Scintillating Fiber Tracker

Supported by 2014 JSA Postdoc Prize
Zhihong Ye
Duke University
JLab User Group Meeting, 06/03/2014
Original Motivation:
The Proton Charged Radius Experiment (PRad) in Hall-B
High resolution, large acceptance, hybrid HyCal calorimeter (PbWO4 and Pb-glass)
Measure GEp within Q2 range of 2x10-4 – 2.0x10-2 GeV2 (lower than all previous (e,p) experiments)
Simultaneous detection of elastic and Møller electrons
Windowless H2 gas flow target
Add a new position detector here
 To increase the resolution at the lowest Q2 points, we
decied to add a new position detector with additional
Thin  Not too much space between Vacuum
Box Exit and HyCal
Minimum radiation materials  Control the
background events at a small level.
 Allow a hole at the center for the electron beam to
Spokesperson: A. Gasparian, D. Dutta, H. Gao, M. Khandaker
go through
Original Motivation:
The Proton Charged Radius Experiment (PRad) in Hall-B
 Possible Candidates of Position Detectors (or Tracking Devices):
Drift Chambers (DC)
• Provide <100 um position resolution; Thin; Widely used;
• No enough time to design, built and test a 1.2 meter x 1.2 meter large DC;
• Hard to produce a hole at the center;
 GEM current selection
 High tracking resolution (<100um) and good timing (~ 10ns); High rate; Insensitive to EM field
 UVa (Nilanga Liyanage’s group) can produce 120cm x 60cm plates;
 A hole can be produced;
 Can be ready before the experiment; Readout electronics are available;
 Scintillating Fiber Tracker (SFT) as a backup due to the lack of time, man-power and experience
• Good position resolution: e.g. 1mm fibers can give as good as ~300um;
• Thin, e.g. 1 mm plastic fiber gives only <0.3% radiation length;
• Replace Veto-Counter to perform precise time-measurement at the same time
• A hole can be easily produced.
• And more advantages!
Scintillating Fiber Tracker: Advantages
 Scintillating Material: emits visible lights via de-excitation when a charged particle deposits its energy through
ionization process;
 Scintillating Fiber (SciFi): A core of scintillating materials with one or several layers of thin cladding with lower
index of refraction;
 Good Time Response: SFT can provide better timing measurement than DC and GEM;
 Without Gas Systems: Unlike GEM and DC;
 Easy Handling: Easily installed, stored and transported; can be used in vacuum or high EM field;
 Easy Analysis: We just need to determine which SciFi is fired (“YES/NO” algorithm).
This new SFT can have a wide application in many projects!
Scintillating Fiber Tracker: Previous Developments
 Existing similar detectors (since 1990s):
 Mainly applied in Medical Imaging (small size):
e.g., Proton Computed Tomography Scanner (FERMILAB-PUB-12-067-E), INFN
 D0 in Fermi Lab: 0.84 mm SciFi + Visible Light Photon-Counter (VLPC)
Four concentric cylinders (Nucl. Phy. B 61B (1998) 384-389)
 KAOS in Mainz: 200cm wide 50cm long 0.25mm SciFi + Multi-Anode PMT
200cm x 50cm, only the vertical plane (C. Ayerbe Gayoso, PhD thesis)
 New detectors under developing:
 LHCb: 300cm long 0.25mm round SciFi+ Silicon Photon Multiplier
250cm x 300cm, 5 super layers, only the vertical plane
Scintillating Fiber Tracker: Our Design
 The new SFT proposed for PRad:
 120cm x 120cm active area
SciFi would be about 1.5m long
Detector Frame
 X&Y position tracking on electrons
Two perpendicular planes, each has two layers of SciFi
 Time measurement on electrons
replacing veto-counter to reject photons
 A hole at the center allowing the beam pipe to go through
Photon-Detectors on one side only
For 1mm SciFi (300um resolution), ~4800 fibers and ~2400 output-channels!
(If combining two-fibers and reading out signal from one-end)
 What we should know before we build:
 What type of SciFi? How many layers?
 How to assemble the SciFi?
 How to mount the SciFi on the supporting structure?
 What type of photon-detector?
Silicon Photon Multiplier (SiPM) or Multi-Anode PMT (MaPMT) ?
 What Read-Out system?
 How to reduce the cost?
Two fibers as
one readout
Prototype Test Project:
The Plan
Propose the project
Prepare Setup
Purchase Samples
Test SciFi
Here we are!
Test SiPM
Purchase &
Assemble SciFi
 The SFT Prototype:
5 cm x 5 cm active area
50 (X) and 50 (Y) read-out channels
200 1.5 meter long SciFi
100 SiPMs
Mounting Frame and Supporting Struecture
Design Mounting
Purchase / Make
(Detectors, PS, PreAmp)
Read-Out System (FastBus, fADC, others?)
Test Tracking Performance (with beam?)
