1 electronics status of the meg trigger system status and plans for daq mscb slow control system...
TRANSCRIPT
1
Electronics
• Status of the MEG Trigger system• Status and plans for DAQ• MSCB slow control system
• Status of the MEG Trigger system• Status and plans for DAQ• MSCB slow control system
2
The Trigger System of the MEG
Experiment
On behalf of
M. Grassi
D. NicolòF. Morsani S. Galeotti S. Giurgola
3
Expected Trigger RateAccidental background and
Rejection obtained by applying cuts on the following variables
• photon energy• photon direction• hit on the positron counter • time correlation• positron-photon direction match
4γ 102~ 97% MeV 45 fE
99%)5.3( 1.2 oo
1-s 20 4
ff
TRfRR
e1.0
4
10 18
sR
16105 sRe
100% 102 nsT5 ; 2 ff
The rate depends on R Re+ R2
ee
e e
4
The trigger implementation
Digital approach – Flash analog-to-digital converters (FADC)– Field programmable gate array (FPGA)
Final system Only 2 different board types Arranged in a tree structure on 3
layers Connected with fast LVDS buses Remote configuration/debugging
capability
Prototype boardCheck of: the FADC-FPGA compatibility chosen algorithms synchronous operation data transmission
5
The board type 0
PMT inputs
LVDS transm.
LVDS receiv.
FADC
FPGA
configuration EPROMS
Differential drivers
package error solved with a patch board
control signals.
6
Prototype system
configuration
Diff. driver Fadc Proc.
Algor.
LVDS Rx
LVDS Tx
Proc. Algor.
Circ.buff
Circ.buff
Circ.buff
Circ.buff
Diff. driver Fadc Proc.
Algor.
LVDS Rx
LVDS Tx
Proc. Algor.
Circ.buff
Circ.buff
Circ.buff
Circ.buff
16 PMT
16 PMT
input
input
output
output
LVDS in
final
Board 1
Board 0
Last BVR conclusionsThe prototype system
met all requirementsIt is available to trigger
the LP in future beam tests
7
Trigger system structure
LXe inner face(312 PMT)
. . . 20 boards
20 x 48
Type1Type1
Type116
4
2 boards
. . . 10 boards
10 x 48
Type1Type1
Type116
4
LXe lateral faces(208 PMT)
(120x2 PMT)(40x2 PMT)
1 board
. . . 12 or 6 boards
12 x 48
Type1Type1
Type116
4
Timing counters(160 PMT)
or(80 PMT)
2 or 1
boards
4 x 48
1 board
4 x 48
2 x 48
2 VME 6U
1 VME 9U
Located on the platform
Type2
Type2
Type2
Type2
Type2
Type2
8
Type 1
Diff. driver Fadc
LVDS TxOpt.
Proces. Algor.
Circ.buff
Circ.buff
Circ.buff
LVDS Tx
Proc. Algor.16
PMT
Type 2 LVDS TxOpt.
Proces. Algor.
Circ.buff
Circ.buff
Circ.buff
LVDS Tx
Proc. Algor.10
type1
LVDS Rx
LVDS Rx
LVDS Rx
9
Software items– New package ISE 6.2– Verilog/schematic implementation– Block transfer in A32D16 format (VME library to be modified)
Hardware items– JTAG programming/debugging through VME by modifying the
Type0– Analog receivers and with DACs for pedestal– FPGA selected: VirtexII PRO
• On Type 1 XC2VP20-5-FF1152 • On Type 2 XC2VP40-5-FF1152
– Other components are fixed: • FADC • LVDS Tx and Rx • Clock distributor• Analog input by 3M coaxial connectors• LVDS connection by 3M cables
– Ancillary logic components and scheme
10
FPGAVIRTEX II - PRO
• easily at 100MHz
• 60% of IO
• 40% of CLB
• 2 PowerPC (not used)
11
Analog receiver
AD5300AD8138
Differential driver
DAC pedestal control
12
Analog receiver
• 16 channels on a type1 board
• 1 unit wide
13
DC/DC converter
• 1.5 Volts
• 3.3 Volts
14
ANCILLARY: TREEANCILLARY: TREE
TRIGGERANCILLARY
#0
START
STOP
SYNC
CLK INT
ANCILLARY#1CLK EXT
ANCILLARY#8CLK EXT
…
START, STOP, SYNC, CLK
………………………………
(… 16)
VMESTOP
15
ANCILLARY: BLOCKSANCILLARY: BLOCKS
4 x SILICON DELAYS:START, STOP, SYNC, CLK
INPUTS & CLK GENCLK - START - STOP - SYNC
TTL2LVDS
4 x 8(16)-LVDS-FANOUT
VME INTERFACE
16
ANCILLARY: INPUTS & CLOCK ANCILLARY: INPUTS & CLOCK GENGEN
10MHz CLK GEN INT/EXT SELECT
INPUT CONNECTOR
LVDS-to-TTL
17
ANCILLARY: SILICON DELAYANCILLARY: SILICON DELAY
18
ANCILLARY: LVDS FANOUTANCILLARY: LVDS FANOUT
19
Full System
2002 2003 2004 2005
Test MilestoneAssemblyDesign Manufactoring
Prototype Board
Final Prototype
Trigger
Full System
Prototype Board
Final Prototype
part.inst.
