Upload
darren
View
21
Download
0
Embed Size (px)
DESCRIPTION
Online Summary. Jean-Sebastien Graulich, Geneva. Achievements Since CM23 Control Room Controls Detector DAQ Open Issues Summary of the summary. Achievements since CM23. Network in MLCR redesigned Target DAQ integrated into Detector DAQ EPICS Archiver and Alarm Handler implemented - PowerPoint PPT Presentation
Citation preview
MICE CM June 2009 Jean-Sebastien Graulich Slide 1
Online SummaryOnline Summary
o Achievements Since CM23
o Control Room
o Controls
o Detector DAQ
o Open Issues
o Summary of the summary
Jean-Sebastien Graulich, Geneva
MICE CM June 2009 Jean-Sebastien Graulich Slide 2
Achievements since Achievements since CM23CM23
Network in MLCR redesignedNetwork in MLCR redesigned Target DAQ integrated into Detector DAQTarget DAQ integrated into Detector DAQ EPICS Archiver and Alarm Handler EPICS Archiver and Alarm Handler
implementedimplemented EPICS Client Server installed in MLCREPICS Client Server installed in MLCR
First steps toward Unified Control SystemFirst steps toward Unified Control System
Progress in Control and DAQ Interface Progress in Control and DAQ Interface Progress in CAM and DAQ Interface with Progress in CAM and DAQ Interface with
Configuration DataBaseConfiguration DataBase Interface between DAQ and G4MICE Interface between DAQ and G4MICE
improvedimproved Problem with Event building understoodProblem with Event building understood
MICE CM June 2009 Jean-Sebastien Graulich Slide 3
MLCR Network upgrade
Craig McWaters and Mike Courthold sorted out the chaotic situation
MiceNet: 172.16.246.XXX used for bothDAQ: 172.16.246.01 -> 172.16.246.99 CAM: 172.16.246.100 -> 172.16.246.253
All DAQ machines moved to MiceNet Structure created for all future
machines Strong safety policy enforced
Strong passwords (ask your MOM) Non standard port for ssh No root ssh
+ We have now a printer in MLCR !
MICE CM June 2009 Jean-Sebastien Graulich Slide 4
EPICS Client / Server Overview
21/04/23 Imperial College 5
Network Status
• Currently have working prototype
– EPICS server connects to PCs via SSH, checks contents of ‘key’ ID file
– Client displays status of all PCs, scans at user-specified period (with ‘check now’ override)
• Need to add service checking & ‘hard’ IOC support
Responsible for System Responsible for EPICS C&M Date Due
Anyone with a PC/IOC in the MLCR/Hall James Leaver (IC) Aug 2009
MICE CM June 2009 Jean-Sebastien Graulich Slide 6
CAM/DAQ Special Case:
Target DAQ Target data is really both CAM and
DAQ Double flow is implemented Based on a single I/O software -> No issue
with format change
Waiting for the target to be online for final tests
Will require dedicated testing runs
MICE CM June 2009 Jean-Sebastien Graulich Slide 7
Archiver and Alarm Handler
MICE CM June 2009 Jean-Sebastien Graulich Slide 8
EPICS Client Server Installed
All client-side applications run on miceecserv Central installation repository greatly simplifies
configuration/maintenance/backup
MOG collates individual applications, applies updates when available from control system ‘owners’
Client control/monitoring GUIs viewed directly on miceecserv, or one of 2 ‘Operator Interface’ PCs
OPI PCs act as ‘dumb terminals’, running displays from miceecserv via SSH
miceecserv miceopi1 miceopi2
EPICS IOC
Portable CA Server
Portable CA Server
EPICS IOC
EPICS IOC
Controls Network
