E. Hazen - Detector Electronics1 Detector Electronics New Developments

E. Hazen - Detector Electronics2 Outline ● Brief summary of detector electronics ● New Stuff: ● Introduction to MicroTCA ● WFD Update ● Tracker Electronics Update ● Plans and Schedule

E. Hazen - Detector Electronics3 Summary of Detector Electronics ● Calorimeter Readout ● ~56 channels per calo station (design for 60) Digitize at 500 MSPS 12 bits linear for 700us Time island processing either in FPGA or CPU/GPU ● Tracker Readout ● 928 straws per tracker station  1ns TDC resolution, ~ 4k single hits per spill Electronics in gas at 1atm ● Fiber Harp Readout ● MuLAN WFD system Firmware mods already completed!

E. Hazen - Detector Electronics4 MicroTCA

E. Hazen - Detector Electronics5 What's New? MicroTCA... ● Derived from ATCA Advanced Telecommunications Computing Architecture ● Lots of jargon, but in a nutshell: ● Fast backplane (min 500 MB/s per slot) vs VME 200MB/s shared ● Lots of telecoms industry support ● Relatively low cost (~$5.5k for powered crate) ● Complicated to use :/ ● ATCA and uTCA widely used: ● SLAC, CERN, DESY others ● Now the baseline for WFD and Tracker readout in g-2 Vadatech VT892 Crate Data Taking in CMS

E. Hazen - Detector Electronics6 g-2 uTCA Crate 12 AMC Slots for WFDs (calo) or TRMs (tracker) Commercial MCH Management Ethernet AMC13 Clocks Fast controls DAQ

E. Hazen - Detector Electronics7 uTCA Dual-Star Backplane MCH 1 Commercial /Std MCH 2 aka “AMC13” Custom design for CMS Bi-directional serial (up to 10Gb/sec) point-to-point links from each AMC to MCH (redundant links to each MCH) g-2 use Fabric A (1 link) 5.0 Gb/s Fabric B (1 link) LVDS timing Fabric D-G (4 links) Spare CLK1 MLVDS clock Note: Interconnections can be customized by the backplane manufacturer inexpensively. Fabric A (1 link) Gigabit Ethernet Fabric B (1 link) Spare Fabric D-G Spare CLK1 Spare

E. Hazen - Detector Electronics8 uTCA for DAQ ● uTCA is missing the following features for DAQ: ● Synchronous controls path ● DAQ path faster than 1GbE ● Two options for adding these features: ● AMC13 module developed for CMS at LHC (by BU) – Existing solution, CMS committed to support – Our choice for g-2 ● MTCA.4 evolving standard (aka “xTCA for Physics”) – Popular at DESY, SLAC but not CERN – No off-the-shelf DAQ modules yet

E. Hazen - Detector Electronics9 AMC WFD Or TRM AMC AMC13 Custom CMS (BU) 3x 10Gb links (protocol t.b.d) 1 GbE via MCH Inputs from Detector DAQ Link 5Gb/s Clock and Timing Gigabit Ethernet IPMI (Management) MCH (COTS) Clock and Timing AMC 12 AMC Modules Wired in “dual star” To AMC13 and MCH MicroTCA connectivity All connections point-point (run in parallel) GbE to every module - control / monitoring - DAQ for testing “Fabric” connections for User applications (in our case, DAQ) - point to point - protocol not defined IPMI to every module - turns on the power! (plus module ID, etc) - required by standard g-2 uTCA Connectivity

E. Hazen - Detector Electronics10 uTCA Development Tasks (required for both WFD and tracker) ● AMC13 firmware customization for g-2 ● Support for larger payloads ● Support for custom processing block – g-2 specific monitoring, histogramming, etc – track fitting? – Calo zero suppression? ● Clock / controls inputs ● R&D project: can use existing fiber input, or need a new mezzanine card? ● DAQ interface ● 10Gb/s DAQ link protocol development ● Plan for non-ZS operation (with ZS as an option) ● Software support ● Control / monitoring software: adapt from CMS effort?

