1 SciFi Results and Comparison Malcolm Ellis, for the MICE Scintillating Fibre Group Abingdon, 31 st October 2003.

1

SciFi Results and ComparisonSciFi Results and Comparison

Malcolm Ellis,

for the MICE Scintillating Fibre Group

Abingdon, 31st October 2003

2

The SciFi GroupThe SciFi Group

JapanK. YoshimuraKEKY. Kuno, H. Sakamoto, A. Sato, M. Yoshida,Osaka

UKP. KyberdBrunelA. Khan, L. TongEdinburghG. Barber, M. Ellis, K. Georgiou, R. Goncalo, K. Long, J. Sedgbeer, A. TapperImperial College LondonP. Cooke, R. GametLiverpool

USA. Bross, J. Estrada, J. Krider,

R. Rucinski, P. RubinovFNAL

D. Kaplan, Y. TorunIllinois Institute of Technology

G. Hanson, A. KlierRiversideX. Yang

UCLA

3

The TrackerThe Tracker

4

The Tracker DocumentThe Tracker Document

Tracker Document asked specific questions, some common to both detectors, others specific to the individual types.

I will first show the various things that have been done in the past few months to allow us to say Yes, and then list those few items we have not answered so far.

5

SimulationSimulation

All software in G4MICE environmentSimulation of nominal configuration

(0.25% dead channels, 7-1 grouping, 1 GHz RF background, etc…)

Position, momentum, x’, y’, t’ resolutions determined

No degradation in resolutions for 100x nominal background rate.

6

Position ResolutionPosition Resolution

•X Resolution: 570 microns

•Y Resolution: 430 microns

7

Momentum ResolutionMomentum Resolution

•PT Resolution: 1.3 MeV/c

•PL Resolution: 5.8 MeV/c

8

X’ Y’ and T’X’ Y’ and T’

• X’ Resolution: 7.7 mrad

•Y’ Resolution: 4.9 mrad

•T’ Resolution: 6.7 x 10-3

9

Pattern Recognition EfficiencyPattern Recognition Efficiency Number of space

points in fitted track (or zero if no track fitted)

Most of the time 5 space points are made and fitted in track

With no dead channels or background: 99.9 ± 0.1 %

Overall tracking efficiency in nominal situation: 99.0 ± 0.5 %

10

Future ImprovementsFuture Improvements

Track fit still needs correct treatment of Energy Loss in Kalman package – this will be added soon

Track fit currently uses space points, made from intersection of three views in a station. We can add individual fibres as measurements, and thus make full use of all available information

Optimise grouping so as to maintain channel count, but increase point resolution at last station (for example).

11

RF TestRF Test

201 MHz RF transmitted in close proximity to the VLPC cryostat.

Transmitting antenna operated horizontally and vertically

Received power measured on scope and compared to Lab G test with the same antenna.

• Pedestal width increased by 0.5 ADC (0.033 PE)

12

SciFi PropertiesSciFi Properties

Clear fibre attenuation length measured: 7.62m

Scintillating Fibre is Kuraray SCSF-3HF Light Yield measured for various

scintillator types, pT + 5000 ppm has highest yield at 2.6 PE per MIP, measured with a PMT

Systematic effect to be understood, however no cross-talk was observed in any of the scintillating fibres.Aluminized tape

22mm

20 fibers

13

Tracker Prototype ConstructionTracker Prototype Construction

Doublets (ribbons, planes, a single view) constructed at Fermilab

Waveguides built and tested at KEK/Osaka Carbon Fibre Supports (IC/Liverpool) Planes aligned and glued to carbon fibre

supports Fibres bundled into groups of 7 Bundles threaded into connector blocks Bundles potted, cut and polished Final assembly and test at Fermilab

14

Cosmic Ray Data TakingCosmic Ray Data Taking

Two 36x36 cm2 scintillators in coincidence form trigger.

Lead sheet used for a momentum filter VLPC timing, pedestal and gains

calibrated prior to waveguide connection Light yield measured Proper efficiency, dead channels, tracking

to come soon…

15

VLPC GainVLPC Gain

pedestal

1p.e.2p.e.

Typical Gain:15 ADC counts/ p.e.

Preamp Gain Setting: 0x350

Gain Setting: 0x150

6 ADC counts/ p.e.

16

Chose events with hits in Chose events with hits in Upper and Lower StationsUpper and Lower Stations

Station A

Station C

X view V view

W viewX view

> 10 counts to select cosmic-ray

17

Light Yield in Station BLight Yield in Station BX view: PT+3HF 5000ppmW view: PT+3HF 3500ppmV view: PT+3HF 2500ppm

W view X view

V viewPeak at 24 counts4 p.e.Mean is 4.6 p.e.

Expectation from beam test in KEK• 2.6 p.e. in 0.35mm scifi• QE of VLPC ~ 4 times larger than PMT• Transport loss in waveguide ~ 40%• assume connection loss ~ 10% at each optical connector 5p.e.•This has a large uncertainty due to PMT calibration

18

Expectation based on D0Expectation based on D0

From measurements with 835 micron fibre and 11.5 m waveguide – expect– 10-12 pe in the current MICE prototype configuration

Why the discrepancy?– Construction process

• Aluminisation, ribbon, bundling, connectorising, cutting, polishing, etc.

– DAQ• Gate timing

Under investigationCheck DAQ timing– Independent measurement of fibre light yield

19

Single Plane EfficiencySingle Plane Efficiency

Tracking not reliable yet, so calculate lower bound on efficiency

Require clear hit in the V and W projections of station B

Count hits above threshold in third (middle) station

Efficiency 95% - consistent with expectations from light yield and threshold setting

20

Dead ChannelsDead Channels

Cosmic Data rate low, only 7000 triggers, however in these, there are on average 1.29 channels per plane which never have the highest hit in the plane in an event.

This gives an upper estimate for the dead channel rate of 0.6%

Note that as with the efficiency, this number has a large uncertainty. We will improve this with tracking.

21

Construction Schedule and CostConstruction Schedule and Cost

Prototype was completed pretty much to schedule and budget.

Assume we will build 3 trackers, to allow contingency for R&D, spares.

First Iteration schedule has delivery on time in 2006 if we start “now”.

Full cost, $1.9M – possible saving of $975k through loan of D0 spares

Excellent opportunity for collaboration with D0, but can’t delay decision long.

22

What were the “No” AnswersWhat were the “No” Answers

Noise on the mains (filterable)– D0 has experience of such noise in the

calorimetry electronics but not in the Fibre Tracker.

– Paul has said we may be able to do this at RAL Didn’t measure performance with known

sources (this is known, so perhaps lower priority)

G4MICE has not yet provided emittance reconstruction

23

AchievementsAchievements

7 planes (in 3 stations) of Scintillating Fibre detector have been successfully constructed and operated at D0.

All simulation and reconstruction software is committed to G4MICE and can be used to check fine details of remaining question marks (PT resolution, etc…)

24

FutureFuture

Open Issues:– Understanding Light Yield– Resolution in reconstructed parameters

and thus emittanceWe plan to produce a document by

the end of the year addressing all aspects (including the open issues)