Readout electronics development for CDC in COMET Phase-I Nakazawa - PowerPoint PPT Presentation

Readout electronics development for CDC in COMET Phase-I Nakazawa Yu Kuno-g 2015.12.25 Year-end joint session 2015 1

Outline • 2 Strategy • Result • Radiation setup Motivation • Readout board (RECBE) • Cylindrical Drift Chamber (CDC) • COMET Phase-I • Year-end joint session 2015 • Introduction • Neutron irradiation test @Kobe Univ. • Issue • Summary

Introduction 3 Year-end joint session 2015

COMET Phase-I Purpose : Search for the charged Lepton Flavor Violation -54 ) @SM extended neutrino oscillation -15 in the COMET Phase-I Detector : Cylindrical Drift Chamber (CDC) & Straw Tube Tracker Signal : ~105 MeV/c mono-energetic electron for an aluminum target Background : Decay-In-Orbit electron & Beam related BG 4 Year-end joint session 2015 • The neutrinoless, coherent transition of muon to electron @J-PARC in 2018 • B.R.(µ -> eγ) ~ O(10 • S.E.S. ~ 3 × 10

CDC Signal 5 Cylindrical Drift Chamber -> Background electron orbit. Muonic Atom Momentum Spectrum expected in the Phase-I DIO Year-end joint session 2015 at IHEP in China • Unstable atom capturing a muon in • Decay In Orbit (DIO) : µ -> eν µ ν e A. Wire • Sense : ø25 um W plated Au • Field : ø126 um Al B. Gas : Helium base (not yet decided) C. Momentum resolution : 200 keV/c D. Readout System : 104 RECBE boards • Mass-production was completed

Readout Board 6 Virtex-5 This talk Evaluation items 8ch/chip Development items Year-end joint session 2015 XC5VLX155T System Clock：128MHz → 120MHz 1. Rewrite the firmware 2. 8ch/chip Aurora：remove Remodel the RECBE developed for Belle-II 3. 1. Noise level 2. Gain linearity (Analog out of ASD) 3. Threshold scan (Digital out of ASD) 4. Time resolution 5. Cross talk 6. TRG rate tolerance 7. Temperature vs Power 1. Data reduction • Throw away the BG data 2. Irradiation tolerance • gamma ray • neutron

Neutron irradiation test @Kobe Univ. 7 Year-end joint session 2015

Year-end joint session 2015 As a result of nuclear reaction with a neutron • Virtex-5 is exposed to a high neutron flux of >10 12 neutrons/cm 2 . (ref. Ye-san’s simulation：28aSN-2) - RECBEs are set at the CDC endplate at the downstream side. • It is known that a firmware is broken by neutrons. So we need re-download the one at intervals. - Single Event Upset：SEU We want to check whether this problem of SEU is serious or not. RECBE • and semiconductor atom, a charged particle or an ion is generated. • These charged particle create electron-hole pairs with drifting. • These charge invert logic in the FPGA What is SEU? Problem in the COMET Phase-I Energy spectrum of neutrons MeV beam 8 Motivation EvtP4[3] htemp htemp Entries Entries 24524 24524 15.12 15.12 Mean Mean 4 10 RMS RMS 23.22 23.22 3 10 2 10 10 1 0 50 100 150 200 250 300 350 400 EvtP4[3]

評価項目 ref : http://iopscience.iop.org/article/10.1088/1748-0221/7/02/C02022/pdf (Univ. of Tokyo) ref. Nakanishi Hitoshi Year-end joint session 2015 9 中性子ビームライン TANDEM Evaluation items En [MeV] ~2 MeV MeV (Tetsuo INADA, 1968) Beryllium Reactions at 1.0 MeV to 3.0 ref：Neutrons from Thick Target at TANDEM Neutron Energy Spectrum Definition of angle 1. Frequency of SEU occurrence TANDEM @Kobe Univ. ( Beam current : 1 microC/s ) ( Distance from target : 10 cm ) 2 6 neutrons/cm Intensity : (1.90 ± 0.13) × 10 Target : Be Beam : ~3 MeV Deuteron Radiation setup 2. A neutron incident angle effect on SEU rate. (θ=0º and 180º) of “Verify” in the iMPACT. JPS 2015 autumn • Download the firmware with the iMPACT（Xilinx Inc.）and a JTAG cable. • Compare a bit stream on the FPGA to a original firmware with using the function • Due to the comparison, count the number of SEUs.

Setup Neutron RECBE (Radially from target) Be Target Deuteron 10 Year-end joint session 2015 ( ✳ ) In this test, we assumed the uniform beam at Virtex-5

11 285 mm Run1 Start 7/28 8:20 Stop 8/2 10:30 Total time 211621 sec Distance Angle Neutrons/SEU 180 º Neutrons 9.25×10 11 SEUs 4934 (1.82 ± 0.03) × 10 8 Rate -1 (neutrons/SEU/Virtex-5) standard deviation. Result : SEU rate at 180º Neutrons/SEU Date and Time Rate 6 × 10 Neutrons / SEU 350 300 250 200 150 100 07.28 09:00 07.28 21:00 07.29 09:00 07.29 21:00 07.30 09:00 07.30 21:00 07.31 09:00 07.31 21:00 08.01 09:00 08.01 21:00 08.02 09:00 08.02 21:00 Date and Time ( ✳ ) Divide a long run into every ~2000sec during measurement. ( ✳ ) Error is statistical error given by the ( ✳ ) Rate is estimated by fitting this graph.