Prototype Test Project:
The Hall-a Laser Lab shared with SoLID-EC test
Prototype Test Project: SciFi Test
 Selection of SciFi:
 Numbers about SciFi claimed by manufactures:
 ~8000(?) photons/MeV for each MIP within a 1mm fiber;
 ~3.1% Trap-Efficiency for Single-Clad (~5.4% for Multi-Clad);
 ~ 3 ns Decay Time;
 ~4 m Attenuation Length (for blue light);
 Position Resolution:
 D
, where D is the diameter of the fiber
Considering the quantum efficiency of photon-detector (<30%), 1-mm SciFi gives <50 p.e. on each end,
but it should be much lower in reality .
 We look for one type of SciFi that has:
Strong Light-Yield, Mechanically Strong, and High Detection Efficiency.
 Option 1 ---Square Fiber
 Option 2 ---Round Fiber
Charged Particle Direction
Good: Smaller Gaps (maximize the detection
efficiency), Easier Align&Assembling
Bad: Shorter Attenuation Length
Good: Longer Attenuation Length
Bad: Larger Gaps, Poor Trap-Efficiency (position
For our SFT with 150 cm fibers, square fiber may be better.
Prototype Test Project: SciFi Test
 SciFi Testing Setup:
The SciFi being testing:
 New Fiber-Samples from Kuraray:
1, x2 SCSF-78MJ , 1mm, Round, 3meters, Multi-Clad
2, x2 SCSF-78MSJ , 1mm, Round, 3meters,
mechanics stronger, Single-Clad (30% less light yield)
3, x2 SCSF-78J, 1mm, Square , 3meters
4, x2 SCSF-78J, 1.5mm, Square , 3meters
 From Hall-D: x8 SCSF-78MJ 1mm, Round, 2 meters
Measuring the Light-Yield and Attenuation
Length for different types of SciFi.
2um 1um
SciFi Polishing Tools
Prototype Test Project: SciFi Test
 SciFi Testing Setup:
1-inch PMT (Hall-C)
Scintillator (HallC)
Ru106 Radiation
We built a 200cm x 20cm Black-Box !
Thank you! Walter Kellner
@Hall-C Machine Shop!
Prototype Test Project: SciFi Test
 SciFi Testing Setup: Quick check
1mm 78MSJ-Round
1mm 78MJ-Round
 Checked the signals with Oscilloscope;
 Will take data with DAQ this week;
~5 p.e.
~8 p.e.
 Hall-D has done many tests with 78MJ
which gives ~ 8 p.e.;
 ~20 p.e. would be a good number to get
high detection efficiency (add two fibers);
 The fibers are needed to be polished with
1mm 78J-Square
1.5 mm 78J-Sqaure
better tools (borrowing a polishingmachine from Hall-D).
~7 p.e.
~10 p.e.
 Hall-D’s experiences and test results can
be adopted!
Prototype Test Project: SciFi Test
 Assembling & Mounting: Just a plan.
 Carl Zorn and Brian Kross, etc. in the Detector Group have given many
 Will learn from Carlos Ayerbe who built the SFT for [email protected]:
 Mark Emamian from Duke is helping the Mounting Frame design.
The plan is divided into groups
Rohacell Foam+Carbon Fiber Foil
Mounting Cookie on each end (Scheme Draw)
 Challenge for us– How to avoid the horizontal SciFis to bend down?
Optical Glue
Solution: Glue them on a plane with Rohacell foam+carbon fiber foils
Problem: Adding more dense materials (potential radiation background)
Prototype Test Project: SiPM Test
SiPM Avalanche Photodiode (APD)
pixels working in Geiger-mode
 Photon Detectors:
1, SiPMs: Silicon Photon Multiplier
 Cheap ~$10 per SiPM+~$10 power supply+~$10 Pre-Amp;
 Large Gain  ~~ x106 ;
 Insensitive to magnet field
 Need a Pre-Amp Design  Hall-D has a very good design
 Gain is temperature-depended
One photon only fire one pixel (unless
cross-talk or dark-current)
 Relatively larger dark current;
 Radiation damage by the neutron background;
 Cross-Talk
Hamamatsu MPPC S12572-100P/50P
Used in
Hall-B &
Hall-D for
Hamamatsu Multi-Pixels Photon Counter (MPPC)
We newly
Prototype Test Project: SiPM Test
 Photon Detectors:
2, MaPMTs (possible candidate)
 More commonly used;
 Multi-channels outputs
 Much cleaner background;
 High radiation tolerance;
 Degraded performance in strong magnet field;
 Cross Talk
 Expensive;
 Our Duke group has a 64ch H8500 MaPMT for test
 We will borrow a 16ch MaPMT from SBS
From Carlos
Ayerbe’s thesis
Fiber+SiPM Mounting Block
Prototype Test Project: SiPM Test
Thank you,
Walter Kellner!
