Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://zebu.uoregon.edu/~uochep/talks/talks07/ecalrd.pdf Äàòà èçìåíåíèÿ: Mon Jul 2 21:46:53 2007 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 10:24:25 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: jet

LCDRD ECal R&D
R. Frey, University of Oregon

·

Physics goals drive the design

·

ECal with scintillator tiles (Project 6.2) Colorado ECal design studies (Project 6.10) Kansas Development of an silicon-tungsten ECal (Project 6.5) SLAC, Oregon, UC Davis, BNL, Annecy

·

·

R Frey

ANL R&D Review

1


Hadronic final states and PFA
Complementarity with LHC: LC should strive to do physics with all final states.
1. Charged particles in jets more precisely measured in tracker 2. Jet energy 64% charged (typ.) Separate charged/neutrals in calor. The "Particle Flow" paradigm · ECAL: dense, highly segmented (an "imaging calorimeter")

R Frey

ANL R&D Review

2


tau id and polarization
· · · Analysis of tau final states can provide crucial information on new physics Important & broad example: The SUSY model leaves fingerprint on tau polarization:

References: M. Nojiri, PRD 51 (1995) E. Boos, et al, EPJC 30 (1993) Godbole, Guchait, Roy, Phys Lett B (2005)
R Frey ANL R&D Review

tan

SPS1a

P

3


tau polarization (contd) - measurement
Separate the important decay modes:

· + + (+o) · + + (+) · + a1+ (+ + -, +o o )
and measure the energy spectrum as done at LEP (ALEPH best by 2â) An important tool to have in the box.

R Frey

ANL R&D Review

4


an imaging calorimeter (contd)
In addition to jets and taus: · Charged particle tracking, especially V0 recognition in silicon trackers · id hadrons which begin showering in the ECal · Photon vertexing
(e.g. GMSB SUSY)

·

o id to improve jet resolution
(G. Wilson, Kansas)

· · ·

final state id, eg electron id in/near jets


(E+ ­ E ) sin /E

Bhabhas, and acollinearity Hermiticity !
R Frey ANL R&D Review



5


Segmentation requirement
· · · In general, we wish to resolve individual photons from jets, tau decays, etc. The resolving power depends on Moliere radius and segmentation. We want segmentation significantly smaller than Rm

Two EM-shower separability in LEP data with the OPAL Si-W LumCal (David Strom) :

R Frey

ANL R&D Review

6


U. Of Colorado R&D (Project 6.2)
· Offset scintillator tiles to improve spatial resolution Proof of principle in simulation with single particles Requires studies of jet reconstruction For application now to scintillator HCal

· SiPM development for scintillator options · Simulation studies for forward calorimetry
SUSY and SUSY background
R Frey ANL R&D Review 7


Investigation of ECAL Concepts Designed for Particle Flow
5 GeV
0

4 mm pixels

RECAL = 1.27m

Project 6.10, PI Graham W. Wilson, University of Kansas ILC Detector R&D Review, Argonne, June 2007
For more details : see talks at LCWS07 (Calorimetry R&D Review and Sim/Reco)

8


Overview
· · Physics-driven ILC detector designs push the calorimetry in new directions. Physics needs:

­ Hermeticity
· Neutrinos, SUSY particles etc

­ Jet energy measurement
· Reconstruct W, Z, h, ...

­ General-purpose performance
· Prepare for the unexpected · Retain reasonable EM resolution, timing resolution. Particle-Flow approach has many open questions and opportunities for innovation

Assuming an excellent tracker, current PFA approaches indicate Ejet resolution has 3 major contributions · 1. Confusion (double counting). · 2. Intrinsic hadronic energy resolution. · 3. Intrinsic EM energy resolution. · This project focusses on investigating approaches which can address these limiting factors. · 1. Larger detector (GLD/LDC like)
­ Cost effective ECAL ­ Investigate ECALs with Si and Scint

·

·

·

3. High granularity ECAL for precision photon measurement
­ Use 0 mass constraint to improve E ­ Only use Si near the front of the ECAL?

