miniTimeCube

University of Maryland

W.F. McDonough, K. Engel, S.A. Wipperfurth

University of Hawaii

R. Dorrill, A. Druetzler, M.J. Duvall, J. Koblanski, J.G. Learned, V.A. Li, L. Macchiarulo, S. Matsuno, S. Negrashov, M. Rosen, M. Sakai, G.S. Varner

Ultralytics

G. Jocher, K. Nishimura

NIST

H.P. Mumm

NGA

S.M. Usman

Applied Antineutrino Physics

Liverpool, UK

December 2, 2016

miniTimeCube (mTC) Concept

• Technology demonstrator:
• Scintillator-based IBD detection.
• Single compact detection volume (2.2 L).
• Portable.
• Fast timing (100 ps) for event reconstruction rather than optics.
• Fast microchannel plate PMTs (MCP-PMTs).
• Multi-GHz sampling electronics.

J. Learned - mTC @ AAP2016 Liverpool

2

12/2/2015

mTC Scintillator

• Eljen Technology, EJ-254
• 1% boron-doped plastic scintillator.
• 0.2% ¹⁰B
• Fast: 2.2 ns decay time
• Dimensions: 13 cm x 13 cm x 13 cm
• IBD neutrons: typically travel ~5 cm in ~10 us.
• Efficiency for neutron interaction in the volume is ~50%.
• Size is such that annihilation gammas deposit, on average, ~1/3 of their energy.

J. Learned - mTC @ AAP2016 Liverpool

3

12/2/2015

mTC Photodetectors: MCP-PMTs

• Photonis Planacon XPS85012
• 25 µm pores.
• 8x8 anode structure.
• Active area ~28 cm².
• Typical gain ~few x 10⁶.
• Transit time spread:
• ~120 ps FWHM.
• ~50 ps σ core.
• Total of 24 MCP-PMTs:
• Photocathode area: 674 cm² (66%).
• 1536 channels.

J. Learned - mTC @ AAP2016 Liverpool

4

12/2/2015

Timing distribution from

IEEE Trans Nucl Sci. 2010 Sep 13; 57(5): 2417–2423.

Partially populated mTC.

Expected IBD Prompt & Delayed Signals

• Simulated prompt and delayed PE distributions, including:
• Scintillator optical properties.
• Optical couplings and optical properties of MCP-PMT glass.
• MCP-PMT quantum efficiency.

J. Learned - mTC @ AAP2016 Liverpool

5

12/2/2015

• Prompt signal:
• Generally thousands of PE (many per channel).
• Long-tail of “missing PE”:
• Escaping positrons increasingly likely at high E.
• Delayed signal:
• Typically < 100 PE (<< 1 PE per channel).

Readout Electronics Requirements

J. Learned - mTC @ AAP2016 Liverpool

6

12/2/2015

• High channel density:
• 1536 channels in a small volume.

​

• Deep buffering to accommodate long potential delay between prompt and delayed IBD signals.
• Many tens of microseconds.

​

• Self-triggering capabilities.
• Delayed signal requires channel-level triggering on single PE signals.

​

• Must preserve fast timing

information from MCP-PMTs.

• Waveform information desired to correct for crosstalk, properly analyze multi-PE events, etc.

*Example MCP-PMT pulse (not Planacon)

Front-end Readout ASIC

J. Learned - mTC @ AAP2016 Liverpool

7

12/2/2015

• MCP-PMT signals digitized by waveform sampling electronics made at UH:

​

• “IRS” series ASICs - switched capacitor array with deep buffering.
• 8 channels per ASIC.
• Sampling rate up to 4 GSa/s.
• Nominal operation at 2.7 GSa/s.
• Buffer depth of 32,768 samples.
• 12 μs at nominal sampling rate.
• Channel-level triggers allow identification of regions-of-interest in analog storage memory.
• On-chip ADC (Wilkinson) digitizes 64-sample blocks in ~few μs.

Front-end Electronics Packages

• IRS ASIC serves 8 channels.

​

• 4 x ASICs per “carrier board” (32 channels).

​

• 4 x carrier boards per “boardstack” (128 channels).

​

• One control board w/ FPGA per boardstack.

​

• One boardstack serves 2 Planacon MCP-PMTs.

​

• 12 boardstacks required to instrument 24 Planacons.

​

• Total envelope per boardstack ~10 cm x 10 cm x 8 cm.

J. Learned - mTC @ AAP2016 Liverpool

8

12/2/2015

• Cabling and interface requirements:
• 2x CAT7, RJ45:
• Trigger and clock.
• Remote JTAG programming.
• 2x LV power cable bundles: 3,4,5 V.
• 1x fiberoptic interface to DAQ system.
• MMCX, SMA for electrical calib signal.
• Input/output cooling connections to chiller plates on boardstack housing.

System Electronics

• Central clock and trigger board:
• Distributed clock synchronizes sampling on all ASICs in the system.
• Receives lower-level trigger primitives from all 12 boardstacks.
• Issues system trigger to all boardstacks.
• Fast calibration pulser with programmable delay for timing calibrations, verifications.

​

• Other custom PCBs:
• High voltage distribution card.
• JTAG programming card to reconfigure front-ends.

​

J. Learned - mTC @ AAP2016 Liverpool

9

12/2/2015

Clock and trigger distribution board

Auxiliary Systems

• Fast laser diode system:
• Variable attenuation for single/multi PE studies.
• Fiberoptic steering to inject at any of 6 faces.
• Data acquisition server.
• Connections to front-end through commercial fiberoptic gigabit Ethernet PCI cards.
• Commercial LV and HV supplies.
• Cooling (required inside radiation shielding cave) provided by external chiller.
• UPS system.

