Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://zebu.uoregon.edu/~uochep/talks/talks05/DS-SiD.pdf Äàòà èçìåíåíèÿ: Sat Dec 3 01:52:10 2005 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 10:24:05 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: http astrokuban.info astrokuban

Silicon-tungsten calorimetry and the SiD forward region David Strom University of Oregon

Thanks to Tom Markiewicz(SLAC), Takashi Maruyama (SLAC), Mary Robinson (UO undergrad)

FCAL Tel-Aviv University

1

19 Sept 05 ­ David Strom ­ UO


Intro duction

· SiD is a silicon based detector ­ Silicon pixel vertex detector ­ Silicon strip tracking for the inner detector ­ Silicon-tungsten ECAL ­ Fine grained (RPC?) HCAL

· Cost optimization results in a relatively small radius of the inner detector of R 1.25m with large 5 T field

See Snowmass talks from Breidenbach and Weerts
FCAL Tel-Aviv University

2

19 Sept 05 ­ David Strom ­ UO


Energy flow drives most of the design Hermeticity is also imp ortant · Pro cesses such as
+- e+e- ~ ~ ~ ~ +0 -0

are tagged using momentum imbalance · Can b e faked by e+e- e+e- + - with missing final state e+ or e- that carry pt

FCAL Tel-Aviv University

3

19 Sept 05 ­ David Strom ­ UO


Quadrant View
8.000 Beam Pipe Ecal 7.000 Hcal Coil 6.000 MT Endcap 5.000 Endcap_Hcal Endcap_Ecal VXD m 4.000 Track Angle Endcap_Trkr_1 3.000 Endcap_Trkr_2 Endcap_Trkr_3 2.000 Endcap_Trkr_4 Endcap_Trkr_5 1.000 Trkr_2 Trkr_3 Trkr_4 0.000 Trkr_5 0.000 2.000 4.000 m 6.000 8.000 Trkr_1

FCAL Tel-Aviv University

4

19 Sept 05 ­ David Strom ­ UO


BR^2 Fixed, Vary R_Trkr
1000

800

600

400

R_Trkr M$ 200 d$/dR

0 0.00 -200 0.50 1.00 1.50 2.00 2.50

-400

-600

FCAL Tel-Aviv University

R_Tr5 r (m) k

19 Sept 05 ­ David Strom ­ UO


Bill Cooper's opening scheme (Endcap not split)

FCAL Tel-Aviv University

6

19 Sept 05 ­ David Strom ­ UO


Si-W Calorimeter Concept ECAL

Inner Tracker

1.25m

Rolled Tungsten
3.6 Meters 1.1-1.3 Meters

Silicon Wafers

Transverse Segmentation ~5mm 30 Longitudinal Samples 1/2 Energy Resolution ~15%/E
FCAL Tel-Aviv University

Polyimide Cables
Layer Assembly

7

19 Sept 05 ­ David Strom ­ UO


Silicon Concept

· Readout each wafer with a single chip · Bump b ond chip to wafer · To first order cost indep endent of pixels /wafer · Hexagonal shap e makes optimal use of Si wafer · Channel count limited by p ower consumption and area of readout end chip · May want different pad layout in forward region

FCAL Tel-Aviv University

8

19 Sept 05 ­ David Strom ­ UO


Silicon Detector Design

· DC coupled detectors (avoids bias resistor work) · Two metal layers

net-

· Keep Si design as simple as p ossible to reduce cost · Cross talk lo oks small with current electronics design · Trace capacitances (up to 30pF) are bigger than the 5 pF pixel capacitance

FCAL Tel-Aviv University

9

19 Sept 05 ­ David Strom ­ UO


Notes on silicon based calorimetry · For large projects, price dep ends on silicon area (p erhaps 3/cm2)

· Ratio of area in a circle to a circle: 3 4

a hexagon inscrib ed square inscrib ed in a 3

3 2

r

r

= 1.29

· Ratio of dead edge (e is edge width) space hexagon to square
4 e 3r = 2 2e r

2 = 0.81 3 detectors is

r/ 2

r

· Area 2
3 3r 2

of current 118cm2

FCAL Tel-Aviv University

10

19 Sept 05 ­ David Strom ­ UO


· Price of mask set for different shap es is likely a fixed cost of ab out 50K USD. · 50K USD is equivalent to approximately 150 wafers at 3 USD/cm2 · Price could b e reduced if sp ecialized pieces where built in-house. · More than one small piece/wafer may b e p ossible