­ ECAL is where showers start, is a big cost driver, and is at · 1,2. Precision timing to resolve 9 the heart of understanding 0 ,n using TOF confusion/reconstruct K L how to design a detector


Example detector model
A radially staggered buildable analog EM calorimeter. High granularity, Tungsten absorber, B = 3T.
R(m) Nlayers X0 Active Cell-size (mm)

frankyaug05, with N. Graf, M. Thomson

EM Barrel 1: 2.10 10 0.5 Si 2.5 â 2.5 â 0.32 EM Barrel 2: 2.13 10 0.5 Si 10 â 10 â 0.32 EM Barrel 3: 2.16 20 0.5 Sc 20 â 20 â 2
Choices made based on 2005 R&D work, driven by making a sensible, robust design with aggressive performance and minimizing Silicon area in a GLD-scale detector. Expect: E/E = 11%/E at low energy
10 50 GeV photon


Using 0 mass constraint to improve energy resolution of prompt EM component of jets
With aggressive design, have demonstrated that 300 m position resolution is achievable for a 1 GeV photon. All results here include the combinatoric issues.

15.7%/E Zqq

12.0%/E
V1.99 assignment algorithm

13.1%/E
V1.99 assignment algorithm
11

Perfect pairing 9.4%/E


Fast Timing / Temporal Calorimetry
Idea: time resolution at below the 100 ps level is now easily achievable with dedicated detectors. Can it be applied in a useful way in an ILC detector ?
Time delay is speed AND trajectory

Can TOF help measure neutral hadrons at low p ? K0 L 100 ps 100 ps n 50 ps 50 ps 25 ps 25 ps

n K
0

L

Can help resolving /±. (PID by TOF possible ­ but redundant with dE/dx in a TPC-based detector). Resolve confusion.

HCAL (LDC DOD)

TOF 12


A Silicon-Tungsten ECal with Integrated Electronics for the ILC (Project 6.5)
Baseline configuration: · transverse seg.: 13 mm2 pixels · longitudinal: (20 x 5/7 X0) + (10 x 10/7 X0) 17%/sqrt(E) · 1 mm readout gaps 13 mm effective Moliere radius

Currently optimized for the SiD concept
R Frey ANL R&D Review 13


Si/W ECal R&D Collaboration
M. Breidenbach, D. Freytag, N. Graf, R. Herbst, G. Haller, J. Jaros Stanford Linear Accelerator Center

· KPiX readout chip · downstream readout
· mechanical design and integration

J. Brau, R. Frey, D. Strom, undergraduates U. Oregon V. Radeka Brookhaven National Lab B. Holbrook, R. Lander, M. Tripathi UC Davis S. Adloff, F. Cadoux, J. Jacquemier, Y. Karyotakis LAPP Annecy
R Frey ANL R&D Review

· detector development
· readout electronics

· readout electronics · cable development · bump bonding
· mechanical design and integration
14


Goals of the R&D
Design a practical ECal which (1) meets (or exceeds) the stringent ILC physics requirements (2) with a technology that would actually work at the ILC. · The physics case calls for a dense (small Rm), highly segmented "imaging calorimeter" with modest EM energy resolution W-Si pixel sampling calorimeter The key to making this practical is a highly integrated electronic readout: readout channel count = pixel count /1000 cost independent of trans. segmentation for seg. > 2-3 mm
· 3.6 mm is current default

·

·

allows for a small readout gap (1 mm) small effective Rm (13 mm) low power budget (passive cooling) handles the large dynamic range of energy depositions (few thousand) This takes some time to develop (getting close).
R Frey ANL R&D Review 15


Silicon detector layout and segmentation
· One KPiX readout chip for the sensor (1024 pixels, 6 inch wafer) · KPiX also being considered for Si tracker
and DHCal with GEMs

· Limit on seg. from chip power (2 mm2 )

(KPiX) Fully functional v1 prototype (Hamamatsu)
R Frey ANL R&D Review 16


Critical design parameter is the gap between layers
· · ·
640 630 620 610 M$ 600 590 580 570 560 550 0.0000

Small gap maintains small Moliere radius Larger Rm larger detector to maintain shower separability cost ! Small gap makes a cost-controlled compact detector practical

