construction of the silicon tracker for the r3b experiment

Post on 18-Apr-2022

2 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

Construction of the silicon tracker for the R3Bexperiment.

M.Borri (STFC)on behalf of the teams at

Daresbury Laboratory, Edinburgh and Liverpool Universities.

Outline:

• FAIR and R3B.

• Overview of Si tracker.

• Mechanics, sensors, ASIC, DAQ.

• Electrical and in-beam tests.

• Conclusion.

FAIR.

• Facilities for Antiproton and IonResearch (FAIR).

• It is an expansion of GSI.

• Secondary beams of highlyunstable nuclei (up to 1 GeV/u).

• Produced by primary heavy ionsbeam hitting production target.

• Separated in-flight within 100’s ofns: Super FRS.

• High-energy branch for secondarybeams up to Bρmax of 20 Tm.

R3B.

• Kinematically complete measurementsof reactions with high-energy radioactivebeams.

• Unprecedented high efficiency,acceptance, and resolution.

• Reactions with relativistic radioactivebeam (R3B).

• Emphasis on nuclear structure anddynamics.

• Located in high-energy branch ofSuper-FRS.

Silicon tracker.

• Precise tracking and vertexing,as well as energy andmultiplicity measurements.

• Conical shape.

• Detect light charged particles(like protons) from the targetregion.

• Double-sided micro-strip siliconsensors wire bonded to acustom made ASIC.

• 30 detectors:6 in inner layer;12 in each of the 2 identicalouter layers.

• Total sensitive area∼5600 cm2.

• 912 ASICs equivalent to116,736 front-end channels.

• Each outer layer: 384 ASICs(49152 chs);Inner layer: 144 ASICs(18432chs).

• Operated in vacuum.

Mechanics.

• Support structure made of Al.

• Cooling system based on Cu (pipes andblocks) and H2O+Glycol (coolant).

• New Cu-blocks: better alignment; goodthermal contact.

• ASIC temperature range: 0-70oC.

• Vacuum chamber made of Al.

• Total volume of ∼0.148m3.

• ∼2mm wall: low material budget (2.2% of X0).

• Reached target vacuum: 1e-5 mbar.

Sensor.

Sensor specificationsType double sided

Bulk doping n-typeBulk thickness 300 µm [ABC] 100 µm [BD]

Strip pitch 50 µmStrip width 38 µmStrip length variable

Strip stereo angle 16.2o

Isolation implant p-sprayGuard rings 9 floating

I-leak per strip 3.25 nA/cm [<100 nA]Strip capacitance 2.3 pF/cm [<80 pF]

• BD detector for inner layer.

• ABC detector for outer layers.

• Starting production from ABC detectors.

• Sensor successfully assembled to carbon-fibre frame.

• It has I-leak<8µA@60V (op.voltage).

ASIC.

• Readout based on a daisy chainsequence.

• Four ASICs are connected in a chain.

• They are not adjacent to each other.

• Rate in each ASIC proportional to theassociated strip length.

• Designed at STFC RutherfordAppleton Lab.

• The R3B ASIC is intended for readingout signals from the N and P strips.

• The chip provides three informations:

1. strip number.

2. energy deposit.

3. time of the hit.

ASIC specifications

Size 13x6 mm2

CMOS process AMS 0.35 µmPower consumption < 1.5 W/ASICChannels per chip 128

Data rate < 5 kHz/channelEnergy range 0-50 MeVTime stamp 200 MHz

ASIC testing.

• Tests performed on bare ASICs.

• Front-end characterization performed withinternal pulse.

• Comparing ASIC with its specifications.

• Particular focus on : threshold tuning;front-end linearity; signal sampling; ADC toenergy.

Channel [a.u.]

0 20 40 60 80 100 120

Pu

lse-

mea

n [

AD

C]

2180

2200

2220

2240

2260

2280

2300PulsePeriod@200us, PulseAmpl@0.5V, PeakTime@8.0us

Asic1

e-mode

h-mode

PulsePeriod@200us, PulseAmpl@0.5V, PeakTime@8.0us

DAQ.

• Control ASICs: set registers,distribute clock.

• Synchronize tracker withother sub-detector systems.

• Transfer data to experimentDAQ (NDAQ).

• MIDAS software: wellestablished in UK nuclearcommunity.

• Total of 3 uTCA crates: 10FEE cards in each crate.

• 1 FEE card for eachdetector.

• Current cable hasmeasurable x-talk, exploringother types.

Detector testing.

