DDAQ Hardware Architecture

DAQ WS02 Feb 2006

Jean-Sébastien Graulich Slide 1

DDAQ Hardware ArchitectureDDAQ Hardware Architecture

o General Architecture

o Tracker Baseline

o TOF Baseline & Options

o EmCal Baseline & Options

o Trigger Baseline

o Summary

Jean-Sebastien Graulich, Geneva

DAQ WS02 Feb 2006


DAQ ArchitectureDAQ Architecture

Trigger distribution

Optical link

Online Monitoring

Tracker EmCal TOFTrigger + Ckovs

Ethernet

Linux PCs

GigaBit Switch

Run ControlEvent BuilderOnline Storage

VME Crates

Remote Storage

DAQ WS02 Feb 2006


Hardware Hardware ConsequencesConsequences

PCs are sitting in the MICE LCRPCs are sitting in the MICE LCR -> VME crates are also sitting in the -> VME crates are also sitting in the

MICE LCRMICE LCRCable transmission of Analog Signal or Digital Signal

MICE layout

Drawing by Tony Jones

DAQ WS02 Feb 2006


Tracker Tracker

Sensors: VLPCSensors: VLPC Front End Electronics (without AFE-Front End Electronics (without AFE-

t):t): Analog to Digital: AFE II boards sitting on the

cryo-cooler Digital Data Buffer: VLSB (512 ch/module, 4

modules/cryo-cooler) Number of ChannelsNumber of Channels

2 x 4096 ADC

DAQ WS02 Feb 2006


Tracker Connection Scheme (half a Tracker Connection Scheme (half a tracker)tracker)

4 VLSB Boards in VME Crate

MICE LCRMICE HALL1 Cryo-Cooler

4 AFE Boards

Fibers from VLPCs

Digital Signal

27+x m (Tracker 1) or 12+x m (Tracker 2)

Question: Where is the AVNET card Question: Where is the AVNET card located ?located ?

MIL1553

DAQ WS02 Feb 2006


TOF Baseline TOF Baseline Sensors: Sensors:

TOF0: Hamamatsu R4998TOF1 and TOF2: R4998 or R7761 (depending on

Magnetic field shielding)

R4998

DAQ WS02 Feb 2006


TOF BaselineTOF Baseline Front End Electronics:Front End Electronics:

Constant Fraction Discriminator CAEN V812 (16 ch, ECL output)

TDC CAEN V1290 (32 ch, ECL or LVDS Input) Number of ChannelsNumber of Channels

112 TDC (48 + 32 + 32)

No ADC in the Baseline

DAQ WS02 Feb 2006


TOF BaselineTOF Baseline

TOF Connection Scheme (1 Station)TOF Connection Scheme (1 Station)

Question: Is CFD going to be precise enough ?Question: Is CFD going to be precise enough ? See other options

3 CFD in VME Crate

BNC Cable (RG213) TriggerLogic

MICE LCRMICE HALL

48 PMTs

1.5 TDC in VMESHV Cable

HV Power Supply

30+x m (TOF 0) or 27+x m (TOF 1) or 12+x m (TOF 2)

DAQ WS02 Feb 2006


Timing is affected by the signal amplitudeTiming is affected by the signal amplitude

2 Usual Solutions: 2 Usual Solutions: Use Constant Fraction Discriminator (CFD) Measure amplitude and correct

TOF OptionsTOF Options

TimeWalk

Time over threshold

Thresh.

Tim

e W

alk

Co

rre

ctio

nExample from Harp

DAQ WS02 Feb 2006



Option 1. Constant Fraction Discriminator (CFD)Option 1. Constant Fraction Discriminator (CFD)The time walk is corrected automatically in the discriminator

Threshold value depends on the amplitude

Advantage: Advantage: Easy CablingEasy Cabling and and No ADCNo ADC

DrawbackDrawback No information at all on the charge (Energy

Deposit) Cost: TDC +Cost: TDC + ~ 125 EUR/ch (CFD)~ 125 EUR/ch (CFD)

PMT CFDBNC Cable

TDC

Trigger Logic

DAQ WS02 Feb 2006



Option 2. Measure the Charge with a QDCOption 2. Measure the Charge with a QDC Advantage: Advantage: More PreciseMore Precise and and