SiD

0.0010

0.0020

0.0030

0.0040 Ga p (m )

0.0050

0.0060

0. 0070

0. 0080
17


readout gap cross section -- schematic

Metallization on detector from KPix to cable

Bump Bonds

Tungsten
KPix Si Detector

Kapton Data (digital) Cable

Kapton

Tungsten

Heat Flow

Gap 1 mm
R Frey ANL R&D Review

Thermal conduction adhesive

18


Conceptual Schematic ­ Not to scale

"Longitudinal" Data Cable

"Transverse" Data Cable

Locating Pins Detectors

Readout Chip "KPix" Tungsten Radiator Data Concentrator

~ 1m

R Frey

ANL R&D Review

19


Energy resolution
· · No physics case has emerged for EM energy resolution better than 0.15/E We have studied how to optimize energy resolution vs cost and Moliere radius

20


KPiX chip Dynamic gain select
Si pixel

One channel of 1024 13 bit A/D

Storage until end of train. Leakage current subtraction Pipeline depth presently is 4

Event trigger calibration

R Frey

ANL R&D Review

21


KPiX Cell 1 of 1024

64-channel prototypes: · v1 delivered March 2006 · v4 currently under test · v5 submitted (June '07) It's a complicated beast ­ will need a v6 before going to the full 1024-channel chip
R Frey ANL R&D Review 22


Dynamic Range
KPiX-2 prototype on the test bench

1 MIP (4 fC)
R Frey ANL R&D Review

Max signal: 500 GeV electron

23


Power

C ol d Tr a i n/ B unc h S t r uc t ur e A ve r a g e Po w e r ( mw )

Ph a s e

C u rre n t (m a )

I ns t a nt a ne ous Ti m e be gi n P o we r ( m w) (u s )

Ti m e E nd (u s )

Du t y F act o r

C o mme n t s

A l l A nal og " o n" H ol d " on" , c ha r ge a m p o f f A na l og pow er d ow n L V D S R ecei v er , e t c D ecode / P r ogr am ADC R ea dout To t a l

370 . 00 85. 00 4. 0 0

93 0. 00 21 0. 00 1 0. 00 3. 00 1 0. 00 10 0. 00 5 0. 00

0 . 00 1, 0 20. 0 0 1 , 021 . 00 1, 2 20. 0 0 1 , 020 . 00 2 00, 0 00. 00 0 1 1 , 021 1 , 220 . . . . 00 2 00, 0 00. 00 10 0. 00 1, 2 20. 00 3, 2 20.

5 . 10E - 03 9 . 95E - 04 9 . 95E - 01

4. 7 0. 2 9. 9 3. 0. 0. 0. 0 0 1 5

P ow er ok w i t h c ur r en t t hr o ugh F E T ' s

00 1. 00 E + 00 00 4 . 95E - 04 0 0 9 . 95E - 04 0 0 1 . 00E - 02

R e c ei v e r al w ay s o n. S eque nc i ng i s v agu e!

1 8 .5 T o ta l p o w e r O K

Passive conduction of 20 mW to module end (75 cm) via the tungsten radiator results in a few °C temperature increase OK !
R Frey ANL R&D Review 24


Noise in KPiX-4

· ·

1 MIP = 3.9 fC meets ECal S/N spec of 8/1 outliers probably due to routing issues
R Frey ANL R&D Review 25


Noise is gaussian

Am

Can set threshold at 0.5 MIP
R Frey ANL R&D Review 26


prototype Si detector studies

R Frey

ANL R&D Review

27


v2 Si detector ­ for full-depth test module

· 6 inch wafer · 1024 13 mm2 pixels · improved trace layout near KPiX to reduce capacitance · procurement in progress (it will take 6-12 months to complete the 40wafer purchase ­ funding limited)

R Frey

ANL R&D Review

28


Readout flex cable (digitized signals, power&control)
· First prototype: 2 stations Buried signal layer between power and ground Wire bond connections No problem for prototypes For ECal: 6 stations: should be OK Would like to determine length limit for next round (vias and multilayers difficult for 1m)

·

R Frey

ANL R&D Review

29


on multiple (2) Si-W R&D efforts
· The CALICE collaboration includes a very significant and well-funded Si-W R&D effort Their effort has focused on developing a test beam prototype using non-ILC technology
· Has collected data at DESY and CERN during the last year

