Jerry Blazey NIU/NICADD Towards A Scintillating (Semi)- Digital Hadron Calorimeter: Progress at NIU/NICADD Jerry Blazey Northern Illinois University

Jerry Blazey Jerry Blazey NIU/NICADDNIU/NICADD

Towards A Scintillating Towards A Scintillating (Semi)-Digital Hadron (Semi)-Digital Hadron

Calorimeter: Progress at Calorimeter: Progress at NIU/NICADDNIU/NICADD

Jerry BlazeyJerry Blazey

Northern Illinois UniversityNorthern Illinois University


LC Activities at NIU/NICADDLC Activities at NIU/NICADD

Scintillator (Semi-)Digital Hadron Scintillator (Semi-)Digital Hadron Calorimeter Simulation and Calorimeter Simulation and Hardware StudiesHardware Studies – This Talk – This Talk

Test Beam Plans for Scintillator Test Beam Plans for Scintillator Hadron Calorimeter & Tail-catcherHadron Calorimeter & Tail-catcher – – Vishnu Zutshi – This SessionVishnu Zutshi – This Session

G4-based Simulation Status & Plans G4-based Simulation Status & Plans – Guilherme Lima– Guilherme Lima – Session 7 Friday – Session 7 Friday 8:308:30

Muon Simulation Development & Muon Simulation Development & StatusStatus – Arthur Maciel – Muon/PID – Arthur Maciel – Muon/PID Session Wednesday 1:00Session Wednesday 1:00


““Generic” Calorimeter Simulations Generic” Calorimeter Simulations First Design & Prototype & Results First Design & Prototype & Results

on Sensitivity and Thresholdon Sensitivity and Threshold Optimization of Unit CellsOptimization of Unit Cells Light Sensor InvestigationsLight Sensor Investigations


A Generic Calorimeter:A Generic Calorimeter: Number of Cells vs. Pion Number of Cells vs. Pion

EnergyEnergy

For a 0.25 mip threshold # cells monotonicallyincreasing with energy for a wide range of cell sizes.

0.25mip threshold

# ofCells

E

100

20


Digital vs. AnalogDigital vs. Analog

Very similar correlations exist for hits or energy Between the EMCAL and HCAL

Energy 10,50 GeV Hits 10,50 GeV

Hits HCAL E HCAL

Hits ECAL E ECAL


Single Particle Energy Single Particle Energy ResolutionResolution

Minimize (EMinimize (Eoo--aaiiLLii))22

aaii calculated for 10 GeV & applied calculated for 10 GeV & applied to all E which is conservativeto all E which is conservative

i=2 for EMCAL & HCAL, also i=2 for EMCAL & HCAL, also conservativeconservative


Single Particle Energy Single Particle Energy Resolution Resolution

Non-projective geometry

For lower energy particles digital approach has superior resolution!

/E

E

0.1

20


Resolution as a Function Resolution as a Function ofof

Multiple Thresholds or Multiple Thresholds or BitsBits

* As in the previous slide, below 20 GeV digital resolution superior to analog.* At all energies, more bits superior.

/E)dig

/E)analog

E

So it works for single particles how about jets?

1

2


Toy Simulation: Toy Simulation: “Recipe” for a Jet“Recipe” for a Jet

Determine resolution independent of algorithmDetermine resolution independent of algorithm For ZZ events PFor ZZ events PTT order stable MC particles, order stable MC particles,

ignore ignore ’s’s For charged hadrons assume perfect energy For charged hadrons assume perfect energy

(from tracker)(from tracker) Smear the energy of other particlesSmear the energy of other particles

– For neutral hadrons use resolutions for charge pions (just For neutral hadrons use resolutions for charge pions (just discussed). discussed).

– For photons use For photons use ~ 17%/sqrt(E) ~ 17%/sqrt(E)

Start with highest pStart with highest pTT particle and cluster in 0.7 particle and cluster in 0.7 conecone

Repeat for remaining particlesRepeat for remaining particles Add individual energies to get jet energyAdd individual energies to get jet energy


ZZ Events: Sanity ChecksZZ Events: Sanity Checks

Stable MC particles

Energy Fractions

Neutral hadron fraction

fraction


Jet E ResolutionJet E Resolution

rms used

/E

Jet E(GeV)

So the idea holds water: At all energies 3x3 single

threshold resolution comparable to analog!


Using full E-flow: Jet Using full E-flow: Jet EErecrec/E/Egengen

~60% better

(Vishnu Zutshi, ECFA-DESY Workshop, 1/4/2004http://nicadd.niu.edu, presentation 0046)

= 0.25 = 0.16

Calorimeter only Eflow


Full Eflow: Jet EFull Eflow: Jet Erecrec/E/Egengen

= 0.17 = 0.16

Eflow digital (2cm2 cells) Eflow analog

Digital approach not yet optimizedbut performance comparable to

analog!


