1. Next generation HCAL prototype Felix Sefkow EUDET electronics meeting UCL, December 5, 2006

2. Felix Sefkow January 29-30, 2007Next HCAL 2 Outline Operational experience Technical design issues Prototype roadmap

3. Felix Sefkow January 29-30, 2007Next HCAL 3 Operational experience DATA MC REALITY

4. Felix Sefkow January 29-30, 2007Next HCAL 4 First lessons Detector is robust and stable, impressively efficient running –Preparation and initial integration in DESY testbeam was extremely valuable Integrated ECAL HCAL TCMT electronics approach proved successful –Common DAQ –Common online monitoring –Common analysis software –“no” detector integration effort SiPMs are still a pioneering technology –no real mass product yet –flexibility and adjustments necessary SiPM noise is just small enough Coherent noise from FEE in first data, already eliminated 6 GeV  + Online event display Front end gain ratio (fast / slow mode)

5. Felix Sefkow January 29-30, 2007Next HCAL 5 More lessons Redundancy proved vital –Observe single photo-electron peaks: –“auto-calibration” –Heavily used: SiPM equalization, scintillator quality control, detector operation, non- linearity, stability monitoring Temperature variations and gradients occurred in the last run period –Exercise monitoring and correction concepts 1.PIN diode monitored LED reference signals 2.Gain measurements 3.Temperature monitoring 0 pixel 1 pixel 2 pixels MIP gain Pixel / MIP (efficiency)

6. Felix Sefkow January 29-30, 2007Next HCAL 6 Detector studies to do next Calibration and calibration uncertainties –What limits the resolution? Dynamic range and non-linearity correction accuracy –Alternative methods? Stability, monitoring and corrections: –Cross-check methods and find most practical Effects of temperature and temperature gradients SiPM long-term studies –Is the response function stable? And: Learn from scintillator ECAL –In the DESY beam next month

7. Felix Sefkow January 29-30, 2007Next HCAL 7 Physics lessons Physics (hadron shower) analysis will provide further input to the design Shower sub-structure resolution, two-particle separation –revisit granularity optimization Neutron sensitivity Semi-digital approach Not yet possible: timing for PFA –Fast digitization (JINR electronics) –Next version of SiPM ASIC from LAL

8. Felix Sefkow January 29-30, 2007Next HCAL 8 Summary so far Proof-of principle prototype (technology & PFLOW approach) Technology –It works! –Successfully scaled up from “minical” by 2 orders of magnitude PFLOW approach –It makes nice pictures –Analysis to come – millions of events are there HCAL 3x3cm 2 tiles ECAL 1cm 2 pads Imaging calorimetry

9. Felix Sefkow January 29-30, 2007Next HCAL 9 Towards a full detector The final HCAL design still needs to be worked out Integration of photo-sensors, electronics, calibration scheme Optimization of granularity –3x3cm 2 : 5 million cells - could be coarser – or finer 1/16 of ½ barrel

10. Felix Sefkow January 29-30, 2007Next HCAL 10 HCAL granularity

11. Felix Sefkow January 29-30, 2007Next HCAL 11 HCAL testbeam prototype  The design is not scalable to a full detector  Front end components not integrated  Electronics not optimized to SiPM signal  Calibration system too complicated  Scintillator layer thickness not minimized  Assembly still quite labour-consuming  A precedence for the electro-mechanical concept of a scintillator calorimeter with integrated photo-sensors does not exist

12. Felix Sefkow January 29-30, 2007Next HCAL 12 Integration issues Tile SiPM optical coupling Tile mechanical positioning SiPM PCB electrical connection PCB size, “stitching” Power pulsing, cooling Power supplies On-detector FE, ASIC control Data concentration, DAQ interface The calibration system g M.Anduze ECAL studies

13. Felix Sefkow January 29-30, 2007Next HCAL 13 SiPMs Critical parameter for SiPM performance is noise above ½MIP threshold –Occupancy < 10 -4 translates into rate < ~ 1kHz Depends on dark rate, inter- pixel crosstalk and light yield; Rule of thumb: N(T) = N dark * (Xtalk) T=LY/2 –E.g. 2 MHz * (0.3) 7 = 400 Hz Light yield determines dynamic range –Minimum 7-8 pixels/MIP Light yield is a property of the SiPM scintillator system SiPM efficiency depends on –Area, geom. pixel density –Spectral sensitivity, Geiger efficiency

14. Felix Sefkow January 29-30, 2007 Next HCAL14 5x5 mm 2 SiPM room temperature SiPM: sensitive area 5.2x4.9 mm 2 Number of pixels 40x40=1600 Active pixel area 100x100  m 2 Period 130x120  m 2 M.Danilov From MEPHI / PULSAR

15. Felix Sefkow January 29-30, 2007 Next HCAL15 The Future of photo-detection Geiger mode silicon photomultiplier Blue sensitive !!! 1 mm SIPD sensitive area CPTA from Photonique, Switzerland Mar. 2005 ● 100/400 pixels Jan. 2006 up to1600 pix Oct. 2006 100/400/1600 pix MPPC from Hamamatsu, Japan http://jp.hamamatsu.com/en/hamamatsu/press/2006/2006_10_26.html Hamamatsu MPPC are on the catalogue! SiPM MEPHI / PULSAR, Moscow SENSL, Irland Is coming…