full.inst.
20
summary
• Components selected • Algorithms implemented• PCB design ready to start
21
Status and plans for DAQStatus and plans for DAQ
22
Domino Chip Principles
DLL
Phase and Frequency Stabilization
Phase and Frequency Stabilization
ExternalCommonReference
Clock
Vspeed8
inpu
ts
shift register
Trigger Signal Sampling
Trigger Signal Sampling
Low-jitterclock
domino wave
FADC
MUX16-bitDAC
uCFreq.Cntr
23
Timing reference
signal
20 MHz clock
PMT hit
Domino stops aftertrigger latency
24
Recovery of Timing
Domino speed stability of 10-3 : 400ps uncertainty for full window25ps uncertainty for timing relative to edge
Domino speed stability of 10-3 : 400ps uncertainty for full window25ps uncertainty for timing relative to edge
50 ns
1) Trigger publishes phase of trigger signal f relative to clock in multiples of 10 ns
1) Trigger publishes phase of trigger signal f relative to clock in multiples of 10 ns
2) Each DAQ card determines and fits “Time-Zero-Edge” in clock signal and uses this as t=0
2) Each DAQ card determines and fits “Time-Zero-Edge” in clock signal and uses this as t=0
3) Measure pulse width of clock to derive domino speed
3) Measure pulse width of clock to derive domino speed
4) Timing of all PMT pulses is expressed relative to t=0 point
4) Timing of all PMT pulses is expressed relative to t=0 point
25
Current readout mode
• First implemented in DRS2• Sampled charge does not leave chip• Current readout less sensitive to cross-
talk etc.
• First implemented in DRS2• Sampled charge does not leave chip• Current readout less sensitive to cross-
talk etc.
write
read
C
. . .
RI
VoutVin
26
DRS2 Chip• DRS2 design
– Up to 4.5 GHz sampling speed– 8+2 channels, 1024 bins deep each– Readout speed up to 100 MHz (?)– Submitted to UMC in Nov. 18th, 58 chips received in Jan.
15th, packaging 3 weeks
• DRS2 chip arrived in Feb. 04– 50 packaged chips (400 channels)– 1.5 GHz – 4.5 GHz sampling speed– Current mode readout works– Jitter estimation: 40ps
• Plans– VME prototype board by Aug. ‘04
• DRS2 design– Up to 4.5 GHz sampling speed– 8+2 channels, 1024 bins deep each– Readout speed up to 100 MHz (?)– Submitted to UMC in Nov. 18th, 58 chips received in Jan.
