres for (cheaper) signal cables Need to study: –Radiation hardness of complete PMT system –Background Cerenkov light in window from primary & secondary particles –Activation –Mechanical design Place single miniature PMTs directly onto the end of each Winston cone? Advantages: –More light because of direct coupling –Lower electronics cost –Swap quartz fibres for (cheaper) signal cables Need to study: –Radiation hardness of complete PMT system –Background Cerenkov light in window from primary & secondary particles –Activation –Mechanical design HAMAMATSU R4296 PMT

14. May 16 th 2006M. Bruschi – CSN1 Roma14/20 Detector installation The installation of the detector early in 2007 is becoming tight. Need to wait for the test beam results. Problems for the electronic design until the detector performances are fixed. Detailed planning available. Still possible, but no contingency Investigating alternative scenarios for a reduced detector (e.g. 1 ring per side). Detector Completion at the first long shut-down (end 2008 ?). Presently under discussion.

15. May 16 th 2006M. Bruschi – CSN1 Roma15/20 Purposes of the LUCID electronics The Cerenkov tubes based ATLAS luminosity detector has the following main goals: 1.To measure the integrated luminosity per BX 2.To measure the luminosity per BX 3.To provide a trigger for interaction (at machine low luminosity) and for diffractive physics Logic blocks needed to achieve these three points: 1) ad hoc front end card 2) ad hoc readout card 3) ad hoc trigger card and a detector acceptance as big as possible

16. May 16 th 2006M. Bruschi – CSN1 Roma16/20 Changes in the electronics design since fall 2005 A very general scheme to fulfill the readout and trigger needs for LUCID was developed in fall 2005 Since then, some important changes were achieved: Rearrangement of the main logic blocks between the detector and USA 15 (the readout will not be placed in an intermediate rack) Front end components (MAPMT, fibers, fed cards, cables) deployment in the nose shielding The front end system is now modular so allowing scalability A readout and trigger scheme for the backup solution is now available

17. May 16 th 2006M. Bruschi – CSN1 Roma17/20 System Architecture FRONT END (FEPCB) Similar to Roman Pot FE: OPERA/MAROC chip Input: MAPMT Output: DIGITIZED INFORMATION on optical Links (~1.5 Gb/s) Analog Information on Coaxial (or diff. on tw. Pair) 19x2 M A P M T GOLLINKS+coaxGOLLINKS+coax LHC sync:TTCrq, CTRL sign.,etc ROS ROD T R I G G. CARDCARD HVFE CONTROL PC

18. May 16 th 2006M. Bruschi – CSN1 Roma18/20 The LUCID FED cards The front end system is now modular so allowing scalability The MAROC Boards will be developed by LUND (thanks!) TX Digital Board GOL, FPGA, TX Connectors Board (HV,LV,RX, TX,analog) MAROC Boards Base board (only connectors and signals routing) ANALOG board (ampl.+line driv.) RX Digital Board TTCRQ, volt. reg.

19. May 16 th 2006M. Bruschi – CSN1 Roma19/20 Conclusions/detector  LUCID can rely on quite settled technologies both on the detector & the electronics side  Detailed studies on the detector in test-beam and laser tests and a lot more is known (simulation, mechanics, light detection, fiber readout, cabling, electronics placement, readout, etc…).  Good chances to improve the light emission by about a factor 6-7.  Improved detector version will be tested in May on bench and in the test beam of June  Large impact of the Bologna Group on all the ongoing activities  Different installation scenarios under discussion. Aim for installation of LUCID in 2007.