EPICS server applications
EPICS client applications
21/04/23 Imperial College 9
Unified User Interface
FNAL BPM Sm all
Large
TO F 0
1
2
CKO V 1
2
Tracker 1 D iffuser
Spectrom eter Solenoid
B-F ie ld P robes
AFEIIts
AFEIIt C ryos
KL Calorim eter
Tracker 2 D iffuser
Spectrom eter Solenoid
B-F ie ld P robes
AFEIIts
AFEIIt C ryos
H2 Absorbers Focus Coils
Hydrogen System
RF Cavities Coupling C oils
RF System
Drive
Contro ller
Beam Loss
Target
Beam line M agnets
Decay Solenoid
Absorbers
FNAL BPM Sm all
Large
TO F 0
1
2
CKO V 1
2
Tracker 1 D iffuser
Spectrom eter Solenoid
B-F ie ld P robes
AFEIIts
AFEIIt C ryos
KL Calorim eter
Tracker 2 D iffuser
Spectrom eter Solenoid
B-F ie ld P robes
AFEIIts
AFEIIt C ryos
H2 Absorbers Focus Coils
Hydrogen System
RF Cavities Coupling C oils
RF System
Drive
Contro ller
Beam Loss
Target
Beam line M agnets
Decay Solenoid
Absorbers
FNAL BPM Sm all
Large
TO F 0
1
2
CKO V 1
2
Tracker 1 D iffuser
Spectrom eter Solenoid
B-F ie ld P robes
AFEIIts
AFEIIt C ryos
KL Calorim eter
Tracker 2 D iffuser
Spectrom eter Solenoid
B-F ie ld P robes
AFEIIts
AFEIIt C ryos
H2 Absorbers Focus Coils
Hydrogen System
RF Cavities Coupling C oils
RF System
Drive
Contro ller
Beam Loss
Target
Beam line M agnets
Decay Solenoid
Absorbers
Large wall-mounted display
Alarm Handler
Message log
Client application
launcher
3 Standard desktop monitors
Client GUIConnected to miceecserv
Connected to miceopi1
Connected to miceopi2
21/04/23 Imperial College 10
C&M Systems OverviewSystem Owner EPICS Developer EPICS Status
Target: Drive Paul Smith; Paul Hodgeson; Chris Booth (UOS) Adrian Oates; Graham Cox (DL) Complete for current Target, but additions required for upgraded system.
Target: Controller Paul Smith (UOS); James Leaver (IC) James Leaver (IC) Not yet commenced.Target: Beam Loss Paul Smith; Paul Hodgeson (UOS); James Leaver (IC) Pierrick Hanlet (IIT) Functionally complete.Beamline Magnets Martin Hughes (RAL) Peter Owens (DL) Complete, with support requirements.Pion Decay Solenoid Mike Courthold (RAL) Adrian Oates; Graham Cox (DL) Complete, with support requirements.FNAL BPMs Alan Bross (FNAL) James Leaver (IC) Complete.TOF Maurizio Bonesini (INFN) Unknown Unallocated.CKOV Lucien Cremaldi; David Sanders (OLEMISS) Pierrick Hanlet (IIT) Not yet commenced.Tracker: Diffuser Wing Lau (OU) Unknown Unallocated.Tracker: Spectrometer Solenoids Steve Virostek (LBNL) Adrian Oates; Graham Cox (DL) DL have completed most of the design drawings, require
~£18K of capital and 0.4 man years of effort to finish the project. Funding has not yet been allocated.
Tracker: B-Field Probes Frank Filthaut (RUN) Frank Filthaut (RUN) Functionally complete.Tracker: AFEIIts Alan Bross (FNAL) James Leaver (IC); Jean-Sebastien
Graulich (UNIGE)Complete, but integration with DATE to be finalised.
Tracker: AFEIIt Infrastructure Alan Bross (FNAL) James Leaver (IC) Thus far, only Wiener PSUs have been identified. Not yet commenced. May have additional requirements.
Calorimeter Virgilio Chimenti (INFN) Unknown Unallocated.H2 Absorbers: Focus Coils Wing Lau (OU) Pierrick Hanlet (IIT) Not yet commenced, manpower unconfirmed.H2 Absorbers: Hydrogen System Yury Ivanyushenkov; Tom Bradshaw (RAL) Adrian Oates; Graham Cox (DL) DL have acquired necessary safety training and have
started evaluating PLC systems. Early stages of development, but fully budgeted for by Tom Bradshaw.
