Report from the Muon Trigger WG

1
Report from the Muon Trigger WG

• Aleandro Nisati
• On behalf of the Muon Trigger Slice Community
• Muon Week, April 18 2007

2
Level-1
3
LVL1 Barrel Efficiency curves for low-pT
thresholds
Standard low-pT thresholds(6, 8, 10 GeV/c)
log scale
pT GeV
pT GeV
pT GeV

• CSC single muons data (Athena release12.0.3)
• LvL1 performances have been studied using
  CommonMuonSlice AANT produced with rel 12.0.6

4
LVL1 barrel Low-pT Trigger rates
Using the Level-1 Efficiency curves we may
estimate the rates with different threshold.
5
LVL1 Barrel Efficiency curves for high-pT
thresholds
Standard high-pT thresholds(11, 20, 40 GeV/c)
log scale
pT GeV
pT GeV

• CSC single muons data (Athena release12.0.3)
• LvL1 performances have been studied using
  CommonMuonSlice AANT produced with rel 12.0.6
• High-pT plateau at 78

6
LVL1 barrel High-pT Trigger rates
Level-1 Efficiency
Preliminary
Inclusive µ cross-section _at_ LHC (prompt µ
and ?/K decay)
7
LVL1 barrel Trigger rates vs. Threshold
The plot shows the expected single muon trigger
rates at a luminosity of 1033 cm-2s-1. The
cosmic configuration is plotted with a threshold
of 3 GeV.
Rate Hz
threshold GeV
8
Endcaps Software Status
Latest TGC(endcap) LVL1 trigger software(towards
13.0.0) TrigT1TGC-00-01-26 - The algorithm
creating CBNTAA is added. - Some minor bugs
are fixed. TGCcabling-00-00-34 - Bug
fixed Effect of these bugs are negligible.
Future plan (once CBNTAA data is available) -
perform rate calculation with the latest trigger
configuration - optimize coincidence window and
investigate position dependency - more detailed
study of LVL1 di-muon trigger.
9
Endcap Trigger Efficiency
We used these data set. - Single muon events
produced with athena 12.0.3(RDO) - Use
TrigT1TGC-00-01-18-10 for LVL1 trigger simulation
6GeV 20GeV
3-station coincidence trigger efficiency of TGC
10
Endcap Trigger Rates
L1033 cm-2 s-1
L1034 cm-2 s-1
11
MuCTPI overlap resolution in the endcap

• We discovered that the sector numbering in the
  TGC and MuCTPI simulations were not consistent
• This led to an increased fake double-count rate,
  since some sector edges were not handled by
  either simulation.
• Compensating for it in the MuCTPI simulation,
  this is how the EC double-count probabilities
  change.
• This results in a 315 Hz ? 37 Hz fake double
  count rate reduction in the endcap. (Preliminary
  result)
• The total fake di muon rate goes from 432 Hz to
  154 Hz

without MuCTPI
with MuCTPI
after the fix
12
MuCTPI overlap resolution in the endcap
() Not using strip masks on sector edges. This
may improve the rejection of double counting.
13
LVL2 Algorithms
14
muFast improvements since last TP week

• mFast resolution improved by the use of LUT for
  separate charge
• 6 GeV resolution moves from 10 to 7
• degradation of the resolution at high-pT due to
  the vertex spread recovered by the use of the MDT
  fit segment from the Innermost Station
• comparison between mFast resolution
• obtained making use of different
• reconstructed variables
• (alpha and beta angles) shows a similar
• behaviour but
• alpha has more coverage than beta
• alpha shows less tails in the resolution
• distributions
• Studies on which variables is best
• to use is still going on
• Studies to improve the timing of the
• calibration access are ongoing too.

DR
15
muFast improvements since last TP Week

• mFast resolution improved by the use of LUT for
  separate charge
• 6 GeV resolution moves from 10 to 7
• degradation of the resolution at high-pT due to
  the vertex spread recovered by the use of the MDT
  fit segment from the Innermost Station
• comparison between mFast resolution
• obtained making use of different
• reconstructed variables
• (alpha and beta angles) shows a similar
• behaviour but
• alpha has more coverage than beta
• alpha shows less tails in the resolution
• distributions
• Studies on which variables is best
• to use is still going on
• Studies to improve the timing of the
• calibration access are ongoing too.

16
muFast Using charge dependent LUT
We use a Look-Up Table to calculate pT from the
angle a. By taking into account the charge
difference in the LUT, we see an improvement in
the overall resolution 10 ? 7
17
muFastMomentum measurement
Sigma of Resolution distributions
Momentum measurement using alpha, beta, radius
and DeltaR Performances obtained using sample
with no spread in Z for primary vertex. (Slight
difference in dataset wrt previous measurement)
As expected momentum measurement from radius and
deltaR are linear wrt momentum on large momentum
range.
Mean of Resolution distributions
18
Accessing the MDT calibration constants at LVL2

• Problem Using MDT calibration a la offline
  takes 4050 of total muFast time
• First Understand the best granularity to access
  the MDT calibration constants (per tube, per
  layer, per station,...)