Channel[a.u.]

0 256 512 768 1024 1280 1536 1792 2048

Std

.Dev

.[A

DC

]

0

1

2

3

4

5

6

7

8

9

10Pulse-std.dev. Vs Channel

N-side

sµPeak-time 1.0

sµPeak-time 4.0

sµPeak-time 7.0

Pulse-std.dev. Vs Channel

Channel[a.u.]

0 256 512 768 1024 1280 1536 1792 2048

Std

.Dev

.[A

DC

]

0

1

2

3

4

5

6

7

8

9

10Pulse-std.dev. Vs Channel

P-side

sµPeak-time 1.0

sµPeak-time 4.0

sµPeak-time 7.0

Pulse-std.dev. Vs Channel

• Test-pulse injected in each individualchannel.

• Input capacitance ∝ strip length.

• I leak ∝ strip length.

• Higher noise for longer strips.

• More uniform and lower noise with highshaping time.

N.B. N-side has a broken daisy chain.

In-beam test @ GSI.

Burkhard Kolb, Phone +49-6159-71 2667, E-Mail: beamtime@gsi.de 10/14/2014 3:48:00 PM

2 / 2014 October 2014 Schedule as of 14-Okt-2014

Week 40 Week 41 Week 42 Week 43 Week 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

U207, Backe/Block, 48Ca (EZR), Y7 SHIPTRAP

U259, Düllmann, 48Ca, X8 U207, Backe/Block, 48Ca, Y7 SHIPTRAP UB04,

Groening/Spiller, U, UNILAC

U259, Düllmann, 48Ca, X8 U282, Ortner/Blazevic, Ca, 3.6MeV/u, pµA, Single Shot

Experiment, 5Hz, Z6

UBIO, Scholz, Ca, X6

U259, Düllmann, 48Ca, X8

U000, Barth/Hollinger, p (MUCIS), UNILAC

S438 Simon, Ca, block sharing, HTC S401Heuser/Kis, Ca, block sharing, HAD

S000,

Spiller, N,

S000, Spiller, U, SIS S411,

Plaß/Purushothaman, U, FRS

S438, Taeib,Simon, U,

Nach HTC,

S411, Plaß/Purushothaman, U,

FRS

S000, Spiller, U, SIS

S438 Simon, Ca, block sharing, HTC S401Heuser/Kis, Ca, block sharing, HAD

S000,

Spiller, N,

S000, Spiller, U, SIS

SMAT, Trautmann, 300-400 MeV/u, Sm, HTA

S401Heuser/Kis, Ca, block sharing, HAD

S000,

Spiller, N,

SMAT, Trautmann, U, FRS

S417, Nociforo/Simon, U, 300-1000

S411, Plaß/Purushothaman, U,

FRS

S438, Taeib,Simon, U,

HTC, FRS

S411, Plaß/Purushothaman, U,

FRS

E082, Litvinov, Sm, FRS-ESR, ESR E000, Steck, U, soweit möglich,

ESR E098/E116, Bräuning-Demian, U, ESR, HTA

• GOAL: study reactions produced by a Ca48 beam hitting a CH2 target.

• MEAN: detect correlations of signals produced by particles in CALIFA and Si-tracker.

• OUTCOME: we acquired two data sets which could enable finding correlations.

• BUT: To achieve this goal we had to adopt several work around to overcome severaldifficulties.

• The scientific goal allowed us to test our detector system in terms of:

1. building.

2. mounting.

3. operating.

4. taking data.

5. analysing data.

In-beam test @ GSI.• 3 detectors to GSI.

• Det101 and Det104 are in front of CALIFA petals.

• Det103 does not have a CALIFA petal behind it.

• Only operated a fraction of Det101.

• Main problems: data corruption, data repetition, highleakage current.

• Problems caused by:cabling (ongoing),firmware (fixed),detector assembly (ongoing).

• Data analysis started.

Exp.S438, cave CRun290 3777

Target Thickness Beam type Beam energyCH2 4.98mm Ca48 550MeV/u

ADC vs TS: Pside (side1)

TS[s]

1048.5 1048.55 1048.6 1048.65 1048.7 1048.75 1048.8

310×

[A

DC

]

500

1000

1500

2000

2500

3000

3500

4000

spill ~25s ADC vs TS: Pside (side1)

Conclusion.

• The R3B silicon tracker is under construction.

• Mechanics, sensors, ASICs and DAQ are in an advanced stage.

• Tests allowed for an improvement of the setup.

top related