E. Hazen - Detector Electronics11 WFD

E. Hazen - Detector Electronics12 WFD in uTCA concept Self-contained Single channel Large buffer (200+ spills!) per channel Module Management Microcontroller (MMC) Developed by U-Wisc Main FPGA with DAQ functions, Gigabit Ethernet Clock synthesizer synch'd to master clock: R&D

E. Hazen - Detector Electronics13 WFD Plans ● Pre-MRI schedule: ● Set up uTCA crate with AMC13: Jan '13 ● Initial 1-channel prototype uTCA card: Jan-Feb '13 ● Develop firmware and test: Feb...? '13 ● Post-MRI schedule: ● Design and Fab 5-channel main board (+ mezzanine?) ● Test and characterize: – clock stability, data integrity and linearity, noise/crosstalk, etc ● If front-end is separate mezzanine, repeat above for FE ● Produce ~ 12 prototypes, full crate tests ● 2nd prototype cycle as required; production and testing

E. Hazen - Detector Electronics14 Tracker

E. Hazen - Detector Electronics15 Key Tracker Electronics Requirements ● Collect straw charge with suitable front-end ● Good match to tube capacitance for efficient charge collection, appropriate preamp peaking time, timing discriminator ● Surplus ASDQs from CDF are a good candidate ● Measure leading edge time on ~ 2k straws to ~1ns ● FPGA TDC is a good candidate (as described by J.Wu et al) ● Readout within 11ns between spills ● Data volume ~ 10k Bytes/fill for one traceback station ● Provide HV ● Fit within gas manifold; minimize power

E. Hazen - Detector Electronics16 Tracker 3D Model (not correct in all details...) Gas input manifold Gas output manifold (and electronics enclosure) Base plate (GND) TDC PCB Front-end PCB ASD chips TDC chips 20 mm

E. Hazen - Detector Electronics17 Chamber End Detail Input Gas Manifold ASDQ chips TDC chip(s) Board-to-board connectors No electrical connection at this end. HV and readout connection at this end. 12 mm

E. Hazen - Detector Electronics18 Tracker Electronics Overview One MicroTCA Crate per tracker station (short cables) -or- one crate per two stations (longer cables) 5 AMC modules control/readout 10 chambers Clock input to AMC13 GbE to DAQ Each chamber has two sets of front-end electronics (top/bottom) Optional 2nd set of modules for 2nd traceback station

E. Hazen - Detector Electronics19 ASDQ 8 ch TDC 16 or 32 channels FPGA Clock Timing/control Strobe Data TDC LVDS Fanout Clock Timing/control TDC Readout Serial data daisy-chain or direct from each TDC Power In +/-3V +3.3V +1.2V Cabling through gas tubing (4) pairs 100 ohm (4) pairs power at V(load) (no regulators) Power ~ 5W max per chamber On-Chamber Electronics Overview HV Current-limit resistors Blocking capacitors

E. Hazen - Detector Electronics20 Off-Chamber Electronics Overview Gas exhaust from tracker With twisted-pair cable inside Electrical feedthru (gas tight, not vacuum) Gas connector box outside vacuum chamber Exhaust gas to gas system External cables for power and clock/data MicroTCA Crate - 12 TRM (Tracker Readout Modules) DAQ Connection Single GbE (copper) per crate Clock / controls One TRM per chamber (total of 9 needed for one traceback station)

E. Hazen - Detector Electronics21 Tracker Electronics Plans (near term) ● Conceptual Design and Specification ● Trial PCB layouts to establish space requirements ● Study power, cooling, mechanics ● Think through off-chamber designs ● Develop TDC / Readout specification ● Functional Prototyping ● Design ASDQ board for next straw prototype ● Develop Prototype TDC Firmware (eval board) ● Connect ASDQ board to TDC eval board, or develop prototype dedicated TDC board ● Test on bench, with source, and at test beam

E. Hazen - Detector Electronics22 Summary ● There are conceptual designs for most systems ● Almost everything is new from the last g-2 ● There is a lot of work to do!