FCAL Tel-Aviv University

11

19 Sept 05 ­ David Strom ­ UO


Possible endcap layout · Requires 15 different mask sets for zero crossing angle · Nearly twice as many are required if luminosity monitor is offset to b e centered on outgoing b eam · Could rotate silicon pattern 45 on alternate layers · Will need 1cm clearance b etween lumi and endcap.
Barrel ECAL Lumi

FCAL Tel-Aviv University

12

19 Sept 05 ­ David Strom ­ UO


Alternative layout
Barrel ECAL Lumi

· Radial design allows for radial pads on endcap · Radial design would require at least 13 different masks · Wedges make p o or use of silicon area on a wafer · For an Octagonal outer edge, many more masks needed

FCAL Tel-Aviv University

13

19 Sept 05 ­ David Strom ­ UO


Crossing Angles ­ Two regimes

· Large crossing angle with separate input and exit ap ertures. Minimum angle limited by the magnet size. · Minimum large angle could b e as large as 20 mrad

· Small crossing angle with shared input and exit ap ertures. Maximum angle limited by magnet b ore (typically 8-9 cm) · Maximum angle is 2 mrad

For physics zero crossing angle is usually preferred. For machine op eration a large crossing angle is b etter.
FCAL Tel-Aviv University

14

19 Sept 05 ­ David Strom ­ UO


Advantages of large crossing angles for machine op eration:

· 10% of 10 MW b eam p ower in photons is brought directly to the b eam dump · Input and output b eam optics can b e adjusted with maximum flexibility · Synchrotron radiation from soft b ends in incoming b eam can b e blo cked without interfering with outgoing b eam · Allows for op eration at s > 1 TeV

· Allows for small b ore ( 2 cm magnets) · Allows for down-stream instrumentation for determining b eam p olarizations
FCAL Tel-Aviv University

15

19 Sept 05 ­ David Strom ­ UO


Disadvantages of small cross angle for machine op eration:

· Requires very large b ore magnets for final elements

· 2 mrad b end imp oses on CM energy limit from synchrotron radiation

· Shared function of magnets leads to less flexibility for the machine

· Large amount of energy dissipated in final magnets

FCAL Tel-Aviv University

16

19 Sept 05 ­ David Strom ­ UO


· Snowmass conclusion (Andrei Seryi)

L High Lumi

20 mrad IR
Low ~2W High ~1000 W

vs
High Lumi

L

2 mrad IR
preliminary & qualitative High Very High To be evaluated

Nominal

Very Low <1W

Very Low <1W

Nominal

Low ~1W

Low ~1W

E 500 GeV 1 TeV 500 GeV 1 TeV
(Numbers in Watts show losses on SC FD magnets)

E

FCAL Tel-Aviv University wg4 1st week summary

· Optimization of design and evaluation will continue, but clear that disrupted beam losses on SC elements limit performance · Better detector hermeticity & background of 2mrad IR comes together with lower luminosity reach · ( 20mrad IR works well with New High L parameters ) ( 2mrad to be evaluated )
17
10

19 Sept 05 ­ David Strom ­ UO Aug 19, 2005


· SiD designed to accommo date a 20 mrad crossing angle

Si D Forward Masking, Calorimetry & Tracking 2005-09-15 20mrad, L*=3.51m
E HCAL C A L Muon Yoke

QD0

SD0

QF1

CRAB Lo-Z

BeamCal LumCal Q-EXT

Support Tube
FCAL Tel-Aviv University

18

19 Sept 05 ­ David Strom ­ UO


Probably b est to center luminosity monitor on outgoing b eam
Vertical 1:5 Horizontal 1:25
0.25