Looking at Sector13 cosmics data (Nov 2006)
accessing t0 and r-t relation per MultiLayer
should be good enough (at least for LVL2 trigger
purposes )
T0 distribution
?4ns
BIL3
BIL2
BIL1
From M. Iodice Muon week talk
Drift length
? 7ns
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
LVL2 mIsol Status

• Algorithm
• Included in 13 nightlies
• Migration to new steering done
• Configurables migration on going (ready in 1
  week)
• Monitoring
• new monitoring module added AANT Histograms
• so far used for CSC studies, eventually will
  provide monitoring histograms for DQ
• Performances/Optimization studies
• Just started
• timing ? cones/cell energy threshold optimization
• To follow
• efficiency for Z/W??????, rejection against bb??X
  BG

23
Event Filter
24
TrigMoore Brief Reminder (12.0.6)

• Two different running modes
• Seeded
• Reconstruction performed only in the geometrical
  regions provided by the RoIs of previous levels.
• Full scan
• Full reconstruction, equivalent to the offline
  working mode

• Seeding Algorithms assume the seed is from LVL2
  or a LVL1 ROI
• Full functionality in barrel and end-caps
• 3 istances of TrigMoore called by the steering,
  for reconstruction in the MS, extrapolation to
  the IP and combination with ID tracks
• TrigMoore attaches to the TE a "TrigMooreFeature"
  for each ROI, accessed by TrigMooreHypo for pT
  test
• TrigMoore records in SG the TrigMooreFeature per
  each ROI and all reconstructed tracks in the
  event in a single container for conversion in
  TrkTrack format and subsequent output in ESD and
  AOD

25
Work Ongoing for Rel 13

• TrigMoore and TrigMooreHypo have been migrated to
  the new trigger steering
• first validation OK more robustness test will
  follow
• Use of the configurables
• EDM Migration (see next slides)
• Use of HLT seeded New Tracking ID for combined
  muons in TrigMuidCombined (see next slides)
• Ongoing work to increase modularity
• present implementation not very flexible
• well have 3 HLTalgos instead of 3 instances of
  TrigMoore

26
EDM Migration

• Need to adapt Moore to be able to easily use new
  pattern recognition algs
• Cosmic pattern recognition
• Local CSC and MDT tracking
• to be able to easily output objects required for
  Calibration/Alignment studies
• use of PRD as common input
• Motivated by Common Tracking for ATLAS and by
  desire for increased commonality in Muon-specific
  reco software
• The current (12.0.6) implementation of TrigMoore
  uses digits (RDO) as input objects. Standard
  muon-sw converters or the production of digits
  from (transient) byte-stream or from RDO are
  used. Need to use Muon PrepRawData as input.

(Trig)Moore
PhiPatternsAlgTool
Phi Patterns
Combine Patterns
Calibrated Segments
Combined Segments
Roads
RZ Patterns
RZPatternsAlgTool
27
Trigger rates
First exercise to look at rates at pT thresholds
different than the typical scenarios 6 and 20
GeV/c
Luminosity set to 1033 cm-2s-1
efficiencies for LVL1 from F. Conventi for 6,
8, 10 GeV/c (in good agreement with our
estimates) our estimates for 11, 20, 40 GeV/c
Typical scenarios mu(6) 3.0 kHz mu(20) 25
Hz
28
Trigger rates
First exercise to look at rates at the EF in the
END CAPS and vs pT thresholds
Luminosity set to 1033 cm-2s-1
Efficiencies for LVL1 from our estimate maybe
slightly inaccurate for the known bugs in
TGCCabling expected effects at EF lt10
Typical scenarios mu(6) 3.1 kHz mu(20) 27
Hz
Barrel EndCap EF Rate _at_ 1033 cm-2s-1
mu(6) 6.1 kHz mu(20) 52 Hz
29
Trigger Rates
Rate for mu(5) at the EF in the barrel use LVL1
efficiency for the barrel with the trigger
re-configured to have 5GeV/c as lowest threshold
Luminosity set to 1033 cm-2s-1
30
Muon Slice Data Quality (II)

• Nothing exists for the moment for Muon Slice DQA
  but
• what implemented for monitoring during 2004 test
  beam (A. Di Mattia for LVL2) and the test of the
  trigger slices on the pre-series machines at
  Point 1 in december 2006 (D. Scannicchio for EF)
  can be a starting point for Data Quality
  Monitoring

e.g. MuFast histos from last technical run
Trigger/TrigAlgorithms/TrigmuFast/src/OnlineSurvey
.cxx
Shows linear distribution between 1/s and pT as
expected
Core of the fit residual matching the resolution
of the single tube 80 mm.
31
Muon Slice Data Quality (III)
e.g. TrigMoore histos for the ongoing technical
run
Trigger/TrigAlgorithms/TrigMoore/src/TrigMooreHist
o.cxx
(here obtained running the jobOptions prepared
for the on-line with a bytestream file containing
50 top events as input muons are selected by the
LVL2 and the EF muon algorithms)