Year-end joint session 2015 7.65×10 9 8/5 9:34 8/6 9:13 Direction 275 mm 285 mm 285 mm Angle 0 º 0 º 0 º Neutrons 2.22×10 10 5.25×10 10 SEUs 8/6 1:23 59 129 23 (3.16 ± 0.23) × 10 8 Rate -1 (neutrons/SEU/Virtex-5) standard deviation. Neutrons/SEU Date and Time Neutrons/SEU Result : SEU rate at 0º 17th COMET Collaboration Meeting (LPNHE Paris) 12 8/4 8:31 Stop 8/5 1:46 Start Run5 Run4 8/3 23:38 Run3 Rate 6 × 10 Neutrons / SEU 1600 1400 1200 1000 800 600 400 200 0 08.03 21:00 08.04 03:00 08.04 09:00 08.04 15:00 08.04 21:00 08.05 03:00 08.05 09:00 08.05 15:00 08.05 21:00 08.06 03:00 Date and Time ( ✳ ) Divide three runs into every ~2000sec during measurement. Total Time 29933 sec 27020 sec 11438 sec ( ✳ ) Error is statistical error given by the ( ✳ ) Rate is estimated by fitting this graph.

Issue 13 Year-end joint session 2015

Issue ref. Abe Shinichiro (Kyushu Univ.) 61st JSAP meeting Estimated SEU rate -1 (sec/SEU/104boards) by the SEU cross section and the neutron energy spectrum in the COMET Phase-I. 0.84 sec/SEU/104boards (θ=180º) A serious problem that reliability of data from RECBEs is lost every 0.84 sec 14 Year-end joint session 2015 from Y. Yang (Kyushu Univ.) Neutron Flux at RECBEs

Strategy • Year-end joint session 2015 15 Not fixed yet (It is expected that MBU rate is about 1/500 of SEU rate.) Measurement of MBU rate. Check the performance of this function. • The next neutron irradiation test is planed Simulation of this function and performance test of new firmware were done. Can’t correct multi SEUs occurred at the same time. (Multi-Bit Upsets : MBU) Detect SEU and correct it. ref. other experiments A. Add the self-repairing function in the firmware. • TANDEM in Kobe Univ. on January 12th - 14th. B. Add extra shielding against neutrons for the FPGA.

Summary 16 Year-end joint session 2015

Summary θ=0º. 17 January 12th - 14th -1 : 0.84 [sec/SEU/104boards] (θ=180º) Year-end joint session 2015 @J-PARC in 2018 -15 A. COMET Phase-I • Search for the neutrinoless, coherent transition of muon to electron • S.E.S ~ 3 × 10 B. Neutron irradiation test @TANDEM (Kobe Univ.) • The number of SEUs with θ=180º is about 1.7 times larger than • SEU rate • Added the self-repairing function in the firmware. C. To-do • The second neutron irradiation test at TANDEM (Kobe Univ.) on 1. Performance test of the self-repairing function. 2. Measurement of MBU rate

18 Year-end joint session 2015

Backup 19 Year-end joint session 2015

Readout Board • • Input ：Clock、TRG • Output：Busy 8ch/chip 8ch/chip Remodel the RECBE developed for Belle-II Aurora：remove • • Rewrite the firmware • System Clock：128MHz → 120MHz Mass-production of RECBEs was completed at IHEP in China 20 FPGA RJ-45 LVDS Optical Fiber • Virtex-5 (XC5VLX155T) • For COMET CDC, rewrite the firmware for Belle SFP slot For downloading the firmware • • SiTCP • Send hit information. (ADC, TDC, etc.) JTAG • RJ-45 • Year-end joint session 2015 • 104 boards：for CDC • 24 boards：for spare • Performance test @IHEP

F.G. ~?m 21 Setup Delay : 500 ns TEST IN CLK TRG Beam safe area Beam area LEMO LEMO LEMO USB LAN LAN Optical Fiber LV Power Supply ~5m LV Cable ~5m JTAG HUB PC Convertor Optical Media (VAIO) PC Downloader G.G. Clock Generator RECBE Year-end joint session 2015

Neutron Flux @TANDEM 22 Year-end joint session 2015 ref. Hitoshi Nakanishi (Univ. of Tokyo), JPS 2015 autumn

Neutron Flux @TANDEM 23 Year-end joint session 2015 ref. Hitoshi Nakanishi (Univ. of Tokyo), JPS 2015 autumn

Neutron Flux @TANDEM 24 Year-end joint session 2015 ref. Hitoshi Nakanishi (Univ. of Tokyo), JPS 2015 autumn

Neutron Flux @TANDEM 25 Year-end joint session 2015 ref. Hitoshi Nakanishi (Univ. of Tokyo), JPS 2015 autumn

Issue Angular dependence 26 This topic is presently under investigation What is thought to be the factor SEU rate dependents on a incident angle of neutrons. Year-end joint session 2015 Rate Inverse 6 10 × Rate Inverse [n/SEU/Virtex-5] 340 320 300 280 260 Rate 180 � = 3 . 16 × 10 8 Rate 0 � 240 1 . 82 × 10 8 ∼ 1 . 74 ± 0 . 13 220 200 180 0 20 40 60 80 100 120 140 160 180 Angle • Structure of the RECBE • Component of the FPGA and the PCB • etc. . .

Geometry of Detector Section 27 Year-end joint session 2015 from Y. Yang (Kyushu Univ.)