performances of the micromegas vertex tracker the … · 2019. 6. 4. · moudden, s. procureur, m....

PERFORMANCES OF THE

MICROMEGAS VERTEX TRACKER AT

THE CLAS12 EXPERIMENT FOR THE 2017-2018 PHYSICS RUN

MPGD CONFERENCE 2019,

LA ROCHELLE, FRANCE

MAXENCE VANDENBROUCKE

ON BEHALF OF THE MVT GROUP AT SACLAY :

D. ATTIE, S. AUNE, J. BALL, Q. BERTRAND, F. BOSSU, G. CHRISTIAEN, M. DEFURNE, J. GIRAUD, R. GRANELLI,, I. MANDJAVIDZE, O. MEUNIER, Y.

MOUDDEN, S. PROCUREUR, M. RIALLOT, J-Y. ROUSSE, F. SABATIE

https://www.google.fr/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&cad=rja&uact=8&ved=0ahUKEwj5wbvmvvvTAhUEnRoKHQQSAG4QjRwIBw&url=http://www.futfanatico.com/2011/10/26/english-youth-academy-extreme-makeover/&psig=AFQjCNFLigjMFCfqSrwjoLZaqAJmi1uqmA&ust=1495267220438088

SUMMARY

The CLAS12 Experiment at Jefferson Lab

The Micromegas Tracker, Forward Detectors, and Cylindrical Detectors

Installation in the CLAS12 spectrometer

First results after data taking2019 MPGD Conference - CEA Saclay - Maxence Vandenbroucke 2

* Serving suggestion

THE CLAS12 EXPERIMENT• Upgrade of the CLAS Experiment at Jefferson lab

• Study of the nucleon structure with ~11 GeV electron beam at high luminosity (1035 cm-2s-1)

• Targets : liquid hydrogen (protons), liquid deuterium (neutrons), other nuclei in the future

32019 MPGD Conference - CEA Saclay - Maxence Vandenbroucke

Micromegas Vertex Tracker (MVT) :

Improve the track reconstruction in the vicinity of the target

Inserted in the 5T solenoid

Used in combination with the Silicon Vertex Tracker (SVT)

Solenoid + Central Tracker

THE MICROMEGAS VERTEX TRACKER


4 m² of Micromegas detectors

DREAM based Front-End Electronics ~ 20k ch.

Remote off-detector frontend electronics connected with 2m micro-coaxial cables

Forward Detectors (Disks)

High particle rate (30MHz)

Resistive strips divided in 2 zones inner/outer

Dimensions: 6x 430 mm diameter disk with a 50 mm diameter hole at the center

Cylindrical Barrel (Curved Tiles)

Low momentum particles => Light Detectors

Limited space of ~10 cm for 6 layers

High magnetic field (5T)

Phase 1 (2016) : 2 Layers (6 Det. of 120°)

Phase 2 (2017) : 6 Layers (18 Det.)

4

5

THE DREAM FRONT-END ELECTRONICS

• Signals are continuously pre-amplified, shaped, sampled at 20-30 MHz and kept in the circular analog memory 512 cells deep

• Covers 16 µs trigger latency

• At each trigger the 4 to 10 corresponding samples are readout and digitized

• Readout does not disturb sampling

• Retained samples are digitally processed

• Pedestal equalization – online

• Common noise subtraction – online

• Zero suppression – online

• Measure charge and time – off-line

• Micro-coax cables – 64 channels – low capacitance 43 pF/m

~1K ch. Detector

16 x 64 ch. Micro-Coax cables (1.5 -2.2m)2 x 512 ch.Front-End Unit2019 MPGD Conference - CEA Saclay - Maxence Vandenbroucke

CLAS12 MM FORWARD TRACKER• 6 layers of Micromegas with 1D strips alternatively

rotated at 0°, 60°, 120°

• 86 mm to 380 mm diameter active area

• Bulk MM + Resistive Layer

• Same detector design for the 6 detectors :

• Dimensions: 430 mm diameter disk with a 50 mm diameter hole at the center; 5mm drift gap

• 100 µm PCB glued on ROHACELL

• 525 µm pitch, with 120 µm between two strips, 1024 strips

• 2 independent resistive strips zones (inner/outer)