 SiPM Test Setup:
High Precision Power Supply (Hall-D)
x2 Low Voltage Power Supplies (Hall-A&-D)
Black Box (from Simona Malace)
Temperature Sensor
SiPMs with Pre-Amp (Hall-D)
SiPMs with Pre-Amp ([email protected])
Understand the performance of the SiPM --- Gain, Noise Level,
Stability with Temperature, ADC & TDC spectra.
Prototype Test Project: SiPM Test
 SiPM Test Setup: (Stepan’s SiPMs+Pre-Amp)
1 p.e.
2 p.e.
3 p.e.
3 p.e.
2 p.e.
1 p.e.
(what we expect to see)
 Not yet seen pretty pattern from scope
 More to learn about SiPM
 Data taking with DAQ will be proceeded soon;
Prototype Test Project: Read-Out System
 Read-Out System of >2400 Output Channels:
1, SiPM (or MaPMT) + FastBus ADC + TDC
Requires a large amount of NIM modules and long delay cables
2, SiPM (or MaPMT) + fADC
Need >20 fADC & VME64 which are rare and expensive
3, A “Cheaper” Solution  EASiROC for SiPM or MaROC for MaPMT
 Developed by [email protected];
 Pre-Amp integrated with adjustable Low/High Gains;
 ADC outputs and TDC outputs;
 One “OR” logic output for triggering; One “SUM” analog output;
 ~$130 for each chip (or <$5 per channel);
 Need an additional readout board (“expensive”)
OMEGA Test Board (USB readout)
Prototype Test Project: Read-Out System
 Read-Out System of >2400 Output Channels:
EASIROC (or the new version called CITIROC)
32 ADC
32ch Inputs with
adjustable High/Low
32 TDC
Logic Output
SiTCP read-out board designed at KEK (TCP/Ethernet 1Gbps )
NIM-based Read-Out Board designed by I. Nakamura (KEK) for J-PAC
 A new MaROC3 with a read-out board
(USB port) has been purchased for
the SFT will be “portable”!
SoLID-EC test; We will study its
performance with SULI students’ help.
 SFT provides a great option to improve the PRad experiment and can be applied to many other projects.
 Prototype Testing Project is undergoing:
(1) It took a few months to prepare the setup due to very limited resources.
(2) Received and receiving many helps from colleagues in Hall-A/B/C/D, Detector Group, Duke Univ, etc.
(3) We have almost everything set up and will have some serious results very soon.
 Near Term goals (not working n full-time):
 Test and choose SciFi;
 Test SiPM and MaPMT
 Design and build the mounting structure
 Assembling the 1.2m x 1.2m SFT is challenging but practicable.
 Three options of the read-out systems are available.
 Highly appreciate your suggestions and helps, and welcome to join.
 I hope one day the full size SFT can be built!
I am grateful to receive many helps from:
Hall-A: Alexandre Camsonne, J-P Chen, Jack Segal, etc
Hall-B: Sergey Boyariov, Stepan Stepanyan, Youri Sharabian, etc.
Hall-C: Joe Beaufait, Mark Jones, Walter Kellner, Simona Malace, Brad Sawatzky, etc
Hall-D: Elton Smith, Yi Qiang, etc
Detector Group: Brian Kross, Wenze Xi, Carl Zorn, etc.
RadCon: Adam Hartberger
Many other colleagues and friends
Special Thanks are given to:
JSA User Board that give me the Postdoc Prize and offer me such a precious opportunity
Hard working Graduate Student: Chao Peng (Duke), Li Ye (Mississippi Statue)
Brad Sawatzky and Yi Qiang who lend me many instruments and help me to complete the setup
Prof. Haiyan Gao, Yi Qiang and Stepan Stepanyan who give me many advices to design and
carry out this project.
Prof. Donal Day, Prof. Haiyan Gao and Doug Higimbotham who provide the reference letters.
And the PRad collaboration & SoLID collaboration.
Cost Estimation of the Full-Size SFT:
• Each Fiber: 1mm width ( round or square ) is $1 per meter.
for 1.2m x 1.2m, we need roughly 2400 1.5m-long fibers for each plane to cover the gaps.
for x-y two planes, 4800 fibers ~ $7.2 K
• Photo-Detector: SiPM module $10 for each channel quoted from Hamamatsu.
Amplifier used in Hall-D: $10 for each channels ( plus Design Fee $???)
Power Supply (~$10 for each channel)
For one-end read-out: 2400 channels x $30 per channel ($72K + engineer design of the Pre-Amp)
• Mounting Frame and Supporting Structure ($???)
• Connectors + Cables + Tools + Supplies ($???)
• ReadOut+DAQ:
From SiPM to raw data: Discriminators, FastBus ADC & TDC (40 cards for each) (or fADC )
OR: EASIROC --- $100 for 32 channels
+ Read-Out Board ( we need to borrow designs and make all by ourselves ~$1500 per board or cheaper)
Total Read-Out: ~$120K

similar documents