EASIROC, an easy & versatile ReadOut device for SiPM

Post on 31-Mar-2023

0 views 0 download

Transcript of EASIROC, an easy & versatile ReadOut device for SiPM

EASIROC, an easy & versatile ReadOut device

for SiPMStéphane CALLIER, Christophe DE LA TAILLE,

Gisèle MARTIN-CHASSARD, Ludovic RAUX

With precious help of :Dominique CUISY, Jean-Jacques JAEGER,

Nathalie SEGUIN-MOREAU, Jean-Luc SOCHA

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 2

CONTENTS

• EASIROC Features

• EASIROC Measurements

• Projects using EASIROC

• Conclusion

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 3

EASIROC FEATURES (BASED ON SPIROC CHIP)

• 32-channel front-end readout• Individual 8-bit DAC for SiPM Gain adjustment• Energy measurement from 160fC to 320pC (1pe to 2000pe @

SiPM gain = 10^6)– 1 pe/noise ratio ~11– Variable gain preamplifier– Variable time constant CRRC² shaper (25 to 175ns)– Common 10-bit DAC for threshold adjustment– 2 multiplexed analog outputs (high gain, low gain) [tri state outputs]

• Trigger output– 1 pe/noise ratio ~24– Trigger on 1/3 pe (50fC)– 32 Trigger outputs– OR32 output– Trigger multiplexed output (latch included) [Tri state output]

• Individually addressable calibration capacitance• Low power : 4.84mW/channel, 155mW/chip

– Unused feature can be disabled to reduce power consumption– Power pulsing facility (idle mode with external signal)

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 4

EASIROC ANALOGUE CORE

Variable LowGain PA (4 bits)

HoldRead

Low GainMultiplexedOutput

Ch31_trigger

Channel 0

Common to the 32 channels

Channel 31

Variable HighGain PA (4 bits)

+

V_thLatch

RS

Channel0_trigger

Discri

OR32

Slow Shaper25 - 175 ns

LG Slow ShaperVariable Shaping

Time (3 bits)

15 nsRead Trigger

MultiplexedOutput

Low GainPreAmp.

HG Slow ShaperVariable Shaping

Time (3 bits)

0.1pF - 1.5pF

1.5pF

10-bitDAC

RS or Discri

0.1pF - 1.5pF

15pF

High GainPreAmp.

HoldRead

High GainMultiplexedOutput

+Slow Shaper25 - 175 ns

3pF

in_calib

Bipolar Fast Shaper8-bit DAC

0-5V

INCh0

ONc15p

3pF

Ctest

32inputs

1input

1output

1output

1output

1output

32outputs

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 5

EASIROC LAYOUT

Technology : AMS 0,35µm SiGe

Die size : 16.6mm²4.157 x 4.013 mm²

Package :– Naked (PEBS)– TQFP160

Dis

crim

inat

ors

Sha

pers

Pre

amps

32 S

iPM

inpu

ts

10-bit DAC

Ana

log

Mem

ory

32 tr

igge

r out

puts

8-bi

t inp

ut D

AC

s

2 MUX charge outputs (LG + HG) OR32 output

1 MUX trigger output

Bandgap

TQFP: heigth=1.4 mm

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 6

TEST BOARD

2 2 ADCsADCs

USBUSB

EASIROCEASIROC

ANALOGUE ANALOGUE OUTPUTsOUTPUTs

HV HV connectorconnector

32 32 SiPMSiPM connectionsconnections

FPGAFPGA

• Testboard allow easy acces to each EASIROC pin• Testboard layout & cabling @ LAL• Firmware using LAL USB interface• Labview software

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 7

SOFTWARE

• ASIC versatility :456 slow control bits• Acquisition system

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 8

Gain and dark rate uniformity correction

High voltage on the cable shielding

The input DACs allow to adjust HV channel by channel by slow control on each SiPM of the detector

2 examples for SiPM connection allowing input DAC use

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 9

INPUT DACs

DAC linearity

00,5

11,5

22,5

33,5

44,5

5

0 50 100 150 200 250 300

DAC

Volta

ge (V

)

-0,04

-0,03

-0,02

-0,01

0

0,01

0,02

0,03

0 50 100 150 200 250 300

DAC linearity on the 32 channels

• Input DAC to optimize SiPM bias voltage (SiPM gain)

