First Measurements with the ATLAS Level-1 Calorimeter Trigger PreProcessor System

Transcription

1 First Measurements with the ATLAS Level-1 Calorimeter Trigger PreProcessor System The ATLAS Level-1 Calorimeter Trigger Collaboration R. Achenbach 1, P. Adragna 2, V. Andrei 1, B.M. Barnett 3, B. Bauss 4, M. Bendel 4, C. Bohm 5, J.R.A. Booth 6, I.P. Brawn 3, D.G. Charlton 6, N.J. Collins 6, C.J. Curtis 6, A.O. Davis 3, S. Eckweiler 4, E. Eisenhandler 2, P.J.W. Faulkner 6, J. Fleckner 4, F. Föhlisch 1, C.N.P. Gee 3, C. Geweniger 1, A.R. Gillman 3, P. Hanke 1, S. Hellman 5, A. Hidvégi 5, S.J. Hillier 6, M. Johansen 5, E.-E. Kluge 1, M. Landon 2, V. Lendermann 1, J.N. Lilley 6, K. Mahboubi 1, G. Mahout 6, K. Meier 1, A. Neusiedl 4, V.J.O. Perera 3, D.P.F. Prieur 3, W. Qian 3, S. Rieke 4, F. Rühr 1, D.P.C. Sankey 3, U. Schäfer 4, K. Schmitt 1, H.-C. Schultz-Coulon 1, S. Silverstein 5, R.J. Staley 6, R. Stamen 1, S. Tapprogge 4, J.P. Thomas 6, T. Trefzger 4, P.M. Watkins 6, A. Watson 6, P. Weber 1, E.-E. Woehrling 6 1 Kirchhoff-Institut für Physik, University of Heidelberg, Heidelberg, Germany 2 Physics Department, Queen Mary, University of London, London, UK Victor Andrei Kirchhoff-Institut für Physik, Ruprecht-Karls-Universität Heidelberg TWEPP 2007, Prague, 04/09/ STFC Rutherford Appleton Laboratory, Oxon, UK 4 Institut für Physik, University of Mainz, Mainz, Germany 5 Fysikum, Stockholm University, Stockholm, Sweden 6 School of Physics and Astronomy, University of Birmingham, Birmingham, UK

2 Overview Introduction Level-1 Trigger & Level-1 Calorimeter Trigger PreProcessor System PPM Production Heidelberg Installation Status & Integration Tests & System (in situ in the ATLAS experiment) 2

3 The Level-1 Trigger System UX 15 (experimental cavern) USA 15 (electronics cavern) ~ 70m E E T Muon detector LVL-1 Latency < 2.5 µs BC rate reduction: 40 MHz 75 khz Level-1 Trigger System Level-1 Muon Trigger Level-1 Calorimeter Trigger (L1Calo) L1Calo Latency ~ 1 µs Central Trigger Processor (CTP) L1-Accept signal 3

4 Analogue calorimeter signals (~7200) The L1Calo Trigger: Architecture PreProcessor 124 modules Digitised transverse energy Cluster Processor 56 modules Jet/Energy Processor 32 modules Merging 8 modules Merging 4 modules Hit counts and E T results to CTP Five main types of custom 9U modules PPM CPM JEM Real-time data path Readout data to DAQ Readout Driver 14 modules Region-ofinterest data to Level-2 Readout Driver 6 modules CMM ROD 4

5 The PreProcessor Module (PPM) (1/2) main component of PreProcessor (PPr) System 124 hardware identical PPMs accommodated in 8 crates each PPM receives and processes 64 analogue calorimeter trigger signals processing of data in custom ASIC (developed in Heidelberg) serial transmission of the digital energy values to CP and JEP (realtime path) 16 x 16 x 16 x 16 x AnIn1 AnIn2 AnIn3 AnIn4 4 analogue inputs provides bidirectional VME interface (set control data, local monitoring) and DAQ/ROD interface (monitoring of the system) 16 PreProcessor MCMs PHOS4 4 x FADCS ASIC CAN Module VME CPLD ReM FPGA TTC Dec. 3x LVDS Serialisers LVDS Cable Driver CAN-bus VME DAQ/ROD Readout path Real-time data path to CP & JEP LVDS output 5

