Proposal for MAPMTs as Photodetectors for the LHCb RICH

transcript

RICH meeting, 3.11.99 F.Muheim 1

Proposal for MAPMTs as Proposal for MAPMTs as Photodetectors for the Photodetectors for the

LHCb RICHLHCb RICH

Proposal for MAPMTs as Proposal for MAPMTs as Photodetectors for the Photodetectors for the

LHCb RICHLHCb RICH

Franz Muheim

University of Edinburgh

on behalf of the MAPMT group

OutlineOutline

Introduction Multianode Photo Multiplier Tubes R&D Results Baseline Design Conclusion

Photo Detector RequirementsPhoto Detector Requirements

Single photon sensitivity (200 - 600 nm) with quantum efficiency > 20%

Good granularity: ~ 2.5 x 2.5 mm2

Large active area fraction: 73% LHC speed read-out: 40 MHz

Photo detector area: 2.9 m2

Options: MAPMT or HPD

MAPMT MAPMT

Combines single photon sensitivity with good spatial resolution

8x8 dynode chains Gain: 3.105 at 800 V

Manufacturer: Hamamatsu 1 mm flange removed,

packing fraction increases by 14 %

Multianode Photo Multiplier Tube

MAPMT R7600-03-M64MAPMT R7600-03-M64

Bialkali photo cathode, QE = 22% at = 400 nm

UV glass window replaces borosilicate, QE dE increased by 50 %

Quantum efficiency

Quartz LensesQuartz Lenses MAPMT active area fraction:

38% (includes pixel gap) Increase with quartz lens

with one flat and one curved surface to 85%

Bench Tests Bench Tests

Collection efficiency not very uniform (~20%)

Gap between pixels: 0.2 mm

40 MHz read-out electronics Average signal/ pedestal width = 40:1 Signal loss: 11.5 % (includes 2.5% for

no multiplication at 1st dynode)

Pixel scan with LED Single channel spectrum (LED)

Test Beam Set-up Test Beam Set-up

Single MAPMT Test BeamSingle MAPMT Test Beam

Air radiatorLensPressure

Run 2059Yes

49 mbar

Run 2057No

49 mbar

Run 2042Yes

960 mbar

Run 2039No

960 mbarNumber ofdetectedphotoelectrons

DataSimulation

0.300.29

0.320.32

1.141.16

0.930.89

Good agreement

Photo electron yield CAMAC electronics RICH 1 prototype

Single MAPMT Test BeamSingle MAPMT Test Beam

Good agreement

Cherenkov angle resolution

CAMAC electronics RICH 2 prototype

– Focal length: 4 m Angular resolution

– 0.27 mrad (data)– 0.26 mrad (MC)

LHC Speed ElectronicsLHC Speed Electronics

LHC Speed F/E ElectronicsLHC Speed F/E Electronics

Test Beam Set-upTest Beam Set-up Cluster with quartz lenses Bleeder board Cluster: 40 MHz Read-out

9 MAPMTs read out with 6 boards

5 threshold cut, common-mode subtracted

Lots of photons, but cross-talk

Cluster TestCluster Test

CF4 Radiator, 700 mbarHV = -1000 V

Probability that pixel y causes hit in pixel x

Cross-TalkCross-Talk

Asymmetric cross-talk (board 9)

Correlated to neighboring APV Sample channels

Not correlated with neighboring pixels in tube

Symmetric cross-talk

Correlated with APV input neighbors (ceramic)

Cross-talk source is electronics

MAPMT do not have large cross-talk

Probability that pixel y causes hit in pixel x

Cross-TalkCross-Talk

Cross-talk correction applied

Observe 6.4 photo electrons per event

Background: 0.41 p.e. Few dead pixels Yield of different tubes

With quartz lenses

Photon YieldsPhoton Yields

0.57 1.07 0.64

1.03 0.00 1.08

0.61 0.88 0.69

Photon YieldsPhoton YieldsNo lenses Quartz lenses

Ratio with/without lenses = 1.45, expected 1.50

Analysis includes– Common-mode subtraction– Cross-talk correction– Background subtraction – Signal loss & dead pixel correction

Results:

Quartz Lenses Yes No

DataSimulation

7.36.9

5.24.6

Good agreement

Preliminary

Single events Number of photo electrons

Charged ParticlesCharged Particles

Charged particles traversing the lens & MAPMT produce background hits

Angle scan

MultiplicitiesMultiplicities

Multiplicity from charged particles – [5..10] for for

most angles– up to 30 for angles

around 45o

For MAPMTs, charged particles are a small background

Magnetic Field TestsMagnetic Field Tests

LED Pin hole mask -metal shield

MAPMT tested with Helmholtz coil 0, 10, 20, 30 Gauss

No ShieldingNo Shielding

MAPMTs are insensitive to transverse magnetic fields up to 30 G

Expect mainly By field

– By = 21 ..27 G (RICH 1)

– By = 150 G (RICH 2) reduce by ~ 15 with shielding of planes

Sensitive to longitudinal fields 10 G, at 30 G lose 50% top or bottom row (18% average)