• 8 bits on 4,5V range => LSB : 20mV

• 10µA sink capability

• Ultra low power (<1µW)

• Linearity ±2%• DAC uniformity

between the 32 channels : ~3%

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 10

ANALOG OUTPUT (CHARGE)

Pedestal uniformity for each Gain output <0.7%

Pedestal removed

High Gain

Low Gain

Hold scanperformed onanalogue data

[using external ADC]

Time (ns)

ADC value

Pedestal dispersion

1,234

1,236

1,238

1,24

1,242

1,244

1,246

1,248

0 5 10 15 20 25 30

Channel

HG VLG V

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 11

CHARGE MEASUREMENT

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 12

ANALOG OUTPUT

• Histogram for 1 to 10 pe- on both gains

Courtesy : Ryotaro HONDA

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 13

THRESHOLD DAC

• 10-bit DAC for Threshold adjustment (trigger)

• 10 bits on 1.3V range => LSB : 1.3mV

• Linearity ±0.3%

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 14

S-CURVES (TRIGGER EFFICIENCY)

OR32 output for 1pe on each channel

Dispersion : <5 DAC unit for 1pe(8fC/DAC unit) [Cf=200fF]

S-curve for 1pe- (160fC)

0

20

40

60

80

100

120

835 840 845 850 855 860 865 870 875 880 885

DAC code

Trig

ger e

ffici

ency

(%)

0

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

Pedestal

Individual pedestal

OR32 output For pedestal

PEBS

Courtesy: Waclaw KARPINSKI

PEBS is a project in Research & Development phase 

The purpose of the experiment is a precision measurement of the electron & positron cosmic rayflux in the energy range from 1 to 2000 GeV.

250 µm

RTWH Aachen

G. Saracino, Orsay, July 2, 2009 the MU-RAY Project: high-resolution muon radiography with scintillators 16

The MU-RAY project: high-resolution muon radiography with scintillators

Courtesy : Giulio SaracinoCourtesy : Giulio SaracinoINFN Napoli

International Collaboration

of physicists, geologists and volcanologists to perform muon radiography of geological structures, Mt. Vesuvius first of all

Des

ign and build muon telescopes

to be operated in difficult environments. Requirements:

modular, light, easy to transport and mount•

little need for maintenance•

very low power consumption•

Develop a methodology and a versatile instrument

Italy: Bologna,Firenze,Perugia,Napoli (INFN and Universities) Istituto Nazionale Geofisica e Vulcanologia)

Japan: Tokyo University and Hearth Research InstituteUSA: Fermilab

Given the mountain topology and the deep crater, there are ≈2 km of rock

to cross!

Location: Cable Cab(Seggiovia) 2.1 km

1.8 kmThe challenge of Mt. Vesuvius

Need for large detector areas!

International Collaboration of physicist,

Active target with MPPC readout

Tohoku

University

+ KEK

Fast time response • Work in a high beam

intensity

Large gain (105106)• Possible to detect 1 photon

Operation in the magnetic field

• Combination of Imaging and Spectrometer

Trigger possibility

Courtesy: Ryotaro HONDA

SIPMED : SIPMED : SIliciumSIlicium PhotomultiplierPhotomultiplier for for bioMEDicalbioMEDical imagingimaging

To To developdevelop a a novelnovel compact compact photodetectionphotodetection systemsystem withwith highhigh energyenergy andand temporal temporal measurementmeasurement capabilitiescapabilities for for radioradio--guidedguided surgerysurgery Task

1 : Characterization

of

single SiPMs

and

SiPMs

matrixes

performances (scintillation and

fluorescent light measurements) under

laboratory

and

real

medical

conditions (e.g.

temperature

up

to 37 °C)

Task

2 : Design and

conception of

a new optimized

SiPM

read-out

electronics

for both

scintillation and

fluorescent light measurement

Miniaturisation of

the

electronic

board

->

Task

3 : Development

and

validation of

new per-operative

prototypes :

