Title: Preliminary Design of Trigger System for BES III
1Preliminary Design of Trigger System for BES III
- Zhenan LIU
- Inst of High Energy Physics,Beijing
- Oct. 14 2001
2Outline
- Introduction
- Estimation of event rate
- The principle of BESIII trigger
- Subsystems Description
- Summary
3Introduction
- BEPCI
- low luminosity(3?10 30cm-2s-1).
- Single bunch mode
- bunch separation 800ns
- triggerlatch-process-decision
BEPCII
- BEPCII
- double ring
- high luminosity(10 33cm-2s-1).
- multi-bunch mode
- bunch separation 8ns
- TRIGGER FOR BESII CAN NOT BE USED IN BESIII
BESIII
4Introduction(2)
- BES III
- new vertex detector
- new MDC
- new TOF
- new EMC(barrel BGO
- endcap CsI)
- superconducting magnet
- new muon counter
- new FEE for all subdetectors
5A NEW TRIGGER SYSTEM MUST BE DESIGNED !
6Estimation of event rate
Purpose To accept all interested events To
rejects as much background as possible
- Good physics events(J/?, ?? at BEPCII/BESIII)
- Bhabha and radiative Bhabha scattering
- Machine background
- Coulomb scattering, Bremsstrahlung with
residual gas - Touscheck scattering
- Cosmic Ray background
7Background from Lost particles
8Background from Lost particles
9Machine background
- Intensity IBEPCII (1.11.1) A, ( 1.1 ? 1013
e ) -
( 40 ? IBEPC) - Lifetime ?BEPCII 3.53.8 hr
-
(0.5 ? ?BEPC) - Lost rate (dn/dt) BEPCII 8.7 ? 108/s
-
(80 ? dn/dt BEPC) - Lost rate(BESIII) (dn/dt) BESIII ? 1.3 ? 107 /s
-
-
? N beam-BEPCII 13MHz (80 ? N beam-BEPC)
10BG measurement of MDC in BES2
11Machine background
- BEPC/BESII
- 15Hz at DAQ
- Single wire of 1st layer of MDC
- BEPCII/BESIII
- 30KHz on single MDC wire at R17cm
- 1500 Hz (DAQ)
FigBackgrounds rate vs beam current
12Cosmic ray background
- Flux Beijing170 m-2 s-1
- N 1500 /s
- BESI/II
- T window 40 ns
- 95 suppressed
- N cosmic 1 Hz
- BESIII
- No time window
- similar geometrical size to BESI/II
- Same of cosmic rays hitting on BESIII
- Using Vertex counter
- N cosmic 20 Hz
13Good physics Bhabha events
- Luminosity LBEPCII 1 ? 1033 cm-2 s-1
- Optimized to J/? and ?? energy
N J/? 1500 Hz
N ? 420 Hz
N BB ? ? BB 550 Hz ( cos? lt 0.95)
14Total event rate
Total event rate good event rate
cosmic event rate
bhabha rate machine background
rate 3000 Hz
15The principle of BESIII trigger
- Trigger requirement
- no dead time or small total dead time
- Flexibility
- Constraints from BEPCII
- 93 bunches in the ring (8 ns separation)
- Latch-process-decision mode not possible in 8ns
Pipeline must be used in FEE trigger
16 Trigger pipeline
- Constraints due to Detectors
- Tdrift MDC 500 ns(C3 H8 small cell), signals
from different collisions mixed together - Trigger pipeline clock
- RF division (fRF499.8MHz)
- 20MHz (fRF/24)
- all trigger promitives are processed every 50ns
Trigger latency
17The principle of BESIII trigger(2)
- Hardware trigger software filter(L2)
- FEE signal splitted
- trigger FEE pipeline
- Trigger pipeline clock 20MHz
- Level 1(L1) 2.4 ?s
- FEE Control Logic checks L1 with pipeline clock
- L1 YES
- moves pipeline data to readout buffer
- L1 No
- overwritten by new data
Time Reference
Detector
0 s
FEE pipeline
Level 1
2.4?s
Readout buffer
switch
Farms
Ev.Filter
(L2)
Disk
BESIII FEE pipeline and Data flow
18Schematic of BES III Trigger
VC
DISC
Hit Count
L0 trigger Logic
FEE
L0P
TOF
DISC
Hit/Seg Count
Global Trigger Logic
MDC
DISC
Track Finder
Track Seg. Finder
DAQ
EMC
BTE Sum
Tile Processor
Tile Sum
Total Ener Sum
L1P
MU
DISC
Mu track
CLOCK
RF
TTC
2.4 ?s
19Subsystem Description
- MDC Trigger
- EMC Trigger
- Trigger Timing and Control
- Global Trigger Logic
20Trigger for MDC
- MDC
- 5740 wires,
- 36 layers in 9 super-layers
- 6 stereo superlayers
- 3 axial superlayers
- Trigger
- Axial superlayers for trigger
- PT gt 150 MeV
- step1 track segment finding
- step 2track-finding
21MDC Tracking
- Track Segment Finder
- Reference layer second layer in each
super-layers - Segment definition
- 3/4 logic in a superlayers
- Track segments
- Superlayer 1 66
- Superlayer 2 162
- superlayer 3 192
- superlayer 4 270
- Track finder
- Reference Segment second superlayer
- (track reference 162)
- Track definition
- all 4 segments in 4 superlayers
22Track Segment finding
23MDC track combination MC Calculation
- MC
- Private code
- random P(0.15-1.547GeV), initial angle in phi
- B 1.0 T
- Ask for all superlayers with segment hit
- combinations for each reference segment 117
- realization with CPLD or FPGA
24EMC(Barrel) trigger
- BEMC
- 19360 BGO crystals
- 88 rings along Z
- 44 rings with 240 crystals(inner)
- 44 rings with 200 crystals(outer)
25EMC(Barrel) Trigger
Tile30(? )?22(z)
?
- Totoal Energy Trigger
- sum of all tile signals
- thresholds
- Segmentation
- 8(?) ?4(z) 32 segments
- Vthl and Vthh
- No. Of segments
- Event time
- Basic Trigger element(BTE)
- BTE1FEE sum of 30 cells
- BTE2FEE sum of 25 cells
- Tile
- Tile1sum of 23 BTEs
- Tile2sum of 23 BTEs
26(No Transcript)
27Trigger Timing and Control(TTC)
- fRF 499.8 MHz
- fRF /12 ? 40MHz
- FEE TDC clock
- fRF/24 ? 20MHz
- trigger FEE pipeline clock
28Global Trigger Logic
- Inputs subdetector primitives
- Time adjustment
- trigger table
- Prescaling of some event types
- System Configuration
- Trigger data readout
29Summary
- Pipeline will be used
- Hardware trigger software filter
- Hardware level 0 level 1
- (L0 is spare signal for FEE)
- LatencyL0- 200 ns, L1- 2.4 ?s
- Monte Carlo simulation undergoing
- backgrounds
- Subsystem trigger scheme
- Collaboration domestic and international
30THANK YOU ! WELCOME COLLABORATIONS!