More recently they have been developing a generation II design · · We decided to directly develop an ILC design (gen. II) Technology was proven in SLD, ALEPH, OPAL lum. calorimeters Many of our design innovations have been incorporated in the CALICE gen. II design Integrated electronics, power pulsing, small gaps, sub-cm transverse segmentation, etc This arrangement has been beneficial for developing a viable ILC ECal design with essentially no redundancy (so far)
R Frey ANL R&D Review 30

·


Si-W (project 6.5) Status Summary
· KPiX readout chip Currently studying v4 prototype (2x32 channels) Submit v5 in next few weeks (4x32 channels)
· Improved biasing of MOS capacitors; new poser bus for comparators · Optimized shaper time constants

·

· ·

Expect to submit 1024-channel KPiX in late Fall or Winter Silicon sensors v2 prototype submitted to industry (40 sensors) Schedule funding limited ­ hope to acquire sensors Fall-Winter Readout flex cable ­ short version for first module OK Bump bonding ­ first trials (UC Davis) just starting

Combine the above: a full-depth, single-wafer wide module Test in a beam: (1) electrons (2008); (2) hadrons with HCal

The R&D leading to an "ILC-ready" Si-W ECal technology is progressing well
R Frey ANL R&D Review 31


Extra stuff...

R Frey

ANL R&D Review

32


Future Si-W Development Milestones
I. · · II. · · III. · IV. V. Connect (bump bond) prototype KPiX to prototype detector with associated readout cables, etc Would benefit from test beam (SLAC?) - 2007 A "technical" test Fabricate a full-depth ECal module with detectors * and KPiX1024 readout * ­ functionally equivalent to the real detector Determine EM response in test beam ­ 2008 Ideally a clean 1-30 GeV electron beam (SLAC??) Test with an HCal module in hadron test beam (FNAL?) ­ 2008-? Test/calibrate the hadron shower simulations; measure response Pre-assembly tests of actual ECal modules in beam ­ >2010 Develop mechanical design, 2008



pending funding

33


R&D Milestones and test beams
I. · · II. · · III. · IV. Connect (bump bond) prototype KPiX to prototype detector with associated readout cables, etc Would benefit from test beam (SLAC?) - 2007 A "technical" test Fabricate a full-depth ECal module with detectors and KPiX-1024 readout ­ functionally equivalent to the real detector Determine EM response in test beam ­ late 2007-8 Ideally a clean 1-30 GeV electron beam (SLAC?) Test with an HCal module in a hadron beam (FNAL?) ­ 2008-? Test/calibrate the hadron shower simulations; measure response Pre-assembly tests of actual ECal modules in beam ­ >2010-?
R Frey ANL R&D Review 34


Longitudinal Sampling
Compare two tungsten configurations: · 30 layers x 5/7 X
0

· (20 x 5/7 X0) + (10 x 10/7 X0)

· Resolution is 17% / E , nearly the same for low energy (photons in jets) · Better for the 20+10 config. at the highest energies (leakage) adopt as baseline

R Frey

ANL R&D Review

35


Electronics requirements
· Signals <2000 e noise Require MIPs with S/N > 7 Large dynamic range: Max. signal is 2500 MIPs (for 5mm pixels) Capacitance Pixels: 5.7 pF Traces: ~0.8 pF per pixel crossing Crosstalk: 0.8 pF/Gain x Cin < 1% Resistance (traces) 300 ohm max Power If < 40 mW/wafer allows passive cooling (as long as power is cycled off between bunch trains) Provide fully digitized, zero suppressed outputs of charge and bx time on one ASIC for every wafer.
R Frey ANL R&D Review

Use DC-coupled detectors: only two metal layers (cost)

·

· ·

·

36


Response of Detectors to 60KeV Gamma's from Am241

Possible ~1% wafer-wafer calibration?
R Frey ANL R&D Review 37


Noise is consistent with expectation from capacitance and series resistance
R Frey ANL R&D Review 38


Backup Slide

39


Summary on potential of 0 massconstraint in hadronic events (s=mZ)
1. Perfect pairing

(uses fit to the error distribution from the fit)

Using fitted of deviation on same 10k events

7.9 15.7 31.0

2. Assignment algorithm 1.99 6.0 6.8 7.5 12.0 13.1 14.8 24.9 26.1 28.7
40