Performance of the Resistive Plate Chambers as LVL1

1
Performance of the Resistive Plate Chambers as
LVL1 ATLAS muon trigger
Michele Bianco INFN Lecce Physics Department,
Salento University on behalf the ATLAS Muon
Comunity
2
Outline

• The RPC and LVL1 muon trigger in the ATLAS barrel
  region
• Cosmics data analysis and results for RPC
  detector
• LVL1 trigger timing and performances
• DCS and monitoring software status
• Conclusions

3
The ATLAS Muon Trigger in barrel region
Resistive Plate Chambers (RPC) will be used as
Muon Trigger Detector in the barrel region (-1 lt
h lt 1)

• More than 1100 RPC units
• 368.416 Read-out channels
• 26 different chambers type
• Total surface 4000 m2

Muon Trigger Segmentation in Barrel region 16
Physical Sectors (Large and Small) 64 Trigger
Sectors 396 Trigger Towers
4
The ATLAS Muon Spectrometer
Muon Chamber during the installation
RPC Chamber
MDT Chamber
5
The ATLAS Resistive Plate Chambers
Gaseous detector, operated at atmospheric
pressure ATLAS RPC works in saturated avalanche
regime Gas mixture C2H2F4 94.7 - C4H10 5 -
SF6 0.3
Each unit contains 2 layers of gas volume. 2mm
gas gap, bakelite resistivity 1-4x1010 ?cm h
and f read-out copper strips panels, pitch
ranging from 26.4 to 37 mm

• Main ATLAS RPC tasks
• Good time resolution for bunch-crossing
  identification ( 1 ns).
• High rate capability to sustain the high
  background level.
• Provide the 2nd-coordinate measurement with a
  8-10mm resolution

6
ATLAS Muon LVL1 trigger strategy

• The COINCIDENCE WONDOWS depends on
• Trigger pT threshould
• ? coorinates
• Muon spectrometer layout

At each strip on pivot plane are associted
COINCIDENCE WONDOWS on HighPT and LowPt planes
RPC3 High-pT
RPC2
RPC2 Pivot
RPC1 Low-pT
Coincidence window
7
ATLAS Muon LVL1 trigger strategy
Muon selection mechanism is based on the allowed
geometrical road (Coincidence Windows)

• Two threshold regimes
• Low-Pt muon trigger (6ltptlt10 GeV) majority 3/4
• High-Pt muon trigger (gt10 GeV) majority 1/2
  Low-Pt

• Low Pt and High Pt trigger are separate but not
  independent.
• Low Pt trigger result is needed for the High Pt
  decision.
• The timing between Low Pt and High Pt has to be
  adjusted depending on the physics (cosmics or
  beam)
• The High Pt PAD routes data out to trigger and
  readout

8
Trigger segmentation

• Organized in 64 trigger sector 32 Side A 32
  Side C
• An Atlas geometrical sector correspond to 4
  trigger sectors
• Each trigger sector contains 6-7 trigger tower
• 1 Trigger Tower 1 Low Pt PAD 1 High Pt PAD
• Each PAD contain 2 ?-CM and 2 f CM
• The overlap of an ?-CM with a f-CM correspond to
  a RoI

9
RPC Detector Analysis Strategy

• In order to ensure redundancy/robustness, a
  twofold strategy are used for RPC detector
  studies
• Exploiting the precise tracking from the MDTs
• Advantage
• extrapolation to RPC layers takes into account
  materials and magnetic field
• precise extrapolation allows to determine
  spatial resolution and to study small local
  effects
• Disadvantage
• applicable only to runs with MDTs on
• presently all RPC hits are used in reco, hence
  a bias is introduced in efficiency measurement
  (will be fixed)
• Using standalone tracking (only RPC)
• Advantage
• Does not depend on MDTs
• Dedicated tracking algo avoids reconstruction
  bias on efficiency (by not using hits of a given
  layer)
• Automatic run at Tier0 facility
• Disadvantage
• Extrapolation precision limited by RPC
  granularity

10
Tracking with MDT, Quality Cuts

• Event selection and track quality
• Events with only 1 track
• c2/d.o.f. lt 20
• At least 2 f hits on track

11
RPC efficiency with MDT tracks
Efficiency distribution HV 9600 V, Vth 1000 mV

• Low Panel Efficiency related to HV channel off
• Efficiency not corrected for dead strips.

