The DarkLight experiment - Thomas Jefferson National ...hallaweb.jlab.org/experiment/APEX/collab15/dlforapex.pdfThe DarkLight experiment Jan C. Bernauer APEX collaboration meeting,

The DarkLight experiment

Jan C. Bernauer

APEX collaboration meeting, April 2015

1

CollaborationR. Alarcon, D. Blyth, R. Dipert, L. Ice, G. Randall

Arizona State University, Phoenix, AZ

B. Dongwi, N. Kalantarians, M. Kohl, A. Liyanage, J. Nazeer

Hampton University, Hampton, VA

S. Benson, J. Boyce, D. Douglas, P. Evtushenko, C. Hernandez-Garcia, C. Keith,

C. Tennant, S. Zhang

Thomas Jefferson National Accelerator Facility, Newport News, VA

J. Balewski, J. Bernauer, J. Bessuille, R. Corliss, R. Cowan, C. Epstein, P. Fisher*, D. Hasell,

E. Ihloff, Y. Kahn, J. Kelsey, R. Milner*, S. Steadman, J. Thaler, C. Tschalär, C. Vidal

MIT, Cambridge, MA

M. Garçon

CEA Saclay , Gif-sur-Yvette, France

R. Cervantes, K. Dehmelt, A. Deshpande, N. Feege

Stony Brook University, Stony Brook, NY

B. Surrow

Temple University, Philadelphia, PA

To come: Johannes Gutenberg University, Mainz, Germany

2

Current and future exclusion limitscurrent future

invi

sible

visib

le

3

Design goals

Measure e−+p −→ e−+p +A′ −→ e−+p + (e+ +e−)100 MeV beam =⇒ below pion threshold, simple finalstateComplete reconstruction of final state: e+e−pair fromdecay, scattered electron, recoil proton=⇒ large acceptance=⇒ proton only has 1-5 MeV =⇒ gas targetResolution of reconstructed A′ of 1-3 MeVFinal state leptons 10-100 MeV =⇒multiple scatteringdominant =⇒minimize material0.5 Tesla solenoidal magnet for momentummeasurement and to bottle up Møller electronsDefinitive measurement in about 1 month: highluminosity: 10 mA @ 1019 Atoms/cm2

=⇒ JLAB LERF world’s only such accelerator

4

JLab LERF

Low Energy Recirculating FacilityEnergy recovering linacup to 10 mA @ 100 MeVMegawatt beam

5

Successful beam test July 2012

Target systemdesigned andconstructed atMIT-Bates R&E

center

4.3 mA, 100 MeV (430 kWatt beam power) transmittedthrough 2 mm hole, 127 mm longMaximum loss of 3 ppm in 7 hoursERL has required stability

Phys. Rev. Lett. 111, 165801 (2013)Nucl. Instr. Meth A729, 233 (2013)Nucl. Instr. Meth. A729, 69 (2013)

6

Phased approach

Phase 1Accelerator studiesMeasure SM physics / detector testPilot DM search

Phase 2Full DM searchInvisible searchFull streaming readout

7

J. Balewski et al., arXiv:1412.4717 [physics.ins-det]

8

Detector setup

Procured existing magnet0.5 Tesla max fieldInner diameter 712mmNow at MIT-Bates

Hydrogen gas targetThickness: 1019 Atoms/cm2

2mm apertures to limit flowMultiple stages of pumping1mm thin berrylium pipeInvestigating jet target

12 ladders, 6 sensors each300 µm thicknessExperience from STAR IST

Thin, high rate capableCylindrical detectorspossible

9

Solenoid magnet

TargetProton silicondetectorLepton tracker4 layers MicroMegasDesign is still in flux!

Detector setup






10

Solenoid magnet

TargetProton silicondetectorLepton tracker4 layers MicroMegasDesign is still in flux!

Detector setup






11

Solenoid magnetTarget

Proton silicondetectorLepton tracker4 layers MicroMegasDesign is still in flux!

Detector setup






12

Solenoid magnetTarget

Proton silicondetectorLepton tracker4 layers MicroMegasDesign is still in flux!

Detector setup






13

Solenoid magnetTargetProton silicondetector

Lepton tracker4 layers MicroMegasDesign is still in flux!

Detector setup






14

Solenoid magnetTargetProton silicondetector

Lepton tracker4 layers MicroMegasDesign is still in flux!