15th, packaging 3 weeks
• DRS2 chip arrived in Feb. 04– 50 packaged chips (400 channels)– 1.5 GHz – 4.5 GHz sampling speed– Current mode readout works– Jitter estimation: 40ps
• Plans– VME prototype board by Aug. ‘04
27
DRS2 chipDRS2 chip
28
DRS2 testsDRS2 tests
29
Sampling Speed Measurement
• Obtained last week with USB-Mezzanine board
• Usable sampling range 0.6 GHz – 4 GHz
• Obtained last week with USB-Mezzanine board
• Usable sampling range 0.6 GHz – 4 GHz
30
Jitter estimation
• Oscilloscope triggered with Domino pulse
• Show 250 turns later• 11ns/250 = 44ps• Should be improved
with better board design
• Oscilloscope triggered with Domino pulse
• Show 250 turns later• 11ns/250 = 44ps• Should be improved
with better board design
31
Analog readout
• 4 pulses, 12ns wide, ~1ns rise time digitized at 2.5 GHz
• Readout at 40 MHz
• Reproduced rise time: 1.2ns
• Tests with FADC will follow
• 4 pulses, 12ns wide, ~1ns rise time digitized at 2.5 GHz
• Readout at 40 MHz
• Reproduced rise time: 1.2ns
• Tests with FADC will follow
32
VME boards with Mezzanine Cards
R. PaolettiINFN PisaMAGIC collaboration
PSIGVME board
33
PSI GVME Board
34
VME Transition Boards
• PMT/DC signals through front-panel connectors to CMC cards with DRS
• Low-jitter clock through front-panel ch. #17
• Trigger, reset, etc through transition board
• Read feedback to trigger thorough transition board
• PMT/DC signals through front-panel connectors to CMC cards with DRS
• Low-jitter clock through front-panel ch. #17
• Trigger, reset, etc through transition board
• Read feedback to trigger thorough transition board
PM
Ts
Clo
ckC
lock
35
DAQ System
PMTActiveSplitter
~11m~3m
area
monitor
~3mDRS
Board(32chn)+ CPU
DC Pre-Amp DRSBoard
(32chn)+ CPU
~7m
SIS3100
5 VME crates
800 + 160
1920
opticalfiber (~20m)
Trigger
GigabitEthernet
Front-End PCs
On-line farm
Rack – PC (Linux)Rack – PC (Linux)Rack – PC (Linux)Rack – PC (Linux)Rack – PC (Linux)
Rack – PC (Linux)Rack – PC (Linux)Rack – PC (Linux)Rack – PC (Linux)Rack – PC (Linux)Rack – PC (Linux)
storage
Fitted data:10 Hz waveform data -> 1.2 MB/sec90 Hz ADC / TDC data -> 0.9 MB/sec
Raw data:2880 channels100 Hz50% / 10% / 10% occupancy2kB / waveform -> 5 x 25 MB/sec.
3+1 VME crates
trigger
ready
backpressure
36
Rack LayoutRack Layout
SplitterTriggerDRSInterface
?
Clock distribution(front)
TriggerBoard readyFast clear
Event counter(transition cards)
DCCalo+TC
37
HVHV
Splitter
Trigger
DAQ Calo+TC
DAQ DC
38
Electronics in B-Field
4.2 6.8 9.1
6.7 18.5 83.8
9.0 33.5433.7 440.331.2 7.5
50.5
73.320.4 56.369.7 19.8
7.8 11.6 11.3
Fringing field measured at πE5 [Gauss]• B. Allongue (PH-ESS Group,
Cern)• Wiener PL500 Power Supply
works up to 300 Gauss (900 with water cooling)
• Fan works up to 80-100 Gauss
• Ordered “normal” crate for tests
• Need air ducts with external A/C if problems arise
• B. Allongue (PH-ESS Group, Cern)
• Wiener PL500 Power Supply works up to 300 Gauss (900 with water cooling)
• Fan works up to 80-100 Gauss
• Ordered “normal” crate for tests
• Need air ducts with external A/C if problems arise
39
Waveform analysis
• Zero suppression in FPGA• Single hit
– ADC/TDC derived in FPGA
• Multiple hit– Waveform compressed in
FPGA (2x12 bit -> 3 Byte)– Waveform fitted /
compressed in PC cluster
• Store ADC/TDC only for “calibration” events
• Store (lossless) compressed waveforms for MEG candidates
• Zero suppression in FPGA• Single hit
– ADC/TDC derived in FPGA
• Multiple hit– Waveform compressed in
FPGA (2x12 bit -> 3 Byte)– Waveform fitted /
compressed in PC cluster
• Store ADC/TDC only for “calibration” events
• Store (lossless) compressed waveforms for MEG candidates
OriginalWaveform
Difference Of Samples
Threshold in DOS
Region for pedestal
evaluationintegration area
ADC1/TDC1
ADC2/TDC2
T
40
