MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 1 DAQ Overview o Terminology o Requirements o Front End Electronics o Trigger System o DAQ Hardware and Software o Event Building and Data Flow o Interfaces: Monitoring, Controls, G4MICE, and Configuration DB o Project Plan: Human resources and Time Line Jean-Sebastien Graulich, Geneva

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 2 DAQ Terminology   ISIS / MICE / DATE Jargon   Isis Cycle: The injection and acceleration cycle of ISIS. It is 20 ms long (50 Hz).   Machine Start or MS: This is the pulse used for the ISIS synchronization   Spill or Spill-Gate: The time window during which the MICE Target is crossing the ISIS beam. The Spill cycle is driven by the cycle of the MICE target   Burst: The ~100 ns time window during which muons can be expected in the MICE detectors. It corresponds to the time it takes for a proton bunch in ISIS to cross through the MICE target   DAQ-Trigger: Signal triggering the readout of the FE-electronics modules of the MICE detectors. By definition, one DAQ-Trigger corresponds to one DAQ-Event   Particle-Trigger: Signal generated when the desired Trigger Condition is met. It is distributed to the sub-detectors Front End Electronics and initiates the digitization of the data therein. By definition, a Particle-Trigger corresponds to one Particle-Event

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 3

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 4

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 5 Requirements   See the full list of system requirement in MICE Note 222.   The main requirement is on the rate: The DAQ system should allow acquiring data for up to 600 Particles in a 1 ms spill repeating at 1 Hz.   Important consequences Detector data Readout must be performed at the end of the spill Data has to be buffered in the FEE This is because the readout of 1 particle event takes several 100 µs…

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 6   Digitization dead time The DAQ dead time after a Particle Trigger should not exceed one burst after the one that generated it. Average Time between 2 muons is 1.7 µs Conversion time for conventional ADC > ~3 µs Critical for EMCal and TOF Solution: Flash ADC after Signal stretching Even if the event buffer was large enough, conventional ADC can not collect 600 muons/ms Requirements (2)

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 7 Charge Measurement in MICE (except Tracker) 100 MHz Flash ADCs (CAEN 1724)   Advantages No Splitter, no delay cable, no discri, no TDC ! Cabling very easy   Need custom made shaper   Strong impact on data size 2 ns rise time Shaper V thr t thr >30 ns rise time By fitting the rising edge, time resolution is much better than the 10 ns of the sampling rate (~1 ns obtained).

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 8 Requirements (3)   Event Size DAQ Event Size is quite large because it corresponds to 600 muons The DAQ system should be able to deal with DAQ event size up to 60 MB (up to 10 MB per sub-event) Means ~ 60 MB/s data transfer rate (readout between spills)

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 9 Requirements (4)   Stability and Reliability   Flexibility   Partitioning Following MICE Installation plan   Allow calibration events between spills   AND maintainable by a single person…

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 10 Front End Electronics o TOF o KL o CKOV o Tracker o EMR o Target Position and Beam Loss

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 11 TOF FEE   Sensor: PMT (Hamamatsu R4998) Signal transmission: Single ended, 50 Ohm, coax cable (RG213)   Number of Channels 40 (TOF0) + 28 (TOF1) + 40 (TOF2) = 108 ch   Main Constrain Time Resolution -> Time-walk correction   TDC CAEN V1290, 32 ch Large Event Buffer ECL input   Discriminator Lecroy Channels ECL output Twisted Pair Input (110 Ohm)   Need Signal splitting for charge measurement

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 12 KL FEE   Sensor: PMT ( Hamamatsu R1355 ) Signal transmission: Differential, 120 Ohm, Twisted pair cable   Number of Channels 42 ch   Main Constrain Charge measurement Time ~ 1 ns   Flash ADC (WFD) CAEN V MS/s, 14 bits Best commercial deal Single Ended, 50 Ohm Input

