So LID DAQ A Camsonne So LID collaboration

So. LID DAQ A. Camsonne So. LID collaboration meeting November 8 th 2013

Outline • • • Experiment requirements PVDIS Calorimeter trigger and readout PVDIS setup, event size and data rates GEM electronics SRS SIDIS event size SIDIS electronics layout MRPC electronics Budget Test stand To do list Conclusion

Experiment requirements • PVDIS – Pion contamination – Trigger rate 30 KHz max per sector – 30 independent sectors • SIDIS – Trigger rate • 60 KHz coincidence e pi • 100 KHz singles

Calorimeter trigger

PVDIS calorimeter readout • Need new development of SD and FADC firmware • 100 K$ for R&D and prototyping

PVDIS calorimeter trigger • Send data from one sector to the neighboring sector

PVDIS setup • Shower : 1830 channels • Preshower : 1830 channels • Light Gas Cerenkov : 270 channels • GEM : 4700 strips • Trigger : Shower and Cerenkov • Rate with 30 ns gate : 3 e 6 * 200 e 3 * 30 e-9= 18 KHz • Rates estimate for 20 KHz

Event size PVDIS • Assume 2 high energy cluster in each sector and all Cerenkov firing for each trigger Detector Channels fired Event size KB EC 28 1. 12 Cerenkov 9 0. 36 GEM 128 1. 3 Total • Data rates at 20 KHz = 56 MB/s 2. 78

SRS

SRS • Switch SRS – Cheaper channel cost – Less crates – Higher data transfer rate

SIDIS • Detectors – Forward • • Calorimeter preshower + shower Scintillator Light Gas Cerenkov Heavy Gas Cerenkov – Large Angle • Calorimeter shower • Scintillator – GEM • 5 planes

SIDIS, J/Psi requirements • SIDIS : 60 KHz coincidence e/pi, 80 KHz singles • J/psi : triple coincidence e , e+, e • Plan – FADC integrated readout – GEM with deconvolution one sample readout • In theory should be able to reach 100 KHz

Event size SIDIS • GEM occupancy with deconvolution : 1. 2 KBytes • Other detectors : ~ 1. 9 KBytes • Data rate 100 KHz with 3. 1 Kbytes event size 310 MB/s Could take more but need reduce with L 3

SIDIS, J/psi Calorimeter electron trigger as PVDIS Send cluster number to GTP Coincidence APV 25 : on chip convolution one sample readout FADC readout : zero suppress 10 % MIP for calorimeter readout time and integral • Might consider waveform for SPD and preshower depends on final trigger rate • • •

SIDIS trigger • Electron trigger – Shower + LC + SPD + MRPC : Forward angle – Shower + SPD : Large angle • Pion trigger – Shower + HC + SPD – Shower • J/psi trigger – Shower ( 3 clusters above threshold ) + LC + SPD + MRPC

SIDIS electronics layout

Electronics layout SIDIS

FADC occupancy

MRPC • Question about electronics and funding • Solution in house / commercial electronics – 2 input registers : 2 K$ – 4 V 1290 CAEN High Resolution TDC : 11 K$ – Price per sector : 44 K$ – Total price : 1. 4 M$ ( only readout no FE ) – Funding request from Tsinghua University – Either use this solution or develop custom module

DAQ cost estimate Module Unit price K$ Number Total (K$) FADC 250 4. 5 282 1, 269. 5 TD 3. 0 16 48 CTP 7. 0 30 210 SSP 5. 0 4 20 GTP 5. 0 1 5 TS 3. 5 1 3. 5 TID 3. 0 61 183 SD 2. 5 31 77. 5 VXS crate 15. 0 32 480 SRS computer 3. 0 30 90 VME CPU 3. 4 32 108. 8 Total 2, 844. 8

Additionnal cost • 100 K$ R&D prototyping SD • 50 K$ shield house electronics • Total around 3 M$

L 3 Farm • Base on Hall D estimate – Driven by PVDIS – Assuming 30 KHz • 390 cores for non tracking option • 1500 cores with tracking

Operation costs • Cables : 300 K$ • Racks : 15 K$ • Tape : 600 K$

Test stand • • Test trigger logic Test max trigger rates Test electronics with real detector signals Test GEM readout max performance • 4 FADC, 1 CTP and 2 VXS crates • 1 SRS from UVA and 1 INFN MPD readout

Things needed for director review • Tes trigger logic and FADC performance • PVDIS – Occupancy in Calorimeter – L 3 data reduction scheme • SIDIS trigger rates – MRPC, LGC, HC , SPD simulation • • Simulation digitization SIDIS GEM occupancy from G 4 SIDIS rates from G 4 simulation MRPC development custom electronics APV 25 high rate, test with SRS L 3 reduction factor and ressource needed test algorithm Ma. PMT test

Conclusion • PVDIS doable and cost reasonable now calorimeter rate down : 18 KHz • SRS : saves cost on crates and per channel, more performant data transfer GEM event size manageable • SIDIS most like should be able to take all single triggers by reading only integral and one sample per trigger for GEMs • Need to test at high rate