Crosstalk Removal to Improve dEdX measurements

1
Crosstalk Removal to Improve dE/dX measurements

• Leo Jenner, UCL

2
In brief

• Crosstalk removal algorithm
• Comparison of MC truth to Tracking Crosstalking
• Some hitches
• Preliminary comparison of MC to data

3
A Cosmic Ray Muon
Crosstalk removal helps with Tracking And is also
important for correcting dE/dX For relative
calibration of the detectors
4
CrossTalk Algorithm

• Work through all the hits in a given pair of
  planes and look at the 8 pixel window around each
  test hit
• Sum the charge and identify the largest hit in
  the window
• The test hit is flagged as crosstalk it its
  charge is less than the sum of the pixels around
  it, multiplied by a user definable factor. This
  is usually set to 1 for beam/cosmic muons since
  low eneryg deposition combined with photon
  statistics can lead to crosstalk with very
  comparable energy to real hits
• The Crosstalker currently runs as a module in
  Chris Smiths CalDetTracker muon tracking code
• Testing is taking place on Mike Kordoskys
  Crosstalking Monte Carlo

5
Number of Crosstalk hits Truth vs Tracked

• The crosstalker finds about 82 of the crosstalks
  hits as Ive opted for a high purity (gt90)

6
Comparison of MC truth to MCTracking

• For the most part there is good agreement

7
Number of Hits Tracked/Truth
A close up of the last few planes Shows the
discrepancy increasing

• But only 0.32 of events go to plane 54 or more

8
Last Plane Hit

• Some of the hits at the end of the track seem to
  go missing after tracking

9
Very curvy muons

• In the third plot along the two blue hits at the
  end of the track are genuine track hits
• Maybe the muon curves sharply to the left and
  travels in the air gap between planes
• The tracker doesnt nominally find the last hit
  because it falls outside of the search window

10
Anomalies

• Entry 2340943 Record 3107 End A Plane 48
  Strip 8 HitBit 786432 PE 5.86536 TrackEnd B
  Plane 48 Strip 8 HitBit 786432 PE 7.3501
• Entry 2340944 Record 3107 eventnum 3107 Entry
  2340945 Record 3107 eventnum 3107 End B
  Plane 54 Strip 10 HitBit 0 PE 0.935773
• Entry 2340946 Record 3107 eventnum 3107

There is an unphysical hit of nearly 1pe its
ignored by the truth (see above) but the tracker
finds it
11
Preliminary comparison to data

• Quite low statistics but it still looks a bit odd

12
Data - Last Plane Hit

• Even with cut 10ltfirst strip hit lt13

PS Muons, Other off momentum muons coming down
the beampipe
13
Conclusion

• More work to do to optimise Tracking and
  Crosstalk removal using Monte Carlo
• Fold crosstalk back into real hits to furter
  improve MC dE/dX
• Apply crosstalk removal to data