microelectronics at in2p3...not reinventing the wheel accumulate know-how building bllocks = an...

22
C. de La Taille Technical Director IN2P3 Microelectronics at IN2P3

Upload: others

Post on 05-Jul-2020

5 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

C. de La Taille Technical Director IN2P3

Microelectronics at IN2P3

Page 2: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

31 may 2013 CdLT HEPIC13 conference

IN2P3, an institute in CNRS

• CNRS :

– Under the authority of the ministry for higher education and research

– 11 500 researchers, 14 200 ITA

– 3 G€ annual budget

– 1 200 laboratories

– 10 thematic institutes, including 2 national institutes : IN2P3 and INSU

• IN2P3 :

– 2 400 CNRS staff, researchers, engineers and technicians; 600 university and other staff

– Running budget from CNRS : 40 M€

– 24 laboratories and platforms

– 40 large international projects

• Similar (but independent) organization for CEA and CEA/IRFU

2

Page 3: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

Répartition des 3058,2 ETP de l'IN2P3

(source : ISIS 2009)

550,9

747,2

526,9

658,0

272,2

302,9

I - Physique des particules et

neutrinos

II - Physique nucléaire et

hadronique

III - Astroparticules

IV - Interdisciplinaire

V - Développemenbts

technologiques

VI - Moyens communs et

management

31 may 2013 CdLT HEPIC13 conference

IN2P3 missions

• Promote and federate research in nuclear physics, particle physics, astroparticle physics

• Coordinate the programs in the name of CNRS and Universities, in partnership with CEA

• Explore

– Particle physics

– Nuclear and Hadronic physics

– Astroparticules and Neutrinos

– Nuclear energy and waste management

– Research and Development of Accelerators

– Instrumentation (new)

– Computing grids

• Bring its competence

– to other scientific domains

– to contribute solving societal problems

• Participate to the formation of students (University, grandes écoles)

• Help the companies benefit from its expertise

LHC - © Cern

3

Page 4: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

CdLT HEPIC13 conference

Instrumentation

• R&D instrumentation

– Photodetectors (PM, SiPM, MCCP…)

– Gaseous detectors (RPCs, Micromegas, TPCs…)

– Semiconductor detectors (Ge, Si, MAPS…)

– Cryogenic detectors (bolometers, LAr…)

– Radiodetection (MHz, GHz…)

– Microelectronics (ASICs)

– DAQ (NARVAL, FASTER, xTCA, …)

– R&D mechanics (cooling, composites…)

• R&D organization

– Transversal thematic networks

– Target next generation experiments

– Centralized funding

• « Microelectronics poles »

4 31 may 2013

Page 5: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

5

Microelectronics at IN2P3

16+1 labs : 60 designers

A huge potential…

…but scattered in 15 labs

National coordination in 2005

Labo

Designers

/Test/CAD

Perma-

nents Contact

GANIL 1 1 P Vallerand

Subatech 2 1 D Charrier

APC 3 2 Ch Olivetto

CENBG 2 2 J-L Pedroza

LPC Caen 2 2 P Laborie

LPNHE 4 3 P Lebbolo

LAL 4 4 D Breton

LAPP 4 4 J Prast

LPSC 5 5 Ch Vescovi

CPPM 9 6 A Calzas

MICRHAU IPNL

LPC Clermont

13 6

7

10 5

5

H Mathez

L Royer

OMEGA LLR

IPNO

LAL_OMEGA

16 1

3

12

11 0

3

8

R Cornat

E Wanlin

G Martin-Chassard

IPHC 23 16 Ch Hu

TOTAL 88 67

LDEF-IRFU 11 9 E Delagnes

UK R&D

@RAL

NL R&D

@NIKHEF

µE pour HEP

@CERN

Universités

INFN +

Univ

31 may 2013 CdLT HEPIC13 conference

Page 6: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

Micro-Electronics Network (2005)

• Goals :

– Facilitate collaboration and exchanges between labs

– Share knowledge and expertise

– Foster a « community belonging »