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 13 Shaper development   Splitter designed in Geneva (P. Bene) Passive splitter, shielded, impedance matched (minimal signal loss) 16 x Lemo, 50 Ohm Input 2 twisted pair outputs, 34 contacts connectors, 110 Ohm to Discriminator 120 Ohm to Shaper   Shaper designed in Sofia (I. Rusinov) 120 Ohm Twisted Pair Input, 34 contacts connectors 16 Lemo, 50 Ohm, single-ended outputs to fADC Single-Ended or Differential mode jumper selectable P-Trg Signal TOF PMT DISCRI Shaper SE Mode TDC Trigger Logic Splitter fADC BNC Cable 50 Ohm KL PMT Twisted pair Cable Shaper Diff Mode fADC

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 14 Shaper and Splitter

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 15 F(t) = offset – norm (  0 / (  0 -  ) ) [ ( e -(t-t 0 )/  -e -(t-t 0 )/  0 ) + (t-t0) /  e -(t-t 0 )/  ] Shaper Output Need for better tuning of baseline restorer Used for individual baseline evaluation - Signal shape well understood - Time resolution better than 1 ns Time (sample)

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 16 CKOV FEE   Sensor: PMT (8”) Signal transmission: Single ended, 50 Ohm, coax cable (RG58)   Number of Channels = 8 ch   Main Constrain Rate (no segmentation) Small charge   Flash ADC (WFD) CAEN V MS/s, 8 bits Single Ended, 50 Ohm Input   No Shaper !

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 17 Tracker FEE 4 VLSB Boards in VME Crate = Data Buffer MICE HALL 1 Cryo-Cooler = ½ Tracker 4 AFE2-t Boards Fibers from VLPCs Digital Signal MIL1553 (Control) Read out PC In Control Room Optical Fiber   Custom Made Digital Data Buffer VLSB = VME LVDS CERDES Buffer MICE defined Data format Measure Both discriminated time and Zero suppressed Charge   = 8198 channels

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 18 EMR FEE   Still in prototyping phase   50 layers x 59 bars = 2950 channels   Digital information only only -> 2950 bits (100 words = 400 bytes) In custom made VME buffer (similar to tracker)   Charge measurement per layer only 50 fADC channel Direct connection, similar to CKOV

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 19 Target Monitoring FEE   The Target system records Target dip trajectory ISIS Beam loss monitors analog signals   Duplicate data flow CAM DAQ   Max 8 channels Read out once per spill   Nuclear Instrument PCI I/O card PCi6254 in target PC   Marginal in term of data size

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 20 Data Size   TOF TDC: Maximum 108 hits, 4 bytes/hit 432 bytes/pt   TOF fADC: 60 samples per channel 13 kBytes/pt   KL fADC:60 samples per channel 6 kBytes/pt   CKOV fADC300 samples per channel, 1 byte/sample 2.4 kB/pt   Tracker: (w/o zero suppression) 5536 bytes per tracker/pt 10.8 kB/pt   TOTAL w/o EMR: ~33 kB/pt 16.5 MB/Spill Maximum 16.5 MB/s   Electron Muon Ranger (Coming up after spring 2010) Digital: 0.4 kB/pt or 2 kB/pt if TDC information fADC: 300 samples/ch, 1 byte/sample: 15kB/pt TOTAL for EMR: 21 kB/pt -> 10.5 MB/Spill -> 10.5 MB/s   TOTAL at completion ~ 30 MB/s

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 21 Trigger System o Particle Trigger o Timing o Logic o DAQ Trigger o Spill Gate

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 22 Particle Trigger   Triggers the digitization of the signals arriving at the Front End Electronics Should arrive a few  s around the signal Distributed to all FEE boards We expect about 500 Particle triggers for 1 DAQ trigger   The timing of the trigger should be given by the burst Delay TOF0, TOF1 TOF2 such that they arrive approximately at the same time in the trigger logic Make the TOF logic pulses ~200 ns long Make the Burst Gate narrow and Delay it such that it arrives more than 100 ns after the TOF signals   All single raw time distribution will be ~ 100 ns wide   Problem for beam commissioning ! Protons have very different timing Will use TOF0 timing until stable muon running is established

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 23 Particle Trigger timing TOF1 Burst Gate TOF0 TOF0  TOF1  Burst Gate 200 ns GVA1 hit Burst Gate (made narrow)

