TOF calibration: status Roberto Preghenella Museo Storico della Fisica e Centro Studi e Ricerche ”Enrico Fermi”, Roma INFN, Sezione di Bologna ALICE Offline Week – CERN, 16/03/2010
TOF data flow RAW DATA 01100110011010 01100110011010 AliTOFDecoder AliTOFDecoder time (TDC bins) time (TDC bins) TOT (TDC bins) TOT (TDC bins) deltaBC (BC bins) deltaBC (BC bins) L0L1Latency (BC bins) L0L1Latency (BC bins) AliTOFRawStream PhysicsDA time (TDC bins) time (TDC bins) TOT (TDC bins) TOT (TDC bins) deltaBC (BC bins) deltaBC (BC bins) L0L1Latency (BC bins) L0L1Latency (BC bins) fTOFsignalRaw (ps) fTOFsignal (ps) ClusterFinder PreProcessor TDCLatency TDCLatency Event-time TDCLatency Calibration FEE config ESD DATA Calibration Event-time OCDB Roberto Preghenella 2
TOF calibration scheme Within the TOF calibration scheme all corrections are applied in AliTOFClusterFinder using the input from raw data and OCDB fTOFsignalRaw = TDC tim e fTOFsignal = fTOFsignalRaw – DeltaBC correction + L0L1 latency + TOF CTP latency – TDC latency window – Calibration – Event-Time fill TDC tim → from decoder → from raw data e DeltaBC correction → from decoder → from raw data L0L1 latency → from decoder → from raw data TOF CTP latency → from OCDB TDC latency window → from OCDB → from FEE configuration Calibration → from OCDB → offline calibration Event-Time fill → from OCDB → online/offline measurement Roberto Preghenella 3
TOF calibration scheme Within this calibration scheme TOF measured time recorded in ESD data (fTOFsignal) is not dependent on the FEE configuration Moreover, TOF time stored in ESD data takes into account the average event-time which is measured online on a run-by-run basis In this way, already at pass1 reconstruction, TOF time is on average consistent with the expected Time-Of-Flight t – t exp = 0 The Time-Of-Flight resolution at pass1 reconstruction is then expected to be σ 2 TOF = σ 2 + σ 2 tim e T0 where σ T0 is related to the beam spread in the z-coordinate Roberto Preghenella 4
Timing-calibration with cosmic rays The baseline calibration has been performed with cosmic-ray data collected during summer 2009 (no magnetic field, TOF trigger) The calibration is based on the time-of-flight τ of muons as measured by TOF τ = t – t 0 and the expected time-of-flight τ (exp) assuming ultra-relativistic muons τ (exp) = L / c using about 100k selected muon events for calibration Roberto Preghenella 5
Timing-calibration with pp data An improvement to the cosmic-ray calibration has been performed with pp data collected last December (pass2). pp calibration aims at determining the parameters which could not be measured with cosmic rays because of the limited horizontal-track events collected. The calibration is based on the time-of-flight τ of matched tracks as measured by TOF τ = t – t 0 and the expected time-of-flight τ (exp) assuming pion-ID where t 0 is the average event-time measured run-by-run Roberto Preghenella 6
PID after Timing-calibration σ 2 TOF = σ 2 + σ 2 tim e T0 σ TOF ~ 190 ps Roberto Preghenella 7
Online avegare event-time TOF Pre-Processor runs at the end of each run and builds the channel-integrated raw-time spectrum: only one histogram no limits on range and binning currently 2000 bins in the range [-24.4,24.4] ns ● read DA raw-data ● perform all needed corrections (consistency with reconstruction) ● subtract the expected time-of-flight (straight track from nominal vertex, ultrarelativistic hypotesis) ● fit the spectrum to measure the time-edge Roberto Preghenella 8
Raw-time spectrum Roberto Preghenella 9
Time-edge measurements Landau fit of the distribution close to the expected position of the edge. Use the maximum bin to correctly steer the fit. The edge is defined at the rising point of the distribution on the basis of landau-fit parameters: t edge = t M PV – 3σ landau Roberto Preghenella 10
Time-edge measurements time-edge Roberto Preghenella 11
Time-edge resolution Time-edge resolution follows 1/√N behaviour N hits > 100k σ edge < 50 ps saturation resolution ~10 ps Roberto Preghenella 12
Online average event-time resolution The online T0-fill method measures the average event-time with good resolution: ~25 ps (20 runs with very different statistics) Roberto Preghenella 13
Offline event-by-event event-time A combinatorial algorithm to measure the event- time on a event-by-event basis using ony TOF data has been implemented. The event-time resolution depends on the number of tracks. Roberto Preghenella 14
Conclusions ● TOF calibration scheme has been updated in order to obtain useful time measurements already at pass1 reconstruction ● the baseline cosmic-ray calibration has been refined with pp data. Though, the optimal calibration is not yet achievable ( ~10 M pp events needed) ● online event-time algorithm implemented for run- by-run measurement of average event-time ● offline event-time algorithm (combinatorial algorithm) is available for event-by-event event-time measurement Roberto Preghenella 15
Recommend
More recommend
