DAQ WS02 Feb 2006 Jean-Sébastien Graulich Slide 1 Does IPM System Matches MICE needs ? Does IPM System Matches MICE needs ? Personal Understanding and Remarks Personal Understanding and Remarks o General Considerations o Front End Boards o Timing Cards and Fan Out o Buffer Cards o Total cost summary o Personal Conclusion Jean-Sebastien Graulich, Geneva
9
Embed
DAQ WS02 Feb 2006Jean-Sébastien GraulichSlide 1 Does IPM System Matches MICE needs ? Personal Understanding and Remarks o General Considerations o Front.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DAQ WS02 Feb 2006
Jean-Sébastien Graulich Slide 1
Does IPM System Matches MICE needs ?Does IPM System Matches MICE needs ?Personal Understanding and RemarksPersonal Understanding and Remarks
o General Considerations
o Front End Boards
o Timing Cards and Fan Out
o Buffer Cards
o Total cost summary
o Personal Conclusion
Jean-Sebastien Graulich, Geneva
DAQ WS02 Feb 2006
Jean-Sébastien Graulich Slide 2
General General ConsiderationsConsiderations
Disclaimer and Acknowledgment: Disclaimer and Acknowledgment:
This is only my personal understanding This is only my personal understanding of the presentations given by of the presentations given by Kwame and Rick. Please react if you Kwame and Rick. Please react if you see mistakes. see mistakes.
This is not a summary of the talks, just This is not a summary of the talks, just a list of things I think it’s important a list of things I think it’s important to keep in mindto keep in mind
Many thanks to Kwame and Rick for Many thanks to Kwame and Rick for their pedagogical effortstheir pedagogical efforts
DAQ WS02 Feb 2006
Jean-Sébastien Graulich Slide 3
General General ConsiderationsConsiderations
The system is designed for a PC based DAQThe system is designed for a PC based DAQ
The system is not yet fully commissioned The system is not yet fully commissioned
4 different cards are needed4 different cards are needed
Among them 2 are PCI cards for which there is no Linux Among them 2 are PCI cards for which there is no Linux driver availabledriver availableAn outdated driver exists for the Data Buffer cards
The Front End Board is the QIE Card. The Front End Board is the QIE Card. It integrates the charge received on the input over a given time (the sampling period) and transfer it to the Buffer Card in a free running continuous mode. No dead time.
Data in coincidence with the “Gate” signal is tagged (one bit in the header)
The Buffer Card is also doing Data SparcificationThe Buffer Card is also doing Data SparcificationIt is possible to store only the data which is tagged as being in coincidence with the “Gate”
It has a very large buffer memory (enough to store 400 ms of continuous sampling at 15 MHz)
The Timing Card is also doing the Control for the QIE The Timing Card is also doing the Control for the QIE CardsCards
It is doing the initialization of the QIE Card, under request from the PC
It provides the sampling clock (here 15 MHz, up to 35 MHz possible)
It has two external TTL (Lemo) inputs. The interpretation of the input signals is determined by firmware. They can be used as gate inputs.
Input Signal
Timing card(PCI)
Timingfanout
QIE cards(16x 8 ch)
Data Buffer(2*8 ch)
(PCI)
Host PC (LabView)
“Gate”
DAQ WS02 Feb 2006
Jean-Sébastien Graulich Slide 4
QIE Cards (FEB)QIE Cards (FEB)
It requires a custom backplane (call it It requires a custom backplane (call it crate) for power supply (6V and 9V). crate) for power supply (6V and 9V). Power consumption is substantial: 12 W / board (1 board is 8 ch)
2 crates are needed for EmCal (16 boards/crate, that is 128 ch/crate) -> ~ 200 W/crate
The input connector is very non-standardThe input connector is very non-standard
The Input signal is between 1 fC and 16 The Input signal is between 1 fC and 16 pC (14 bits)pC (14 bits)For EmCal we want flat cable connectors and signal attenuation. Ideally, we would implement there the transformers doing analog splitting with one output going to the QIE and one output coming out to the discriminators.
-> Some redesign has to be foreseen. Not trivial because the card is already crowded. We could go to 9U
Choosing less radiation resistant components could lead to some savings in Cost and Power consumption
DAQ WS02 Feb 2006
Jean-Sébastien Graulich Slide 5
Timing CardsTiming Cards
Could/Should be used as it is Could/Should be used as it is Including the LabView User Interface for the Control
Firmware has to be adapted to our Firmware has to be adapted to our trigger schemetrigger schemeOne input signal would be the normal gate. The firmware used for the calibration setup looks similar to what we’ll need.
The second input could be used for measuring the background induced by muon decay (random gate In Spill but out of Burst) without having to readout the tracker.
DAQ WS02 Feb 2006
Jean-Sébastien Graulich Slide 6
Timing Fan out CardTiming Fan out Card
Not a mere fan-outNot a mere fan-out Input an output connectors are not the
same (Cat5 input, Cat3/4 output
Could/Should be used as it is Could/Should be used as it is A simpler version can be done for a test if we need only one channel -> Simple connector adapter
Requires power supply tooRequires power supply too
DAQ WS02 Feb 2006
Jean-Sébastien Graulich Slide 7
Buffer CardsBuffer Cards
Could/Should be used as it is Could/Should be used as it is One part (Optical Interface) is not available on the market anymore -> Some simple redesign is needed
There is probably no need for new There is probably no need for new firmware firmware
Requires 64-bit 66MHz PCIRequires 64-bit 66MHz PCI That is a dedicated PC
Not compatible with the VME/PCI32 Interface CAEN V2718!
VME/PCI64 Interface exists. ~ same price but
it’s annoying to have different interfaces.
There is no driver for LinuxThere is no driver for Linux -> Large amount of work required to make it compatible with the DDAQ System.
DAQ WS02 Feb 2006
Jean-Sébastien Graulich Slide 8
Cost & Work Cost & Work EstimateEstimate
There is an oral agreement that Fermilab would endorse the Engineering There is an oral agreement that Fermilab would endorse the Engineering work involved work involved
I am not able to estimate that workload anyway
Enough QIE chips and GOL chips are available for free Enough QIE chips and GOL chips are available for free We would have to pay only for the parts and the board production