1) a compact positron probe

2) a mono-pixel

fluorescent probe

3) a large field-of-view

and

ultra-compact

intraoperative

gamma-cameraCourtesy: Laurent MENARDIMNC

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 19

EASIROC SUMMARY

• 1pe- -> 2000pe- charge measurement

• 100% Trigger efficiency @ 1/3photo-electron

• Versatility & easy use– Testboard & Software ready to use, embedding acquisition system

• Very low power consumption– Full power pulsing capability– Unused stages can be shut down

• Large scale production in 2010– Already used in experiments (astrophysics, vulcanology, nuclear

physics, medical imaging)

• Any extra information available on http://omega.in2p3.fr

BACKUP SLIDES

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 22

• Reticle size : 18x25 mm2– 50-55 reticles/Wafer– 25 wafers

• Final arrangement:

– « Calice » chips produced:

• 7 Hardroc 2b => ~9000 chips• 1 Spiroc 2a => ~1250 chips• 1 Spiroc 2b => ~1250 chips• 1 Skiroc 2 => ~1250 chips

– Additionnal chips produced:

• 1 Spaciroc : JEM EUSO experiement=> ~1250 chips

• 1 Maroc 3 : for PMT readout=> ~1250 chips

• 3 Easiroc : for PEBS experiment=> ~3750 chips (5000 received)

• Production run for CALICE chips• => cost reduction for CALICE

Engineering run (2010)

SPIROC2BAnalog HCAL(SiPM)36 ch. 32mm²June 07 / June 08 / March 11

HARDROC2BDigital HCAL(RPC, µmegas or GEMs)64 ch. 16mm²Sept 06 / June 08 / Sept 09

SKIROC2ECAL(Si

PIN diode)64 ch. 64mm²March 11

E-CAL

H-CAL (analog or digital)

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 23

SPIROC : One channel schematic

50 -100ns

50-100ns

Gain selection

4-bit threshold

adjustment

10-bit DAC

15ns

DAC output

HOLD

Slow Shaper

Slow Shaper

Fast

Shaper

Time measurement

Charge measurement

TDC ramp

300ns/5 µs

12-bit Wilkinson

ADC

Trigger

Depth 16

Depth 16

Depth 16

Common to the 36 channels

8-bit DAC

0-5V

Low

gain Preamplifier

High gain Preamplifier

Analog memory

15pF

1.5pF

0.1pF-1.5pF

Conversion

80 µs

READ

Variable delay

0.1pF-1.5pF

IN

Discri

Gain

Flag TDC

DAQASIC

Chip ID register

8 bits

gain

Trigger discri

Output

Wilkinson ADC Discri

output

gain

Trigger discri

Output

Wilkinson ADC Discri

output

..…

OR36

EndRamp

(Discri

ADC Wilkinson)

36

36

36

TM (Discri

trigger)

ValGain

(low

gain or high

Gain)

ExtSigmaTM

(OR36)

Channel 1

Channel 0

ValDimGray

12 bits

Acquisition

readoutConversion

ADC

+

Ecriture RAM

RAM

FlagTDC

ValDimGray

12

8

ChipID

Hit channel

register

16 x 36 x 1 bits

TDC rampStartRampTDC

BCID 16 x 8 bits

ADC rampStartrampb

(wilkinson

ramp)

1616

ValidHoldAnalogbRazRangN

16ReadMesureb

Rstb

Clk40MHz

SlowClockStartAcqt

StartConvDAQb

StartReadOut

NoTrig

RamFull

TransmitOn

OutSerie

EndReadOut

Chipsat

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 25

SPIROC ReadOut: token ring

Acquisition A/D conv. DAQ IDLE MODEChip 0

Chip 1 Acquisition A/D conv. DAQ IDLE MODEIDLE

Chip 2 Acquisition A/D conv. IDLE MODEIDLE

Chip 3 Acquisition A/D conv. IDLE MODEIDLE

Chip 4 Acquisition A/D conv. IDLE MODEIDLE DAQ

1ms (.5%) .5ms (.25%) .5ms (.25%)

1% duty

cycle 99% duty

cycle

198ms (99%)

• Readout architecture common to all calorimeters• Minimize data lines & power

5 ev

ents

3 ev

ents

0 ev

ent

1 ev

ent

0 ev

ent

Chip 0 Chip 1 Chip 2 Chip 3 Chip 4

Data bus

ILC beam

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 26

ILC beam structure and SPIROC running modes

Acquisition1ms (.5%)

