R&D status of a large HAPD T. Abe, H. Aihara, M. Iwasaki, H. Nakayama, T. Uchida (Univ. of Tokyo) M. Tanaka, (KEK) Y. Kawai, H. Kyushima, M. Suyama (HPK) M. Shiozawa (ICRR) 12/3/2018

Introduction We believe HAPD is the photon detector for the next generation water-Cherenkov detector (i.e. Hyper Kamiokande, 2km detector for T2K). We have developed large (13-inch) HAPD. In this talk, we present the R&D status of the HAPD. 12/3/2018

Hyper Kamiokande 20 times larger than Super K Low cost Quality control 48m×50m×250m×2 Total mass ~ 1Mton Low cost Quality control Single photon sensitivity Good time resolution 12/3/2018

HAPD HAPD uses avalanche diode as e multiplier. # of parts < 1/10 of PMT 12/3/2018

Theme 190ps Excellent HAPD performance than PMT’s σ 190ps Excellent HAPD performance than PMT’s Single photon sensitivity Single photon time resolution ~ 190ps (s) Readout system development dedicated for HAPD Possible application to other field 12/3/2018

Principle of HAPD Operation PMT Ne/sNe=67 Total Gain:~105 Bombardment: x4500@20kV Avalanche Gain: X30 1st Dynode Gain: x5 Ne/sNe=2.2 Incident Light Incident Light Photoelectron Photoelectron Large gain @ 1st mult. stage  good single photon detection 10~20kV Dynodes Avalanche Diode No Dynodes (No TTS)  large HAPD with good time resolution Total Gain:~107 Total Gain:~105 12/3/2018

Property of HAPD Single photon sensitivity Good time resolution large gain at the first stage of multiplication Good time resolution No TTS in dynodes Cost reduction and easy Quality control simple structure without dynodes Lower gain Need a dedicated readout system 12/3/2018

HAPD system QP.H. TTime 13-inch HAPD Pre-amplifier ASIC (Low noise) (Q, T) QP.H. TTime ≥1Gsps MOF 13-inch HAPD Pre-amplifier ASIC (Low noise) Waveform sampling (Fast sampling) +Digital filter (Noise Suppression) HAPD performances are evaluated using this system 12/3/2018

13inch HPD performances Energy measurement related items Comparison between HAPD and PMT Dynamic range Position dependence of relative gain Timing measurement related items Comparison & # of P.E. dependence HV dependence Dark rate 12/3/2018

HPD and PMT HPD（HV=12ｋV bias=330V) PMT(13inch) 1P.E. 2P.E. 3P.E. 4P.E. 12/3/2018

Dynamic range (no preamplifier) 3000p.e. equivalent input 110 HV=+20kV, Bias=290V for Sample A 370V for Sample B 390V for Sample C Light Source: Pulsed Laser (PW: ~70ps, λ: ~400nm) 100 90 Output/input [%] 80 70 60 :Sample A :Sample B :Sample C -3% Deviation at Output Charge of 100pC (~3000p.e. input signal intensity equivalent at total gain of ~2x105) 50 40 1 10 100 1000 12/3/2018 Output Charge [pC]

Position dependence of gain for photons<5 18cm ~65deg. Gain uniformity:s~2% 12/3/2018

Time resolution (large HAPD) Transit time variation Timing resolution Center 45 deg 70 deg 12/3/2018

Timing resolution 190ps Spot Laser illuminated HV=12ｋV bias=330V Better st No dependence of # of p.e. 190ps 12/3/2018

Photoelectron transit time Center Bias=290V Light Source: Pulsed Laser (PW: ~70ps, λ: ~400nm) 45 deg 70 deg Position dep. of Transit time < 1 nsec 12/3/2018

Dark rate Dark rate quickly goes down below ~17 kV. Dark count : 15kHz Sample#1 20inch PMT dark rate @ room temp Sample#2 After modification of HPD, Dark rate HPD@20kV~20inch PMT Dark count : 15kHz 12/3/2018 HAPD