Hardware Prototypes:Hardware Prototypes:Stack, Layer, & Unit CellStack, Layer, & Unit Cell

Clear FiberMPTM


Cosmic Data with PMT Cosmic Data with PMT ReadoutReadout

~11 p.e. peak = 1MIP


Efficiency and Noise Rejection

0%

20%

40%

60%

80%

100%

120%

Number of MIPs

Pe

rce

nt

Efficiency

Noise Rejection

0.25 MIP threshold: efficient, quiet


Cell Response Uniformity Cell Response Uniformity & Dispersion& Dispersion

0

0.2

0.4

0.6

0.8

1

1.2

0 10 20 30 40 50

POSITION OF Sr-90, MM

NO

MA

LIZE

D R

ESPO

NSE

Column1Column1

MeanMean 1562.5061562.506

Standard ErrorStandard Error 24.5264724.52647

MedianMedian 1557.961557.96

ModeMode #N/A#N/A

Standard DeviationStandard Deviation 115.0394115.0394

Sample VarianceSample Variance 13234.0513234.05

KurtosisKurtosis -0.05291-0.05291

SkewnessSkewness 0.3349390.334939

RangeRange 444.52444.52

MinimumMinimum 1386.471386.47

MaximumMaximum 1830.991830.99

SumSum 34375.1434375.14

CountCount 2222

Cell-to-cell ~ 7%Uniformity ~ 3%


Fiber Response Fiber Response Dominates:Dominates:

Dispersion ~ 6% Dispersion ~ 6%Histogram

0

2

4

6

8

10

12

14

16

-1.3

0E-0

6

-1.2

6E-0

6

-1.2

2E-0

6

-1.1

8E-0

6

-1.1

4E-0

6

-1.1

0E-0

6

-1.0

6E-0

6

-1.0

2E-0

6

-9.8

0E-0

7

-9.4

0E-0

7

-9.0

0E-0

7

Bin

Freq

uenc

y

Frequency

Column1

Mean -1.14599E-06Standard Error 6.53057E-09Median -1.1464E-06Mode -1.10134E-06Standard Deviation 6.59555E-08Sample Variance 4.35012E-15Kurtosis 0.427712361Skewness 0.44298345Range 3.4981E-07Minimum -1.27961E-06Maximum -9.298E-07Sum -0.000116891Count 102Largest(1) -9.298E-07Smallest(1) -1.27961E-06Confidence Level(95.0%)1.29549E-08


Other Uniformity Other Uniformity MeasurementsMeasurements

-0.200

0.000

0.200

0.400

0.600

0.800

1.000

1.200

0 10 20 30 40 50

POSITION OF THE Sr-90, MM

NO

RM

ALI

ZED

RE

SP

ON

SE

0.000

0.200

0.400

0.600

0.800

1.000

1.200

0 10 20 30 40 50 60

POSITION OF THE Sr-90, MM

NO

RM

ALI

ZED

RE

SP

ON

SE


Relative Response Relative Response MeasurementsMeasurements

CellCell GrooveGroove AreaArea ResponseResponse

HexagonHexagon SigmaSigma 9.49.4 1895.31895.3

SquareSquare SigmaSigma 9.49.4 1665.81665.8

SquareSquare SigmaSigma 66 1740.51740.5

HexagonHexagon SigmaSigma 66 1743.81743.8


SquareSquare StraightStraight 9.49.4 1523.41523.4

SquareSquare StraightStraight 44 1618.61618.6

SquareSquare StraightStraight 9.49.4 861.5861.5

HexagonHexagon StraightStraight 9.49.4 900.9900.9


Since light ample, can optimize for ease of construction


Surface Surface Treatment/WrappingTreatment/Wrapping

TyvekTyvek PaintPaint VM 2002VM 2002 MylarMylar CM590CM590 CM500CM500 Alum FoilAlum Foil

1.001.00 0.890.89 1.081.08 0.830.83 0.280.28 0.440.44 0.630.63

UNPOLISHED TOP ANDUNPOLISHED TOP AND

POLISHED BOTTOMPOLISHED BOTTOMPOLISHED TOP ANDPOLISHED TOP AND

POLISHED BOTTOMPOLISHED BOTTOMUNPOLISHED TOP ANDUNPOLISHED TOP AND

UNPOLISHED BOTTOMUNPOLISHED BOTTOM

0.980.98 1.001.00 1.021.02

Paint easy, little light lossPaint easy, little light loss


Miscellaneous Miscellaneous Measurements:Measurements:

area, goove type, profile, area, goove type, profile, source, glues, fiberssource, glues, fibers

6cm6cm22/9cm/9cm22 Straight/SigmaStraight/Sigma Square/HexSquare/Hex Extruded/CastExtruded/Cast

0.950.95 0.950.95 0.940.94 0.70.7

After/Before glueAfter/Before glue EJ500/BC600EJ500/BC600

(optical glues)(optical glues) Y11/BCF92Y11/BCF92

1.151.15 1.01.0 3.13.1

NICADD scintillatorNICADD scintillator1mm round Kurray

0.8 mm

square B

icron

Can tune light yield with fiber typeCan tune light yield with fiber type.