Differential DRS channels
write
Vin
write
Vin +
Vin -
DifferentialDriver
cross-talk
signalscancel
Measured for 1ns rise-time:• 6% neighbor• 2% next nb
Measured for 1ns rise-time:• 6% neighbor• 2% next nb
41
DRS3
• DRS2 can probably initially be used for DAQ (have 400 channels, can produce 400 more)
• DRS2 limitations– Only two channels are fully differential (others show
larger crosstalk)– Some tests remain to be done …
• New DRS3 design– All channels differential– Additional shielding between channels (ground bond
wires)– Reduced readout time (5x) minimized dead time– Internal cascading allows for n x 1024 sampling bins
• DRS2 can probably initially be used for DAQ (have 400 channels, can produce 400 more)
• DRS2 limitations– Only two channels are fully differential (others show
larger crosstalk)– Some tests remain to be done …
• New DRS3 design– All channels differential– Additional shielding between channels (ground bond
wires)– Reduced readout time (5x) minimized dead time– Internal cascading allows for n x 1024 sampling bins
42
Plans
• DRS2 VME prototype board Aug. 04• Measure all parameters (cross-talk, resolution,
stability)• Produce VME boards and equip with DRS2 chip
(400 chn + 400 chn ?), install as much as possible in are in summer 2005
• Design DRS3 in parallel• Mass production of DRS3 in fall 2005• Replace installed DRS2 with DRS3
• DRS2 VME prototype board Aug. 04• Measure all parameters (cross-talk, resolution,
stability)• Produce VME boards and equip with DRS2 chip
(400 chn + 400 chn ?), install as much as possible in are in summer 2005
• Design DRS3 in parallel• Mass production of DRS3 in fall 2005• Replace installed DRS2 with DRS3
43
DRS (DAQ)
2002 2003 2004 2005
Test MilestoneAssemblyDesign Manufactoring
DRS2
DRS2 test board
DRS3
VME boards
400 chn 400 chn
Mass Production
3000 chn
DRS1
2nd Prototype
Tests
Boards & Chip Test
DRS1
DRS2
Full System
installation
OptionalDRS2 production400 chn
44
Slow Control System
• Unified control system for– Cryogenics (temperature, pressure,
valves, etc.)– Environment (temperatures, crates)– Drift chamber gas system (pressure,
mass flow, temperature)– High Voltage (Calo+TC+DC, ~1000
chn.)
• System must be fail-safe
• Unified control system for– Cryogenics (temperature, pressure,
valves, etc.)– Environment (temperatures, crates)– Drift chamber gas system (pressure,
mass flow, temperature)– High Voltage (Calo+TC+DC, ~1000
chn.)
• System must be fail-safe
45
Slow ControlHV
PC
RS
232
12345
Temperature, pressure, …
GP
IB
Valves
??? 15° C
heater
PLC
12:30 12.3
12:45 17.2
13:20 15.2
14:10 17.3
15:20 16.2
18:30 21.3
19:20 18.2
19:45 19.2
MIDASDAQ
Eth
ern
etTerminal Server
46
Slow Control BusHV
Temperature, pressure, … Valves
heater
MIDASDAQ
47
• CAN, Profibus, LON available• Node with ADC >100$• Interoperatibility not guaranteed• Protocol overhead• Local CPU? User programmable?• How to integrate in HV? (CAEN
use CAENET)
• CAN, Profibus, LON available• Node with ADC >100$• Interoperatibility not guaranteed• Protocol overhead• Local CPU? User programmable?• How to integrate in HV? (CAEN
use CAENET)
Field Bus Solutions
48
Hardware Overview
• 8051-compatible C with– ADC 12-bit– DAC 12-bit– Flash EEPROM– Timers– Watchdog– Temperature sensor– UARTs– Up to 100MHz clock
speed• External signal conditioning
if needed• Serial communication
• 8051-compatible C with– ADC 12-bit– DAC 12-bit– Flash EEPROM– Timers– Watchdog– Temperature sensor– UARTs– Up to 100MHz clock
speed• External signal conditioning
if needed• Serial communication
49
RS-485 bus
• Similar to RS-232 but– Up to 256 (1/8 load) units can
be connected to a single segment, use repeater for more
– Address space for 65536 nodes– single line, half duplex– differential twisted pair– Segment length up to km (20 m
