Upgrade of the ATLAS MDT Readout and Trigger for the HL-LHC

The luminosity increase at the LHC in Phase-II calls for an in-depth redesign of the MonitoredDrift Tube (MDT) chamber readout. One reason for this is the high rate of primary detectorsignals which requires increased readout bandwidth all along the data path. The second reason isdue to a new concept for the formation of the Level-1 muon trigger (L1), which aims to improvethe selectivity for muons with high transverse momentum (p$_T$ ). While in the present system theL1 trigger is exclusively based on specialized trigger chambers (RPC in the barrel, TGC in theendcap), which deliver tracking information with very limited accuracy, the L1 trigger will, inPhase-II, also consider tracking information from the MDT: the time resolution of the triggerchambers, able to tag the beam crossing (BX), will be combined with the high spatial accuracy ofthe MDT.In the new L1 trigger scheme, the selection of MDT hits corresponding to a given BX, will bebased on a Region of Interest (RoI) provided by the trigger chambers. Considering only MDThits in the RoI will result in a large reduction of the relevant data volume. The selected MDT hitswill be tested for consistency with the track direction, as defined by the RoI. If confirmed, theMDT hits are used for a better p T definition.The rate of background hits from γ and neutron conversions is several orders of magnitude higherthan the one from muon hits. The selection of muons against this background ("hit extraction")requires high-speed processing power, capable to supply the relevant information to the MuonCentral Trigger Processor (MUCTPI) inside the latency of 10 μs.