Black : no ladders

Charging up effect studies with X-Rays

Resistive Strips w/o interconnections (ladders)

FORWARD DETECTORS IN COSMIC TEST BENCH

2019 MPGD Conference - CEA Saclay - Maxence Vandenbroucke 7

FORWARD DETECTOR COSMIC RAYS TEST• 6 Detectors fully operational with no current on the resistive layer after

many cleaning procedures

• 6 Detectors have been delivered to J-Lab in Sept. 2016

• Radiation length of 0.70% X/X0 => To be be lowered for the next run

• Close to full efficiency (98%) in the active area

• Resolution better than 200 µm (limited by tracking of the test-bench)

• Time Resolution better than 20 ns (same)

8

Efficiency map

Cut for eff. calc


CLA

S12

BA

RR

EL D

ETEC

TOR

S


CLAS12 BARREL DETECTORS PRODUCTION

• Total of 6 layers segmented in phi (3 x 120° sectors) = 18 detectors total

• 6 Different detector’s radii

• 2 different types (C and Z types)

• Material (PCB/Bulk + Drift) from the CERN Workshop

• Assembly to cylindrical shape at Saclay

• Test and Characterization at Saclay before shipping to J-Lab

• 8-9 days to assemble one detector + 1 week of test

“C” Barrel“Z” Barrel

Layer Production ch. Radius Length Width

CR4-C 3 + 1spare 896 146mm 712mm 302mm

CR4-Z 3 + 1spare 640 161mm 712mm 333mm

CR5-Z 3 + 1spare 640 176mm 712mm 364mm

CR5-C 3 + 1spare 1024 191mm 712mm 396mm

CR6-Z 3 + 1spare 768 206mm 712mm 427mm

CR6-C 3 + 1spare 1152 221mm 712mm 459mm

CR6-C new 3 spare 1152 221mm 712mm 459mm


COMPACT DESIGN


INTEGRATION OF ONE 120° SECTOR


OPERATION IN CLAS12 AND THE 5T MAGNETIC FIELD

• Small volume for instrumentation (6x15mm)

• Remote off-detector frontend electronics using 2.2m long coaxial cable + DREAM FEE

• No fan for cooling

• EM Shielding challenging

• Lorentz Angle

• Slow gas to reduce drift velocity

• Small drift gap

• High electric field

• 6kV/cm for C

• 5kV/cm for Z

• Degradation of spatial resolution

• Effect depends on the charge

of the particle

• Modify transverse diffusion

13

Detector Radius (mm)

CR6C 222.53

CR6Z 207.54

CR5C 192.65

CR5Z 177.57

CR4Z 162.56

CR4C 147.57

825V on 3mm 3825V on 3mm Effect on Ion Backflow

=> Clas-note 2007-004: Simulations of Micromegas detectors for the CLAS12 experiment (S. Procureur)

[V/mm]driftE0 100 200 300 400 500 600 700 800 900

Lo

ren

tz a

ng

le [

de

g]

0

10

20

30

40

50

60

70

80

90Hall B data, B=1.4T

Hall B data, B=2.8T

Hall B data, B=4.2T

Magboltz, B=1.4T

Magboltz, B=2.8T

Magboltz, B=4.2T

Lorentz Angle Vs HV


BARREL DETECTORS: PERFORMANCE WITH COSMIC RAYS

14LAYER 6 LAYER 5 LAYER 4

• 6 Layers of cylindrical detectors divided in 120° sectors = 18 Micromegas tiles

• Bulk + Resistive Micromegas

• Less than 0.5% of a radiation length per layer

• Cylindricity measured to be precise up to ~2mm in radius

• Resolution better than 200µm per layer with cosmic rays

• Time resolution of ~25ns with cosmic rays


FIRST INTEGRATION AT J-LAB


In 2016 Successful integration at JLAB of ~1/2 of the layers together with the Silicon tracker and joint test with cosmic rays

Residuals of MM : ~ 400 µm for Z det.~ 5 mm for “C” det.


