Telescope Plans Proposed Work Packages for the Timepix Telescope Richard Plackett CERN 8 th December 09 Outline Telescope Outline Timeline and Dependencies Technical Details and Progress ID: 238151
Download Presentation The PPT/PDF document "Timepix" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Timepix Telescope Plans
Proposed Work Packages for the
Timepix
Telescope
Richard
Plackett
CERN,
8
th
December
09Slide2
OutlineTelescope OutlineTimeline and DependenciesTechnical Details and ProgressMore Technical DetailsFurther Technical Details
Questions?!
I will discuss the DUT systems in far less depth than they deserve…
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide3
Telescope OverviewCompletely compatible with AIDA proposalTasks to complete the
Timepix
Telescope
Timescale : To work in April
Telescope ArmsRELAXD with Timepix (NIKHEF/PAN)Interface
BoardNew Chip BoardsMechanicsServices – PSU, timing outputs, cooling
PMTs/ScintillatorsPixelman/other software (Prague)
Timing UnitFPGA Development BoardEnclosure (NIM)
Programmable Switch?Software
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide4
Very Parallel ProjectIndependent Packages (i.e. will work with current telescope)RELAXD ReadoutTiming UnitNew Chip Boards
Pixelman
software updates
Coupled Packages (need other parts to be useful)
Mechanics needs RELAXDServices heavily linked to MechanicsScintiallators linked to Mechanics and Timing unitTiming unit software linked to timing unit
Programmable switch linked to RELAXD and Timing UnitDUT(s)Mechanics
Electronics and Readout
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide5
Timing and ResponsibilityPackage
Time
Depends on..
Who
Arm~3 months
RELAXD2 months
NIKHEFInterface Board
2 monthsRELAXDCERN (Medipix)
Chipboard2 months
CERN (Medipix)
Mechanics2 monthsRXD,
Int, ServServices
2 monthsRXD, Int,
MechPixelman1 month
CERN (Medipix)PMTs2 weeksMech
,
Timing
Timing
Unit
~3
months
FPGA
Program
3 months
Enclosure
2 weeks
FPGA program
Software
2 months
FPGA program
DUT Systems
Santiago
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide6
TimelinePackage
Dec
Jan
Feb
MarchArm
AssembleRELAXD
Interface BoardBegin Design
Finalize DesignProductionDeliveryChipboard
DesignProductionDeliveryMechanics
SpecifyDesignProductionDelivery
ServicesSpecifyDesignProductionDelivery
PixelmanWorkingWorkingComplete
PMTsPurchase
Timing Unit
FPGA ProgramSpecify
Program
Program
Test
Enclosure
Specify
Purchase
Construct
Software
Specify
Program
Program
Test
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide7
Timepix
Telescope Overview
RELAXD
Timing Unit
DUT+ readout
RELAXD
Timing Signals collected by a timing unit provide flexible software synchronisation
Telescope arms and timing unit read back to PC via Ethernet
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide8
Arms
RELAXD
Interface
Power
Cooling
Photomultipliers
Enclosure – light shading but not light tight
Local 12V Power Supply
Exchange of NIM and
ethernet
signals
New Chipboards
New Interface Board
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide9
Timing UnitProvides Syncronisation signals depending on operation modeSends and Records shutters and scintillator coincidencesUsed to integrate timing of external devices
Will drive Data back over Ethernet
Control will
eventually be integrated into Pixelman
pluginCould be combined with programmable switch into a single box or module
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide10
Flex or Cable link to RELAXD interface board
‘Finger’ Board PCBs
Data IO – Backward compatibility with current chipboards SCSI connector
Link to current MUROS and USB systems for testing
Long distance
rad
hard link to a readout system for SPS
Robust
Samtec
connector incorporating HV
Low mass ‘finger’ to reduce scattering with chip cutout
Accurate mounting holes
CERN
rad
hard power regulators
Removable test pulse mezzanine board
80mm?
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide11
RELAXD150Hz quad readout under development by NIKHEFMedipix2 and Medipix3 readout demonstratedTimepix readout and rate optimisation
underway
External shutter workingPrototype at CERNPower Regulator Board generates digital analog and sensor voltages from 12V supply
Interface Board/Quad
Readout Board with Ethernet
Power Regulator Board
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide12
Interface and Flexible ConnectorsRELAXD designed to read a Quad not 4 independent chipsNeed an interface and short flexible linksNeeds to be completed rapidly as potential problems with signal transmission need to be
overcome
CERN group will take this on as very coupled with the chipboard design.
