Gianluca Aglieri Rinella1 - PowerPoint PPT Presentation

About This Presentation
Title:

Gianluca Aglieri Rinella1

Description:

Development and Implementation of the Level 0 Pixel Trigger System for the ALICE experiment Gianluca Aglieri Rinella1 On behalf of the ALICE Pixel Trigger Project – PowerPoint PPT presentation

Number of Views:60
Avg rating:3.0/5.0
Slides: 26
Provided by: gianl9
Category:

less

Transcript and Presenter's Notes

Title: Gianluca Aglieri Rinella1


1
Development and Implementation of the Level 0
Pixel Trigger System for the ALICE experiment
  • Gianluca Aglieri Rinella1
  • On behalf of the ALICE Pixel Trigger Project
  • 1CERN, European Organization for Nuclear Research

2
Outline
  • ALICE experimental apparatus and the Silicon
    Pixel Detector
  • Fast-OR signals
  • Triggering with Fast-OR signals
  • Pixel Trigger System
  • Description
  • Features
  • Measurements

3
The ALICE experiment at LHC
L3 solenoid magnet
  • Heavy ions collisions
  • Quark gluon plasma
  • Proton-proton collisions
  • pt 0.6 GeV/c

4
Silicon Pixel Detector
  • 120 detector modules (half staves)
  • Installed in the experimental cavern

5
Readout pixel chips and Fast-OR
  • Each half stave (120)
  • 2 silicon pixel sensors (160x256 pixels of 425x50
    um2)
  • 10 readout pixel chips (32x2568192 channels,
    13x14 mm2 bump bonded to sensors)
  • 1 optical readout Multi Chip Module
  • Pixel chip prompt Fast-OR
  • Active if at least one pixel hit out of 8192
  • Transmitted every 100 ns
  • 10 on each of 120 optical links (1200)
  • Low granularity, low latency pad detector
    1200 pads of 13x14 mm2

Silicon Pixel Detector
800 Mb/s G-Link
Readout MCM
Half stave
141 mm
Sensor
1
Sensor
Pixel chips
6
Triggering with SPD
  • Pre-process low latency Fast-OR signals and
    generate input to contribute to the Level 0
    trigger decision
  • Proton-proton
  • Minimum bias
  • High multiplicity trigger
  • Heavy ions
  • Selection of impact parameter
  • Algorithms
  • Boolean functions (AND/OR) of 1200 Fast-OR bits
  • Topological coincidence trigger (Global OR,
    Vertex)
  • Occupancy (multiplicity)
  • Implementation on FPGA

7
Requirements and constraints
  • Requirements
  • Extract and process 1200 Fast-OR signals from 120
    optical links operating at 800 Mb/s
  • Latency 800 ns
  • Allow user selection of processing algorithm
  • Constraints
  • No interference on data readout
  • System location and space occupation

8
40 m fibers 200 ns propagation delay
9
Pixel Trigger System architecture
To DAQ in control room
1200 bits _at_ 10 MHz
CTP
Processing
Fast-OR extraction
Optical splitters
120 G-Link
Pixel Trigger electronics
350 ns
200 ns
225 ns
25 ns
800 ns
  • Processing time lt 25 ns
  • Bottleneck data deserialization and Fast-OR
    extraction

10
Developments
  • Advanced 12-channel parallel optical fiber
    receiver modules
  • Devices customized by Zarlink
  • 1310 nm, single mode
  • Experimental validation
  • Space saving
  • G-Link protocol deserializers on programmable
    hardware
  • Implemented and tested with advanced FPGAs
  • Not fulfilling latency requirement
  • Dedicated deserializer ASIC

11
Pixel Trigger system electronics
  • 9U VME size processing board (BRAIN)
  • Main processing FPGA (960 user I/O pins, 1513
    BGA)
  • 2x5 receiver boards (OPTIN) connected as
    mezzanine boards

400 mm
  • High speed optical interfaces
  • Detector Data Link
  • Timing Trigger and Control

Large I/O Virtex4
Optical TxRx
360 mm
  • Data flow parallelism (1000 lines)
  • Double Data Rate
  • Digitally Controlled Impedance
  • 800 impedance matched lines

TTCRx
Virtex4
USB
DDL SIU
JTAG
12
OPTIN receiver board
  • 12 channels
  • Parallel optical receiver module
  • 12 closely packed G-link deserializer ASICs

