CMS%20Upgrade%20Requirements%20and%20Upgrade%20Plans

1
CMS Upgrade Requirements and Upgrade Plans

• Hans-Christian Kästli, PSI
• 11.6.2008
• CHIPP Workshop on Detector RD

2
Contents

• Introduction Motivation
• Part I Subsystems
• Trigger
• Tracker
• Calorimeters Muon system
• Common Projects
• Part II Phase I barrel pixel upgrade
• Contains material from J. Nash, G. Hall, R.
  Horisberger

3
A realistic (?) upgrade scenario
J. Nash
4
Requirements from Physics

• Performance _at_SLHC need to be at least as good as
  now
• Especially true for forward jet tagging and
  b-tagging
• pT resolution ok as is
• IP resolution ok, but need to be kept at this
  level at SLHC (radiation damage reduces tracker
  resolution)

CMS upgrade plans
11.6.2008 3/32
5
Organisation
CMS upgrade plans
11.6.2008 4/32
6
The CMS Experiment
7
TRIGGER
8
Triggers

• Level 1 trigger consists mainly of
• Muon triggers from individual stations (local)
• Calorimeter triggers
• e/g isolated ECAL trigger towers, no HCAL energy
• t isolated ECAL HCAL trigger tower
• jets ECAL HCAL energy in cone
• Higher level trigger
• Combine local muon to global muons
• Combine muon tracks with tracks from inner
  tracker to improve pT resolution
• Correlate calorimeter trigger with tracker
  information (e/gt/QCD discrimination )
• At SLHC most of these triggers at level 1 will
  exceed any realistic trigger rate.
• ? There is a severe trigger problem for CMS at
  L1035cm-2s-1 !

9
Level 1 Trigger at SLHC
Level 1 trigger has no discrimination for
pTgt20GeV/c
(MHz)

• and jet rates
• Problem isolation of calorimeter towers
• Same problem for electron trigge

Muon trigger rates Problem insufficient pT
resolution from muon system
10
Level 1 Trigger at SLHC

• Guided from high level trigger at LHC
• Muon system needs tracking information on level 1
• Need trigger layers/trigger information from
  tracker
• pT measurement-gt outer radius
• Need correlation between muon and tracker
• Electron trigger needs tracking information on
  level 1
• Z-vertex information for cluster-track matching
• g rejection -gt innermost tracking layers
  (conversion)
• Need correlation between calos and tracker
• t trigger needs tracking information on level 1
• Jet rejection -gt complete tracking
• Need correlation between calos and tracker
• ? Tracking triggers absolutely needed
• But cannot do everything. Need priorities and/or
  compromises
• Very challenging, only vage ideas around

CMS upgrade plans
11.6.2008 9/32
11
TRACKER
12
Tracking _at_ LHC
pT and transvers impact parameter resolution

• CMS has an all-Si tracker with
• 3 pixel barrel layers
• 10 strip barrel layers (4 stereo)
• 2 pixel endcap disks
• 39 strip enddisks (22 stereo)
• The expected performance is excellent as
  indicated on the right

13
Material Budget

• In very central region 0.3 radiation length
• A lot of materials from cables in forward
• region!
• ? In future difficult to bring the power in
• Support structure contributes a lot, but
• cannot gain much in inner part
• Except fewer layers
• less material, less power, less costs
• Why are there so many?
• Serious concern about pattern recogni-tion with
  so few points (unprecedented design and
  environment)
• Probably can be relaxed (although still no full
  simulation of loopers and we havent seen data
  yet!!)

14
Going to SLHC
25 ns bunch structure
15
Tracking with 500 min bias events
Inner layers of strips reach 30 occupancy on
every xing!

• Study of current CMS tracker for Heavy Ion events
• Track density very similar to 50ns running
• dnch/dh/crossing 3000
• Tracker occupancy very high
• Need more pixel layers/shorter strips
• Tracking is possible (thanks to pixel!)
• When tracks are found they are well measured
• Efficiency and fake rate suffer
• CPU Intensive

Pixel layers
Momentum Resolution
Impact Parameter Resolution
16
Key Areas for Tracker RD

• Reducing and delivering power at lower voltage
• DC-DC conversion, serial powering
• and removing it, with limited material
• CO2 cooling, other cooling schemes
• Ensuring sufficiently rad hard sensors are
  available
• and establishing the manufacturing capability
  with industry
• A new readout control system using modern
  technologies
• radiation hardness may be a minor point this time
• Developing components to build one or more layers
  which can contribute to the L1 trigger
• these layers must also be light and
  power-efficient
• Designing a tracker layout from these which will
  meet the physics objectives (last, but by no
  means least)