J. Learned - mTC @ AAP2016 Liverpool

10

12/2/2015

mTC chiller

Laser coupling to mTC

Laser diode, attenuators, face selection

Ongoing Calibrations

• Calibrations required:
• ASIC-level timing calibrations.
• Waveform sampling technique results in non-uniform delays
• Channel-level gain maps.
• Laser injection at single PE levels.
• Cosmic ray muons:
• Calibrate light yields.
• Validate timing calibrations and MC optical models.

​

J. Learned - mTC @ AAP2016 Liverpool

11

12/2/2015

Measured muon event

Best fit reconstruction

PRELIMINARY

PRELIMINARY

Preliminary relative gain map

Reconstructed muon parameters

Deployment at NCNR

• mTC will be deployed at NIST Center for Neutron Research (NCNR), near Washington DC
• 20 MW research reactor.
• ~5 meters from split-core.

🡺 ~1 detected neutrino per day.

• Shielding cave required to limit neutron, gamma backgrounds.

J. Learned - mTC @ AAP2016 Liverpool

12

12/2/2015

mTC in various venues

Background Simulations

• Geant4 simulations used to model muon, neutron, and gamma backgrounds.
• With and without shielding enclosure:

J. Learned - mTC @ AAP2016 Liverpool

13

12/2/2015

• Shielded rates are used to estimate accidental coincidences that can fake our IBD signal:
• Accidental rate estimated from simulation: ~1 / day
• This corresponds to an SNR of ~1:1.
• Correlated backgrounds need further study.

Unshielded HPGe gamma background spectra

Reconstruction & Efficiency

• Preliminary reconstruction criteria developed to study performance on simulated IBD events:
• Both prompt and delayed vertices must be > 5 mm from any wall.
• Reduces effective volume by 20%, from 2.2 L to 1.7 L.
• Time between prompt, delayed signals:
• Minimum: 50 ns.
• Maximum: 12 μs.
• Energy:
• Prompt: 1 – 8 MeV.
• Delayed: 40 – 400 keV.
• Detected PE:
• Prompt: 20 – 10,000
• Delayed: 20 – 400

​

🡺 Overall efficiency ~30%.

J. Learned - mTC @ AAP2016 Liverpool

14

12/2/2015

*Results shown in this talk are for a maximum likelihood fit to two point sources to the prompt, delayed vertices.

Reconstruction Efficiency vs. ν Energy

Simulated Energy Resolution (1)

• Expected energy resolution, weighted by reactor spectrum:

​

• ~11% including tail.

​

• Tail largely correlated with higher energy events where positron is not fully contained.

J. Learned - mTC @ AAP2016 Liverpool

15

12/2/2015

Simulated Energy Resolution (2)

• Poorer energy resolution observed with reconstructed vertices near wall.
• Best energy performance is ~11% at 3-4 MeV.

J. Learned - mTC @ AAP2016 Liverpool

16

12/2/2015

Simulated Vertex Resolution & Pointing

• Distributions of error on reconstructed prompt, delayed vertices.
• Prompt vertex resolution can suffer due to low light at low energies, incorrectly modeled positron track length at higher energies.

J. Learned - mTC @ AAP2016 Liverpool

17

12/2/2015

• Reconstructed vertices are used to point back to incoming neutrino and study angular resolution.
• σ not well defined, so study distributions over cos(θ).
• We also define vector SNR:
• Mean displacement between vertices / uncertainty in one of the dimensions.

Directionality Performance Comparisons

• Comparison of directionality performance with other simulated detectors:

​

• Double CHOOZ and (hypothetical) TREND directionality from:

arXiv:1307.2832.

​

• mTC shows good potential for directional reconstruction.

​

• Initial study with Li-doped mTC shows promise for further investigation.

J. Learned - mTC @ AAP2016 Liverpool

18

12/2/2015

mTC Cave Installation (April 2016)

mTC Workstation

Completely Remote Operation Capable

​

Laser, Muon and Source Calibrations

• Detector-wide gain maps
• Laser injected at single PE levels
• Muon track test fits
• Adjustment and testing of trigger parameters
• Data runs with γ sources, shield door open and closed

PRELIMINARY

A scan with the “simple trigger”

Muon Track Reconstructions in Matlab

• A simulated muon track for comparison

• Example of a real, reconstructed muon track through the mTC
• Muons can deposit a lot of Energy ~30 MeV

mTC Schedule and Outlook

• mTC electronics was upgraded.
• Upgraded amplifier chains and ASICs address SNR, trigger efficiency, timing precision and stability.
• Construction of shielding cave complete April 2016.
• Installed mTC inside the shielding cave and began operations May 2016.
• Published Rev. Sci. Instrum. paper on detector [RSI 87, 021301 (2016)]
• Data pub next early 2017.

J. Learned - mTC @ AAP2016 Liverpool

23

12/2/2015

Upgraded electronics timing distributions

mTC in (mostly) assembled cave

σ ~ 33 ps

Summary of mTC

Completed NIST Run in October 2016 and returned equipment to UH

Currently doing tests and PMT studies in new lab at UH

Finish analysis and publish data coming soon.

Soon (month) to make neutron (Ca) source runs.

- 2017 Repurpose mTC Electronics for LAPPD tests, Neutron Camera and NuLat