Instrumentation Frontier Community Meeting – Snowmass Process 2013
GEM Upgrade
for
CMS Forward Muon System
Marcus Hohlmann
(an IF-EF liaison for gaseous detectors)
Florida Institute of Technology
Argonne National Laboratory - Jan 11, 2013
MOTIVATION & INTRODUCTION
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Premise for CMS GEM upgrade
•
CMS was designed with a “hermetic and redundant muon system” –
Joe Incandela, CERN “Higgs Discovery” Event, 7/4/12
•
But: CMS currently has the least redundancy in the most challenging muon
region, i.e. at |η| > 1.6:
Bakelite RPCs descoped in high-η region (lack of rate capability);
only Cathode Strip Chambers currently present
•
Long-term functioning of the muon system into LHC Phase II (beyond Long LHC
Shutdown 3) is of vital interest for CMS. Use Phase I to evaluate muon
technology for Phase II.
•
The high-η muon region in particular will need robust and redundant tracking
and triggering at the anticipated increasingly higher muon rates
 Additional muon detectors with high spatial and temporal
resolution in the high-η endcap region could bring benefits
in triggering, reconstruction, and ID for muons: → GEMs
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
CMS GEM Endcap Chambers
The currently un-instrumented high- RPC region of the muon endcaps
presents an opportunity for instrumentation with a detector technology that
could sustain the radiation environment long-term and be suitable for
operation at the LHC and its future upgrades into Phase II: GEM Detectors
GE1/1
in nose of
first Endcap Yoke
GE3/1
GE4/1
GE1/1 simulation
geometry
CMS Detector
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Integration into CMS
Four superchambers in
their final position
on endcap yoke
Installation sequence:
Superchamber
(Two Triple-GEMs)
A. Conde Garcia
1/11/2013
Mounting on yoke disk
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Expected Benefits for Reconstruction & Trigger
Expected gains in momentum resolution at high-pT
Staving off looming muon trigger inefficiencies at high-
Simulation
Expected CSC inefficiency
at PU=400 due to Cathode
LCT - Anode LCT timing
mismatching
for one GEM station
A. Safonov (Texas A&M)
Acceptance impact:  distribution of 4 muons in H → ZZ → 4µ
1.6<<2.4
M. Maggi (Bari) – GEM Workshop 3
Paolo Giacomelli (Bologna) & Markus Klute (MIT) – GEM Workshop 3
Strip Readout granularity: # GEM strips / # RPC strips (orig. TDR)
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
1/11/2013
CONSTRUCTION
OF FULL-SIZE GEM PROTOTYPES
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Evolution - GEM foil stretching
Current state-of-the-art: Self-stretching assembly sans spacers (CERN)
Readout PCB
Tightening the
horizontal screws
tensions the GEMs
& seals gas volume
GEMs
Drift electrode
Detector base pcb
2012
only glue joint in assembly
Allows re-opening
of assembled
detector for repairs
if needed.
R. De Oliveira, CMS-GEM/RD51Workshops
CERN & U. Gent
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
3rd GE1/1 Prototype: “Self-Stretched - Sans Spacer”
No spacers in active volume
GEM foil with inner & outer frame
GEM foil in inner frame assembly
GEM tensioning
Vias for strips sealed
w/ kapton
compact HV divider
-sector with
384 radial
readout strips
(12.4 cm long)
Inside of readout board with O-ring seal
1/11/2013
HV noise
filters
Chamber closed by
readout board with
Panasonic connectors
for frontend electronics
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
CERN
2012
Full-size GE1/1 Detector Prototypes
 GEM active area: 990 mm
 (220-445) mm
 Single-mask technology
 1D radial strip read-out with
3  8  128 = 3,072 channels
 35 HV sectors
 3/1/2/1 mm gap sizes
 Gas mixtures:
Ar/CO2 (70:30; 90:10)
Ar/CO2/CF4 (45:15:40;
60:20:20)
 Gas flow ≈ 5 l/h
New self-stretching technique has been applied to the full-size CMS GE1/1 GEMs
4 prototypes produced & tested at CERN in 2012
1 prototype to be assembled & tested at Fl. Tech in early 2013
5-8 new final prototypes to be produced for installation of 4 during first LHC long shutdown (LS1)
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Next: No Gluing & PEEK frame
Panasonic readout connector
Readout PCB
Outer frame
On-foil protection resistors
GEM foils
Use O-ring to seal
outer frames to drift plane;
removes ALL glue joints
Now PEEK
Inner frame
Drift Board
As a lot of dust was released when inserting screws into FR4
frames; CERN has replaced FR4 by PEEK for inner frames.
PEEK is one of the best polymers in terms of:
-radiation tolerance
-mechanical properties
-outgassing
-chemical resistance
Rui De Oliveira
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
LHC Long Shutdown 1
Integration & Installation of 2 GE1/1
superchambers = 4 GE1/1 chambers
P5 CAVERN UXC
Proposed to CMS Upgrade Management
ME2/1
GE2/1
Objective: Participate in CRAFT 2014
YOKE
CMS
ME1/1
GE1/1
GE1/1
ME1/1
Proof of
Concept
4 GE1/1 GEM chambers in LS1
Measure in situ:
• Rates, Background/Noise, Stability,
Uniformity, Efficiency
• Spatial resolution
– In actual high-η environment
– In actual magnetic field
• Split signal to CSC and participate in CMS muon trigger and reconstruction
• Install new pre-production trigger motherboards on chambers that overlap with GEMs
• Prove that the electronics design is working and demonstrate in situ that we can operate
CSC TMB with GEM input in various operating regimes
• Reduce CSC X-Y ambiguity and ghosts
• Once we go back to beam operations, demonstrate the above again, this time measuring
muon trigger rates and efficiency with and without GEMs.