17
Tracker related RD Projects
Proposal title Contact Date Status CMS ref
Letter of intent for Research and Development for CMS tracker in SLHC era R Demina 14.9.06 Approved 06.01
Study of suitability of magnetic Czochralski silicon for the SLHC CMS strip tracker P Luukka, J Härkönen, R Demina, L Spiegel 31.10.07 Approved 07.06
RD on Novel Powering Schemes for the SLHC CMS Tracker L Feld 3.10.07 Approved 07.01
Proposal for possible replacement of Inner Pixel Layers with aims for an SLHC upgrade A Bean 31.10.07 Approved 07.07
RD in preparation for an upgrade of CMS for the Super-LHC by UK groups G Hall 31.10.07 Approved 07.08
The Versatile Link Common Project F Vasey, J Troska 11.07 Received 07.12
3D detectors for inner pixel layers D Bortoletto, S Kwan 12.07 Received 07.13
Proposal for US CMS Pixel Mechanics RD at Purdue and Fermilab in FY08 D Bortoletto, S Kwan 12.07 Received 07.15
RD for Thin Single-Sided Sensors with HPK M Mannelli 7.2.08 Received 08.01
An RD project to develop materials, technologies and simulations for silicon sensor modules at intermediate to large radii of a new CMS tracker for SLHC  F Hartmann, D Eckstein  6.3.08 Received 08.02
Development of pixel and micro-strip sensors on radiation tolerant substrates for the tracker upgrade at SLHC M de Palma 9.4.08 Received
18
Working Group Organisation

• These goals motivated the RD structure
• active for 12-18 months

19
Barrel Strawman Designs
Strawman A Geometry Perturbation of current
tracking system 4 Inner pixel layers, 2 strixel
2 short strip layers (TIB), 2-strixel 4 short
strip layers (TOB)
Strawman B Geometry Design radically different
from current tracker Super-layers, each with two
doublet layers (integrated tracking/ triggering
layers) 3 inner Pixel layers
Short strips
Superlayer
Pixel layers
Strixel layers, could be doublets
Mini-strip or Pixel doublets
20
Strawman B triggering local data reduction

• High pT tracks point towards the origin, low pT
  tracks point away from the origin
• Use a pair of sensor planes, at mm
  distance
• Pairs of hits provide vector, that measure angle
  of track with respect to the origin
• Keep only vectors corresponding to high pT tracks
• Stacks of 2 sensor pairs, at cm distance
• Redundancy
• Track stub provides higher resolution local pT
  measurement
• Two level data reduction
• But Material budget? How light can we build
  this? Power?
• Triggering needed, but mustnt geopardize
  tracking

a
CMS upgrade plans
11.6.2008 19/32
21
CALORIMETERSMuon System
22
Hadron Calorimeter

• Forward calo may be blocked by potential changes
  to the interaction region. This has a direct
  impact mainly in the case of looking for WW
  scattering
• Upgrade proposal for new trigger/DAQ readout
• New SiPM with higher gain/less noise
• Allows to
• add timing information
• make longitudinal segmentation to reduce
  sensitivity to min bias
• Both Calorimeters suffer degraded resolution at
  SLHC
• affects electron ID, Jet resolutions

CMS upgrade plans
11.6.2008 21/32
23
Electromagnetic Calorimeter

• ECAL
• Crystal calorimeter electronics designed to
  operate in SLHC conditions
• VPT in Endcap and Endcap crystals themselves may
  darken at SLHC
• Very difficult to replace
• At SLHC, per Trigger Tower, per crossing, 12 ?
  expected (? rate in ECAL 2.4 MHz/cm2), ltPTgt 3
  GeV ?No empty ECAL towers
• The readout of every crystal is technically
  possible, by increasing the readout rate from 0.8
  Gbits/s to 14 Gbits/s for 25 crystals.
• full calorimeter information could be used in the
  level 1 trigger decision
• Proposal by R. Rusack, Proposal for ECAL RD
  (Phase II), DPG 27-FEB-08
• Physics benefits arguable but definitely not
  proven
• Major effort to remove the supermodules, remove
  the LV cables and replace the front-end boards,
  check and re-install.
• Requires un-cabling of the tracker to remove the
  ECAL supermodules
• Radiation levels for all this work

CMS upgrade plans
11.6.2008 22/32
24
Muon System

• System front end electronics look fairly robust
  at SLHC
• Cathode Strip Chambers/RPC Forward
• Drift Tubes /RPC Barrel
• Trigger electronics for the muon systems would
  most likely need to be replaced/updated
• Some Electronics is less radiation hard (FPGA)
• Coping with higher rate/different bunch crossing
  frequency
• May have to limit coverage in ??(? gt 2) due to
  radiation splash
• This effect will be known better after first data
  taking, potential additional cost of chamber
  replacement

CMS upgrade plans
11.6.2008 23/32
25
Infrastructure Modifications Yoke

• Reinforced Shielding inside forward muons
• up to h2
• automatically implies replacement of inner CSC,
  RPC

• Supplement YE4 wall with borated polythene
• Improve shielding of HF PMTs
• possibility of increased YE1-YE2 separation to
  insert another detector layer?