13
BRAIN and OPTIN boards
14
Pixel Trigger system crate
  • Installation, integration and commissioning with
    SPD foreseen by the end of the year

15
Control and configuration
  • Status monitoring and control on ALICE DDL
    communication layer
  • User selection of different processing algorithms
  • Download of configuration file into local SRAM
    memory
  • Reconfiguration of the processing FPGA
  • Interfaces to several ALICE subsystems

Experiment Control
Central Trigger Processor Control
Silicon Pixel Detector Control
DDL
Pixel Trigger servers (control room)
16
Latency measurement
  • Serialization, deserialization, processing
    latency
  • Hardware emulator of the Silicon Pixel Detector
    as data source
  • Overall latency 800 ns

215 ns
17
Conclusions
  • The ALICE Pixel Trigger system allows to use the
    prompt Fast-OR outputs of the Silicon Pixel
    Detector for the Level 0 trigger decision
  • Minimum bias and multiplicity trigger
  • ALICE the only experiment at LHC to include the
    vertex detector in the first trigger decision
    from startup
  • The Pixel Trigger system
  • Challenging design including innovative
    developments
  • Highly compact, dense and parallel architecture
  • 800 ns latency
  • Hardware ready
  • Firmware and software development ongoing

18
Spare slides
19
References
  • References
  • ALICE collaboration, ALICE physics Performance
    Report, CERN-LHCC-2003-049, J. Phys., G30 (2004)
    1517-1763
  • J. Conrad et al., Minimum Bias Triggers in
    Proton-Proton collisions with the VZERO and
    Silicon Pixel Detectors, ALICE Internal note,
    ALICE-INT-2005-025, 19/10/2005
  • G. Aglieri Rinella et al., The Level 0 Pixel
    Trigger system for the ALICE experiment, Journal
    of Instrumentation JINST 2P01007
  • A. Kluge, The ALICE Silicon Pixel Detector
    front-end and readout electronics, NIM A 560
    (2006) 67-70

20
Angular resolution
21
Trigger algorithms
  • Combinational (boolean AND/OR) functions of 1200
    Fast-OR bits
  • Occupancy (multiplicity)
  • Coincidence trigger (topology)
  • Not possible iterative algorithms on data set
  • Example vertex trigger
  • Pseudo-Tracklet one chip hit on inner and one on
    outer layer, in line with region /-10 cm around
    vertex
  • Chip map for pixel trigger electronics calculated
    from simulation (L11,L21), (L12, L22), , (L1n,
    L2n)
  • FPGA looks for at least 1 out of 11000
    pseudo-tracklets
  • Processing time 12.4 ns (Xilinx ISE)
  • 4 of FPGA resources (Xilinx ISE)
  • FPGA counts how many out of 11000 tracklets are
    present
  • 27 ns processing time (Xilinx ISE)
  • 5 of FPGA resources (Xilinx ISE)

22
Multiplicity trigger
Jan Fiete Grosse Oetringhaus
23
Radiation effects
  • Radiation Results of the SER Test of Actel,
    Xilinx and Altera FPGA instances, iROC report,
    2004
  • Failure In Time (FIT) errors in 109 hours with
    neutron flux of 14 cm-2hr-1
  • SEFI Single Event Functional Interrupt
  • SEU Single Event Upset (configuration)

FIT SEFI SEU
Altera EP1C20 453
Xilinx XC3S1000 320 1240
Xilinx XC2V3000 1150 8680
  • Neutron max fluence 2.0 108 cm-2 (10 y)
  • Morsch, Pastircak, Radiation in ALICE Detectors
    and Electronic Racks, ALICE-INT-2002-28
  • Central Trigger Processor using SRAM based ALTERA
    Cyclone EP1C20

Errors in 10 years SEFI SEU
Altera EP1C20 7
Xilinx XC3S1000 5 18
Xilinx XC2V3000 17 124
24
ALICE trigger parameters
Level 0 Level 1 Level 2
Last trigger input at CTP (ms) 0.8 6.1 87.6
Trigger output at CTP (ms) 0.9 6.2 87.7
Trigger input at detectors (ms) 1.2 6.5
Rate (Hz) 1000 40-800
25
The ALICE experiment at LHC
L3 solenoid magnet
Dipole magnet
  • Heavy ions collisions
  • Quark gluon plasma
  • Proton-proton collisions
  • pt 0.6 GeV/c
Write a Comment
User Comments (0)
About PowerShow.com