0.2

0.15

0.1

0.05

0

-0.05

-0.1

-0.15

-0.2

-0.25 0 2

FCAL Tel-Aviv University

19

19 Sept 05 ­ David Strom ­ UO


How do es a kink in the solenoid effect the background?
100 20 mrad + DID 20 mrad 2 mrad 80 20 mrad + Kink solenoid pointing incomin beam 20 mrad + Kink solenoid pointing outgoing beam

Energy in BEAMCAL (TeV)

60

40

Energy In BeamCal (TeV)
20

0 0.0 0.5 1.0 1.5 2.0 2.5 3.0

Beampipe Radius (cm)
FCAL Tel-Aviv University

20

19 Sept 05 ­ David Strom ­ UO


· Detail showing clearance b etween lumi and endcap. · Note that HCAL is centered on the detector axis. · Some electrons will b e tagged by the HCAL · Must supp ort luminosity monitor with a minimum of material outside of detector
17.4 cm

FCAL Tel-Aviv University

21

19 Sept 05 ­ David Strom ­ UO


Lumi Detector Geometry · It is essential to survey the detector at the micron level with cosmic ray muons or test b eam. Electronics must have MIP sensitivity even if it is not needed by the luminosity measurement MIP sensitivity needed for p ossible muon veto (See Graham Wilson's Calorimeter talk). · Detectors should fit on a single wafer
8 inch wafer

· SiD geometry Rmin 8.7 cm ( 50mrad) Rmax 24.7 cm( 150mrad) 8 inch wafers would b e needed

64 radial pads

Moliere Radius

· Rate at 500 GeV is 8 bhabhas bunch train ­ Inner radius could b e much larger and 6 inch wafers used
FCAL Tel-Aviv University

22

19 Sept 05 ­ David Strom ­ UO


Electronics · The ECAL electronics (Kpics) have chips with 1024 channels and store up to 4 hits/bunch train:

reset 10 pF 0.5 pF

Threshold 1

+ -

reset holdoff x4

1V

+ -

digital memory

Scale Select (1 bit)

Detector
Charge Amplifier

shaper analog memory Charge (12 bits)

+Bias

Threshold 2 beam crossing counter

+ digital memory 13

beam crossing (13 bits)

FCAL Tel-Aviv University

23

19 Sept 05 ­ David Strom ­ UO


A note on a shaping times · For the ECAL calorimeter, maximum leakage current from radiation is exp ected to b e < 10nA. Gives negligible contribution to noise (< 500 electrons) · In the lumi-detector, this numb er could b e a factor of 30 higher or more. · At 300nA, 1 µs integration, shot noise is 1400 electrons. · A shorter shaping time may b e needed.

FCAL Tel-Aviv University

24

19 Sept 05 ­ David Strom ­ UO


· Calorimeter of buffers for capacitors for same area as

approach could b e extended to include a large numb er each channel, e.g. 128 Channels Chip each with 128 storing charges for up to 128 bunches would fit into the is presently used for the calorimeter:

Single Cell Layout
FCAL Tel-Aviv University

25

19 Sept 05 ­ David Strom ­ UO


· A more likely scenario would b e to readout every bunch crossing. This would require a different digitization technique, either SAR or a pip elined ADC with 12 stages. 128 channels would probably fit into the space presently used by the 1024 · Assume Successive Approximation ADC with 12 bits + range, digitizing at 3MHz (internal clo ck is 36MHz). Data rate is 576 MBytes/s/chip during bunch train ( 3.0MBytes/s sustained) · On-detector electronics cost will b e dominated by development costs (very similar to run needed for test b eam) · Won't save much money by reducing channels/wafer · Power consumption should b e reasonable, but no design yet for co oling in the endcap in SiD. LDC will b e easier.

FCAL Tel-Aviv University

26

19 Sept 05 ­ David Strom ­ UO


Conclusion · SiD thinking ab out forward region is in very early stages ­ your input is needed and welcome · Precision luminosity measurement with cross angle seems p ossible · Main lumi worry is p otential background from machine and physics backgrounds Main overall challenge is to engineer luminosity monitor and endcap without large areas with dead material · Many costs in the forward region will b e fixed by R&D rather than part count ­ keep this in mind when designing detectors

FCAL Tel-Aviv University

27

19 Sept 05 ­ David Strom ­ UO