ETD/Online Report D. Breton, U. Marconi, S. Luitz WS goals session – Elba Meeting May 31 th 2010 ETD/Online - Super. B Elba Workshop - May 2010
Technical choices for the TDR : front-end • There is a certain amount of common questions for the different front-end electronics: • Radiation • We need a radiation map in order to fix the rules for the electronics mitigation • We have to be sure that the technologies chosen can handle radiation • We have to define rules for selecting the commercial components • Power supplies • Do we intend to use radiation-tolerant supplies on the detector (short cables but more expensive supplies) or remote standard supplies (cheaper supplies but long cables) ? • Do we want a common supply policy at the detector level ? • Grounding and shielding • We have to define the rules • Our preference is a strong common ground • Electronics integration on the detector • When designing the system, it is important to think of: • How easy it will be to change a board (access, connectors, …) • Ensuring there will be no extra pressure on the connectors (for instance avoiding using connectors to mechanically hold boards ETD/Online - Super. B Elba Workshop - May 2010
Technical choices for the TDR : trigger • Baseline: – re-implement Ba. Bar L 1 trigger with some improvements • Shorter latency (~4 us instead of 12 us) • Higher sampling frequencies (DCH and EMC) • 2 -d map for calorimeter • Possible additions – SVT trigger – Bhabha Veto • Challenge – To keep the event loss due to dead time below 1% => a maximum of ~70 ns “per-event dead time” is allowed in trigger and FCTS • Other considerations – What goes in L 1, what in L 3, what’s the optimum? • See trigger session (ETD 2) – First meeting of “Trigger Interest Group” ETD/Online - Super. B Elba Workshop - May 2010 150 k. Hz Exponential Inter-arrival time pdf.
Technical choices for the TDR : common items (1) • FCTS • Well on tracks. No major open question. Will most probably be a custom crate (see talk on Wednesday) • Links • Clock distribution has to be perfectly safe • The solution chosen will have to guaranty that clock will always be present despite radiation events • Control link • We have to validate the commercial chipset for radiation: but it has to be the one we will buy! (actually true for all components) • We have to define the command frame length and the reasonable overhead necessary for redundancy • We have to define the format of the trigger primitives sent back from the FEE • Readout link • We have to validate the commercial chipset for radiation • We have to refine the link effective payload in order to allow the subsystems to finely tune their number of links ETD/Online - Super. B Elba Workshop - May 2010
Technical choices for the TDR : common items (2) • Common FEE • We have to define the depth of the derandomizer buffers, located between the latency buffers and the readout links • ECS • Well on tracks. We take benefit there of the ongoing developments for LHCb. • ROM • A new R&D has been launched, implying: • Building and testing an optical to PCIe interface board to get readout data • Writing the Linux drivers for PCIe • Evaluating the optimal number of input and output boards which can be hosted in a single mother board • Testing the transmission rate/bandwidth achievable by using standard/commercial 10 Gb. E. ETD/Online - Super. B Elba Workshop - May 2010
Technical choices for the TDR : Online • Well under control • Event Management and Event Builder • Define interaction between FCTS, Event Builder and HLT • What information needs to be broadcast to the ROMs, what to the FEE to ensure consistent event build – has implications for FCTS/ROM design • All other components of the Online system are expected to be commodity / offthe-shelf • Only notional technical choices will be required for the TDR to get to a reasonably accurate estimate of cost and manpower, such as: • HLT farm architecture (storage, network) • Detector Control System • Computing infrastructure for the Online system • Data quality monitoring ETD/Online - Super. B Elba Workshop - May 2010
Скачать презентацию ETD Online Report D Breton U Marconi S Luitz
aa9eae62e735bbc67f04b6b548dec5d5.ppt