MVT ARRIVES AT CLAS12 IN FALL 2017

ENVIROMENTAL CHALLENGE

• Unexpected high background increased the silicon vertex tracker leakage current and force to increase cooling in the central tracker

• First cooling brought humidity issues fixed by nitrogen flushing

• A few detectors had to be replaced

• Micromegas tiles have been operated below freezing with no further issues


Temperature recorded on the detectors for 2018-2019

Sensors on BMT

Humidity on barrel detectors

DETECTOR CURRENT

• Detector current vs beam luminosity


Up to 2µA per Micromegas at nominal beam intensity

DETECTOR OCCUPANCY WITH BEAM


a)

c) d)

Hit occupancies for C-tiles at 2.2 GeV (left) and 10.6 (right). The elastic recoil protons are responsible for the large excess of events at 2.2 GeV, between strip number 400 and 500. The cross sections is too small at 10.6 GeV to see the protons.

Occupancies of FMT (left) and BMT (right) as a function of the beam current. Half of the FMT disks has a lower occupancy since only their inner region is active.

Forward Barrel

CENTRAL TRACKER– 1 TRACK EVENT


Silicon

MM

TOF

Neutron Det.

April 2019 - 10.2 GeV electron on LH2 target

CENTRAL TRACKER– 2 TRACKS EVENT



CENTRAL TRACKER– 3 TRACKS EVENT



EFFICIENCY AND HV PLATEAU SCAN


• Tracking algorithms are not final yet=> Very Preliminary Results

• 90%-100% efficiency reached in physics data taking conditions

• Working point for the mesh HV at ~500V in Ar:Iso 90:10

STATUS

24

• The CLAS12 experiment MM Vertex Tracker has been assembled and delivered to Jefferson Lab is on its way after 10 years since the first proposal

• Resistive Micromegas have been build and characterized

• 18 + 6 (spares) cylindrical MM for the CLAS12 barrel detectors

• 6 disks for the Forward Micromegas Tracker

• The Micromegas detectors have been taking data since fall 2017

• At nominal luminosity, detectors operate with a 2µA current on the mesh

• Unstable environmental conditions (temperature and discontinuous gas flow) damaged a few detectors that had to be placed

• Since December, we have stable condition and no detector issue

• Now CLAS12 is off for the HPS experiment (Heavy Photon Search)

• Fall 2019 : LD2 data taking

• Winter 2019-2020 : replacement of the central tracker by a radial TPC using the same mechanics and DREAM electronics (BONUS)

• 2021 : Central tracker is put back and data taking with nuclear target (carbon, …)


OUTLOOK

www.cea.frwww.cea.fr

THE MVT GROUP AT SACLAY:

D. ATTIE, S. AUNE, J. BALL, Q. BERTRAND, F. BOSSU, G. CHRISTIAEN, M. DEFURNE, J. GIRAUD, R. GRANELLI,, I.

MANDJAVIDZE, O. MEUNIER, Y. MOUDDEN, S. PROCUREUR, M. RIALLOT, J-Y. ROUSSE, F. SABATIE

Contact : [email protected]

REQUIREMENTS

• Compact tracker => Cylindrical detectors

• High magnetic field => High lorentz Angle => small gap, slow gas, high drift field, high gain,

deported electronics

• Low Energy proton/electron => Low X0

• High rate in the forward region => Low spark rate, fast gas

High gain, low spark => Resistive MM

Cylindrical Det. => Bulk

Low X0 => Light material, glued


CYLINDRICAL MICROMEGAS

Electric leak testSegmentation and preparation Gluing of the side carbon ribs on circular shape

Gluing of additional ribs Setting drift plane Gluing of the drift plane2019 MPGD Conference - CEA Saclay - Maxence

Vandenbroucke

27

• 3D probing machine measuring points :

• 270 points on top (drift side)

• 120 points under (readout side)

• Cosmic rays data for cross-check

List of measured points

X Y Z

PT-Arceau1_1 -3.588 231.995 -0.005

PT-Arceau1_2 28.993 231.995 33.071

PT-Arceau1_3 64.992 231.994 57.126

…

CLAS12 BARREL : GEOMETRY

Mechanical precision up to ~2mm in radius

28

Crosscheck using cosmic rays data reconstruction


PRE-PRODUCTION FORWARD DISK FOR CLAS12: ISSUE WITH RESISTIVE LADDERS

• 2 pre-production (2015) detector tested, One was not ok :

