CMS HL-LHC EM Calorimeter Upgrade

CMS HL-LHCEM Calorimeter Upgrade

M. Hansen, CERNfor the CMS ECAL upgrade project(s)

M. Hansen, CERN. [email protected]

Content

Legacy system design and Status Requirements Barrel Calorimeter electronics upgrade End Cap calorimeter upgrade Conclusion

CMS HL-LHC EM Calorimeter Upgrade @ ACES2014


Legacy ECAL Barrel FE electronics system- Design and Status

Modularity 25 channels = 1 trigger tower

Features Trigger Primitive generation Pipeline, event buffer in FE

Status 25/2592 with issues 13/2592 irrecoverable


STRIPASIC

HSSer.STRIPASIC

STRIPASIC

STRIPASIC

STRIPASIC

HSSer

TCPASIC

DAQASIC

FENIX

FENIX

FENIX

FENIX

FENIX

FENIX

FENIX

GOL

GOL

800Mbps

800Mbps

Trigger etc.ADC

QuadADC (5x)

PreAmp

3 gain MGPA (5x)PD1

APD1; G=50

PD2APD2; G=50

PWO crystal4p.e. / MeV to APD pair

VFE card (5x)

FE card

Low voltage regulator board – 10 linear rad tol regulators

Dynamic range 40MeV – 1.5TeV, 40ns RC-CR

ASICS in 0.25um tech

LegacyECAL Barrel trigger tower

CMS HL-LHC EM Calorimeter Upgrade @ ACES2014 M. Hansen, CERN. [email protected]

GOH (GOL hybrid)

CCU (Slow control)Front end chip

QPLL (clock cleaning)

Low voltage regulatorsVFE

FE

Mother board

LegacyECAL Supermodule integration &

Slow control 72 legacy towers in SuperModule

36 SM in total Slow and fast control in 8 rings

Single tower of fibre failure recovery; very few issues to date


Noise vs ageing Electronic noise vs ageing and

eta APD dark current predicted to

increase with integrated luminosity – confirmed ->

Large Extrapolation suggest noise increase by a factor 10 after 3000fb-1 Roughly corresponds to 400

MeV/channel TBC by measurements

The feasibility to operate the ECAL barrel colder and thereby decrease the APD dark current noise is under study


CMS ECAL Barrel Anomalous events:Spikes

APD

Crystal

HCAL

(2)

Hadrons interacting with the APD’s causing anomalous high E deposits

Hadrons come from primary interaction and backsplash


LegacyReal Time Spike rejection Algorithm

An EM shower deposit E in more than one crystal A spike typically occurs only in one APD


eta strip number: 1 2 3 4 5

OR result: 1 (shower-like)

OR result: 0 (spike-like)

EM shower Spikecrystal above threshold

crystal below threshold

eta strip number: 1 2 3 4 5

0 1 0 0 01 1 1 0 00 1 0 0 00 0 0 0 00 0 0 0 0

0 1 0 0 0

channels above threshold

strip LUT result:

0 0 0 0 00 1 0 0 00 0 0 0 00 0 0 0 00 0 0 0 0

0 0 0 0 0

channels above threshold

strip LUT result:

strip LUT configuration: 00000,00001,00010,00100,01000,10000 → output=0 all other combinations → output=1

TCP ASIC → OR of 5 strip LUT results

ϕ ϕ


Spike rates L1 Spike rejection in 2012

O(95%) Rate sustainable up to LS3

with somewhat raised channel threshold

Legacy Spike Rejection Algorithm is sensitive to PU (see plot ->)

Exploring improvements having access to full granularity of data in the trigger path Single crystal readout

Exploring pulse-shape variables to provide an additional efficiency/rejection safety margin Analogue Signal ProcessingCMS HL-LHC EM Calorimeter Upgrade @ ACES2014 M. Hansen, CERN. [email protected]


ECAL barrel Phase 2 UpgradeRequirements

Requirements Trigger rate up to 1 MHz

Legacy max ~150 kHz Trigger latency up to 25 us

Legacy max ~5 us Full installation during LHC Long Shutdown 3 Maintained or improved reliability and availability Improve the EB spike mitigation

High on the wish list Decrease the Low Voltage Current delivered to the Front End

system in order to decrease the physical volume required for services

Solid failure mitigation scheme E.g. avoid dependence between neighbours



Phase 2 ECAL Barrel FE electronics system- Design idea

Modularity 1 channel for readout and trigger As legacy for services (bias, LV)

Features Trigger-less Streaming

Requires 10 Gbps GBT 4 fibers/readout unit


HSSer

HSSer

HSSer

GBT

GBT

GBT

ADC?PreAmp?PD1

APD

PD2APDPbW crystal

VFE card

FE card

PreAmp/ADC?

QIE?