– Increase visibility of in2p3 in µelec

• Strengths

– Common Cadence tools, managed at national level

– Few selected technologies (130 nm, SiGe 350 nm) : sharing of expertise

– Annual meeting, good community knowledge

• Weaknesses

– Many subcritical groups re-inventing the charge preamp

– Human resources managed at local level

31 may 2013 CdLT HEPIC13 conference 6

Page 7: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

7

An

TD

C 3

50

ps

DLL

32

@ 6

25

ps

Slow Control

Transmission LVDS, 8b/10b

Zero Suppress Clustering

DLL 32 195 ps

PLL 400MHz

Dig

TD

C 1

2.5

ns

Temps PLL, DLL, TDC

FEAFS Cms

Si Strips Tracker

SUZE STAR/EUDET Pixel sensor

8b/10b 160MHz

I2C

JTAG

DLL 32 195 ps

Converters DAC

Régulateurs Band Gap

Mémoires Analogiques Numériques

Amplis de charge shapers

14 bits DAC 15 MS/s 20 mW

FP CSA Dyn 104, Rate 10kHz

CSA Dyn 103 Rate 100 kHz

Cryogenic 2°K 128x7 bits Memory

14 bits DAC 5 MS/s 7 mW

12 bits DAC 5 MS/s 5 mW

16x32 Dice latch Memory

CSA 1.2-10fC CCT 10ns 200µw

10 GS/s An Memory

BG 0.18µm

Reg 3,3v

100mA 4% LR

Reg 3,3v 10mA 10% LR

SCATS SuperB

FEAST SNEMO

IMOTEP IMABIO

PARISROC PM2

Mimosa26 EUDET ULTIMATE

STAR

ASIC64 MIMAC

FEI4 ATLAS

SQMUX128 QUBIC

NECTAR CTA

MICROROC CALICE

HARDROC CALICE

AGET ACTAR

LONAMOS AERA

Synergie 4: Briques de Base (3)

commercial IPs Building Blocs

Obsolescence of technologies Proven solutions

Architecture: SoC Detailed design

Finalize full circuits Validate prototypes

Never identical Duplicate

Accumulate know-how Not reinventing the wheel

Building bllocks = an oxymoron

•Bonn • LVDS/SLVS pads

• PLL

• Gigabit driver

•AGH • LVDS/SLVS pads

• PLL

Common work on 65nm CMOS : proposal for IP blocks from WP3 groups

•INFN • Discriminator

• Charge injection circuits for Front-End calibration

• n-bit D/A Converter (current steering topology)

• I/O PADs (LVDS2CMOS, CMOS2LVDS)

• Technology characterisation,

•IN2P3 • Bandgap - Service serial link (LPNHE)

• OTA - Preamp (pixels LAPP)

• DAC (LAL) - SPI serial interface

• ADC (LAL,(LPSC], LAPP) - I2C (IPNL/LAPP)

• Monitoring ADC (CPPM) - JTAG controller (LAPP)

• Temp. sensor (CPPM) - I/O cells

• Generic amplifier (CPPM) - PLLs (TOT, pixels, LAL,

LAPP) AIDA WP3 Annual Meeting – DESY – March 29, 2012 31 may 2013 CdLT HEPIC13 conference

To be renewed/shared/documented…

Page 8: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

CdLT HEPIC13 conference

Examples of chips at IN2P3

• MAPS sensors at IPHC (Strasbourg)

• ROC chips at OMEGA (Orsay)

• Chips at MICHRAU (Lyon-Clermont)

HARDROC2 SPIROC2 MAROC3 SPACIROC SKIROC2 MICROROC1

PARISROC2

OMEGAPIX

MIMOSTAR

Chip dimension: ~2 cm²

MIMOSA26 Pixel array: 576x1152

Chip dimension: ~ 3 cm² MIMOTEL

M18 LUSIPHER

8 LARZIC FEAFS (IC) CHOCAPIC FATALIC SICASIC HOCTASIC 31 may 2013

Page 9: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

31 may 2013 CdLT HEPIC13 conference 9

Page 10: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

1

0

Network transverse to instrumentation D

-Gaz

eux

Ph

otD

C

ryo

gén

ie

Sem

iCo

n

Cold CCD ASPIC LSST

LarTPC LARZIC T2K

Bolomètres SQMUX128 QUBIC

Bolomètres CRYOxx (Irfu) IXO/ATENA

Si PIN Diode SKIROC eCal Calice

PMT Mx SPACIROC Jem-Euso

SiPM SPIROC AHCal Calice

PMT Mx MAROC ATLAS

PMT PARISROC PMm2

Si PIN Diode CALORIC

PMT HOCTASIC

PMT SCATS (timing)