• High current due to a contact in the active area (can’t burn it with sparks)

• Current flows from the contact to ground (black dots)

• Large impacted zone due to ladders

• Drift electrode glued (intervention impossible)

• Carbon frame for gas distribution

Before the current appeared With current

40mm

0.5mm

Resistive StripsS. ProcureurS. Procureur

29

Efficiency maps

Current without ladders


PRE-SERIE BARREL DETECTORS TEST

Zoom on a part of the CLAS12 Barrel with thermal cam, with HV on and current of about 300 mA

Solutions :

No resistive interconnection (ladders)

Aluminum frame at the gas inlet

More ground connections

“C” Barrel detectors had no problem !

Fixing is possible by filing the area with a drop of polymer

Efficiency Map

without current with currentposition of the noisy strip

position of resistive contacts

30

S. Procureur S. Procureur

Fixing Procedure : A lot of HV Test -> Soft Cleaning (aspiration, antistatic roller)

imaging the problem -> HV + current (10-500µA) and IR imaging Cleaning -> water cleaning (karcher) -> drying (air+oven)

-> Sodium Chlorate 60ºC -> TrashOr if possible -> passivation with a drop of glue


NEW CONNECTION SCHEME,NEW PROBLEMS

SILVER PASTE ISSUE

• The CR4Z layer has been the first produced using this method

• => All 4 of them died after ~2 weeks of tests


=> Thermal imaging shows that high current appeared on the silver paste connection between the resistive layer and the PCB

TIME RESOLUTION

• Time Resolution with resistive detectors is ~25ns instead of ~15ns (depending on the conditions)

• Time resolution not critical for the CLAS12 barrel

• Inhomogeneities over the detector surface observed

Time Vs Det. X 2D Time distribution

Solutions : Electronic parameters optimization (~3ns) - done 1D (2D?) corrections (~5ns) but difficult with cosmic rays


THE ASACUSA MICROMEGAS TRACKER


[V]MESH

HV

410 420 430 440 450

Effic

iency

0.75

0.8

0.85

0.9

0.95

1

AMT Detector

RminC

RmaxC

RminZ

RmaxZ

2

Front-end crate

Cold head

Readout

connectors

Cold head

L = 61.5 cm

Micromegas

Anti Helmhotlz

coils

FIG. 2. Technical drawing of the AMT detector installedaround the cent ral CUSP trap. The detector in the drawing

is surrounded by the magnet ic coils, which are drawn witht ransparent rendering. The two cold heads, used for the cryo-

genic t rap system, on the two sides of the t rap are also visible.

and at 7 cm with several thermal isolat ion layers in be-tween. A B = 0-4 T magnet ic field with double-cusp con-figurat ion is provided by ant i-Helmholtz coils, as shownin Figure 3 together with the posit ion of the AMT detec-tor. The bore radius of the shield of the magnet is at 10cm.

-400 -300 -200 -100 0 100 200 300 400

-3

-2

-1

0

1

2

3

4

Bz

[T]

Z [mm]

FIG. 3. Schemat ic view of the nominal cusp magnet ic field

configurat ion on axis (R= 0 mm) and at the Micromegas de-tector layers (R= 78.5 mm, R= 88.5 mm).

The main design parameter of the detector was to re-solve ant iproton-nucleon annihilat ion vertex posit ion insuch an environment with a resolut ion of σver t ex ' 1 cm,in order to be able to dist inguish between annihilat ion

events occurring near the cent re of the trap and on thet rap elect rodes at 4 cm radius. The major limitat ionin the possible resolut ion is the mult iple Coulomb scat -tering inside the various t rap materials. In order to es-t imate the resolut ion the elect rodes, vacuum chambers,thermal isolat ion materials and the lever arms of eachlayers were taken into account . The overall gaussian un-certainty of a single t rack point ing at a cent ral vertexis est imated to be σCoul .

hi t ' 2 mm for a charged pionwith p = 600 MeV/ c momentum, and σCoul .

hi t ' 4.5 mmfor the same part icle with p = 300 MeV/ c momentum.From GEANT4 Monte Carlo simulat ion [8] of ant ipro-ton annihilat ion events, combined with t rack and vertexreconst ruct ion, we obtain an overall vertex resolut ion ofσver t ex ' 7 mm, which value is within the target resolu-t ion.