Good Charge MeasurementGood Charge Measurement DrawbackDrawback

General MICE ADC problem (No suitable QDC exists yet!) Requires Splitters More cabling

Cost: TDC +Cost: TDC + ~ 125 EUR/ch (DISC)~ 125 EUR/ch (DISC) ~ 200 EUR/ch (QDC)~ 200 EUR/ch (QDC)

~ Delay Cables (~20 m/ch)~ Delay Cables (~20 m/ch)(Assuming free splitters)(Assuming free splitters)

Gate Signal

PMT

DISCRI

CableDelay

TDC

Trigger LogicSplitter

QDC

BNC Cable

DAQ WS02 Feb 2006



Option 3. Measure the Charge with any available Option 3. Measure the Charge with any available SolutionSolution

Same Advantages and Drawbacks as 2.Same Advantages and Drawbacks as 2. Cost: TDC +Cost: TDC + ~ 125 EUR/ch (DISC)~ 125 EUR/ch (DISC)

~ ~ ?????? EUR/ch (ADC) EUR/ch (ADC)~ Delay Cables (~20 m/ch)~ Delay Cables (~20 m/ch)

(Assuming free splitters)(Assuming free splitters)

DAQ WS02 Feb 2006



Option 4. Measure the Charge with Time Over Threshold Option 4. Measure the Charge with Time Over Threshold PrinciplePrinciple

Advantage: Advantage: Easy CablingEasy Cabling and and Charge Measurement Charge Measurement (not linear though)(not linear though)

Good PrecisionGood Precision for Time Walk Correction for Time Walk Correction

NINO = ASIC developed by CERN/MIC for Alice TOFNINO = ASIC developed by CERN/MIC for Alice TOF Discriminator with ToT “proportional” to the Charge (at low Discriminator with ToT “proportional” to the Charge (at low amplitude)amplitude) Proven technique, providing 50 ps resolution Proven technique, providing 50 ps resolution (IEE TNS Vol 51-5 (2004) 1974-(IEE TNS Vol 51-5 (2004) 1974-1978)1978)

Very Cheap: 2 EUR/chVery Cheap: 2 EUR/ch

Drawback Drawback Requires Board development !

Cost: TDC +Cost: TDC + Discri. board production (~50 EUR/ch) Discri. board production (~50 EUR/ch)

PMT NINO TDC

Trigger LogicWidth ~Charge

Rising edge + Trailing edge Time Measurement

DAQ WS02 Feb 2006


EmCal Baseline EmCal Baseline Sensors: Sensors:

Hamamatsu R1355 (100 in hand, Ludovico is trying to find the other 140…)

Modified for differential output (100 Ohms)

DAQ WS02 Feb 2006


EmCAl BaselineEmCAl Baseline

Front End Electronics:Front End Electronics: Splitters Discriminators TDC: CAEN V1190 (128 ch, ECL or LVDS input) ADC: No Baseline defined yet

Number of ChannelsNumber of Channels 240 ADC 120 TDC in the BaselineTime information is useful for

Hit Position information in the direction // to the slab Suppression of muon decay background (the muons

are stopped in the EmCal and ready to decay when the next muon arrives, disturbing PID)

-> My personal feeling: TDC needed for all the channels

DAQ WS02 Feb 2006


EmCal BaselineEmCal Baseline

EmCal Connection Scheme (Half EmCal)EmCal Connection Scheme (Half EmCal)

Twisted Pair Cable

MICE LCRMICE HALL

120 PMTs Splitter

SHV Cable

HV Power Supply

12+x m

4 Discriin VME

Charge measurement

1 TDC in VME

Transformer

Delay

DAQ WS02 Feb 2006


EmCal OptionsEmCal Options

TransformerTransformer Needed to adapt the differential output of the

PMt (twisted pair, 100 Ohm) to the Coaxial (Lemo, 50 Ohm) of the Splitter.

SplitterSplitter It is possible to reuse active splitters

developed in Sofia for HARP. 200 channels available

If we have enough charge, the transformer can also be made as a passive splitter with 1 differential input and two coaxial ouputs.