HAPD vs. PMT Parameters* Developed HPD (13-inch HPD) 13-inch PMT (R3600-02 for SK) Order of Gain 105 107 Single Photon Time Resolution (s) 190ps 1400ps 2300ps Single Photon Energy Resolution 44% (preliminary) 70% 150% Pulse Response Rise Time 1ns 6ns 10ns Pulse Width 2.2ns 20ns Transient Time 12ns 100ns 95ns Dynamic Range (Signal Intensity in p.e.) 3000 p.e. 2000 p.e. 1000 p.e. 12/3/2018

HPD electronics status f 1Gsps AMC FPGA Analog memory cell (AMC) + FPGA for Digital Filter Pre-amplifier ASIC Under development t f (Q, T) ≥1Gsps MOF 12/3/2018

ASIC preamplifier for low noise 12/3/2018

ENC (noise) and impulse response Rise time=5.8ns Dynamic range:2V Power consumption:~4mW/ch (depends on driving capability) S/N=59 for the HAPD gain of 2X105 12/3/2018

Achievable ENC S/N=2X105/2.2X103 S/N~100 is attainable!? Quadruple input transistor amplifier Detector Capacitance=40pF S/N=2X105/2.2X103 S/N~100 is attainable!? 12/3/2018

AMC + slow FADC vs. fast FADC Ref. ; A. Kusaka Master Thesis 2004 U. of Tokyo 12/3/2018

AMC operation principle 12/3/2018

AMC pulse response test 5ns/Div. Test Signal Input ~10ns 100us/Div. Output (with 100kHz Readout Clock) ~12counts under 1ns sampling period Sample and Hold Trigger Readout Clock: Much Slower Than Sampling Speed 12/3/2018 Prototype AMC

AMC performance (measured) Sampling speed=1GHz Uniformity of sampling speed < 100psec Dynamic range : 3.5V 9bit equivalent resolution 50mW/ch Depth 64 AMC 12/3/2018

Depth 512 AMC We develop a depth 512 AMC which has enough depth to measure entire signal shape of HAPD. 512nsec 12/3/2018

FPGA for DSP So far, we evaluate this part using Offline computer analysis or Mathlab/Simulink + Signal Master (general purpose DSP prototyping hardware) We start to design a FPGA board dedicated for Digital Signal Processing with Giga-bit Ethernet output. 12/3/2018

Possible application to other fields We are seeking possible HAPD application to other fields. Excellent single photon sensitivity of HAPD might be useful for environment science. We try to measure photons from photo protein (single photon events). 12/3/2018

Filter output of photo protein events The origin of photon source is same as fireflies. 40msec 12/3/2018

Photon measurement from photo protein Large HAPD can measure single photon events by photo protein source. noise signals 12/3/2018

Summary We develop a 13inch HAPD and confirm Excellent single photon st~190ps. Excellent single photon sE~44%. Total gain of > 2X105. We develop a dedicated readout electronics for the HAPD. HAPD would be useful for other fields. 12/3/2018

Backup slides 12/3/2018

HPD system in the next step In the next step (after several modifications) Stability for temperature (for general purpose) Evaluation of long term stability Optimization of manufacturing process for production cost Photocathode formation Low cost material Quality check during production …. Customization of electronics/system for HK Multi-SK Large-SK HK 12/3/2018

Two directions Development of common devices Similar to SK On HPD 1Gsps AMC MOF t f 1Gsps AMC MOF Digital data through Ethernet Ethernet Development of common devices After determination of timing I/F(i.e. trigger, clock ..) Multi-chip module or deep submicron ASIC (blue rectangle part) will be developed as a building block of the readout system. 12/3/2018

EB and Avalanche Gain AD Bias=30V(fixed), HV=Swept Gain=~4700 @20kV HV=+10kV(fixed), Bias=Swept Gain=~4700 @20kV Gain=~50 @390V 12/3/2018

Incorporated Avalanche Diode and its C-V Characteristics Capacitance: ~40pF over bias voltage of 120V (assembled on the base) Assembled Bare Chip Backside Illumination Avalanche Diode Effective Area: 5mm in dia. 12/3/2018

Collection Efficiency as a function of Magnetic Field (at HV of +20kV) Photoelectron Collection Efficiency and Effect of Magnetic Field (Simulation) Photoelectron Collection Efficiency as a function of HV (No Magnetic Field) Collection Efficiency as a function of Magnetic Field (at HV of +20kV) 12/3/2018

Mathlab/Simulink + SignalMaster 12/3/2018