I I I . T H E A SA CU SA M I CROM EG A S T R A CK ER

A . M icr om egas det ect or

The AMT is a t racker detector with two half-cylinderlayers of Micromegas and with one full cylinder layer ofplast ic scint illator bars (providing t rigger signal) sand-wiched in between, all curved to fit into the cylindricalst ructure of the ASACUSA double-cusp trap. The over-all length of the detector is 61.5 cm with an act ive areaof ⇠ 40 cm. The inner and outer Micromegas layershave radius of 78.5 mm and 88.5 mm, respect ively. TheMicromegas detector operat ion principle is illust rated inFig 4. Elect rons are released, as a result of a passageof a charged part icle through a few mm wide drift re-gion, and they drift to a thin amplificat ion region sepa-rated by a transparent mesh elect rode. The high elect ricfield in the amplificat ion gap provides elect ron mult ipli-cat ion. The amplified signal is collected by micro st ripswhich are sampled by custom built front -end electron-ics. In case of the presence of external magnet ic field thedrift ing elect rons are under the influence of an addit ional~F = q(~v⇥ ~B ) Lorentz force.

The st ructure of the detector, from bot tom to top, ismade of a 250 µm FR4 printed-circuit board base, with5 µm thick gold-coated copper anode strips on top of thePCB. The amplificat ion gap of 128 µm is followed vert i-cally by a 25 µm thick stainless steel woven micro-mesh(wire of 18 µm and opening of 45 µm), supported by anarray of pillars. The pillars have diameter of 300 µm andpitch of 2 mm, and addit ionally smaller diameter pillarswereadded with diameter of 150 µm, between each largerpillar, in order to support the curvature of the micro-mesh. The fract ion of the surface area covered by thepillars relat ive to the total act ive area is ' 2 %. Abovethe micro-mesh a drift gap follows with a thickness of 3mm. The structure is closed with a 250 µm kapton driftcathode with 5 µm of copper on both sides, to provideexternal shielding. The st rip layout of the Micromegasis illust rated in Figure 5. In the case of AMT one layer

33

Characterization at Saclay :

Rmin C [strip]0 50 100 150 200 250 300 350 400 450 500

Rm

ax

C [

str

ip]

0

50

100

150

200

250

300

350

400

450

500

Rmin C [strip]0 50 100 150 200 250 300 350 400 450 500

Rm

ax

C [

str

ip]

0

50

100

150

200

250

300

350

400

450

500

2 x 6 Months of smooth data taking !

Effect of the magnetic field (80º Lorentz angle at low drift field)

B Field ONB Field OFF

S. ProcureurS. Procureur

AMT – MICROTPC ALGORITHM


Resolution Vs Track Angle

Antiproton distribution(calibration data)

Filtered distribution

34

Fitted signal in the drift space

The MicroTPC algorithm uses the time information on multi-strip clusters to extrapolate the track angle :

MicroTPC algo on anti-p data :

Track angle from a single layer barrel

tim

e

space

µTPCCut

CORRELATION BETWEEN SVT AND MVT

| PAGE 35

DAQ including both SVT and MVT data is working

Mapping and Geometry understoodTrack reconstruction with cosmic rays is working

Correlation bet. SVT tracks and BMT in Micromegas coordinates

Residuals* : ~ 400 µm for Z det.~ 5 mm for “C” det.

* MC simulations show that 400µm residuals correspond to a 350µm resolution


2D EFFICIENCY MAPS

| PAGE 36


sector1detectorBLayer2

layer1

sector2detectorA

sector3detectorC

Detectors’ efficiency with cosmic rays tracks from the SVT

No major defect in BMT detectors

SVT “ghost” due to shallow tracks in the SVT not reconstructed properly at that point

Eff. Vs HV (gain)

DREAM FRONT-END ELECTRONICS

2019 MPGD Conference Vandenbroucke

37

performances of the micromegas vertex tracker the … · 2019. 6. 4. · moudden, s. procureur, m....

Documents