ATLAS Preliminary
HV 9600 V Vth 1000 mV
Efficiency vs sector
ATLAS Preliminary

• Cluster size for h and f panels
• h view cluster size is a little bit lower wrt f
  view. This is as expected, due to difference in
  detector costruction

12
RPC StandAlone Track Quality

• Pattern recognition seeded by a straight line,
  which is defined by two RPC space points.
• RPC space points not part of any previous tracks
  and inside a predefined distance from the
  straight line are associated to the pattern.
• From cosmic data about 95 percent of events
  have at least one RPC track.
• Applying a quality cut of chi2/dof lt 1 about 70
  of events have at least a good tracks and 10
  with more than one.
• The detection efficiency is measured by repeating
  6 times the RPC tracking.
• The layer under test is removed from the pattern
  recognition and track fitting.

13
RPC StandAlone Track Quality

• 70 Events with at least a track after cuts on
  c2/d.o.f. lt 1
• Efficiency is measured by repeating 6 times the
  RPC tracking.
• Monitoring of Time Tracks Residual, any cuts
  applied on time residual up to now

14
RPC StandAlone Tracking Results
RPC Efficiency measured for all strips panels,
with the RPC standalone tracking dead strips not
removed.
HV 9600 Volts, Vth 1000 mV. Average
Efficiency 91.5 Fitted Efficiency 94.4
ATLAS Preliminary
RPC panel noise distribution measured for all
strips panels, with the RPC standalone package,
ATLAS Preliminary
HV 9600 Volts, Vth 1000 mV.
15
Other Off-line StandAlone Monitoring Results
Rocks concrete layers
ATLAS
Cosmics muon map reconstructed by Off-line RPC
standalone muon monitoring extrapolated to
surface . The tracking is based only on RPC space
points, which are defined by orthogonal RPC
cluster hits. Main shafts and elevator shafts are
clearly visible.
16
RPC trigger coverage status
Trigger Coverage gt 97 5/396 Trigger towers
with readout problems Few other holes due to HV
problems (recoverable changing trigger majority)
17
LVL1 trigger timing and performances

• A correct timing-in means that we will trigger
  the µ, with the desired Pt, emerging from the IP
  at given BC and we will stamp it with the correct
  BC ID.
• The timing-in of the trigger requires to correct
  for
• The delay due to the propagation along cables,
  fibers and to the latencies of the different
  elements.
• The Time of Flight, i.e. the physics to select,
  needs to know the physical configurations
  (cosmic, beam).
• The strip propagation is relevant for the trigger
  time spread ( max 12ns ), read out cable were
  optimized to reduce this spread.
• All these delays have to be corrected in the
  pipelines of different element.

For a good detector timing is necessary to
ensure the correct alignment of ? Layers within
the same CM ? Views (f CM - ? CM) within the
same PAD ? Towers (PADs) within the same Trigger
sector ? Trigger Sectors wrt each other
Local alignment
Global alignment
18
Time alignment inside LowPt trigger tower
Distribution of the relative time between RPC
layers of Low Pt non-bending view coincidence
matrix delivering one and only one hardware
trigger in the event.

• Time alignment inside Low Pt trigger towers in
  phi view with cosmic data.
• Entries are not a track time residuals.
• The time is relative to the layer nearest to the
  IP.
• HV 9600 V, Vth 1000 mV

19
Time alignment inside Sector Logic and between
Sector

• The misalignment between trigger sectors is the
  combination of the delay and time of flight.
• With cosmics is very difficult to disentangle the
  2 components using only RPC.
• The best way to check it is to use only pointing
  tracks (known time of flight) and look at
  relative alignment.
• Dedicated runs were taken using Transition
  Radiation Tracker (TRT) as source of external
  trigger (its small radius allow to select
  pointing tracks easily).
• Misalignment between trigger tower inside same
  Trigger Sector and misalignment between
  different Trigger Sector have been significantly
  reduced via an iterative procedure.

