July 15-17, 2008ALICE EMCal Meeting, Nantes 2 APD: Ordered US, 1300 Catania (Hamamatsu) –Quoted Delivery (as of June 3): Strip Module Internal Components: Status HPF/CERN (total 5K pcs) HPI/INFN (total 1.3K pcs) July450 pcs on 7/14450 pcs on 7/7 300 pcs on 7/28300 pcs on 7/22 August300 pcs on 8/25300 pcs on 8/6 September150 pcs on 9/22250 pcs on 9/ pcs--> Checking
July 15-17, 2008ALICE EMCal Meeting, Nantes 3 Preamplifier: Ordered US+EU (Gnomes Design, Japan - PHOS supplier) –Quoted Delivery (as of July 1): Strip Module Internal Components: Status Terry さんから納期短縮のご依頼 理解しました。 現在 部品の納期確認作業を行っております。 残念ながら 部品調達には 時間がかかります。 今回の基板は ハロゲンフリー基板を予定しています。 ハロゲンフリー基板の特性上 基板製造、組立作業を 全部品の入手後 今年度分を一括生産する予定です。 現時点では 最速の納品でも 8 月中旬以降の予定です。 ご要望ん従い 出来るだけ早く納品したいと思います。 もうしばらく お待ち下さい。 納品可能な日時分かりましたらご連絡いたします。 Mr. Terry. The delivery date confirmation work of parts is being done now. To our regret, it takes time to the part procurement. This substrate schedules a halogen free substrate. After it obtains it of entire goods, current year is scheduled to be produced in the lump the substrate manufacturing on the characteristic and the assembling work of a halogen free substrate. The delivery of goods of the velocity is a schedule now after the middle of August. I want to follow and to deliver demand as soon as possible. Please wait for a while. When the date that can be delivered is understood, I will report.
July 15-17, 2008ALICE EMCal Meeting, Nantes 4 Transition Cards: –PCBs Ordered US+EU (144 for 2 SM) Quoted Delivery July 4 Components ordered, most received Molex cables –Ordered 7000 (2304 for 2 SM) Expected delivery late September - Long story –Initially given 2 week delivery –Delay to receive quote due to PHOS non-payment –Delay due to non-compliance with CERN standards (Halogen) waiting for quote with compliant halogen- free cables –Quoted delivery now 12 weeks - due to materials? Strip Module Internal Components: Status
July 15-17, 2008ALICE EMCal Meeting, Nantes 5 Electronics: Testing/Calibration Test APD+preamplifier (Houston, Catania) –Initial APD gain vs Voltage measurements, QA Test FEE cards after production (CERN) –Confirm all channels have proper FADC outputs –Confirm APD bias control works properly –Measure Hi/Low amplifier channel gains Calibrate Super Modules (Yale, Grenoble) –Measure APD voltage-gains for each channel –Obtain pre-calibration with cosmics –Determine APD operating voltage for each channel No special test fixtures will be built. Instead we will use EMCal FEE setups as test station(s). Test stations at: –Houston, Catania (exists) for APD+preamplifier tests –CERN (exists) for FEE/TRU/crate tests –Yale, Grenoble for SM test and calibrations
July 15-17, 2008ALICE EMCal Meeting, Nantes 6 APD Test Setup Houston setup: –LED Driver, FEE Card, T-Cards, FEE Cables, GTL bus, RCU, DCS, SIU, DRORC, PC, LV cables, fiber optic, ethernet+DAQ tested at CERN, now at Houston –Chiller ordered, mechanical pieces under construction Completely implemented already at Catania Need to design/build APD lightguide gluing fixtures.
July 15-17, 2008ALICE EMCal Meeting, Nantes 7 SM Calibration Setup System components (Yale, Grenoble) –DAQ + 2 DRORC cards (for readout of up to 4 RCUs) –Mini-Crate with FEE+GTL bus + RCU + DCS + SIU + Trigger card + cables (LV, Fiber, ethernet) for testing –For SM: LV for RCU+FEE, DDL fiber, Signal cable+connectors (to be made for 1SM each), Temperature sensors + cables (to be made 3SM each) –ELMB system with distribution blocks for 8 T measurements in assembly area + 8 T measurements on SM –LV supply for full SM (for Yale, September for Grenoble) Full system setup and tested at Grenoble Ready to ship to Yale (confirm RCU working)
July 15-17, 2008ALICE EMCal Meeting, Nantes 8 EMCal SM Readout ~3.5m 3 2 FEE Crates 3 Shielded Ribbon Cables per Strip Module (round!) - APD bias, preamp LV and signal. Ground shielding at APD and/or FEE end? One 3mm Fiber per Strip Module for LED monitoring (WSU) –1m (7x1mm) fiber extension to 3mm fiber - Catania proposal - Yes! Fabricate at WSU? –One monitoring fiber per 3mm fiber Readout via photodiode+preamp+T-card - 2/SM Mount fixture design? Location? 8 Temperature Sensors per SM - Where? –T sensor cable to ribbon cable block Between FEE crates
July 15-17, 2008ALICE EMCal Meeting, Nantes 9 EMCal FEE Crate From Didier Gabrielle
July 15-17, 2008ALICE EMCal Meeting, Nantes 10 EMCal FEE Installation in ALICE Readout electronics of a SuperModule of 24x48 towers sits in a water-cooled enclosure attached to the SM. Attached during SM assembly and used for calibration. Two FEE crates (2 RCU) each with 2 GTL branches with 9 FEE cards. A group of 12 FEE cards will be connected to a TRU to form a Trigger Region. Readout electronics/LED system
July 15-17, 2008ALICE EMCal Meeting, Nantes 11 EMCal FEE Crate Status Design complete All commercial card guides and supports delivered Prototype crate built at CERN including cooling manifold –Cooling manifold tested with PHOS negative pressure cooling water system No leaks! All copper panels uniformly at 18 o C (temperature of cooling water) with ambient T=27 o C After heating copper panel locally to 60 o C it returned to T=18 o C in 90 seconds. Plan –Build cooling manifolds for all 22 FEE crates at CERN –Make parts for 4 FEE crates (2 SMs) at CERN –Make parts for other SMs at ?