SuperB

PMT MAPPRA64 POLAR

PMT Mx IMOTEPAD64

Biomed

CPS ALICE MFT

PIXSIC

Biomed

CPS Mimosa2 ALICE ITS

HPS FEI3 ATLAS

HPS FEI4 ATLAS

HPS XPAD Biomed

CPS PHASE STAR/FOCAL

CPS Mimosa26 EUDET/CBM

CPS LUCIPHER Biomed

3DIC Pixels TSV/Stacked CAIRN OMEGAPIX FE_TC4 ILC /ATLAS

SiStrip SICASIC Biomed

SiStrip FP CSA

S3

CPS Mimosa 32 ALICE ITS

HPS FEI5 SATLAS

CPS ULTIMATE STAR

CPS Mimosa 31 ILC

PMT NECTAR(Irfu)

CTA

Rad

iod

etct

ion

Antenne LONAMOS AERA

RPC/µMegas HARDROC Calice

µMegas MIcroROC eCal Calice

µMegas ASIC64 MiMac

TPC AGET(Irfu) TPC2-BX

µMegas Idfx-LXe(Irfu

/Sub) Biomed

Drift Chamber FEAST (timing)

SNemo

Gaz Sensor CHOCAPIC Transfert

µMegas Dream(Irfu)

CLAS12

31 may 2013 CdLT HEPIC13 conference

Page 11: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

11

MAPS: A Long Term R&D

STAR 2012 Solenoidal Tracker at RHIC

EUDET 2006/2010

Beam Telescope

Spinoff: Interdisciplinary Applications, biomedical, space …

Main objective: ILC, with staggered performances MAPS applied to other experiments with intermediate requirements

ILC >2020

Internatinal Linear Collider

CBM 2017 Compressed Baryonic Matter

EUDET (R&D for ILC, EU project)

STAR (Heavy Ion physics)

CBM (Heavy Ion physics)

ILC (Particle physics)

HadronPhysics2 (generic R&D, EU

project)

AIDA (generic R&D, EU project)

FIRST (Hadron therapy)

ALICE/LHC (Heavy Ion physics)

EIC (Hadronic physics)

CLIC (Particle physics)

SuperB (Particle physics)

31 may 2013

Page 12: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

12

Foundries for HEP

• A solid know how in 350nm

– 350nm AMS SiGe, CMOS, CMOS epi, CMOS Hres epi

– 350nm XFAB CMOS, CMOS epi, CMOS Hres epi

• An investment in 130nm with CERN

– IBM (CERN)

– Global Foundries, Tezzaron 3DIC

• Now at a turning point

– 65nm TSMC (CERN)

– 180nm TOWER (CPS CERN)

– 180nm AMS, XFAB SOI HV

• Higher costs call for a better organization

– More training

– More concentration

– Better concertation with international efforts

31 may 2013 CdLT HEPIC13 conference

Page 13: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

CdLT HEPIC13 conference

« Microelectronics poles »

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

readout

Conversion

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)

16

16ValidHoldAnalogb

RazRangN

16ReadMesureb

Rstb

Clk40MHz

SlowClock

StartAcqt

StartConvDAQb

StartReadOut

NoTrig

RamFull

TransmitOn

OutSerie

EndReadOut

Chipsat

DAQASIC

Chip ID register 8 bits

gain

Trigger discri Output

Wilkinson ADC

Discri output

gain

Trigger discri Output

Wilkinson ADC

Discri output

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

readout

Conversion

ADC

+

Ecriture

RAM

Conversion

ADC

+

Ecriture

RAM

RAMRAM

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)

16

16ValidHoldAnalogb

RazRangN

16ReadMesureb

Rstb

Clk40MHz

SlowClock

StartAcqt

StartConvDAQb

StartReadOut

NoTrig

RamFull

TransmitOn

OutSerie

EndReadOut

Chipsat• Motivation :

– Continuous increase of chip complexity (SoC, 3D…)

– Minimize interface problems

– Reduce development times

• Importance of critical mass

– Daily contacts and discussions between designers

– Sharing of well proven blocks

– Cross fertilization of different projects

• Creation of poles with critical mass (~10 persons)

