1

Production of 500 Pixel Hybrid Photon DetectorsProduction of 500 Pixel Hybrid Photon Detectors

for the RICH counters of for the RICH counters of LHCbLHCb

Thierry GysThierry GysCERN CERN –– Geneva Geneva –– SwitzerlandSwitzerland

on behalf of the on behalf of the LHCbLHCb--RICH groupRICH group

44thth International Conference onInternational Conference onNew Developments in Photon Detection New Developments in Photon Detection

1919--24 June 200524 June 2005BeauneBeaune -- FranceFrance

T. Gys – Pixel-HPD’s - Beaune 2005 2

Talk layoutTalk layout

The LHCb experiment, the RICH1 and RICH2 detectors and their photon detector requirements

The pixel Hybrid Photon Detector

Production flow-chart

Pre-series tubes and test beam results

Local magnetic shielding, distortions and calibrations

Conclusions

T. Gys – Pixel-HPD’s - Beaune 2005 3

The The LHCbLHCb experimentexperiment

Single-arm spectrometer for precision measurements of CP violation in B-meson systemSearch for signals of ‘New Physics’ beyond Standard ModelParticle identification over momentum range ~2-100 GeV/c,using 2 Ring Imaging CHerenkov (RICH) detectors with silica aerogeland 2 fluorocarbon gas radiators

T. Gys – Pixel-HPD’s - Beaune 2005 4

The RICH1 detectorThe RICH1 detector

HPD plane:7 columns,

14 tubes each

Magnetic shield boxVerticalX-section

(Be)

(glass)

T. Gys – Pixel-HPD’s - Beaune 2005 5

The RICH2 detectorThe RICH2 detector

HPD plane:9 columns,

16 tubes each

Magnetic shield box

Horizontal X-section

Al superstructure

(glass)

(glass)

T. Gys – Pixel-HPD’s - Beaune 2005 6

RICH photon detector requirementsRICH photon detector requirements

Good sensitivity over the wavelength range 200nm to 600nm (aerogelresponse and scattering, and chromatic dispersion in gases)

Large detection area of 2.6m2 with active area fraction of ~65%

Position-sensitive with fine granularity of 2.5mm x 2.5mm (optimum of pixel vs chromatic vs emission point errors)

Single photon sensitivity ⇒ good signal-to-noise ratio*

Fast response for LHC bunch-crossing rate of 40MHz*

Radiation tolerant (3krad per year)*

* see talk of K. Wyllie

T. Gys – Pixel-HPD’s - Beaune 2005 7

Main features (already presented during Main features (already presented during previous previous BeauneBeaune Conferences:Conferences:

Close collaboration with industry

Quartz window with thin S20 photo-cathode:•Typical ∫ QE dE > 0.7eV

Cross-focussing optics (tetrode structure):•De-magnification by ~5•Active diameter 75mm

⇒ 484 tubes for overall RICH system•20 kV operating voltage (~5000 e– [eq. Si])

32×32 pixel sensor array (500μm×500μmeach)Encapsulated binary electronics readout chip

PixelPixel--HPD descriptionHPD description

http://www.cern.ch/~gys/LHCb/PixelHPDs.htm

Vacuum photon

detector

Anode

T. Gys – Pixel-HPD’s - Beaune 2005 8

HPD flow chart: anode partHPD flow chart: anode part

Readout chip (IBM - F)

Detector chip (Canberra - B)

Ceramic carrier (Kyocera - JP)

High T bump-bonding (VTT - FIN)

Brazing (DEP - NL) and gold-plating (CERN)

Packaging (HCM - F)

Wafer probing

Assembly probing

Anode testing

Visual inspection and plating control

T. Gys – Pixel-HPD’s - Beaune 2005 9

Completed anode partCompleted anode partDie-attach process

@ 410°C

T. Gys – Pixel-HPD’s - Beaune 2005 10

HPD flow chart: HPD tube part (DEP)HPD flow chart: HPD tube part (DEP)

Tube body assembly

Photo-cathode deposition and vacuum sealing

Anode incoming inspectionand testing

Anode testing Final HPD testing

HPD cabling and potting

QE measurement and anode testing

Vacuum bake-out@ 300°C

T. Gys – Pixel-HPD’s - Beaune 2005 11

Completed HPD tube + QACompleted HPD tube + QA

Thorough HPD testing

LHCbLHCb quality assurance:quality assurance:

2 photon detector test facilities

1 HPD per work day per facility over production period of ~18 months

Production rate of 30 HPD’s per month over production period of ~18 months

T. Gys – Pixel-HPD’s - Beaune 2005 12

HPD preHPD pre--series (9 tubes): performance resultsseries (9 tubes): performance results