July 15-17, 2008ALICE EMCal Meeting, Nantes 12 EMCal FEE Crate Status Intermediate PCB - backplane of FEE crate with connectors to ribbon cables from SM; ordered for all SM –4 versions: upper & lower, 9 and 10 FEE connections –PCBs ordered, delivery any day now –Connectors ordered, partially delivered GTL bus - (delay due to check & modification of Serial I/O lines) –PCBs quotes being received - order in next days –Components ordered, partially delivered RCU boards - Need 4 for 2 SMs –Not clear if will be available order for 80 new RCUs has not yet gone out DCS boards - No problem to get 4 for 2 SMs, –May be a problem to get 22 for EMCal SIU boards - No problem FEE LV bus in crate: cable+connectors purchased (all SM) and made for 2 SM (Cal Poly, San Luis Obispo)
July 15-17, 2008ALICE EMCal Meeting, Nantes 13 FEE Card Status FEE Cards - for all SMs (396 needed) –PCBs quotes from HUST and ELTOS (lowest) waiting for third quote, then place order –“Reject” cheap ALTRO chips tested at LUND, 599 ALTROs for EMCal (150 FEE boards) –GTL chips already purchased –FEE component price search underway by CERN Maybe a mistake?! –Plan to have FEE cards complete by late September Would like them sooner of course…
July 15-17, 2008ALICE EMCal Meeting, Nantes 14 LED Monitoring System Status LED system criteria: –Prototype system performed well but too bulky. Need 24 LEDs per SM issues: double-pulses, reproducibility –Ideally, should vary LED delays to match LED pulse timing to physics signals –Amplitude variability would be a plus, but not at the price of stability
July 15-17, 2008ALICE EMCal Meeting, Nantes 15 LED Monitoring System Status Catania solution: Use essentially the same prototypes packed into a high density copper box with one fan-out and 8 LED drivers per box. –Prototype built and tested Meets space restrictions to fit 3 boards into gap between 2 FEE crates Good performance, as expected, (no double pulsing…) Good electrical isolation - probably cannot do better Possibility to add additional cable delay of (n-1) ns * 20cm/ns to each channel within box. –Two additional driver boxes being constructed from existing drivers (for Grenoble, Yale) to be ready to use when SM assemby and testing begins
July 15-17, 2008ALICE EMCal Meeting, Nantes 16 LED Monitoring System Status New LED driver solutions: –Solution A: Ambitious: Build new 8 or 12 LED channel boards with amplitude and gain control via DAC DAC control via modified FEE for LED readout to use FPGA on FEE to control LED boards Need to prototype - might expect problems Need to modify FEEs and develop FPGA code for control –Solution B: Better: Build new 8 or 12 LED channel boards with exactly same functionality of prototypes Cheaper to build than individual drivers, less mechanical work, more elegant Some risk of problems due to increased cross-talk Possible to add additional cable delay of (n-1) ns * 20cm/ns to each channel?
July 15-17, 2008ALICE EMCal Meeting, Nantes 17 LED Monitoring System Status Need to design/build new RCU trigger boards –“Existing” RCU trigger boards have only trigger input to trigger DAQ externally via the RCU. Basically a NIM to TTL convertor to give input signal of proper amplitude and width to RCU FPGA input. Used in beam test and systems being shipped to Houston,Yale,etc…. –In ALICE, the LED system will be triggered by Pre-Pulse trigger, only available via TTC, I.e. available in RCU Use RCU FPGA output to trigger LED drivers Positive output only - not NIM signal - –Need to modify LED trigger fanout boards –Need new boards at time of SM installation Design will be started when Valencia student arrives (September)
July 15-17, 2008ALICE EMCal Meeting, Nantes 18 LED System: Other Issues Decision to build 1m long 7x1mm extensions to 3mm fibers –Yes, at WSU Location of fibers for monitoring LED output –At Strip module distribution on back of SM (default) –Or at coupling of extension to 3mm fiber near FEE crate Tests underway by Francesco Readout of monitoring fibers: 24 LEDs using 1 FEE card. –Photodiode+preamp+molex+T-card+ribbon to FEE –Either on back of SM, or at FEE crate (see above) If on SM: must be <~3cm thick for clearance during SM insertion –Need to design T-card+preamp+APD+fiber fixture T-card? –Default: Use normal EMCal T-card (12 of 16 channels used) –New T-card: Make a new 12-channel T-card as compact as possible
July 15-17, 2008ALICE EMCal Meeting, Nantes 19 EMCal Trigger Status See talks by Hans and Olivier Summary: Production of prototype TRU end of ~July Production of prototype STU end of ~July Prototype tests August-September Production in Fall Try to have TRUs installed at SM assembly sites for first SMs –If not, then should be installed in FEE crate prior to installation on CALFrame.