tested)– MSCB system: 115kbit
– Single Master – Multiple Slaves (like USB)
– Power through bus (10-wire flat ribbon)
– PC USB interface
• Similar to RS-232 but– Up to 256 (1/8 load) units can
be connected to a single segment, use repeater for more
– Address space for 65536 nodes– single line, half duplex– differential twisted pair– Segment length up to km (20 m
tested)– MSCB system: 115kbit
– Single Master – Multiple Slaves (like USB)
– Power through bus (10-wire flat ribbon)
– PC USB interface
50
Generic node SCS-200• C8051Fxxx Micro
controllers with 8x12 bit ADC, 2x12 bit DAC, digital IO, 8051 C and 32kB Flash Memory
• RS-485 bus over flat ribbon cable
• Powered through bus• Costs ~CHF 50• Piggy back board for
signal conditioning cards• 32 kB for real time C
programs
• C8051Fxxx Micro controllers with 8x12 bit ADC, 2x12 bit DAC, digital IO, 8051 C and 32kB Flash Memory
• RS-485 bus over flat ribbon cable
• Powered through bus• Costs ~CHF 50• Piggy back board for
signal conditioning cards• 32 kB for real time C
programs
51
2 Versions
• Generic node with signal conditioning• RS232 node with protocol translator• PC connection to parallel port (USB
planned)• Integration on sensors, in crates
BUS OrientedBUS Oriented
Crate OrientedCrate Oriented
• 19” crate with custom backplane• Generic node as piggy-back• Cards for analog IO / digital IO / °C /
220V• Used in 3 experiments at PSI
Can be mixed
52
SCS Nodes•SCS 210: RS232 I/O•SCS 300: Centronics I/O (14 bit digital)•SCS 310: GPIB (IEEE-488) I/O•SCS 400: 8 chn. Thermocouple + 4 chn.
digital output (PWM), temp. regulation in SW
•SCS 500: 8 chn. differential analog input0…10V, 0…1V, 0…100mV, 0…
10mV-10…10V, -1…1V, -100mV…100mV0…100mA, 0…10mA, 0…1mA15V power, Lemo or screw terminal
•SCS 600: 8 chn. digital output 0…30V, 1ALEDs and buttons on front panel220V “power box”
•SCS 700: 8 chn. PT100/PT1000, 8 bit digital output
•SCS 800: 8 chn. capacitance meter•SCS 900: 8 chn. 24-bit ADC -10V…10V
8 chn. 16-bit DAC -10V…10V
•SCS 210: RS232 I/O•SCS 300: Centronics I/O (14 bit digital)•SCS 310: GPIB (IEEE-488) I/O•SCS 400: 8 chn. Thermocouple + 4 chn.
digital output (PWM), temp. regulation in SW
•SCS 500: 8 chn. differential analog input0…10V, 0…1V, 0…100mV, 0…
10mV-10…10V, -1…1V, -100mV…100mV0…100mA, 0…10mA, 0…1mA15V power, Lemo or screw terminal
•SCS 600: 8 chn. digital output 0…30V, 1ALEDs and buttons on front panel220V “power box”
•SCS 700: 8 chn. PT100/PT1000, 8 bit digital output
•SCS 800: 8 chn. capacitance meter•SCS 900: 8 chn. 24-bit ADC -10V…10V
8 chn. 16-bit DAC -10V…10V
53
Software overview
mscb.dll
msc.exeCommand line
interface
VI
LabViewApplication
VI VI VI
LPT USB
PC
SCS 250
SCS xxxFramework
User code
RS485
SCS 300
54
Remote access
mscb.dll
msc.exeCommand line
interface
LPT USB
PC
RS485
mscb.dll
msc.exeCommand line
interface
VI
LabViewApplication
VI VI VI
PCTCP/IP
Speed (commands/sec, 115kBaud):
~1500 local~900 remote
Speed (commands/sec, 115kBaud):
~1500 local~900 remote
SCS 250
55
Protocol
• Asynchronous 345 kBaud / 115kBaud• 16-bit addressing (65536 nodes)• CRC-code for error detection• Optional acknowledge• Concept of channels and configuration
parameters (256 each per node)• Optimized protocol: 1500 cmd/sec @ 345kB
• Asynchronous 345 kBaud / 115kBaud• 16-bit addressing (65536 nodes)• CRC-code for error detection• Optional acknowledge• Concept of channels and configuration
parameters (256 each per node)• Optimized protocol: 1500 cmd/sec @ 345kB
command channel value CRC
write data
node
param1
param2
param3
channel1
channel2
channel3
ADC
ADC
port
command LSB MSB CRC
address command
1 Byte
command CRC acknowledge
56
Node communication
msc.exe
LabView
57
Labview control of Large Prototype
Labview control of Large Prototype
58
How to make it fail-safe?