Generally well above specs

See next picture

Quantum efficiency

79-89%Typ. 85%P.e. detection efficiency

<10-3Max. 10-2 rel. to signal

Ion feedback rate

0.03–3kHz/cm2Max. 5kHz/cm2Dark count rate

< 1uATyp. 1uA @ 80V bias

Leakage current

Typ. 1200e-Typ. 160e-

<2000e-<250e-

Min. thresholdNoise

>99%>95%Pixel responseResultsSpecificationItem

T. Gys – Pixel-HPD’s - Beaune 2005 13

HPD preHPD pre--series: QE curvesseries: QE curves

Typical and minimum QE specs based on HPD prototype results

T. Gys – Pixel-HPD’s - Beaune 2005 14

Beam Pipe

Radiator vessel

HPD housing

HPD column mechanics

Mirror

Thin Al foil

Test beam withTest beam witha prototype RICHa prototype RICH

Beam

6 HPD’s on 3 columns

T. Gys – Pixel-HPD’s - Beaune 2005 15

CherenkovCherenkov rings in Nrings in N2 2 focussed on each HPDfocussed on each HPD

3500

0

10GeV/c e-

10GeV/c π-

Light yields agree to within 10% of

expectations

T. Gys – Pixel-HPD’s - Beaune 2005 16

0 10 20 30 40 50 60 70 80 900

10

20

30

40

50

60

70

0

5000

10000

15000

20000

25000

Number of hits: 3297589

Number of events: 99477

Run Number: 1113

10 GeV/c pion

Left Centre RightColumn number

Row number

CherenkovCherenkov rings in Crings in C44FF10 10 radiatorradiator

0 10 20 30 40 50 60 70 80 900

10

20

30

40

50

60

70

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Number of hits: 48Number of events: 1

Run Number: 1113

10 GeV/c pion

Left Centre RightColumn number

Row number

Very clean rings

DAQ display showing hits on anode planes (pixel chips)Single event Aggregate events

0 G

20 GRICH 1

HPD shields

Photocathode plane

To avoid image loss and minimize distortions, local shielding of HPD’s required to reduce B field below 10G (1mT) inside HPD volume

140mm

0.9mm

Simulated magnetic field inside a Mu-metal array enclosed in RICH1 shielding box

Local magnetic shieldingLocal magnetic shielding

Mu-metal shield grounded and insulated with ~250μm-thick Kapton foil

T. Gys – Pixel-HPD’s - Beaune 2005 18

Magnetic field not uniform and varying from tube to tube⇒ with test pattern, reconstruct pixel hit – photon hit position correspondence for

each HPDDistortion correction must not degrade pixel size error

B=0 B║ 30 G (3mT axial) B┴ 50 G (5mT transverse)

Magnetic distortionsMagnetic distortions

Test pattern measurements with locally-shielded HPD

T. Gys – Pixel-HPD’s - Beaune 2005 19

Parameterization, test pattern and protocolParameterization, test pattern and protocolRadial distance of hit on chip vs radial distance of LED source on entrance plane

y = -0.0016x2 + 0.2425x

y = -0,001x2 + 0,2043x

y = -0,0004x2 + 0,1772x

0

1

2

3

4

5

6

7

8

0 5 10 15 20 25 30 35 40 45Entrance plane radius [mm]

Rad

ius

on

chip

[mm

]

Radius on chip [mm] - 5.0 mT

Radius on chip [mm] - 3.0 mT

Radius on chip [mm] - 0 mT

P li (R di hi [ ] 0

0 GLarger central light spot

Smaller spots at increasing distances from the central

B field calibration protocol:• Projection of a static pattern on the detectors

plane in the experiment (proposals for RICH1 and RICH2)

• Combined approach: parameterization, sampling and filling of a look-up table

• Additional use of data for monitoring

Non uniform radial dilatation: 2nd order polynomial fit.

Non uniform rotation (S-distortion): 3rd order polynomial fit, (a1=0)

B║ 1, 3 and 5mT

Radius on photo-cathode

Radi

us o

n ch

ip40mm

8mm

T. Gys – Pixel-HPD’s - Beaune 2005 20

ConclusionsConclusions

Production of 500 pixel-HPD’s on-going

HPD pre-series performance good

Test beam confirm HPD lab tests

Magnetic distortions and calibrations: protocol exists and tested; Mu-metal insulation concept and implementation OK;

Electronic aspects: see talk of K. Wyllie