CMS upgrade plans
11.6.2008 24/32
26
COMMON PROJECTS
27
Common Projects

• There are many common RD topics which must be
  pursued by ATLAS and CMS
• A Joint meeting was held last year to discuss
  possible joint RD
• ACES meeting - http//aces.web.cern.ch/aces/
• Will repeat later this year
• Topics include
• Link Technology
• Versatile link project (CERN) rad hard physical
  network layer for optical links
• GBTproject (CERN) gigabit optical link for
  data/control/trigger
• ASIC technology
• 130nm technology
• Power distribution
• Funding from EU
• Rad hard DC-DC converter (CERN)
• Small contribution from PSI on-chip DC-DC
  conversion
• Cooling
• Tests on CO2 cooling hosted by CERN cryolab
• Radiation/shielding issues

CMS upgrade plans
11.6.2008 26/32
28
Part IIBarrel Pixel System for Phase I
29
Phase 1 Tracker Upgrade

• In 2013 need to replace gt 50 of pixel barrel
  modules probably optical-links due to radiation
  damage. Pixel module fluence limit 6x1014
  cm-2 (TDR)
• Barrel mechanics constructed for quick
  insertion/deinstallation
• Radiation damage activation probably require a
  complete rebuild of barrel pixel
• Services (cables, fibres and cooling tubes) will
  have to be re-used. No replacements, no
  additions.
• Rest of tracker will be unchanged (long shutdown
  would be needed to access strip tracker)
• PSI/UNIZ/(ETHZ) are interested in building a new
  barrel pixel detector for the phase I upgrade
• Several possible scenarios presented by Roland
  Horisberger
• http//indico.cern.ch/conferenceDisplay.py?confId
  28746
• No official proposal yet.

30
Options for Barrel Pixel
Readout analog 40MHz analog 40MHz
analog 40MHz analog 40MHz m-tw-pairs
digital 320MHz m-tw-pairs digital 640
MHz m-tw-pairs
Power as now as now as now as now as
now DC-DC new PS
Cooling C6F14 C6F14 CO2 CO2
CO2 CO2
Comment as 2008 Data loss reduction Large
gain in material budget Simplifies module
design Test structures designed Uunlikely for
2013 (services)
Pixel ROC PSI46 as now 2x buffers 2x
buffers 2x buffers 2xbuffer, ADC 160MHz
serial 2xbuffer, ADC 160MHz serial
Option 0 1 2 3 4 5
Layer/Radii 4, 7, 11cm 4, 7, 11cm 4, 7,
11cm 4, 7, 11cm 4, 7, 11cm 4, 7, 11,
16cm
31
Option 0 Simple Replacement

• Pro
• Data taking and physics output of CMS experiment
  is not disturbed
• Pixel detector is well calibrated, aligned,
  efficiencies are studied, algorithms are stable
• No big brainpower has to be drafted for rebuild
  ? focus on physics output of CMS
• Time schedule and costs are likely under control

• Con
• Rebuild detector conceived in 1997
• Detector designed for Luminosity of 1x1034 and
  will develop substantial inefficiency for 4cm
  layer at 2x1034
• We did a reasonable job in material budget
  (1.92/Layer at h0) but could be improved,
  especially in eta region 1.4-2.3
• Present pixel needs considerable algorithmic know
  how to deal with gain variability of optical
  chain and event decoding

32
Option 2 CO2 cooling

• Use bi-phase CO2 cooling and benefit from reduced
  material budget !
• allows long cooling loops (2-3m) with very
  small diameter pipes (1mm) for thermal loads of
  100 W
• Present C6F14 monophase has parallel cooling
  pipes with manifold and large diameter silicon
  hoses for feed and drain in front of FPIX
  tracking region.
• New CO2 allows serialized pipes without pressure
  drop problems and therefore reduces resident
  cooling liquid by large factor.
• Density of liquid CO2 is 1.03 g/cm3 compared to
  1.76 g/cm3 of C6F14
• Needs considerable engineering support from CERN
  for CO2 cooling plant.