20. May 16 th 2006M. Bruschi – CSN1 Roma20/20 Conclusions/electronics  The design of the front-end electronics has been started (although we need a confirmation of the performances of the detector)  The front end electronics (both for the baseline and the backup solution) should be reasonably ready for february 2007  The project is now modular and can easily adapted in case we will decide for the LUCID readout backup solution  The development of the readout and trigger electronics could be a bit relaxed since we can use already available pieces of DAQ to give a luminosity measurement as soon as the front end electronics will be available  First version of MAROC chip including modifications for LUCID submitted on March 6 th

21. May 16 th 2006M. Bruschi – CSN1 Roma21/20 Backup slides on detector

22. May 16 th 2006M. Bruschi – CSN1 Roma22/20 LUCID position: MBTS TILE η MAX = 6.073 η MIN = 5.374 Detector goals:  provide a relative luminosity measurement (integrated and bunch by bunch)  Provide η-coverage for diffractive physics

23. May 16 th 2006M. Bruschi – CSN1 Roma23/20 Parametrization of PMT spectra

24. May 16 th 2006M. Bruschi – CSN1 Roma24/20 Improvements: Winston cone

25. May 16 th 2006M. Bruschi – CSN1 Roma25/20 Improvements: new fibers

26. May 16 th 2006M. Bruschi – CSN1 Roma26/20 Fiber efficiency A transmission has been made on the fibers used in the TB. The quartz fibers have a better transmission than the plastic fibers but this is compensated for by a larger numerical aperture of the plastic (0.51) compared with the quartz (0.37).  Same number of p.e. in the test beam

27. May 16 th 2006M. Bruschi – CSN1 Roma27/20

28. May 16 th 2006M. Bruschi – CSN1 Roma28/20 BACKUP SLIDES on electronics

29. May 16 th 2006M. Bruschi – CSN1 Roma29/20 12 m 24 m 12 m 2 m (wall) Rack? 12 m 8 m (MAPMT out) From MAPMT out to USA15 Total Cable Length: 8+12+12+24+12+2 = 70 m Inside USA15: <25 m 80 m < Total Cable Length < 100 m

30. May 16 th 2006M. Bruschi – CSN1 Roma30/20 Baseline Readout Scheme LUCID 1 Nose Sh. USA15 ~4m ~80-100 m Optical fibers MAPMT MAROC DRIVERS COAX (or tw. pair) + Optical fibers READOUT TRIGGER Total number of channels: 168x7x2=2352 for 19x2 MAPMT

31. May 16 th 2006M. Bruschi – CSN1 Roma31/20 Electronics and Cables placement

32. May 16 th 2006M. Bruschi – CSN1 Roma32/20 Electronics GOLGOL First version of MAROC chip including modifications for LUCID submitted on March 6 th (thanks to the ORSAY group!)

33. May 16 th 2006M. Bruschi – CSN1 Roma33/20 General Comments The TDC in the readout block will be probably be replaced by a simple gated coincidence In case the LUCID readout backup solution would be preferred, then: - The MAROC chip will not be used - The FED cards will contain only the ANALOG board - The STRU unit of the readout card will be simplified 

34. May 16 th 2006M. Bruschi – CSN1 Roma34/20 Single Tube Readout Unit for PMT (1 channel) 15 ns int 10 ns reset 25 ns time LHC Clock LHC Int. Time ADC GATE TDC START to the trigger unit Fan Out TDC GI + ADC Multiplicity per Tube LUT 8 8 3 PMT output STRU Discr. (Prog. Thr +NR) TDC START STOP ADC GATE

35. May 16 th 2006M. Bruschi – CSN1 Roma35/20 Possible Initial DAQ scheme for MAPMT readout (digital part) SBCSBC LTPLTP TTCVITTCVI TTCEXTTCEX TTCOC CTP Filar Cards (PCI IF) ROS Ethernet Link To LUCID FED 2 Server Computers from LUCID FED

36. May 16 th 2006M. Bruschi – CSN1 Roma36/20 COST ESTIMATE (must be updated With MAROC boards Cost ~ 10k€)

37. May 16 th 2006M. Bruschi – CSN1 Roma37/20 FED TIME SCHEDULE for FED

38. May 16 th 2006M. Bruschi – CSN1 Roma38/20 Single PMT readout The candidates PMT are Hamamatsu R2496 with quartz window < 15  A (could be 30  A with active divider) with HV=1250 V Number of photons @ 1 bar 400 (MC/8 Feb. 2006) QE ~ 20% 15  A>400x0.2x1.6x10 -19 xG PMT x40MHzx0.3  G PMT ≤10 5  QMAX~1.3 pC  (signal duration 10 ns)  Vsig ~ 13 mV

