Pulse Analysis for ATLAS MDT Twin Tubes

1
Pulse Analysis for ATLAS MDT Twin Tubes

• Alex Koutsman
• Supervisor dr. ir. Harry van der Graaf
• 18-05-2005

2
Large Hadron Collider (LHC)

• Operational in 2007 at CERN
• 14 TeV proton-collisions
• 1148 TeV lead-nuclei collisions
• Higgs search, B-physics
• Physics beyond the Standard Model

3
ATLAS

• 1) Inner Detector
• 2) Calorimeters
• 3) Muon Spectrometer
• Muon path bent by the magnetic field
• Amount of curvature momentum
• Muon Spectrometer
• 1200 Muon Chambers with Trigger Chambers (RPCs)
• 240-432 Monitored Drift Tubes (MDT) per Chamber
• In total 370 000 MDTs in ATLAS

4
Operation principle of a MDT

• Aluminium tube filled with gas (radius 15mm)
• Wire at positive high voltage (3200 V)
• Ionisation by passing particle
• Electron-clusters form and drift to the wire
• Response Electronics (ASD)

5
MDT

• Drift-time drift-radius
• Target precision of a single MDT radius 80 µm
• Measurement with the whole chamber new precision
  30 µm in the bending direction
• Diffusion ? Focussing

6
Twin Tubes

• Two tubes interconnected at the high voltage end
• Read-Out on both ends
• Difference in time gives position along the tube
• Built-in delay in HV-jumper (6 ns)
• Shaping by the wire
• Amplitude loss
• Leading edge slope less steep

Prompt or Twin Tube !!!
7
Why Twin Tubes?

• PRO
• 1) Measurement hit along the tube (non-bending
  direction)
• 2) Simplification pattern recognition
• 3) Resolving ambiguities from multiple tracks
  (essential)

• CONTRA
• 1) Higher occupancy of the MDTs (2x)

OUTER CHAMBERS
8
Cosmic Ray Stand

• Twin-tube connectors installed
• Scintillators for trigger

• ATLAS Data acquisition
• -Threshold crossing time
• -Signal amplitude
• -Trigger reference time

9
Experimental Setup

• Trigger unit
• Momentum cut muons 1 GeV
• Three positions along the tube (x-axis)

Accepted angle!!!
10
Data Acquisition
muon
Negative Part of the Differential Signal
Ethernet
Digital Scope
Scintillators
Logical AND
TRIGGER
Logical AND

• LabVIEW
• Set Channel parameters
• Set trigger parameters
• Set time parameters
• Set data recording parameters
• ASCI ? ROOT-histograms (Active MDT events)

Logical OR
11
Good Twin Tubes Events
12
Drift-time spectrum

• Top scintillator as reference time
• Compare with threshold crossing time for MDTs
• Prompt or Twin Tube!!!

13
Amplitude Spectrum

• Integrate signal charge over 15.5 ns
• Discriminate prompt and twin tube
• Landau distribution

14
Noise Level

• Noise spectrum first 400 ns
• Mean distribution
• Sigma distribution

15
Rt-relation

• Drift-time to drift-radius
• Homogenous irradiation
• Integrate drift-time spectrum

16
Mean Signal vs Radius

• Get drift-time borders from the Rt-relation
• Add all signals in a radius-bin
• Normalize with the number of events

Diffusion ? Focussing
17
Slope at Threshold vs Radius

• Set fitting range at threshold crossing 3ns
• Fit a line to measure the slope
• Slope-distribution per Radius-bin

Prompt vs Twin Tube
18
Risetime vs Radius

• Risetime Vmin/Slope
• Find First Minimum (Vmin)
• Slope calculation between 0.2Vmin and
• 0.8Vmin

19
Smoothness Selection

• First Minimum
• Find First Minimum (FM) within 120ns after
  threshold crossing
• Check for Minima OR Inflection Points (IP)
  between 0.9FM-Value and FM-Value

Around 35 dont make it through the
selection!!! More Low Amplitude events are
cut off!!!

• Select
• If FM or IP is between threshold crossing and
  0.9FM, then event is BAD!!!
• Check for CLEAR Minima with Amplitude gt
  0.7FM-Value

20
Risetime vs Radius
21
Time Resolution

• Fit a gauss to the time difference distribution
  resolution sigma (st)
• Cosmic Ray Stand resolution 1 ns
• Strong selection of events
• Track fitting
• Slewing correction
• No secondary hits

22
Shadowing Particles

• Leading edge of the muon signal is screened
• Shadowing particles
• d-electron (d-ray)
• Photon
• Shower particle

• Clean up and select similar events

23
Smoothness Selection
24
Mean Range Selection

• Get Minimum of Mean Signal
• Deselect all events outside Vmin Vminmean
  CMRS

Varying the Vminmean resolution deteriorated
25
Vmin Ratio Selection

• Ratio
• Fit a gauss through the selected distribution
• Deselect events with ratio outside CVminRS X
  sratio

26
Prompt Change Selection

• Some strange events secondary particles
• Check if the prompt changes
• Use two thresholds for certainty

Cut off are 2-3 of events that get through the
other selections!!!
27
Charge Selection

• Resolution highly dependent on amplitude
• Deselect events if one channel amplitude is below
  the mean
• Resolution improvement of 20

28
Time Resolution vs x

• Tails especially big for events at the Read-Out
  end
• Difference with Cosmic Ray Stand
• 1) data
• 2) spread in accepted angle

29
Time Mean vs x

• Take mean for each x-position
• Fit a line
• 2Slope gives the signal speed in the tube
• Extrapolating to maximum length of 4.9m gives the
  HV-jumper delay

30
Conclusions and Outlook

• Twin Tube concept offers an adequate measurement
  of the hit position along the tube
• The signals of the twin tubes are as expected
• Drift-time spectrum
• Amplitude spectrum
• Low noise
• Slope vs Radius
• Risetime vs Radius
• Time resolution

• The ATLAS setup will reach better resolution, as
  on average a muon hits 6 tubes in a chamber
• Slewing correction
• Comparison to Cosmic Ray Stand data and to the
  simulation
• Qualitative and quantitative

http//www.nikhef.nl/ajkoutsm/TwinTubes.ps
31
BACKUP SLIDES
32
Amplitude Ratio

• The further the twin tube signal has to travel
  the further away the ratio goes away from unity

33
Slewing
RO-end

• Time Difference vs Integrated Charge
• Profile histogram
• Slewing is the greatest for RO-end, as expected
• Tails have a big influence on the distribution

Middle
HV-end
34
Electronics Scheme

• MDTs read out before the discriminator stage
• Trigger top AND bottom OR

35
3300 Volts

• 3300V has more tails for all x-positions
• Preliminary results are not better then 3200V
  (also with Selection)