1. AHCAL – Electromechanical Integration EUDET annual meeting – Paris8.-10. Oct. 2007 M. Reinecke

2. Outline Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 -AHCAL - Mechanical and Electrical Integration -Light Calibration System – First Results -Integration Studies

3. AHCAL Half Sector - Integration Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 AHCAL Slab 6 HBUs in a row HBU HCAL Base Unit typ. 12 x 12 tiles SPIROC typ. 4 on a HBU HEB HCAL Endcap Board Hosts mezzanine modules: DIF, CALIB and POWER HLD HCAL Layer Distributor

4. Mechanical Constraints Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 HCAL: 2 x 8 Sectors 2,432,000 Tiles Requirements for a HCAL Base-Unit (HBU) from the Barrel‘s mechanics: -As large as possible (assembly time) -As thin as possible (barrel diameter) -Easy de-/installation of single units (repair) -Rail System needed (Sector walls ?) -Minimize dead area

5. HBU – Cross Section Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 Sector wall Reflector Foil 100µm Polyimide Foil 100µm PCB 800µm Bolt with inner M3 thread welded to bottom plate SiPM Tile 3mm HBU Interface 500µm gap Bottom Plate 600µm ASIC TQFP-100 1mm high UV LED 500µm Component Area: 900µm high HBU height: 6.1mm (4.9mm without covers => absorber) Absorber Plates (steel) Spacer 1.7mm Top Plate fixing

6. HBU – PCB Layer Structure Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 -6 layer design with cut-outs for ASICS and connectors -75  Lines for high-gain SiPM setup -Two signal layers for impedance-controlled routing -Total height (PCB + components): 1.5mm -Two companies agreed on structure at reasonable costs!!

7. Ligth Calibration System (LCS) Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 SiPM response strongly depends on temperature and bias voltage. LCS (based on UV LEDs) needed for: -Calibration (ADC counts per PE) -Gain Monitoring Two different concepts under investigation: - Quasi-Resonant LED driver setup on DIF, fibers into AHCAL gaps (see: our Prague colleagues, I. Polak et al.) - One LED per tile, direct coupling without fibers (currently tested at DESY)

8. LED Testboard (LCS Test) Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 Test LED integration into HBU (LCS): -Crosstalk of driving circuit to SiPM? -Integration to PCB / coupling to tile? -Connector test: stability, number of connection-cycles?

9. LED Integration Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 LEDs radiate through holes in PCB Two different types of LEDs assembled on PCB (top) 1 Cent

10. LED optical output Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 Pulse Generator Output (inv), 2.5V pre-bias of LED LED optical output, measured with PMT H9858-01 (Hamamatsu) in 50 , Vc=0.8V => Gain ≈ 3*10^(4) A/W Our LEDs are very fast !!! 30nW peak 5ns

11. LED driving circuit on HBU Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 LED: Ledtronics SML0603-395-TR Transistor: Infineon BFR340F (npn) Charge on C drive defines current through LED

12. Response from Tiles Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 SiPM 1 SiPM 2 SiPM 4 Trigger WLS decay time? SiPMs coupled to oscilloscope (50  ), LED1 amplitude controlled by VCALIB(DC). Only LED1 active 30ns

13. Estimate Crosstalk Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 Scope in averaging mode Crosstalk ≈ 2.5% electr./optical ?? Light in Tile 1 Crosstalk to tile 2 and tile 4 ≈40mV ≈1mV

14. Preliminary LCS Status Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 GOOD Critical - The LEDs can be assembled automatically without problems. All of the 5 tested LEDs survived the assembling. -The LEDs are very fast. -The driving circuit works concerning speed and amplitude. The amplitude can be controlled in wide range by VCALIB (6-9V) -The driving voltage (base of transistor) has to be large (1V step). -The sensitivity to changes of the driving amplitude is high. Next steps -Connect the LED testboard to the ASIC‘s testboard (VFE ASIC) in order to measure dynamic range, crosstalk and linearity w.r. to VCALIB and LED uniformity.

15. Testboard II : ASIC + Integration Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 SPIROC Testboard (HBU prototype): -Assembly (Tiles, PCB, ASICs, LEDs), Cassette Construction -Performance in the dense HBU setup: Noise, gain, crosstalk, power and signal integrity -DAQ Interface -LCS with LEDs on board. Tile integration to HBU : see M. Danilov‘s talk (alignment pins)

16. Testboard III : Power-System Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 Test Power-Ground System (2.20m): -Oscillations when switching? -Voltage drop, signal integrity (traces, connectors)? -SPIROC performance @ far end (blocking caps sufficient)? ‚Layer Concentrator‘ (DIF*) SPIROC Testboard Extenders (Power-Gnd, Traces)

17. Conclusions Mathias ReineckeEUDET annual meeting – Paris8.-10. Oct. 2007 -First ideas about the next generation AHCAL develop to a promising concept. -Feasibility of many design aspects (e.g. PCB structure) have to be proved. -Testboard Design I (LCS) is alive now! -Testboard II (HBU prototype) design starts in spring 2008. -Testboard III (power plane test) runs in parallel (beginning of 2008). -Mechanical engineering of absorber stack and HBU cassette is starting!