RF Cavities: Coupling Coils Derun Li; Steve Virostek (LBNL) Pierrick Hanlet (IIT) Not yet commenced, manpower unconfirmed.RF Cavities: RF System Andy Moss (ASTeC) Dimity Tettyleman (LBNL); Adrian
Oates; Graham Cox (DL)Early stages of development, but fully budgeted/accounted for by Andy Moss.
DATE Status Jean-Sebastien Graulich (UNIGE) James Leaver (IC); Jean-Sebastien Graulich (UNIGE)
EPICS server + client GUI complete, but client-side integration with DATE required.
Network Status Anyone with a PC/IOC in the MLCR/Hall James Leaver (IC) Prototype complete, requires increased functionality.
21/04/23 Imperial College 11
Tracker: Spectrometer Solenoids
• Work currently halted due to budget constraints
• 3 options– Allow DL to complete project
• Requires ~£18K capital + 0.4 man years effort
– Take DL’s current design & complete within the collaboration
• Requires ~£18K capital + ~£15.2K vxWorks developer licence + 0.6-0.8 man years effort
• Insufficient MICE manpower available…
– Discard DL’s design & start over within the collaboration
• Unknown capital requirements (likely ~£18K)
• Requires ~1.5 man years effort
• Insufficient MICE manpower available…
Responsible for System Responsible for EPICS C&M Date Due
Steve Virostek (LBNL) Adrian Oates; Graham Cox (DL) Possibly Sep 2009
MICE CM June 2009 Jean-Sebastien Graulich Slide 12
Detector DAQ Issues
Three priorities were defined in CM23
Limit on Data Size Event Building CAM/DAQ Interface
Limit on Data Size From both DAQ and Monitoring First action on the data size itself ! Simply reducing the number of samples per
fADC channel -> factor of ~5 Requires fADC firmware upgrade First attempt failed
MICE CM June 2009 Jean-Sebastien Graulich Slide 13
Event building problem
Vassil Verguilov developed software to investigate the problem
The Trigger Time Tag and Bunch ID from the TDC can be used to understand what happened
Each TDC records Trigger Time Tag (TTT)
27 bits 800 ns LSB ~ 100 s full range
Bunch ID (BID) 12 bits 25 ns LSB ~ 50 s full range Clock running Independently
It is possible to retrieve the absolute time of the particle trigger
W.r.t. to board reset (boards not yet synchronized) With some ambiguities… and complications …
MICE CM June 2009 Jean-Sebastien Graulich Slide 14
CAM/DAQ Interface
Two ways link EPICS should know the DATE status A summary of EPICS data should be inserted
in the online data stream The run should stop automatically when the
CAM goes in severe alarm state
Date Status Monitoring James Leaver has provided
EPICS Server, EPICS Client,Test Software, Status Display
Next step: A small piece of code reading real status from DATE
MICE CM June 2009 Jean-Sebastien Graulich Slide 15
CAM data Online
Principle re-discussed CAM data appended at the end of the run file
vs CAM data inside the Spill Data
The issue: backward/forward compatibility of data format
Solved by James Leaver An XML format description file will be provided
and appended to the run file Require some “intelligent” unpacking code
CAM data will be stored Spill by Spill(Took just ~4 years to understand that AB was
right…:-)
MICE CM June 2009 Jean-Sebastien Graulich Slide 16
Trigger IssuesTrigger Issues
No modification since last CMNo modification since last CM The effect of trigger condition choice The effect of trigger condition choice
on TOF calibration has been clearly on TOF calibration has been clearly understoodunderstood
Flexibility has a huge cost ! Flexibility has a huge cost ! TOF Calibration constants depend on the TOF Calibration constants depend on the
trigger sourcetrigger source Do we really need so many trigger conditions Do we really need so many trigger conditions
??
Decision on the future of GVA1Decision on the future of GVA1 It will stay in place at least until Step 2 is
complete
MICE CM June 2009 Jean-Sebastien Graulich Slide 17
Online SoftwareOnline Software
Unpacking library has been upgradedUnpacking library has been upgraded New fADC (V1731) for CKOV Some bug fixed
Interface with G4MICEInterface with G4MICE Vassil Verguilov has optimized the Date Reader for both
clarity and efficiency Spill structure implemented in G4MICE (for scalar data,
number of particle triggers per DAQ event, etc.) David Adey reproduced online histograms with G4MICE.