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
GEM DAQ Prototype System for LS1
DAQ also with link to new
CSC Trigger Mother Board
Gigabit Link Interface Boards
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
P. Aspell
CURRENT R&D:
ZIGZAG STRIP READOUT
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Previous Work @ BNL
Zigzag strips:
Previous exp. studies show <100 µm resolution with 2 mm strip pitch is possible:
100
y
avalanche
Position Error [µm]
80
60
40
20
0
-203000
3500
4000
4500
5000
5500
6000
6500
7000
-40
-60
-80
-100
x (measured coordinate)
BNL
Reconstructed Position [µm]
Concept:
• Charge sharing among adjacent strips allows quite sensitive position-interpolation in x-direction
• We are sacrificing the measurement of the 2nd coord. (y) to gain precision in the 1st coord. (x)
• CMS GE x/1 detectors are currently intended for 1D-coordinate measurements, so the zigzag
approach is applicable to these detectors
7/24/2012
1/11/2013
GEM Upgrade for CMS
HEP
Forward
Division
Muon
Seminar,
System
ANL
- M.
- Marcus
Hohlmann,
Hohlmann
Snowmass IF meeting, ANL
Zigzag strips vs. straight strips
Pitch
[mm]
Typical Resolution
[µm]
Zigzag strips & analog r/o
2.0
80
Straight strips & VFAT
(current design, short end)
0.6
300
Improvement factor
w/ zigzag strips
3.33
3.75
Can reduce # of &
readout channels
(and electronics
cost) by 70% of
current design
Improve resolution
by factor 3-4
A “figure of merit”: 3.33 × 3.75 = 12.5
~ Potential for order of magnitude improvement over current design
 Well worth a try!
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
R&D: Zigzag strips to reduce readout channel
count while maintaining high spatial resolution
55Fe
@ FIT
spectrum
Stand. CERN
10cm × 10cm
Triple-GEM
Gas
gain
1/11/2013
2 mm
CAD Design by C. Pancake, Stony Brook
2 mm
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
2012 CMS/RD51 beam test @ SPS
Zigzag Resolution
= /2 = 73 µm
150 GeV/c
µ &  beams
M. Staib (FIT)
June 2012
looking for hit in 5mm window centered on track
M. Staib (Fl. Tech)
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Next steps for zigzag r/o
• Develop successively larger zigzag strip
readout boards:
– 30cm × 30cm (Detector already assembled)
– 100cm × 45cm CMS GE1/1
– GE2/1 prototype? (1.5-2m long)
• Test performance
– Spatial resolution in magnetic field
– Timing resolution achievable with analog readout
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Some thoughts on…
THE “BIGGER PICTURE”
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
GEM Evolution: Problems & Solution
Problem:
Solution:
• Experiments to face high
rates at LHC (HERA-B)
→ First MPGDs invented:
high readout granularities
with microstrips and small
pads lowering occupancies
• Aging encountered in
original Micro-Strip Gas
Counters (MSGC’s)
→ Move avalanches away
from materials (microstrips)
towards empty space (hole):
Gas Electron Multiplier
• MPGDs show sizable
spark rates
→ Distribute gain over
several elements:
Double-GEMs, Triple-GEMs
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
GEM Evolution: Problems & Solution
Problem:
Solution:
• GEM detector size limited
to ~ 40cm × 40cm by
alignment of two masks
during production
• Detector construction
using external stretching
of GEM foils and epoxy
glue (curing time) slows
down assembly
• Spacers eat into GEM
detector acceptance
• Rising demand for foils
→ Develop single-mask
production process; allows
large-area GEMs (currently
up to 1m)
→ Devise “self-stretching
sans spacer” assembly
technique (CMS GEMs)
→ Ditto…
→ Upgrade CERN workshop
NB: ALL solutions so far have basically come from CERN; US by far not a leader!
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
GEM Evolution: Problems & Solution
Problem:
(Future) Solution:
• Large-area GEM
detectors, e.g. in muon
systems, require large
number of electronics
channels and big DAQ
systems  rising cost
→ Scalable Readout System
→ Zigzag strips?
→ Frontend electronics with
much higher chan. integration
(à la KPiX); could we do, say,
4k ch. on a single chip!?
→ Work directly with industry?
Involve NL’s?
• One dimension of
chambers limited by width
of Cu-Kapton foil base
material (~60cm)
• Industrial foil production
notoriously problematic
(Tech Etch, New Flex)
1/11/2013
→ Ditto; joint ventures
between industry & HEP
beyond SBIRs?
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Brainstorming for “2020”
• In a phone meeting, someone asked the important question “Can MPGDs
complement silicon as vertex/tracking detectors in highest-rate environments
to save cost?” Electron-Ion Collider detector designs in NP
• Can we automate chamber assembly (use robotics)? Cheaply massproduce a “standard” large-area GEM detector (akin to the CERN standard
10cm × 10cm GEM detectors used for R&D)?
• Can we marry commercial flexible foil circuits and GEM foil technology?
Put cheap surface mount readout electronics directly on r/o strip foil or on a
GEM foil? => Save chip bonding, connectors, cables, i.e. cost.
(see also Julia Thom’s talk on novel technology in EF sessions)
• Can we make MPGDs much more attractive to commercial applications
(medical imaging, homeland security,...) so we can get cheap industrial
mass production going?
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL
Thank you for your time!
1/11/2013
GEM Upgrade for CMS Forward Muon System - M. Hohlmann, Snowmass IF meeting, ANL