With With -Metal Shielding-Metal Shielding

-metal: t = 0.9 mm Extension:

d = 10, 13, 32 mm Reduced loss at 30 G

– 7 .. 25 % worst row (d=10,13 mm)

– no structure (d = 32 mm)

Expect low Bz field

– Bz = 0 .. 5 G (RICH 1)

– Bz = 0 .. 38 G (RICH 2) reduce by ~ 15 with shielding of planes

MAPMT R & D SummaryMAPMT R & D Summary

Successfully tested close-packed 3x3 array of MAPMTs– Quartz lenses work as expected

Measured photon yield in agreement with simulation

Demonstrated 40 MHz read-out

Commercial MAPMT fulfilsLHCb RICH specifications

Baseline DesignBaseline Design

Pointing geometry 4x4 array,

1024 channels Bleeder board with

8 F/E chips -metal shield Pixel size at lens:

3.0 x 3.0 mm2

Filling factor: 0.79

Tilted Modules

Half PlanesHalf Planes

RICH 1 RICH 2

= 440 mrad

5 columns

10 rows

Total 232 modules, 3504 tubes

Outermost modules only partially equipped

= 240 mrad

6 columns

11 rows

Little distance between Vertex tank and RICH 1 Tracker 1 must also fit

Cooling for 8.8 W /module

IntegrationIntegration

MAPMT pitch: 26.7 mm Module pitch: 108.8 mm -metal shield: 0.4 mm Mounting frame: carbon

fibre, G10

MAPMT geometry RICH 1 integration

Vertex tank

RICH 1

F/E Electronics F/E Electronics

Characteristics– Spread: 3 – Signal/pedestal width: 60:1– Dynamic range: 3 photons

5000 … 1’560’000 e Attenuation: 6 F/E chip input:

– Noise/ dynamic range: 833 ... 260’000 e– ADC bits: 9

Occupancy: 3 %

Single photon signal: 300’000 e

F/E Electronics BaselineF/E Electronics Baseline

APVm chip not suitable (shaping time) SCTA128 is baseline (analogue) Changes necessary to existing chip

– Back-end (32 multiplexing), same as for the vertex detector

– Gain adaptation for MAPMT signals, attenuation, additional work

Alternative: BEETLE chip – when it becomes available

L0 & L1 ElectronicsL0 & L1 Electronics Level 0:

– # of modules / # of chips: 232 / 1856– # of channels per module: 1024– # of channels total: 224256– # of data links: 7424

Level 1:– Bandwidth (3% occ.) 85/7.7 Gbits/s

with/ without Zero suppression – # of VME modules: 78– # of multiplexers: 5

PerformancePerformance

Performance study Guy Wilkinson Preliminary results:

– Identification efficiencies: 86 %, K : 87 %

– Fake rates: 1.4 %, K : 3.0 %

ScheduleSchedule

Photo detectors must be ready by 1/7/2004

Testing takes 2 years

Must place order by 1/3/2001

Photo detectors are on critical path

F/E electronics design by 1/10/2000

1/7/2004today

MAPMT Test StationMAPMT Test Station

Automated test bench– LED or Laser light source– Optical stages – Measure gain of each tube, pixel scan– HV scan

QE measurements (~10% of tubes)– Monochromator– Calibrated standard

Measure Cherenkov light (~10% of tubes) source in quartz bar & MWPC

MAPMT Costs MAPMT Costs

Unit cost Cost Subtotal Tube [kSFr] [kSFr] [kSFr]

– MAPMT 0.931 3262

– Lenses 0.070 245 3507

Level 0– F/E chip, hybrid 0.200 371

– Motherboard 2.000 464

– TTC, DCS, Data Links 135 970

Level 1– 9U VME boards5.000 390

– Links, crates, MUX, RU, etc 199 589

Total cost: 5066

Risk AssessmentRisk Assessment

Baseline design is very close to what we have already tested MAPMT photon yield and resolutionLHC speed read-out electronicsClose packing (quartz lenses)

Commercial photo detectorPossible delays, Note: LHCb is tomorrowManpower Cost

Risk AssessmentRisk Assessment

PerformancePhoton yield, angular resolutionCharged particlesMagnetic stray fields

Electronics Adapt F/E chipNot on critical path

StabilityRadiation damage HV

Risk / ImprovementsRisk / Improvements

Further improvements possible– 2-3% higher QE is possible (manufacturer)

9 -14 % more photons– Incorporate lens into vessel window

8% more photons– Optical coupling between lens & MAPMT

8% more photons– Use 4 threshold cut

Smaller signal loss– Binary electronics Cost savings

is a viable choice

as photo detector

ConclusionsConclusions Results of the MAPMT R&D program

– Device performs according to specifications

– LHC speed read-out demonstrated Baseline design presented

– Mechanics, Electronics, Integration – Schedule, Cost

Proposal for MAPMTs as Photodetectors for the LHCb RICH

Documents