Detector setup






15

Solenoid magnetTargetProton silicondetectorLepton tracker4 layers MicroMegas

Design is still in flux!

Detector setup






16

Solenoid magnetTargetProton silicondetectorLepton tracker4 layers MicroMegas

Design is still in flux!

Detector setup






17

Solenoid magnetTargetProton silicondetectorLepton tracker4 layers MicroMegasDesign is still in flux!

Kinematics

, Can detect full final state, Large acceptance/ How to trigger?

Signal tracks do not reach beyond lepton trackerCan not identify reaction on trigger levelCan not trigger on ”3 particles”: background tohigh

18

Background rates: Elastic scattering

90 patchesindependent2D strip readout>20 tracks every 25 ns frame

19

Normal electronics

Trigger

APV

64..256 Channels

ADC

APV/DREAM/... multiplex N channels to 1 ADCTheoretical maximum readout rate: 1/N of ADC clock

20

Streaming front end electronics: NSA Scale

ADCADCADCADCADCADCADCADC

FPGA

Continuous readout80k ch, 40 MSps @ 12 bit =⇒ 4.8 Terabyte/s∼ 80 M ch, 40 KSps=⇒Listening to every German citizen in CD quality.Zero suppression: 250 Gigabyte/s

21

Streaming back end electronics

Transport data from FEE to CPU farmSolve transposition problem (”Event building”)

Data aggregated per channelMust be processed by time slice

Common problem at intensity frontier

Solve once and reuseOpen design

wire protocolhardware

Use standard hardware

cheapereasier to extend

22

Streaming back end electronics

Transport data from FEE to CPU farmSolve transposition problem (”Event building”)

Data aggregated per channelMust be processed by time slice

Common problem at intensity frontier

Solve once and reuseOpen design

wire protocolhardware

Use standard hardware

cheapereasier to extend

23

Phase 1

Beryllium −→ Aluminum pipeCylindrical MicroMegas −→ planar GEMsOnly one proton detector ladderOnly partial streaming readout

24

Phase 1 current design

proton, pT = 8 MeV

e−, pT = 20 MeV

e+, pT = 30 MeV

e−, pT = 10 MeV

Tripple coincidence to select QED background / A′

candidate events at reasonable ratesInvestigate if we can run double coincidenceRun APV in quasi-streaming mode - free runningtrigger, mutiple frames

25

Summary

Phase 1:

Proof of target conceptSM processes: Møller, elastic scatteringDM search mock upPrototyping of streaming readoutWill run 2016

Phase 2:

Full acceptance with full streaming readoutVisible and invisible search

Low energy, intense beam physics

Intense Electron Beams WorkshopCornell University, June 17-19, 2015

http://www.classe.cornell.edu/NewsAndEvents/IEBWorkshop

26

Backup slides

27

Current status: Hardware

Xilinx Virtex-7 development board: VC 707XC7VX485T-2FFG1761Gigabit Ethernet + SFP/SFP+FMC connectors

TI ADS5295 evaluation module + adapter board8 channels80 MSPS / 12 bit

Signed up for Xilinxuniversity program

XUP donatedhardware &software

28


Xilinx Virtex-7 development board: VC 707XC7VX485T-2FFG1761Gigabit Ethernet + SFP/SFP+FMC connectors

TI ADS5295 evaluation module + adapter board8 channels80 MSPS / 12 bit



29


Xilinx Virtex-7 development board: VC 707

XC7VX485T-2FFG1761Gigabit Ethernet + SFP/SFP+FMC connectors

TI ADS5295 evaluation module + adapter board

8 channels80 MSPS / 12 bit



30

Current status: FPGA firmware

Milestones:SetupEthernet send/receive

OSI layer 2

Readout of ADC:

Full data of 1 ch.8 ch. with zerosuppression

Planned:

Partial streamingreadout for DL phase 1Full streaming readoutfor DL phase 2

31



OSI layer 2

Readout of ADC:


Planned:


32



OSI layer 2

Readout of ADC:


Planned:


33

Documents

The DarkLight experiment - Thomas Jefferson National ...hallaweb.jlab.org/experiment/APEX/collab15/dlforapex.pdfThe DarkLight experiment Jan C. Bernauer APEX collaboration meeting,