Interface Board
Readout Board
Power Regulator
Flexible Couplings
Timepix
Chip Boards
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide13
ARM MechanicsRequired to mount, protect and position four telescope planesChipboards, RELAXD , Services and PMTs / Scintillators need to be integratedChipboard mounting needs to be accurate
Mounting to external hardware to be as universal as possible
Needs to be relatively robust
External alignment targets and references to chip position (clear plastic targets?)Needs to shade from
light but not be ‘dark’Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide14
Services (electronics etc for Arms)The Arm Mechanics need to support some servicesCooling – fan cooling will require airflow, possibly temperature monitor and external port?Timing Signals – one or more timing signals from chipboards/interface routed to external port
Trigger/shutter signals to RELAXD will require external ports
Power supply 240V mains to 12V dc for RELAXD external power input port
Ethernet – RELAXD ethernet
may need to be rerouted to external port depending on design of mechanicsLots of small things heavily dependent on RELAXD and MechanicsNot all a strongly required but will make it into a nice package
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide15
PMTs + ScintillatorsEach arm ought to be capable of mounting a PM pair to provide fast timing and ‘triggers’Will take HV from NIM crate and send standard PM signals backCoincidence area to be compatible with Telescope Active area
Relatively little work as mostly off the shelf parts
Consider making separate mechanics add-on for arms
Coupled to mechanics and timing unit (to some level)
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide16
DUT IssuesWill require a stage system similar to the current one to rotate and align DUTsSome programming work needed on current system to make it safe to use with various mechanicsMechanics development for new sensors and chipboardsIntegrating readout for
Timepix
/Medipix3 based DUTs easy with 3
rd RELAXD board, Beetle based devices will be harder and will make heavy use of timing unit for
synchronisationCold box for irradiated DUTs and integration with arms…Multiple DUTs for fast changeovers as we hope to be running ~100 times faster than last summer…
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide17
Timing Unit FPGA BoardLooks possible with FPGA Development board Already purchased Stratix2 with Nios2 embedded processor architecture to drive IP linkNeed to provide accurate timing information
20ns quite easy, 1ns possible
Better than 1ns needs to drive CERN Nino chip (future)
Ethernet ReadoutEventually will drive
shutter signals to control system remotely from pixelmanProbably 3 months work to get to working system
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide18
Timing Unit SoftwareSome simple software required to save data received from timing unit for basic operationMore advanced software to control modes of system easily should probably be integrated to Pixelman as a
plugin
Need to define modes of operation early to ease development...
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide19
Pixelman UpdatePixelman is being updated by Medipix group to better handle multiple chipsDisplay – docking multiple previews and controlsOperating system compatibility – Windows, Mac, Linux
Data File Structure – fewer files for each run
Moving to RELAXD is also a benefit as only 2 systems to start
Some specific Testbeam Plug-in would be helpful but not essential
Work is well underway now and is being finalised and tested
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide20
Programmable Switch or PC ?It would be beneficial to only run one ethernet cable back to the readout PC in the barracksRELAXD and Timing Unit will probably require a programmable gigabit switch to send data as
ethernet
protocols are not very advanced
Programmable switches like this can be purchased off the shelf but we need to investigate more to ensure this solution would
workWould it be better to just have a PC with 4 Ethernet cards in the area as before?
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide21
Question? - Operational ModesThere are a number of ways we could run the timing unit, we can process scintillator coincidence data in real timeWe can drive the shutters with a signal generated by the FPGA
We can
suppress
data outside beam spillWe can run for a predetermined number of triggers and close the shutterWe can run a burst of triggers and then wait
We can run with an external shutterWe can program delays and set modes remotelyThese options give a large number of permutations which need to be discussedWhat do we want to do with the system?
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
ModesSlide22
ConclusionsThere is still a lot to doMany tasks are decoupled from one another
Getting the full thing for April will be very hard work
At all times we will always have what we had before (and more)
We cant go backwards
Delivering any of the decoupled systems brings a benefit to the quality of the system and the data.
Overview
Arms
Chipboard
RELAXD
Interface
Mechanics
PMTs
DUTs
Timing Unit
Pixelman
Data Flow
Modes