Trigger Time read-out for, each trigger tower,
along RPC trigger sectors.
RPC trigger distribution wrt TRT trigger signal.
20
Trigger Road Analysis
RPC spatial correlation between trigger strip
(Pivot) and confirm strip (LowPt) in phi view
for a programmed trigger road in cosmics data. It
is possible to see the trigger road projective
pattern by the deviation of the data points from
the dashed line. Strip number 0 corresponds to
the center of the geometrical sector.
Along the phi view (non bending view), trigger
road are used to reduce the background, requiring
pointing tracks.
21
Detector Control System ( DCS ) overview

• DCS system
• Controlling the detector power system (chamber
  HV, frontend LV)
• Configuring and/or Monitoring the frontend
  electronics
• Reading/Recording non event-based environmental
  and conditions data
• Adjusting operations parameters to ensure
  efficient detector operation
• Controlling which actions are allowed under what
  conditions to prevent configurations potentially
  harmful for the detector
• Hierarchical approach
• Separation of frontend (process) and supervisory
  layer
• Commercial SCADA System
  CERN JCOP Framework
• Muon specific developments, Scalable,
  Distributed
• Performance monitoring
• Monitoring and historical trend for all monitored
  quantities.
• Data Quality Assessment automatically generate
  and transferred in a dedicated Data Base.

22
DCS overview

• Overview of the whole detector via FSM PS, Gas,
  Env. Sensors, DQ.
• Alarms and watchdogs (safety scripts) for
  unattended operation Mainframe connections, HV-
  GAS Igap
  currents.
• Global Switch ON/OFF via FSM command for LV
  system.

Advanced shifter and expert operations
interfaces

• Gas channels, Stations status.

• LVL1 crates.

• DQA Monitoring.

23
Off-line Monitoring at Tier0

• A software package to debug, monitor, and asses
  data quality for the RPC detector, has been
  developed within the ATLAS software framework.
• Run by run, all relevant quantities
  characterizing the RPC detector are measured and
  stored in a dedicate database.
• These quantities are used for MonteCarlo
  simulations and off-line reconstruction by
  physics analysis groups.
• The code was developed using C objet oriented
  framework and it is configurable via Python
  script.

Three algorithms have been developed inside the
RPC monitoring package to completely monitoring
the RPC detector RPC, RPCLV1, MDTvsRPC
24
Data Quality framework
The status of ATLAS data taking is evaluated
based on information from the data acquisition
and trigger systems (TDAQ), and the analysis of
events reconstructed online and offline at the
Tier-0, constituting the Data Quality Assessment
or DQA. DQA comprises data quality monitoring
(DQM), evaluation, and flagging for future use in
physics analysis
RPC has three different have three different
sources of DQA DCS, On-line and Off-line
monitoring
In DCS threshold on active fraction of the
detector is applied to generate the DQ
Assesment. On-line and Off-line monitoring use
the ATLAS DQM Framework to generate the DQ
Assesment, it allow to apply automatically
pre-defined algorithm to check reference
histograms. DQA results grouped as the DAQ
partition are collected in specific DB. The
DataQuality Off-line is totally based on RPC
off-line monitoring performed at Tier0
25
Conclusion

• RPC detector have been installed and commissioned
  since long time.
• Long Cosmic Data Taking allowed to perform a
  complete detector characterization.
• Two different Off-line strategies of detector
  performance analysis has been developed to
  assure a complete characterization.
• Offline RPC monitoring fully integrated in ATLAS
  Software Framework.
• Detector behavior during the run is fully
  monitored via DCS system.

26
Backup slides