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 24 TOF0 Logic  TOF0 has 40 channels 10 horizontal slabs 10 vertical slabs 2 PMts per slab Because of hardware constraints, We also define TOF0_H: OR over 8 central H slabs TOF0_V : OR over 8 central V slabs TOF0_EXT: OR over the 4 frame slabs TOF0 signal is the OR over all the slabs of the AND of the two pmts of each slab

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 25 TOF0 Logic implementation TOF1 will be included in the trigger in a similar way

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 26 Trigger Electronics  Crate Layout DAQ Trigger distribution Scalars and particle trigger NIM Logic TOF discriminators and trigger CAMAC Logic Interface w/Target & Spill Gate Shapers GVA Discri, KL cosmics trg TOF & CKOV KL

MICE Ready Target Ready RF Ready DAQ Ready Gated Machine Start Spill Request Target Trigger Protons on target RF Trigger RF Power DT Gate DAQ Trigger Target Delay RF Delay DT Delay 20 ms Extraction Validated Machine Start MICE Systems synchronization

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 28 DAQ Trigger Four types of DAQ triggers - PHYSICS - CALIBRATION - Start of Spill - End of Spill Event Type is tagged in the data header Used to impose a well- defined sequence of events and online check of synchronisation

Fixed delay ~ few ms Fixed delay Software Check: No overlap with SOS busy (otherwise stop with error) Depends on Data Size ~ 1 s VMS Particle Triggers Fixed width ~ 1 ms Physics Event DAQ Trigger Fan out to LDCs EOS Trigger SOS Trigger DAQ Idle = DAQ Ready DAQ trigger distribution Target Trg

CAEN V977 Channels 0-7 configured as Flip-Flop, reset by software

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 31 DAQ Spill Gate  Detailed documentation of the procedure to adjust the DAQ Spill Gate is available Machine start after delay in the target system Spill Gate is always aligned to extraction

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 32 Machine Start DAQ Gate ~ 3.5ms Hits in GVA1 DAQ Spill Gate Sync

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 33 DAQ Hardware and Software o Hardware overview o Software Framework o Event Building o Data flow

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 34 D-DAQ Hardware Overview

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 35 Software Framework   MICE has decided to use the ALICE DAQ software framework The ALICE DAQ software framework, DATE, will provide us with the necessary EventBuilder tool: Subevents collected by different processors have to be synchronized and put together before storage DATE has much more functionalities than what MICE needs Previous Experience (from HARP) Experiment could be used   Agreement with ALICE collaboration included 6 months hands on training with real contribution to the software (2006)   Currently using DATE version 6.05 (20/06/2007) Historical reasons Upgrade in July and then froze it until the end of the experiment

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 36 DATE Vocabulary   LDC : Local Data Concentrator The PC connected to the VME crate via the PC-VME Interface   GDC : Global Data Collector Event Builder   Event DATE Event = DAQ Event !!! It contains data for several Particle Events (~500)   Event Type Tag attached to the event depending on which trigger receiver‘s input has been used

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 37 DATE Readout Process   Two processes running in each LDC The readout process waits for a trigger, reads out the front-end electronics, and fills a FIFO buffer with the sub- event data The recorder process off-loads the FIFO and sends the sub- event data to one (or several) GDC over the network   Each LDC contains a set of Equipments Equipment =~ 1 Vme board (in MICE) Each equipment has its own set of routines for its initialization and readout. Adding an equipment is done without recompiling all DATE Equipment configuration data is saved in MYSQL database (but not archived)

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 38 DATE Readout Algorithm 5 user routines have to be implemented (XXX is the name of the equipment) ArmHwXXX Executed at the beginning of the Run Allows initialization of the board AsynchReadXXX Executed constantly even when there is no trigger Don’t use ! EventArrivedXXX Used only if the equipment needs to trigger the readout ( Trigger Receiver) ReadEventXXX That is the readout itself DisArmHwXXX Executed at the end of the Run   General algorithm for equipment readout:

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 39 Readout Code   Available and tested for TDC V1290A (TOF) FADC V1724 (TOF and KL) FADC V1731 (CKOV and EMR) Scaler V830 Trigger Receiver I/O V977 NI I/O PCi 6254 (Target) VLSB (Tracker) Trailer (special equipment handling the release of the busy)   Todo EMR Front End Board

