CALICE Summary of 2006 testbeam for the ECAL

1
CALICESummary of 2006 testbeam for the ECAL

• Anne-Marie MAGNAN
• Imperial College London
• On behalf of the CALICE-ECAL Collaboration

2
Content

• Introduction
• Briefly, what does the ECAL look like ??
• The three 2006 test beam periods and their
  objectives
• Data taken
• gt 5 TBytes on disk, 65 million physics events
• Electron and hadron events
• A few event displays.
• Noise issues
• A perfect detector would be amazing !
• Conclusion
• A successful experience

3
The Electromagnetic Calorimeter prototype

• W layers wrapped in carbon fibre
• 3 modules with different tungstene thickness,
  total 24 X0.
• PCBSi layers8.5 mm

• ECAL prototype

• 30 layers of variable thickness of Tungsten
• Active silicon layers interleved
• Front end chip and readout on PCB board
• Signals sent to DAQ

4
Overall objective of 2006 beam test for ECAL

• Characterize detector performances.
• Test and tune the simulation. Once we trust the
  simulation ? optimisation of the ILC detector.
• Identify hardware problems to correct them before
  building the whole so-called EUDET module.
• Test particle flow algorithms.

5
2006 beam test periods
Installation _at_ DESY . First cosmic events in the
tent.
Feb
DESY TB area, 1-6 GeV electron beams, with only
24 layers . 14 days in total.
May
July 27th Aug 8th
ECAL CERN period cancelled !! No beam, SPS
fault, not ours....
AHCALTCMT period transformed into 5 days
ECAL5 days AHCAL stand alone 3 days, courtesy
of ATLAS, for ECAL stand alone.
Aug 25th Sept 6th
Oct 11th Oct 30th
Combined ECALAHCAL TCMT , with ECAL-AHCAL _at_
6cm
6
DESY TB setup
Tracker 4 drift chambers ECAL 24 layers, 5184
channels First tests of ECAL-HCAL combined
acquisition
7
DESY Data Samples
? Energy and angle scan completed, with 200k
events or more per point.
8
CERN installation

• ? Particle ID Cherenkov counter, 1bit signal
• ? Tracker 3 XY proportional chambers (MWPC)
• ? Calorimeters
• ECAL 30 layers, 6480 channels
• HCAL 15 modules, 3240 channels
• TCMT 8 modules, 160 channels

9
CERN- August period in figures

• 30 Millions of Muon events for calibration.

10
A few hadron runs combined with AHCAL and TCMT
11
CERN October period

• Add another 27 Million physics events !
• 42M of muons for calibration
• 14M only for HCAL calibration
• 3.5 M of HCAL only runs
• 1.2 M only dedicated to beam tuning
• Detail of the 27 M
• - Another 3.8 M dedicated to electron studies
  between 6 and 50 GeV
• - And 23 M of hadronic events between 6 and 80
  GeV

12
Electron runs
Electron combined runs Electron combined runs Electron combined runs
E (GeV) e (kEvts) e- (kEvts)
6 208 128
8 218
10 152 172
12 211
15 476 124
16 310
18 303 231
20 390 210
30 409
50 305

• ? One point overlap with DESY TB at 6 GeV.

13
Hadron runs
Combined pion runs Combined pion runs Combined pion runs
E (GeV) p (kEvts) p- (kEvts)
6 480 1800
8 1800
10 960 1800
12 1600
15 720 1600
18 1700
20 770 1600
30 3300
40 1400
50 1500
80 1800
With ECAL and HCAL at 6cm from each other this
time !
14
Summary of events

• Size on disk 40 kB/evt
• 65M events 2.5 TB for CERN Physics runs
• 70 M 3 TB for muon calibration runs

• See Romans Talk
• on Wednesday.

15
Electron showering in the ECAL
16
An hadronic event
10 GeV pion event, taken Oct 16th 2006
17
A few problems, of course

• Pedestal instabilities results in noisy
  layers, with a correlated noise over a whole PCB
  of up to 50 of a MIP !!!
• Not understood yet ! Currently investigating on
  an hardware point of view.
• A preliminary procedure to correct this effect
  event by event is applied in the data
  reconstruction, but need further studies.
• the pedestals are measured at periods throughout
  the run so this effect can be tracked.
• This concerns only less than 3 PCBs over 30.
• Square events appearence of hits around some
  wafers, in the guard ring zone.

18
Pedestal instabilities ...
A Good PCB
Ex Muon run (ECAL threshold 0.5 MIP)
Noise
A PCB with unstable pedestals
Noise
19
Square events ...

• Proportion of events concerned
• high energy data (electrons 45 GeV) 4
• low energy data (electrons 3 GeV) lt 10-3.

20
Conclusion

• A successful experience !!
• Lots of data taken, stable detector running,
  people on-site to solve the problems as soon as
  they appeared.
• Electron data samples from 1 to 50 GeV ready for
  analysis.
• Further plans TB in May-June next year
  (currently in negotiation with CERN) with
  possibility to rotate both ECAL and HCALTCMT.
• Fermilab with AHCAL in fall 2007, and DHCAL in
  winter 07/08.

21

• Thank you for your attention.

22
CERN installation
Muon 1 Muon 2
beam
400 m
11000
all in mm
Beam instrumentation 1) 500 m beam line after
Be trg magnets, collimators, secondary trg,
abs 2) Cherenkov detector for e/p separation lt 40
GeV 3) 3 x/y pairs of Multi Wires Proportional
Chambers (MWPC) with double readout, multi-hit
capability 4) veto counter, r/o analog amplitude,
to separate multi-particle events 5) trigger
system 2) 3) 4) 5) are integrated in the DAQ
and read out event by event
p event
23
DESY installation
24
View of an ECAL board
25
Beam quality issues _at_ DESY
26
The data taken
CERF period parasitic muon high intensity, wide
distribution ? Very important for calibration !!!
Combined run, goal ECAL EM program - e 10-45
GeV, from 50 GeV beam, with 0,10,20,30 deg -
small samples of p 30-80 GeV too large distance
ECAL-AHCAL
up to 10 million events in one day!
AHCAL stand alone, ECAL removed - 1 day _at_ 10 GeV
secondary beam tested p / e 6,10,15,20 GeV -
3 days _at_ 50 GeV secondary beam e 10-45 GeV and
p 30-80 GeV
All what was collected in the ECAL run 60 GeV
secondary beam, tested e 10-45 GeV and p 30-80
GeV.
!!! large fraction of time invested in beam
tuning !!!
? 3 additional days courtesy of ATLAS AHCAL
and TCMT out of beam line, ECAL re-installed for
high statistics low energy runs (thanks to all
voluntary shifters)