Commissioning of the ALICE HLT TPC and PHOS

Commissioning of the ALICE HLT, TPC and PHOS systems Kalliopi Kanaki University of Bergen 1

Norwegian participation in: § § § High Level Trigger (HLT) Time Projection Chamber (TPC) PHOton Spectrometer (PHOS) Online calibration framework Front-End Electronics (FEE) for TPC/PHOS ALICE GRID computing 2

Aims of HLT § Lossless data compression to match data acquisition and mass storage bandwidth § Online event reconstruction → fast pattern recognition (online tracking) § Physics triggers (e. g. open charm, J/ψ, jets, dielectrons) → event characterization § Rates of 1 k. Hz for p+p and 200 Hz for Pb+Pb (central) 3

HLT Processing Data Flow DAQ: DAta Acquisition ECS: Experiment Control System DCS: Detector Control System 4

The TPC Detector 845 < r < 2466 mm Drift length 2× 2500 mm 557568 readout pads Ongoing Norwegian contribution: § Online cluster finding & tracking § Online krypton gas gain calibration 5

Online HLT Tracking – Dataflow Cluster Cluster Finder Finder 36 Sectors Global Merging Track Finder/Fitter further analysis… 6

Online Cluster Finder § § ADC sequences above threshold as input Simple sequence matching between neighbouring pads Centroids calculated as weighted mean of ADC values Assigned 3 D coordinates Deconvolution scheme: split clusters at local minima of charge distributions along time and pad direction 7

Online Tracking Algorithm real space § 3 D space points as input § Conformal mapping conformal space § “Follow-your-nose”: build tracks from outer to inner TPC-radius 8

HLT Cluster Status § Full connectivity / all detectors included § Additional CPU power is expected 9

Benchmarks on HLT cluster preparation for p+p runs § Setup for 1 sector § 6 Cluster. Finders running in parallel on 6 CPUs § 1 Tracker running on 1 CPU § Results (with current configuration): § Cluster. Finder: 240 Hz § Tracker: ~200 Hz (design value 1 k. Hz) § Significant improvement expected with additional CPU power 10

Event Display for Primaries HMPID TPC PHOS offline HLT 11

HLT and Offline Efficiency for Primary Particles Overall Efficiency: § HLT: 95% § Offline: 94% § HLT and offline analysis deliver comparable results § Running HLT code in Ali. Root framework → comparison with full generator events (Pythia) 12

Krypton Gain Calibration § Kr decay products (e-, γ) deposit known energy § Dedicated cluster finder because of larger clusters § Energy calibration possible 41. 6 ke. V 29 ke. V 9. 4 ke. V 12. 6 ke. V 19. 6 ke. V 14

The PHOS Detector § Information about the initial phase of the collision § High granularity electromagnetic calorimeter split in 5 modules § 17920 detection channels of leadtungstate crystals (Pb. WO 4) § Optimized for measurements of γ (0. 5 - 10 Ge. V/c), π0 (1 - 10 Ge. V/c) and η mesons (2 - 10 Ge. V/c) 15

PHOS - HLT Operation Principle PHOS DAQ 2 channels per crystal extract amplitude & timing information Shower & μ reconstruction HLT for PHOS Online monitoring & calibration Offline condition database MIPS deposit ~210 Me. V in one crystal length → peak in the energy distribution 16

Online Calibration Framework Motivation: § Several ALICE systems (DCS, DAQ, ECS, Offline) provide data, settings and tools required for producing and monitoring calibration data § System interfaces for data exchange § Provide similar environment to online Analysis Software / Detector Algorithms (DA) as in offline 17

DCS (Detector Control System) ECS DCS values (Experiment Control System) Control DAQ Calculated values ECSproxy DCS-portal (Pendolino, FED) DDL Processed (Data Acquisition) events Trigger decisions Online event monitoring Ali. Eve HLT Event data FEP FEE (Front-End. Electronics) DAQ (Data Acquisition) cluster (ALICE Event Monitor) HOMER OFFLINE-portal Calibration data (Shuttle, Taxi) New calculated calibration data OFFLINE OCDB (Offline Conditions DB) Shuttle FEP = Front-End-Processor, DDL = Direct Data Link HOMER = HLT Online Monitoring Environment including ROOT 18

Summary § § § TPC/1 st PHOS module operational, cosmics run PHOS module calibrated with cosmics HLT for TPC and PHOS operational for first physics runs Similar efficiencies for primaries in offline and HLT Kr gas gain calibration for TPC implemented Most online calibration interfaces implemented and tested & Outlook. . . § § TPC/PHOS are being connected/commissioned Integrated test with lasers/cosmics in December Kr calibration test in February On the way to p+p… 19

Backup slides 20

Calibration Input & Output Input Output DCS SCADA Ali. En Taxi (Grid) Taxi-portal HLT cluster Front End Devise FED -portal API Calibration settings OCDB (Offline) DA Pendolino Amanda DCS Archive DB Pendolinoportal Data preparation current DCS conditions Calibration& Conditions OFFLINE OCDB Shuttleportal Meta Data DB Shuttle FXS Data Subscriber portal Monitoring (Ali. Eve) Homer 21

PHOS Commissioning § PHOS cosmics run § Calibration of crystal gains using cosmic muons § MIPS deposit ~210 Me. V in one crystal length → peak in the energy distribution § Equalization of electronic gains § Absolute energy calibration § HLT for PHOS § Data compression § Data integrity checks § Event display § Histogram plotting 22