Micromegas for the Central Tracker
Sébastien Procureur
CEA-Saclay
Micromegas and CLAS12
- 3 double layers of cylindrical MM (Barrel)
- 3 double layers of flat MM (Forward)
Micromegas
~ 4 m²
CLAS collaboration meeting, 09/03/2011
S.Procureur
The MM challenges (Cerchi dell'Inferno)
“Abandon all hope, you who enter here”
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Cylindrical Micromegas
Make use of the « bulk » technology (2006)
Mini: 4 mm
2 to 4 mm
128 mm
Photoresist
border
PCB with strips
Photoresist
amplification
spacer (~300 µm)
Mesh
→ more robust
→ PCB can be thin
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Cylindrical Micromegas
Performance compared to thick flat MM using cosmics
Thick detector
Thin detector
→ similar performance as thick detectors
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Spatial resolution in 5 T
X tiles of Barrel Micromegas are sensitive to the Lorentz angle of drifting electrons
x = h tanθ = h v B / E
h
→ minimize h (but less signal)
→ use heavier gas (but more sparks)
x
→ increase E field (but lower transparency)
→  ~ 220 µm if  can be lowered down to 20°
Garfield simulation
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Spatial resolution in 5 T
Test to validate Garfield simulation with a Micromegas in dvcs magnet (Hall B)
→ use of a focused UV laser to extract electrons
from the drift electrode
Garfield validated,  can be as low as 20°
P. Konczykowski et al., NIM A612 (2010), 274
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Large area detectors
Full size Y prototypes have been built at CERN
→ doubled area compared to COMPASS
→ Cdet ~ 25 nF, Cstrip from 90 to 120 pF (can be reduced)
Scan with Fe55 source
Micromegas
→ 90% of strips OK (1st proto!)
→ ready to build them at Saclay
CLAS collaboration meeting, 09/03/2011
S.Procureur
Electronics
The limited space requires an off-detector electronics → long cables
→ initial cables were 160 pF/m (FLEX)
→ Irakli found 70 pF/m (Hitachi)
→ Ccab = 105 pF ~ Cstrip
→ Significant effect on S/B (~50%)
→ Needs 10 V more to be compensated
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Electronics - recent developments
 Detector cables: Hitachi 50 pF/m cables expected
on March 15th
→ check ability to withstand sparks
→ Goal: 40 pF/m cables
Asic DREAM0: Magnitude versus input capacitance
1,9
1,8
1,7
detector capacitances up to 200 pF
→ ENC of 2200e for 150 ns peaking time
→ Expected S/N : 30 – 40 depending on gain
1,6
Magnitude ADC input (V)
 Dream: reworked input stage adapted to high
1,5
1,4
1,3
AGET
modif
1,2
1,1
1
0,9
0,8
0,7
0
20
40
60
80
100
120
140
160
180
200
Input capacitance (pF)
 Test bench: Dream carrier board in place & route
→ Up to 7 Dreams
→ One Dream on a remote board
 Firmware: in progress
→ Test bench and front-end unit firmware
→ Goal: estimate FPGA resources needed
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
DREAM and test bench schedule
• February: Dream carrier board place & route started
• End of March
– Submission of Dream
• Backup date: end of April
– Production of the Dream carrier board
• April: Dream carrier board in test
• May: Dream test bench ready
– Final adjustments for firmware and software
• June: packaged Dream ready for tests
– T2-T3 – Dream tests
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Spark rate studies
Activity started in 2009
→ simulation: try to relate sparks with large
energy deposits with Geant4 (Gemc)
→ spark condition: Nel ~ 107 (Raether)
→ Quantitatively reproduces (few) existing data
→ Explains gas effect & give predictions (bulk)
S. Procureur et al., NIM A621 (2010), 177
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Spark rate studies
Tests at CERN/SPS, October 2009
→ 6 different detectors in 150 GeV pion beams
→ Effect of a 1.5 T  magnetic field
→ 1st test of a MM+GEM detector (D. Neyret)
→ 1 week of data
→ bulk ~ non-bulk
→ No strong effect of B field
S. Procureur et al., submitted to NIM (Feb. 1st 2011)
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Spark rate studies
Tests at JLab/Hall B, July 2010
→ 1 MM & 1 MM+GEM in FROST setup
→ Effect of a 5 T // magnetic field
→ 2.5 days of data
→ simulation ~ OK
→ x10 with 5 T field
→ 100 with GEM foil
B. Moreno et al., submitted to NIM (Mar. 8th 2011)
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Spark rate studies
Tests at CERN/PS, August 2010
→ 12 detectors in π+ or π- beam
→ beam momentum tunable between 0.2 and 3 GeV/c
→ 2 MM+GEM to understand spark rate suppression
→ 2 weeks of data
→ peaks in spark rates with π+
→ confirms GEM effects
G. Charles et al., submitted to NIM (Feb. 25th 2011)
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Spark rate – last mysteries
2 observations cannot be reproduced by the naive simulation
→ effect of the longitudinal 5 T field
→ spark suppression with the MM-GEM detectors
→ strong hint for significant effect of transverse diffusion
 new spark condition: critical surface charge density
→ explains all effects seen with MM-GEM:
→ large suppression with moderate GEM gains
→ effect of trans. diff. only at high GEM gains
→ change of slope at high GEM gain
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Integration
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Forward Tagger with Micromegas?
Project to equip the FMT with central pixels for small angle e- detection
- add ~ 6k channels to the FMT (electronics?)
- add 2-4 MM layers in front of calo

φ
p
Very large background due to Moeller
All hits
Selected hits
z
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Schedule
Conclusion
Micromegas
CLAS collaboration meeting, 09/03/2011
S.Procureur
Additional slides
Working point
Parameter
Barrel (Z&Y)
Forward
Remarks
Drift space
2-3 mm
5 mm
Gas
Ar+10%iC4H10
Ne+iC4H10
Fields ratio
5.5-6
50
Mesh Transparency
40%
40%
Electron
Transparency
40-45%
~100%
Gain (effective)
5000
3000
Detection Efficiency
>90%
>95%
Measured @ B=0(X)
Tracking Efficiency
>93%
>90%
Simulation (tracking)
Detectors / layer
3
1-2
Background noise/
Detector
2 MHz
12 MHz
Geant4 estimate
Sparking rate
1/s
0.4/s
Geant4 estimate
Space resolution
220&100 µm
80 µm
Garfield estimate
Lorentz angle
20°
0°
Garfield estimate, validated
Measured @ B=0