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 40 DATE Data Format   The data sent by the equipment is just wrapped with a LDC header (+ a GDC header if used)   The data format in the payload is defined by the manufacturer of the equipment ! (we will stick to 32 bits words)   DATE Header format defined in a header file event.h This file contains all the information the offline codes needs to know about DATE Data from the equipment ->

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 41 Interfaces o Online Monitoring o Reconstruction o Controls o Configuration Data Base

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 42 Online Monitoring   Interface is based on DATE monitoring facility   A simplified version independent from DATE (running only from files) exists   Online monitoring process produces histograms and makes then available on demand on a ROOT socket Socket handling implemented has a thread to avoid clashes with DATE monitoring semaphore system   Online Monitoring GUI is just a ROOT macro allowing the user to request histograms   See Linda’s talk

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 43 Reconstruction   Unpacking has to take into account the DATE format   Common library for online monitoring and Reconstruction   Dependency limited to the event.h file   All implemented in C++

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 44 Control And Monitoring   Two ways Interface CAM should know the DATE status A summary of CAM data should be inserted in the online data stream The run should stop automatically when the CAM goes in severe alarm state   An EPICS Client uses the DIM layer of DATE to make the status available   A special DATE equipment will be implemented to readout some CAM data identified as relevant for Offline Analysis Currently under development The same equipment will stop the run if a dedicated CAM variable is set   More details in James’ and Pierrick’s talks

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 45 Configuration Database   The Equipment Parameters (configuration of the VME boards) will be stored in the configuration database   A run configuration file (in xml format) will be saved in the CDB each time a run is started The CAM monitors DAQ status When the run starts, the CAM calls an API function saving the configuration file automatically   More details in David’s talk

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 46 Project Plan o Organization and Resources o Time Line

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 47 Organization  Online Group  Responsibilities Detector DAQ : J.S Graulich (80%) (Group Leader) Control and Monitoring: J. Leaver (50%) P. Hanlet (20%) Online Reconstruction: L. Coney (20%) MLCR Manager: C. MacWaters (20%)  Other Contributors Vassil Verguilov (DAQ) David Forrest (Config DB) Henry Nebrensky (GRID – Data Flow) Mike Courthold (Network Supervisor)

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 48 Time Line   Project started from scratch in July 2005   First data taken acquired in March 2008   The DAQ is running stable with some limitations Trigger selection by hardware Limit on DAQ Event Size is currently 30 MB   July 2009: Integration of Tracker in DAQ Include CAM Data in Online data stream (at the Spill level)   Current Priority: Problem with DAQ trigger distribution system Cause Event building desync Need to include that in the Online Monitoring   Include EMR   Include RF System !

MICE DAQ and Controls ReviewJean-Sebastien GraulichSlide 49 The Cat only grinned when it saw Alice. It looked good- natured, she thought: still it had VERY long claws and a great many teeth, so she felt that it ought to be treated with respect. `Cheshire Puss,' she began, rather timidly, as she did not at all know whether it would like the name: however, it only grinned a little wider. `Come, it's pleased so far,' thought Alice, and she went on. `Would you tell me, please, which way I ought to go from here?' `That depends a good deal on where you want to get to,' said the Cat. `I don't much care where--' said Alice. `Then it doesn't matter which way you go,' said the Cat. `--so long as I get SOMEWHERE,' Alice added as an explanation. `Oh, you're sure to do that,' said the Cat, `if you only walk long enough. ' Alice felt that this could not be denied, so she tried another question. `What sort of people live about here?' `In THAT direction,‘ the Cat said, waving its right paw round, `lives a Hatter: and in THAT direction,' waving the other paw, `lives a March Hare. Visit either you like: they're both mad.' `But I don't want to go among mad people,' Alice remarked. `Oh, you can't help that,' said the Cat: `we're all mad here. I'm mad. You're mad.‘ `How do you know I'm mad?' said Alice. `You must be,' said the Cat, `or you wouldn't have come here.'