6 The PreProcessor Module (PPM) (2/2) 10-bit digitisation (40 MHz) with exact sampling on the signal s peak (25 ns delay, in steps of 1 ns PHOS4) experiment specific algorithms put into ASIC synchronisation of pulses originating from the same event to the same bunch-crossing clock tick BC identification of the E T deposition per trigger channel and of the corresponding bunch-crossing, for all pulses in the linear range & saturated region (FIR Filter, PeakFinder) noise suppression, pedestal subtraction, fine-calibration of extracted E T using lookup table data preparation for transmission to L1Calo processors (jet-element formation, multiplexing) conditioning of the analogue input 10-bit pulse digitisation at 40 MHz (FADC) BC identification & fine-calibration (ASIC-LUT) 6

7 Production Heidelberg (1/2) PPM production is completed 124 modules needed by the full-coverage of the experiment 36 spare modules all modules are tested in Heidelberg before sending to CERN Single Board Tests (bring each module into operation) Initial preparation (optical inspection, power up tests w/o daughterboards, etc) Operational tests (check conditioning & digitisation of the analogue input signal, verify ASIC algorithms, etc) LVDS output Front view PPM analogue signals Rear view ROD Test Card 7

8 Production Heidelberg (2/2) Single Board Tests (cont d) ROD Readout data tests verify the ROD buffer formation & content using both the VME & DAQ readout interfaces Real-time (LVDS) data tests check quality of LVDS data after transmission over 15m long cables LVDS cables in USA15 are 11m long Full Crate Tests ( burn-in of a week or more with 16 PPMs in a standard PPr crate) repeat all functional tests performed during the Single Board Tests 8

9 PPr Installation Status in USA15 so far 2/3 of the PPr system installed will be completed during autumn 2007 installation of analogue cables from receiver stations to the PPr system is completed installation of LVDS cables from PPr system to the L1Calo Processors (CP, JEP) nearly completed 9

10 Tests In Situ in the ATLAS Experiment: PPM Temperature Monitoring monitor the MCM temperature of each PPM mounted in each crate used digital stress patterns to define high-activity in ASIC very dense assembly, air flowing is not completely homogeneous in general temperatures on chip are even in the worst case below 60 C (safe operating condition) temperature stable in time 10

11 Tests In Situ in the ATLAS Experiment: Connectivity (1/4) check signal channel connectivity between the calorimeters (Tile, LAr) and L1Calo Trigger System automated tool to test a large number of channels in parallel using a small number of checking steps TTC UX 15 USA 15 Tile CIS, LAr Pulser System L1 Trigger VME & Ethernet 11

12 Tests In Situ in the ATLAS Experiment: Connectivity (2/4) Low amplitude signal Channel not fired pulse height Good channel (central part) Good channel (outer part) 12

13 Tests In Situ in the ATLAS Experiment: Connectivity (3/4) Typical minor problems seen in tests Cables 3 & 4 were found to be interchanged at the output of the receiver system Input cables on one PPM Cable 4 Cable 3 Cable 2 Cable 1 Dead channels upstream of PPM (to be repaired) Power supply off (large region) pulse height 13

14 Tests In Situ in the ATLAS Experiment: Connectivity (4/4) C3 C4 Mistake was corrected by swapping the cables at PPM input C3 C4 BEFORE AFTER 14

15 More Tests In Situ in the ATLAS Experiment Digital connectivity LVDS cable transmission (400 MBit/s) from PPMs to JEP tested for half of the PPr system no error observed Signal quality tests at different energies (including saturation), with different patterns from Tile & LAr calorimeters to be done as soon as the integration of the system is completed Participation to ATLAS System Integration Run (M4) 4/8 Ppr crates were integrated in the run 15

16 Conclusions PPM production is completed (160 modules) More than half of the PPr system is already installed in USA15 to be completed during autumn 2007 Installation of analogue cables is completed and installation of LVDS cables is nearly completed Connectivity tests: ok, so far only done for large parts of the barrel region, no serious problem seen More tests are foreseen during and after the installation of the system 16