Summary STS working group

1
Summary STS working group

• Design optimization
• STS (Strip) layout
• RD (MAPS)
• Working packages

2
Manpower

• STS activities momentarily not backed by EU
  funding
• FP6 I3 (SOSDET)
• RD on monolithic (MAPS, amorphous Si)
  andultra-thin hybrid (Giga-tracker) sensors.
• not funded in favor of diamond ? standard
  technologically!!
• FP6 Design close call
• STS foreseen for second round (i.e. 2005)
• now cancelled in favor of proposals from first
  round
• Total manpower in STS project of O(1 FTE)

3
Generic simulation

• Design as of now
• 7 stations at 5,10,20,40,60,80,100 cm from target
• 100 mm (1,2) and 200 mm (3-7) thick
• 10, 20 mm resolution in both coordinates
• Needed for proposal
• Simulation including pad, strip structure
• Optimized configuration

4
Design optimization
STS group
Tracking groups
Mainframe
Algorithms
Digitizers
MAPS Michaels talk
Design Optimization
Strip Tatianas talkValery
Final configuration
5
Si strip STS_4 Layout
20 cm
Valery Savielev, Obninsk State Univ.
-20 cm
Read out
6
Si strip STS_6 Layout
40 cm
Valery Savielev, Obninsk State Univ.
4cm
- 4cm
-40 cm
Read out
7
Digitization result
Tatiana Galatyuk, GSI/Kiev
3-d STS
5-th STS
4-th STS
number of hits
7-th STS
6-th STS
number of zone
8
RD

• MAPS IReS, LEPSI, (GSI)
• radiation tolerance
• readout speed
• Driven by common interest between NLC and CBM
  community
• Strip detectors MEPHI, MSU
• Thin double-sided sensors
• Read-out (FEE) chip
• no activity yet due to missing funding

9
Successor1 before and after 1MRad X-Rays
Observations (Priliminary) No significant
influence on gain
T -15C t 200µs
Charge collected in 1 pixel ADC
No significant charge loss
Michael Deveaux, GSI/IReS
Charge collected in 9 pixels ADC
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
10
Smart Detector Reticle
Matrix of Pixel
Band Height of sensitive part
Sensitive part
B1
Control ELN
ADC
Sparcification
Memori-sation
Band Height of non-sensitive part
Control and processing part
B2

• Control and address Electronics
• Fast signal processing and data sparsification ?
  Signal digitalisation
• Storage threshold values needed during data
  sparsification
• Surface ratio (B2/B1) must be as small as
  possible

Hayet Kebbati, GSI/IReS
11
Towards a Design Proposal
12
The end ...
13
Choice of technology

• Availability
• Minimize own development effort (manpower)
• LHC silicon trackers needed 800 man years RD
• Conservative fall-back solutions for future
  technology options

14
Problems / Open questions

• Material budget
• increase to balance position resolution!
• identification of conversion electrons
• Efficiency?
• Realistic simulation for track fragment
  reconstruction
• Efficiency
• Fake rate
• Magnetic field
• Close pair recognition in the inner tracking
  section
• Efficiency?
• d-electrons
• Vertex reconstruction
• fast station needed in vacuum?

15
Oleg Rogachevski, conformal mapping

• 5,10,20,40,60,80,100 cm, 3 first stations in
  vacuum, 100 mm (1-2) and 200 mm (3-7) thickness.
• no delta electrons
• Hit multiplicity for secondaries peaks at 1-2
  hits, 40 of the hits in the first two stations
  have no other hits and cannot be reconstructed
• Efficiency drops from 0.95 to 0.7 below 10 GeV,
  efficiency gt 95 if 7 hits are observed.
• Matching with inner track segments, which are
  incomplete, problematic!!!

16
Sascha Ierusalimov, STS track fitting

• Generic simulation
• Uniform field used
• 100 mm thickness throughout, 20 mm resolution
• best resolution between 1 and 10 GeV 0.7 1
  if all hits are present.
• Vertex reconstruction
• small systematic shifts lt 1 mm
• s 16-21 mm
• Results from realistic simulation (GEANT-3)
• Problems
• broken tracks
• 2-hit tracks?
• close tracks?

17
G. Osakov, track reconstruction a la LIT

• Old geometry used
• on average 1.62 ghost tracks
• efficiency around 80-90 only

18
Pavel Akishin, Polynominal approximation

• Best resolution reached in this approach. 0.4

19
Pavel Zrelov, Kalman filtering

• 2-dimensional projection
• Very precise starting points needed
• 20 mm resolution assumed
• Parabola used as track model between the detector
  stations 5-7

20
Joachim Gläß, Fast tracking

• Fast Hough transform used
• Efficiency only above 1 GeV (cut)
• e amounts to 0.95
• Large ghost rate
• Needed clock cycle 150 MHz
• Total number of processing units 200