July 15-17, 2008ALICE EMCal Meeting, Nantes 20 Electronics Services EMCal HV and LV –Strasbourg purchases HV crate + 4 modules (1 module per EMCal “quadrant” –ISEG Crate + 1 module purchased for CERN beam tests –2 more modules purchased LV Wiener –One 0-15V 12-channel for EMCal purchased –One 0-8V 6-channel per quadrant - 2 purchased –One 0-8V 12-channel for SM assembly at Yale and Grenoble - 2 purchased
July 15-17, 2008ALICE EMCal Meeting, Nantes 21 Other work during access: Services Installation of LV and HV supplies Busybox Cables –HV cable connectors + distribution panels –Ethernet connectors + patch panel –Change TRU LV cables (probably) –LV and HV cables for LED Drivers –Cables to BusyBox
July 15-17, 2008ALICE EMCal Meeting, Nantes 22 Extras
July 15-17, 2008ALICE EMCal Meeting, Nantes 23 FEE / Trigger / DAQ Overview 9 FEE + 1 Trigger Region Unit (TRU) setup/readout via GTL bus. Readout Control Unit (RCU) controls FEE on up to 2 GTL bus branches. Detector Control System (DCS) RCU daughter card for FEE setup (e.g. APD bias) Data to DAQ via Detector Data Link on RCU - passed to High Level Trigger. Same ReadOut for TPC/PHOS/EMCal
July 15-17, 2008ALICE EMCal Meeting, Nantes 24 Trigger / DAQ Connectivity PHOS Trigger Region Unit (TRU) provides one L0 and 3 L1 triggers on shower energy (4x4 towers) - and 0 trigger. L0 and L1 triggers must be ORed before input to CTP. An L1 EMCal jet trigger will require a new post-TRU trigger module - Summary Trigger Unit (STU). PHOS Trigger OR TOR-> EMCal STU PHOS Trigger OR TOR-> EMCal STU
July 15-17, 2008ALICE EMCal Meeting, Nantes 25 EMCal LED Gain Monitoring System LED system needed for gain adjustment and gain monitoring. One fiber per module ( shown ) excites WLS bundle - low efficiency. 12 modules (fibers) per strip module fed by one 3mm fiber from remote LED to strip module. Prototype LED driver used in test beam.
July 15-17, 2008ALICE EMCal Meeting, Nantes 26 STU specification : L0 trigger : OR of the 32 L0 calculated by the TRU. L1-gamma trigger: Energy summed over sliding window of 4x4 towers (2x2 fast OR) and compared to a multiplicity (V0 det) dependent threshold. L1-jet trigger : Energy summed over a sliding window of n*n subregions and compared to a multiplicity dependent (V0) threshold. STU TRU L0 L1-gamma L1-jet 32 TRU Multiplicity from V0 4 diff pair Ethernet CAT7 cable TRU is clocked by the BC clock forwarded by STU (40.08MHz) TTC link To CTP DDL DCS Put trigger data in the data stream on L2a (via DDL) LVDS links Summary Trigger Unit (STU)
July 15-17, 2008ALICE EMCal Meeting, Nantes 27 STU Design : Closely follows Trigger OR (TOR) design (shown). TOR used by Grenoble in “LoopBack” test to demonstrate LVDS serial data link feasibility. Main differences with TOR –TTC for better clock signal for serial clock –DDL link to insert trigger data into data stream –Virtex5 firmware Plan to have TOR as STU as backup, and STU as TOR upgrade for PHOS. Summary Trigger Unit (STU)
July 15-17, 2008ALICE EMCal Meeting, Nantes 28 Jet Trigger New EMCal development (CERN, Jyvaskyla, Grenoble) –Upgrade of PHOS TRU FPGA from VIRTEX2 to VIRTEX5: programming, prototypes, tests. (MIE, CERN, Jyvaskyla ) Backup plan: PHOS TRU could be used for EMCal, but upgrade to VIRTEX5 will simplify implementation and reduce production cost. –TRU upgrade layout completed (CERN, Jyvaskyla) –Design, build, program, and test Summary Trigger Unit (STU) module. –Layout of STU just begun –TRU and STU prototype tests expected for May/June