– Orsay (OMEGA)

– Clermont-Lyon (MICHRAU)

– Strasbourg (IPHC)

13 31 may 2013

Page 14: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

CdLT HEPIC13 conference 14

OMEGA group

• Mutualized ASIC design team formerly hosted at Orsay

• Specialized in low power, low noise, high speed design (SiGe)

• Very high level of integration : System on Chip (SoC)

• Design for calorimetry, astrophysics, medical imaging…

31 may 2013

HARDROC2

SPIROC2

MAROC3

SPACIROC

SKIROC2

MICROROC1

PARISROC2

Page 15: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

CdLT seminar at KEK 15

PMm² for large area photodetection

• Photomultiplier Array Integrated in SiGe Read-Out Chip

– Replace large PMTs (20’’) by arrays of 16 smaller ones (12’’) , PMm2 project

– Smart photodetector

– 16 independent channels

– Auto-trigger at 1/3 p.e.

– Charge (300 pe) and time (1ns) measurement (10-12 bits)

– Water tight, common high voltage for PMTs

– Data driven : « One wire out » for DATA and power supplies

Page 16: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

Evolution

• Creation of a national lab : OMEGA “Organization for Micro-

Electronics desiGn and Applicatiions” (2013)

– Difficulties with the hosting lab (LAL)

– Local management with national commitments, managing priorities

– Team had to be preserved

– Becomes an independant lab

• Hosted at Ecole Polytechnique : « 3 axes »

– Research : close to IN2P3 labs (LLR, LAL, IPNO…)

– Teaching: « electrical engineering », spatial

– Industry: startup Weeroc, : Astrium, Siemens…

• Missions :

– Design ASICs for IN2P3 physics experiments

– Participate to teaching and technology trransfer

16

Page 17: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

31 may 2013 CdLT 17

+ National ASIC lab

Page 18: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

Conclusion

• An active micro-electronics network

– Common Cadence tools, common technologies

– Sharing of expertise (to some limits)

– Sharing of building blocks (to some limits…)

– Thematic schools => community building

– National coordination (IN2P3)

• A few large groups (“poles”)

– Michrau, Omega, Saclay, Strasbourg

– Critical mass ~10 people

• Open issues / material for discussion

– Specialized groups vs collaborative design

– Propietary dsign vs sharing building blocks

– Technology transfer vs academic collaborations

– Versatility vs dedicated design

– Measurements and Integration

31 may 2013 CdLT HEPIC13 conference 18

Page 19: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

Backup slides

31 may 2013 CdLT HEPIC13 conference 1

9

Page 20: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

HARDROC2 for RPC readout

• HARDROC2: 64 channels (RPC DHCAL)

– preamp + shaper+ 3 discris (semi digital readout)

– Auto trigger on 10fC up to 20 pC

– 5 0.5 Kbytes memories to store 127 events

– Full power pulsing => 7.5 µW/ch

– Fully integrated ILC sequential readout

– 10 000 chips produced to equip 400 000 ch

– SDHCAL technological proto with 40 layers (5760 HR2 chips) built in 2010-2011.

• Successful TB in 2012 : 40 layers with Power Pulsing mode

@IPNL

Cosmic hadronic shower

Blue : 150 fC

Green : 2 pC

Red : 18 pC

X

Z Y

Z

X

Y

CdLT seminar at KEK 20

Page 21: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

CdLT seminar at KEK 21

SPIROC

Pedestal

50fC

injected

36-channel S-curves: trigger efficiency

versus threshold (1 LSB = 2 mV)

Qinj (fC)

50 fC

50 % Trigger efficiency point vs Qinj

Trigger at 50fC matched

(1/3 photon-electron at 106)

MIP response in DESY

6 GeV electron testbeam

©M. Reinecke (DESY)

SiPM SPECTRUM with Autotrigger

linearity using the auto gain mode

and internal ADC

Page 22: Microelectronics at IN2P3...Not reinventing the wheel Accumulate know-how Building bllocks = an oxymoron •Bonn •LVDS/SLVS pads •PLL •Gigabit driver •AGH •LVDS/SLVS pads

CdLT HEPIC13 conference 22

Backup slides •