39. May 16 th 2006M. Bruschi – CSN1 Roma39/20 stru 1 stru 2 stru 20 SUM_OUT 1_1 SUM_OUT 1_2 SUM_OUT 2_10 GOL 1 GOL 2 GOL 9 GOL RX LVDS S/P 3 3 3 LVDS 1_1 (to trig. unit) LVDS 1_2 (to trig. unit) LVDS 2_1 (to trig. unit) LVDS 2_2 (to trig. unit) stru 2 TTCRQ i.f. opt. lnk from TTCEX VME P1 VME I.F DPRAM CTRL LOGIC 6 Bytes EVENT BUFFER s-LINK to ROS from CTRL LOGIC 160 MB/s s-LINK Busy LUCID ROD BOARD (22 units + spares) – VME 9U Analog_In 1 Analog_In 2 Analog_In 20 GOL_In 1 GOL_In 2 GOL_In 9 ~200 Bytes/ev

40. May 16 th 2006M. Bruschi – CSN1 Roma40/20 Signal Buffer TTCRQ i.f. opt. lnk from TTCEX VME P1 VME I.F CTRL LOGIC s-LINK to ROS 160 MB/s s-LINK Busy LUCID TRIGGER BOARD (1 unit + spares) – VME 9U Detector1Detector1 Detector2Detector2 LVDS 1_1 LVDS 1_2 LVDS 22_1 LVDS 22_2 1 2 43 44 LVDS 23_1 LVDS 23_2 LVDS 44_1 LVDS 44_2 1 2 43 44 Signal Buffer FPGA based TRIGGER PROCESSING UNIT 44 ser. Inp 594 bit/BX 44 ser. Inp 594 bit/BX to the L1 trigger ~200 Bytes/ev Algorithm: MC simulations are needed

41. May 16 th 2006M. Bruschi – CSN1 Roma41/20 MAPMT readout The candidates MAPMT are Hamamatsu H- 7546-03 with UV glass (or quartz?) < 100  A (8 stage boosted divider) with HV=1000 V Number of p.e. per channel: 5 (100  A/64)>5x1.6x10 -19 xG PMT x40MHzx0.3  G PMT ≤1.6x10 5  QMAX~0.13 pC  (signal duration 10 ns)  Vsig ~ 1.3 mV

42. May 16 th 2006M. Bruschi – CSN1 Roma42/20 MAROC analog output characteristics 25 mV/160 fC  typical output from single fiber (130 fC) ~ 20 mV Typical output from 7 fibers ~ 140 mV

43. May 16 th 2006M. Bruschi – CSN1 Roma43/20 Cabling The readout of the front includes both analog and digital signals transmission For the analog ones: single ended on coaxial (but: differential on twisted pairs under consideration) For the digital ones: optical fiber is the baseline

44. May 16 th 2006M. Bruschi – CSN1 Roma44/20 Cables Summary Globally we will have 2x19=38 MAPMT

45. May 16 th 2006M. Bruschi – CSN1 Roma45/20

46. May 16 th 2006M. Bruschi – CSN1 Roma46/20

47. May 16 th 2006M. Bruschi – CSN1 Roma47/20

48. May 16 th 2006M. Bruschi – CSN1 Roma48/20

49. May 16 th 2006M. Bruschi – CSN1 Roma49/20 General Considerations-III For the description of the readout electronics I will refer essentially to the baseline of the detector described in the LOI Detector: formed by two parts each one consisting of 200 Cherenkov counters (tubes) 5 layers/section x 40 tubes/layer x 7 fibers/tube x 2 sections = 2800 fibers Signal: Prompt particles coming from the IP (primaries) will traverse the full length of the counter and generate a large amplitude signal in the photo-detector Background I: Particles originating from secondary interaction of prompt particles in the detector material and beam-pipe (secondaries) are softer and will traverse the counters at larger angles (multiple reflections), with shorter path lengths  Background I significantly smaller than signal Background II: Particles crossing the readout fibers will produce light only on the crossed fibers  Background II will have different pattern of hit fibers wrt signal