16. Felix Sefkow January 29-30, 2007Next HCAL 16 Directly coupled MPPCs Light yield not yet optimized –Reflector, coupling Noise and Xtalk are small, but still need 7-8 px/MIP for good MIP eff Minimum dynamic range: see testbeam 400 1600 5mm tiles MPV ~ 4

17. Felix Sefkow January 29-30, 2007Next HCAL 17 MPPC with present FLC_SiPM ASIC Dynamic range and gain OK Other sensors (MPI,..) to be verified

18. Felix Sefkow January 29-30, 2007Next HCAL 18 SiPM scintillator coupling Uniformity to be re-addressed, not as good as with fibre Also studied: 3mm thin tiles Tile 30x30x5 mm 3 SiPM Tile 30x30x5 mm3 E.Tarkovsky Tile with diagonal fiber Tile with arch fiber Standard 5 mm thick tile with arch fiber More efficient SiPMs: simpler coupling, thinner tiles

19. Felix Sefkow January 29-30, 2007Next HCAL 19 EUDET: Calorimeter electronics Electronics –Integration is key –Digital part next to sensitive analogue FE –Power pulsing, stability HaRDROC –64 ch digital HCAL chip –Under test SKIROC –36ch ECAL chip –At foundry (0.35 AMS) SPIROC –36ch analoge (SiPM) HCAL chip –Under design More versions in the pipeline

20. Felix Sefkow January 29-30, 2007 Next HCAL 20 Prospective for A-HCAL SiPM Chip Similar developments for AHCAL Chip fully dedicated to SiPMs developped after ECAL chip Internal DAC for SiPM gain adjustment (5V range) Auto-trigger (fast shaper + Discriminator) Internal TDC, 1 ns step Internal 12 bit ADC Power pulsing T&H x1 Variable gain Preamplifier Discri TDC 12-bit ADC 8 bit DAC (0-5V) in Fast Shaper Shaper t p ~30-40ns Auto-trigger 12-bit DAC Threshold Capacitance for AC coupling … Analogue Memory Charge Ouput Time Ouput C. De La Taille

21. Felix Sefkow January 29-30, 2007Next HCAL 21 EUDET: (calorimeter) DAQ Scalable DAQ system –Commercial hardware where possible –Prototype for full detector and useable in test beam e.g. off-detector receiver: off-the-shelf

22. Felix Sefkow January 29-30, 2007Next HCAL 22 HCAL electronics integration Typical layer 2m 2 2000 tiles 38 layers 80000 tiles FEE: 32 ASICs (LAL) (64-fold) 4 readout lines / layer Layer data Concentrator (control, clock and read FEE) Module data concentrator Instrument one tower (e.m. shower size) + 1 layer (few 1000 tiles) To DAQ EUDET: Mechanical structure, electronics integration: DESY and Hamburg U

23. Felix Sefkow January 29-30, 2007Next HCAL 23 Integration issues Tile SiPM optical coupling Tile mechanical positioning SiPM PCB electrical connection PCB size, “stitching” Power pulsing, cooling Power supplies On-detector backend, ASIC control Data concentration, DAQ interface The calibration system M.Goodrick g M.Anduze ECAL studies

24. Felix Sefkow January 29-30, 2007Next HCAL 24 Temperatures Preliminary analytic calculation (Peter Goettlicher) –25μW/channel from the ASIC with 1% power pulsing –15μW/channel from SiPM dark currents –Additional infrastructure not yet included –Heat conduction in stainless steel Gradient: 0.3 K / 2.2 m –Air gap ASIC – steel plate: max ΔT = 0.2K Time constant 6 d –Monitoring after shutdown, negligible after short break To be verified with realistic model calculation

25. Felix Sefkow January 29-30, 2007Next HCAL 25 HCAL layer 2m x 1m, 2000 tiles 16 front end boards –32 ASICs a 64 channels 1 concentrator board –8 ASICs per FPGA –Few calibration units Real HCAL will not all be powers of 2 ! 8 x 16 tiles 24 x 48 cm 2 x 16 8 ASICs per FPGA Layer concentratorPower supply ( + few calibration units) nsec timing

26. Felix Sefkow January 29-30, 2007Next HCAL 26 Calibration system Challenges met by the present system: –Large dynamic range –PIN readout via SiPM ASICs –Light distribution uniformity –Short, large pulses Experience will tell how to simplify Test stand at Prague J.Cvach

27. Felix Sefkow January 29-30, 2007Next HCAL 27 Much simpler calibration If saturation stable  smaller dynamic range required …  smaller amplitudes, few LEDs for many SiPMs If gain monitoring works (“auto-calibration”)  eliminate high amplification PIN readout for LED monitoring If lower noise of SiPM and pre-amp  relaxed requirements on light distribution uniformity To be seen…

28. Felix Sefkow January 29-30, 2007Next HCAL 28 Scalable prototype roadmap Milestones on the horizon 1.Concept2007 2.Design 2008 3.Construction2009 Input to mechanical concept –Photo-sensor developmentIndustry –Photo-sensor scintillator couplingus started –Electronics to photo-sensor coupling next –Electronics integration; cooling?started –Electronics and DAQ architecturestarted –Calibration conceptparallel Ingredients to overall concept –Testbeam experiencestarted

29. Felix Sefkow January 29-30, 2007Next HCAL 29 Summary  HCAL still needs to be invented  EUDET schedule is a challenge  Concept and schedule largely driven by physics prototype testbeam  Started on schedule - real data are rolling in  Operational experience rapidly accumulating  R&D roadmap taking shape  Practical task sharing between institutes to be worked out