A/D conv..5ms (.25%)

DAQ.5ms (.25%)

1% duty cycle

IDLE MODE

99% idle cycle

198ms (99%)

time

Time between two trains: 200ms (5 Hz)

Time between two bunch crossing: 337 ns

Train length 2820 bunch X (950 µs)

Acquisition1ms (.5%)

A/D conv..5ms (.25%)

DAQ.5ms (.25%)

1% duty cycle

IDLE MODE

99% idle cycle

198ms (99%)

time

Time between two trains: 200ms (5 Hz)

Time between two bunch crossing: 337 ns

Train length 2820 bunch X (950 µs)

Two

orders

of

magnitude saved

on the

consumption

by using

the

ILC beam structure and

the

power

pulsing

Acquisition A/D conv. DAQ IDLE MODEChip 0

Chip 1 Acquisition DAQ IDLE MODEIDLE

Chip 2 Acquisition IDLE MODEIDLE

Chip 3 Acquisition IDLE MODEIDLE

Chip 4 Acquisition IDLE MODEIDLE DAQ

A/D conv.

A/D conv.

A/D conv.

A/D conv.

Acquisition A/D conv. DAQ IDLE MODEChip 0

Chip 1 Acquisition DAQ IDLE MODEIDLE

Chip 2 Acquisition IDLE MODEIDLE

Chip 3 Acquisition IDLE MODEIDLE

Chip 4 Acquisition IDLE MODEIDLE DAQ

A/D conv.

A/D conv.

A/D conv.

A/D conv.Chip 0 Chip 1 Chip 2 Chip 3 Chip 4

5 ev

ents

3 ev

ents

0 ev

ent

1 ev

ent

0 ev

ent

Data bus

Readout based on token ring mechanism initiated by DAQ

One data line activated by each chip sequentially

Readout rate few MHz to minimize power dissipation

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 27

OTHERS

• Default value for Slow Control– Chip ready to use after a simple SC reset (except

Threshold DAC)

• Debug embedded feature :– Analog Probe system allow to monitor each critical point

in the chip• 32 PreAmplifiers High Gain output • 32 PreAmplifiers Low Gain output • 32 Slow Shapers High Gain output • 32 Slow Shapers Low Gain output • 32 Fast Shapers output

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 28

S-CURVES (TRIGGER EFFICIENCY)

OR32 output for 1pe on each channel

Dispersion : <5 DAC unit for 1pe(8fC/DAC unit) [Cf=200fF]

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 29

INPUT PREAMPLIFIERS

• Bi-gain low noise preamp– Low noise charge preamplifier

capacitively coupled = voltage preamplifier

– Gain adjustable with 4 bits common to all preamps : Cf=0.1, 0.2, 0.4, 0.8 pF

– Positive input pulse– Power : 2 mW (unpulsed)

• High gain :– 15pF coupling capacitor– 8 mV/pe in High Gain– Noise : 1.4 nV/sqrt(Hz)

• Low gain at preamp level– 1.5pF coupling capacitor– 0.8 mV/pe, MAX : 2000 pe (300pC)

[@Cf=400fF]

Low GainPreAmp.

0.1pF - 1.5pF

1.5pF

0.1pF - 1.5pF

15pF

High GainPreAmp.

0.15

0.10

0.05

0.00

Vout

(V)

500400300200100Time (x10-9s)

Cf=100fF Cf=100fF Cf=200fF Cf=300fF Cf=400fF Cf=500fF Cf=600fF Cf=700fF Cf=800fF Cf=900fF Cf=1pF Cf=1.1pF Cf=1.2F Cf=1.3pF Cf=1.4pF Cf=1.5pF

High gain charge output at different Cf Qinj= 1.5pC

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 30

TRIGGER TIME WALK

trigger time walk vs injected charge

0

2

4

6

8

10

12

0 1000 2000 3000 4000 5000

charge (in fC)

time

(in

ns)

callier@omega.in2p3.fr - NDIP 2011, Lyon, France, 05/07/2011 31

ANALOG CROSSTALK

• Very low electronic cross-talk : 0.3% (long distance cross talk due to slow shaper voltage reference: If this voltage decoupled with 100µF, it becomes negligible ~0.04%)