Analysis is under way See Software session for more details
Know issues (No change since CM23)Know issues (No change since CM23) Online monitoring GUI tends to get stalled at the end of a
run Need better Online plots for fADC
Current plots found not efficient in detecting broken channels
MICE CM June 2009 Jean-Sebastien Graulich Slide 18
Front End ElectronicsFront End Electronics
TOFTOF Cabling for TOF TDCs synchronization is finished (Internal
clock and trigger distribution). Will need some dedicated beam time for testing TOF2: We miss some electronics.
Shapers/Splitters: Production finished. Under evaluation TDC: One board missing FADC: 3 boards missing
EMR: Electron Muon Ranger (aka SW)EMR: Electron Muon Ranger (aka SW) Prototype being tested by Michela Prest and Erik Valaza
MICE CM June 2009 Jean-Sebastien Graulich Slide 19
Schedule Schedule MilestonesMilestones
For DDAQFor DDAQ Target DAQ included in Detector DAQTarget DAQ included in Detector DAQ mid Februarymid February DAQ migration to the DAQ/Control networkDAQ migration to the DAQ/Control network March 06March 06 Installation of the Online Reconstruction FarmInstallation of the Online Reconstruction Farm March 06March 06 Upgrade of fADC firmwareUpgrade of fADC firmware March 06March 06 -> July 09 -> July 09 TOF2 Shaper/Splitter ProductionTOF2 Shaper/Splitter Production mid Marchmid March need need
testingtesting CAM data in Online Data StreamCAM data in Online Data Stream end of April end of April -> June -> June
0909 Tracker integrated in DAQ and OLMTracker integrated in DAQ and OLM end of April end of April -> July -> July
0909 TOF TDC Clock SynchronizationTOF TDC Clock Synchronization end of April end of April -> July 09-> July 09 DAQ reviewDAQ review May 2009May 2009 ->-> June 09June 09 Burst Gate Signal in the Trigger SystemBurst Gate Signal in the Trigger System need beamneed beam
Approval of SW/EMR Front End ElectronicsApproval of SW/EMR Front End Electronics In Proto phaseIn Proto phase Production of SW/EMR Front End ElectronicsProduction of SW/EMR Front End Electronics September 2009September 2009
21/04/23 Imperial College 20
Items Which Require Action!
• Must find resources within MICE community to complete EPICS C&M systems for– Time of Flight System
– Diffuser
– Calorimeter system
• Must resolve issue of funding for DL’s work on the Spectrometer Solenoids
• PH’s contract expires very soon…– He is essential to success of Online Group
– If he is not reemployed, we won’t have: Alarm Handler, Channel Archiver, remote parameter monitoring, C&M systems for CKOV, Focus Coils, Coupling Coils, etc.
MICE CM June 2009 Jean-Sebastien Graulich Slide 21
SummarySummary-------------- DDAQ --------------------------------- DDAQ -------------------
Network in MLCR is under ControlNetwork in MLCR is under Control Target DAQ is integrated into online data streamTarget DAQ is integrated into online data stream Interface between DAQ and G4MICE improvedInterface between DAQ and G4MICE improved Problem with Event building understoodProblem with Event building understood DAQ review will happen on Thursday June 4 DAQ review will happen on Thursday June 4
*2009**2009*
-------------- CAM ----------------------------------- CAM --------------------- Infrastructure design has been re-exanimatedInfrastructure design has been re-exanimated EPICS Client Server installed in MLCREPICS Client Server installed in MLCR EPICS Archiver and Alarm Handler implementedEPICS Archiver and Alarm Handler implemented
First steps toward Unified Control SystemFirst steps toward Unified Control System Progress in Control and DAQ InterfaceProgress in Control and DAQ Interface Man power is critical – we should not let go the Man power is critical – we should not let go the
rare & precious expertise we have nowrare & precious expertise we have now