Material distribution budget for 3 barrel layers
C6F14cooling
CO2cooling
33
Option 2 Weight of one half barrel
BPIX now BPIX CO2 cooling Empty
mechanics 1103 g 550 g 94g ,
1.5mm pipes 384 Module 2.27 g each 872 g
same 384 Signal cable (21.7mg/cm)
167g same 384 Power cable (10.6 mg/cm)
82g same 384 Power plug (42
mg/plug) 16g same 32 Print (15.61g
each) with red power cable 499g
same Cooling (C6F14) in tubes, manifolds pipes
810 g 83 g 1.45mm
pipes 2x1024mm Silicon tube C6F14 in side
372g 5 g serial piping
Total 3921g 2274 g
Average material budget gain 1.72 , but in region
h 1.4 2.3 much more
CMS upgrade plans
11.6.2008 32/32
34
Option 3 Replacing Kapton signal cables

• use m-twisted pairs 2x125mm of enameled Copper
  Cladded Aluminum (CCA) wires could provide a
  viable signal cable alternative for Kapton cables
• Why?
• Kapton cable can only bend in one plane.
• O(1000) modules result in very complex,
  expensive, laboursome endflange prints, which
  contribute considerable in material budget in h
  range 1.6
• Kapton cable length limited to lt 40cm
  expensive
• Each 300mm plug is small light, but 800
  soldered onto PCB with 21 traces in the sensitive
  tracking region are definitely not light !
• Benefit?
• modules with long pigtail cables (1.2m) allow
  transmission of analog output signals without
  impedance breaks.
• More freedom in bending cables in all directions
• Omit endflange print ? no soldering, simpler
  mechanics endflange, no PCB, no strong mechanics
  supports of PCB for plugging forces
• Can move optical links with auxiliary chips
  further out (50cm) to high h - range and remove
  material budget from sensitive tracking region

35
Option 3 Weight of one half barrel
BPIX now BPIX CO2 cooling Empty
mechanics 1103 g 550 g 94g ,
1.5mm pipes 384 Module 2.27 g each 872 g
same 384 Signal cable (21.7mg/cm) 167g
14g 4x(2x125m) CCA 384 Power
cable (10.6 mg/cm) 82g 68g
5x250m CCA 384 Power plug (42 mg/plug)
16g 0g 32 Print (15.61g each)
with red power cable 499g 32g radial
cables Cooling (C6F14) in tubes, manifolds
pipes 810 g 83 g 1.45mm
pipes 2x1024mm Silicon tube C6F14 in side
372g 5 g serial piping
Total 3921g 1624 g
gain 2.4x
Large material budget gain for h 1.4 2.3
? forward impact parameter
CMS upgrade plans
11.6.2008 34/32
36
Summary and Conclusions

• Luminosity will increase gradually
• Due to radiation damage pixel needs to be
  replaced in 2013 anyway
• Aim for pixel re-design in Phase I
• Several options sketched
• with large gain in material budget
• possible within reasonable time scale/costs
• Phase II upgrade
• Calo muon system minor upgrades needed
• Trigger need tracking information on level 1
• Tracker completely new tracker needed to cope
  with high data rates. Extremely challenging
  project
• Increasing channel number
• Must reduce power
• Must provide trigger information
• Should reduce material budget
• Have to reuse present services
• Time is short. Schedule has to be harmonized with
  ATLAS !

CMS upgrade plans
11.6.2008 35/32
37
Backup slides
38
Input to LoI preparation
Examination will show this is an aggressive
schedule When is t 0?
Constraints on LoI TDR submission dates come
from several places..
Input to TDR preparation
NB Remaining RD shows time left after known
within RD phase
Geoff Hall
SLHC UG WG June 2007
37
May 2008
G Hall
37
39
Timescales

• Plan still seems a reasonable estimate, given 1
  year shift
• Examples of a few predecessor activities
• Prototyping of readout ASICs, including any
  needed for triggering
• Development of sFEC, control system sFED
• Prototype sensors with major manufacturers
• Provisional mechanical support design
• Development and evaluation of prototype modules
• Definition of cooling scheme
• Definition and proof of power distribution scheme
• Development of cost model, TDR cost estimate
• including establishing funding, preliminary
  negotiations with vendors
• 6 months for TDR preparations and writing
• Submission of Phase II TDR end 2012
• With approval and launch of approved construction
  project mid-2013

Possible LoI 2010 Phase I TDR?
38
May 2008
G Hall
40
BPIX Options for 2013 replacement/upgrade
Option 0 1 2 3 4 5
ltX0 gaingt 1 1 1.7
2.4 3.1 1.6
Costs 4.5 MCHF 5.0 MCHF 5.4 MCHF 5.4
MCHF 5.9 MCHF 9.8 MCHF 0.4 MCHF
Weight 3921 g 3921 g 2274 g 1624
g 1267 g 2400 g estimate
Start Date Aug 10 Nov 09 Nov 09
Nov 09 Dec 08 not possible for 2013
Comments ROC wafers exist new ROC wafers
0.4 MCHF for CO2 plant - - - new ROC
TBM HDI mod. pxFED pxFEC DC-DC
converters new LV Power Supplies