• Robust protocol– CRC code– Automatic reconnect
• UPS– SC crates on battery (~30 min.)– Use Laptop for control PC
• Redundancy– Operate two completely independent
nets– Switch between nets on failure
• Robust protocol– CRC code– Automatic reconnect
• UPS– SC crates on battery (~30 min.)– Use Laptop for control PC
• Redundancy– Operate two completely independent
nets– Switch between nets on failure
59
Redundancy
Temperature, pressure, … Valves
uC uC uC
uC uC uC
Switch box
ControlPC1
ControlPC2
• If uC fails, use other system• If PC fails, use other system (PC-PC watchdog)• For critical valves, use two (parallel or serial)• Avoid single point of failure !• Test failures
• If uC fails, use other system• If PC fails, use other system (PC-PC watchdog)• For critical valves, use two (parallel or serial)• Avoid single point of failure !• Test failures
System 1
System 2
Ethernet
60
New HV DesignVin
Vout
• • • 12 times • • •
3000V
0-3000V
ADC24 - bit Micro
Controller
DAC16 - bit
• Microcontroller optically decoupled from HV side
• Higher ADC+DAC resolutions
10 mV accuracy• Stable operation in lab
(weeks)• Newer test results will be
presented in review
• Microcontroller optically decoupled from HV side
• Higher ADC+DAC resolutions
10 mV accuracy• Stable operation in lab
(weeks)• Newer test results will be
presented in review
61
Integration of Slow Control into DAQIntegration of Slow Control into DAQ
• Run parameters will be written to MySQL database• Calibration parameters recalculated after every run and stored in DB• More frequent database update possible• Use of MIDAS history system (widely established since many years)
• Display through Web interface
• Gif images generated dynamically in memory
• Command line query with mhist
• Write speed: 2000 events/s
• Query over one month in ~10sec
• Display through Web interface
• Gif images generated dynamically in memory
• Command line query with mhist
• Write speed: 2000 events/s
• Query over one month in ~10secCombines Slow Control,
monitoring and calibration data
Combines Slow Control, monitoring and calibration data
62
Summary Midas Slow Control Bus
• 256 nodes, 65536 nodes with one level of repeaters• Bus length ~500m opto-isolated• Boards for voltage, current, temperature, Digital IO, 220V• Readout speed: 700 channels / sec. @ 115kBaud• C library, command-line utility, Midas driver, LabView
driver• Nodes are “self-documenting”• Configuration parameters in EEPROM on node• Node CPU can operate autonomously for interlock and
regulation (PID) tasks (C programmable, floating point library)
• Nodes can be reprogrammed over network
http://midas.psi.ch/mscb
• 256 nodes, 65536 nodes with one level of repeaters• Bus length ~500m opto-isolated• Boards for voltage, current, temperature, Digital IO, 220V• Readout speed: 700 channels / sec. @ 115kBaud• C library, command-line utility, Midas driver, LabView
driver• Nodes are “self-documenting”• Configuration parameters in EEPROM on node• Node CPU can operate autonomously for interlock and
regulation (PID) tasks (C programmable, floating point library)
• Nodes can be reprogrammed over network
http://midas.psi.ch/mscb
63
Conclusions
• Trigger system is on schedule• DRS2 chip works• Partial electronics installation until
mid-2005 planned• Slow control system works (as
tested in ~10 installations at PSI and TRIUMF)
• Trigger system is on schedule• DRS2 chip works• Partial electronics installation until
mid-2005 planned• Slow control system works (as
tested in ~10 installations at PSI and TRIUMF)