9.6Gbps

4.8/9.6Gbps

9,6Gbps

4.8Gbps

Controls

Data stream + control handshake

Data stream

Data stream

LV regulator board, e.g 3 rad tol DCDC


Phase 2 ECAL Barrel FE electronics system- Design idea

Modularity 1 channel for readout and trigger As legacy for services (bias, LV)

Features Trigger-less Streaming

Requires 5 Gbps GBT 6 fibers/readout unit


HSSer

HSSer

HSSer

HSSer

HSSer

GBT

GBT

GBT

GBT

GBT

4.8Gbps

ADC?PreAmp?PD1

APD

PD2APDPbW crystal

VFE card

FE card

PreAmp/ADC?

QIE?

4.8Gbps

4.8Gbps

4.8Gbps

4.8Gbps

4.8Gbps

Controls

Data stream + control handshake

Data stream

Data stream

Data stream

Data stream

LV regulator board, e.g 5 rad tol DCDC


Current EB R&D plan for TPVFE

Develop a VFE chip set and Carrier Develop the optimal scheme for spike detection

At the source, i.e. in the very front end amplifier and ADC In the back end exploiting fine granularity of channel level data

Define the best compromise between noise and power consumption

Considering the increased APD noise due to radiation damage Considering the possibility to operate the ECAL barrel below room temperature

Optimise sensitivity to out-of-time Pile-Up Optimise sampling rate and pulse shaping time for HL-LHC conditions



Current EB R&D plan for TP Develop the upgraded front end power system demonstrator

Develop a carrier for e.g. the DCDC module described last Tuesday afternoon by Federico Faccio

Evaluate with legacy front end system Look out for issues with Noise and Thermal management

Develop a front end card Demonstrator Streaming data to the back end Look out for issues with Thermal management

Study existing CMS back end cards and define requirements for the future ECAL back end system E.g. cards developed for the CMS level 1 trigger phase 1

upgrade



CMS EB upgrade Working Groups WG1: Evaluate Scope of Upgrade with regard to Risk and

Schedule W. Funk, M. Hansen, et al.

WG2: Develop motivation and Evaluate options for upgrading the VFE and the LVR M. Dejardin et al.

WG3: Evaluate the desired Functionality and characteristics of replacement FE card A. Singovsky, J.C. Silva, et al.

WG4: Detector Monitoring D. Bailleux, P. Gras, et al.

Group of electronics developers is growing


ENDCAPS SECTION


CMS Endcap calorimeter upgrade

ECAL and HCAL Endcaps are suffering radiation damage

Planning re-build or fully replace ECAL and HCAL Endcaps for HL-LHC

Three proposals on the table


HCAL

ECAL

ES


End-cap calorimeter upgrade Proposal 1Shashlik ECAL


The Shashlik proposal comprises ECAL and a re-built full HCAL endcap

Readout chain very similar to ECAL barrel Hl-LHC upgrade and HCAL legacy / Phase 1 ~ECAL barrel system size: 61k channels

SiPM location alternatives In front of HCAL: InGaP PM and 3m analogue cable Behind HCAL: SiPM and 3m fiber carrying Shashlik light

ADC?PreAmp?

PDSiPM

PreAmp/ADC?

QIE10

Electronics location(s)

Intermediate Layer behind HCAL:- E.g. Igloo3 radtol FPGA- 10Gb optical link drivers- Services -- Power -- Controls

OFF-DETECTOR


End-cap calorimeter upgrade Proposal 2Combined Forward Calorimeter


PD

SiPM

QIEx, TDC

OFF-DETECTOR

Scintillating fiber (Slow)

Quartz fiber (fast)

PD

SiPM

QIEx, TDC

ROC, e.g. Domino Ring Sampler; (“Sparsified” 5 GHz sampling)

FPGAe.g. Igloo 3

CFC Comprises full ECAL and full HCAL endcap Essentially three ~ECAL barrel sized readout systems

Scintillating, quartz, timing Fine timing data stored in ROC

Read out after L1A

Electronics located behind HCAL:- E.g. Igloo3 FPGA- 10Gb optical link drivers- Services -- Power -- Controls

L1A


End-cap calorimeter upgrade Proposal 3High Granularity Calorimeter

HGC Comprises full ECAL and half HCAL endcap Combined with half re-built legacy HCAL Endcap

“Coarse” trigger primitives sent off detector Full granularity data stored in ROC

Read out after L1A


64 Chan FE ROC

TriggerPrimitives5Gbps e-link

RO/CTRL30k modules in 41 planesAbsorber interleaved

Intermediate Layer:Behind HCAL:- Map on 10Gb optical links- Services -- Power -- Controls (Fast and Slow)

OFF-DETECTOR


Conclusion

Described the Legacy CMS ECAL Electronics system design and Status

Described few issues to be addressed with the upgrade of the CMS ECAL electronics system

Listed a set of CMS Requirements, general to CMS phase 1 and Phase 2 upgrades

Listed future R&D for the ECAL Barrel Calorimeter electronics upgrade

Briefly described options for the Endcap calorimeter upgrade

All in all, interesting time ahead of us


Date post:	20-Feb-2016
Category:	Documents
Upload:	ura
View:	56 times
Download:	1 times

CMS HL-LHC EM Calorimeter Upgrade

Documents