NICADD Extruder @ NICADD Extruder @ FermilabFermilab


Thickness Tolerance: 2-3%Thickness Tolerance: 2-3%Response Depends weakly on Response Depends weakly on

Thickness: ~20%/mmThickness: ~20%/mm

NORMALIZED CELL RESPONSE OF Cs-137

y = 0.8426x + 0.1376R2 = 0.9852

0.75

0.95

1.15

1.35

1.55

1.75

0.75 0.95 1.15 1.35 1.55 1.75

CELL THICKNESS NORMALIZED TO 3 MM

RE

SP

ON

SE

NO

RM

AL

IZE

D T

O 3

MM

CE

LL

3mm4mm

5 mm

Extruded Tile

4.78

4.8

4.82

4.84

4.86

4.88

4.9

4.92

1 2 3 4 5 6

Position in 20 cm steps

Thick

ness

in m

m Side oneSide two

Thickness not an issue


Optimum CellOptimum Cell

Hexagonal or SquareHexagonal or Square 4 - 9 cm4 - 9 cm22

Straight GrooveStraight Groove High efficiency fiberHigh efficiency fiber Glued Fiber and Painted SurfaceGlued Fiber and Painted Surface Extruded (cut costs) @ 5mmExtruded (cut costs) @ 5mm

But a bigger question is the light sensor: But a bigger question is the light sensor:

PMTs costly, bulkyPMTs costly, bulky

we have been investigating APDs, MRS, Si-PM…we have been investigating APDs, MRS, Si-PM…

My currentguess…


Hamamatsu Avalanche Hamamatsu Avalanche Photo-DiodesPhoto-Diodes

Gain for Hamamatsu APD for different light wavelengths at 18 ºC

1.0

10.0

100.0

1000.0

100 150 200 250 300 350 400

Bias Voltage, V

Ga

in

486nm 565nm for 587nm 660nm


Cosmic MIP with Cosmic MIP with Avalanche Avalanche

Photo-DiodePhoto-Diode

Hamamatsu S8550


MetallicMetallicResistiveResistiveSensors Sensors

LED signal

0

200

400

600

800

1000

1200

1 33 65 97 129

161

193

225

257

289

321

353

385

417

449

481

513

ADC counts

Even

ts


Cosmics with MRSCosmics with MRS

Range of working points

y = 24.255x - 1180.5

R2 = 0.9969

0

10

20

30

40

50

60

49.8 50 50.2 50.4 50.6 50.8 51

Voltage (V)

Avera

ge m

inu

s p

ed

esta

l

~5 PE


Si-PM’s (mounted on cell?)Si-PM’s (mounted on cell?)

Ru106, Si-PMT, 51 Volts, ~6 PE

0

100

200

300

400

500

600

700

800

900

1000

1 56 111 166 221 276 331 386 441 496 551 606 661 716 771 826 881 936A D C C HA N N EL

MEPHI sample, Courtesy of B.Dolgoshein


Cosmic Data with Si-PMCosmic Data with Si-PMN

umbe

r of

P.E

.

Comparable to PMTComparable to PMT


Tabulated Studies/SpecsTabulated Studies/SpecsDevice HAMAMATSU

APD MRS or SiPM

VLPC PMT

Input sig. from MIP (photons)

60

60

60

60

Photo Electrons

48 77 48 77--1100 Gain 400 ? 10E(6) 10E(5) 10E(6) APD output Charge (fC)

3

1152

768

1152

S/N(room T)

~ 5.5 Est. ~ 3 real ~ 8.1 (10oC)

~ 8 meas.* >10 (9K) meas.**

> 10 meas.***

* B. Dolgoshein An Advanced Study of Silicon PM ICFA IB 2002

**A. Bross et all. Fermilab FN 0733 2003

*** Rykalin V. NICADD presentation http://nicadd.niu.edu 2002 Believe <$10/unit in bulk for SiPMBelieve <$10/unit in bulk for SiPM


Scintillator DHC Scintillator DHC ConclusionsConclusions

Simulations indicate approach competitive Simulations indicate approach competitive with analog approachwith analog approach

Prototypes indicate there is sufficient Prototypes indicate there is sufficient sensitivity (light x efficiency) & uniformity.sensitivity (light x efficiency) & uniformity.

Now optimizing materials & construction to Now optimizing materials & construction to minimize cost with required sensitivityminimize cost with required sensitivity

SiPM and MRS look very promisingSiPM and MRS look very promising

All-in-all looks like a very competitive option….All-in-all looks like a very competitive option….

We’ll be moving towards the next prototypeWe’ll be moving towards the next prototype

Documents

Jerry Blazey NIU/NICADD Towards A Scintillating (Semi)- Digital Hadron Calorimeter: Progress at NIU/NICADD Jerry Blazey Northern Illinois University