DAQ WS02 Feb 2006


EmCal Charge Measurement EmCal Charge Measurement OptionsOptions

Option 1. Conventional gated QDCOption 1. Conventional gated QDC AdvantagesAdvantages

Familiar, proven design Standard VME board, well documented Support from the company (CAEN), the equipment

have a long live after MICE DrawbacksDrawbacks

The module does NOT exist yet, prototype could be delivered at the end of the year

CostCost Transformers

+ Discriminators (125 EUR/Ch)+ TDC ( 60 EUR/ch)

+ QDC (200 EUR/ch) + Delay Cables

DAQ WS02 Feb 2006



Option 2. QIE (TeV IPM)Option 2. QIE (TeV IPM) AdvantagesAdvantages

Probably no need for delay cable The transformer can probably be included in the

front end board (if we re-layout the board…) Could be used for TOF (or CKOV) charge

measurement DrawbacksDrawbacks

Development Risk Not VME readout, makes the DAQ more tricky.

Probably OK but who knows… CostCost

+ QIE FEB (??? EUR/ch)+ PC BC (??? EUR/ch)+ Discriminators (125 EUR/Ch)+ TDC ( 60 EUR/ch)

DAQ WS02 Feb 2006



Option 3. 200 MHz Flash ADCs Option 3. 200 MHz Flash ADCs

AdvantagesAdvantages No Splitter, no delay, no discri, no TDC ! Cabling very easy Equipment has a long after MICE life

DrawbacksDrawbacks Transformer -> Shaper Time resolution ?

CostCost Flash ADC (450 EUR/ch) + Shaper ( ??? EUR/ch) + VME Crate (6 kEUR)

2 ns rise time

ShaperVthr

tthr >30 ns rise time

If we can fit the rising edge, time resolution can be much higher than the 5 ns of the sampling rate.

DAQ WS02 Feb 2006


Readout Trigger Readout Trigger BaselineBaseline

6 Event Types:6 Event Types: Normal Event (RF ON or RF OFF) Start of Spill

make sure the DAQ is ready for Normal Event End of Spill

should Contain the relevant MCM data Empty Event (Pedestals) Pulser Events Cosmic/Source Events

1 trigger receiver input per Event Type1 trigger receiver input per Event Type 1 trigger receiver output per Event Type 1 trigger receiver output per Event Type

(busy) + 1 common(busy) + 1 common Trigger receiver is a simple VME IO Register Trigger receiver is a simple VME IO Register

(4 kEUR/Unit)(4 kEUR/Unit)

DAQ WS02 Feb 2006


Particle Trigger Particle Trigger BaselineBaseline

3 Particle Trigger conditions for Normal Events3 Particle Trigger conditions for Normal Events Beam: Burst x TOF0

Timing given by TOF0

Upstream: Beam x TOF1 Timing given by TOF1

Traversing: Upstream x TOF2 Timing given by Upstream

TOFn is defined by a twofold coincidence of the OR of the 12 (8) slabs of a single layer

Should be possible to add Traversing Muon, using Should be possible to add Traversing Muon, using EmCal informationEmCal information

If it’s really only these 4 conditions, there is no need If it’s really only these 4 conditions, there is no need for PLU. One simple I/O register is enoughfor PLU. One simple I/O register is enough

Option: CMAC PLU (available in Geneva). Requires CAMAC-VME or CMAC PCi Interface

For Empty and Pulser Events, the particle trigger is For Empty and Pulser Events, the particle trigger is generated by the DAQ systemgenerated by the DAQ system

For Cosmic Events, the particle trigger is set up locally For Cosmic Events, the particle trigger is set up locally by the subsystem who wants it by the subsystem who wants it

DAQ WS02 Feb 2006


SummarySummary

Tracker Baseline has to be updated Tracker Baseline has to be updated for AFEtfor AFEt

TOF Baseline is well defined and TOF Baseline is well defined and several options will be testedseveral options will be tested

EmCal Baseline is not clear (Charge EmCal Baseline is not clear (Charge measurement)measurement)

Trigger Baseline is well defined, no Trigger Baseline is well defined, no problem expected thereproblem expected there

DDAQ Hardware Architecture

Documents

Transcript of DDAQ Hardware Architecture