Speaker
Description
At the high-luminosity upgrade of the LHC (HL-LHC), the electromagnetic calorimeter of CMS (ECAL) will have to cope with an increase in the number of interactions per bunch crossing and radiation levels. CMS implements a sophisticated two-level triggering system composed of the Level-1, instrumented by custom-designed hardware boards, and a software High-Level-Trigger. The off-detector electronics has been redesigned with increased capabilities, exploiting the full granularity of the calorimeter at Level-1. The talk focuses on the new design and its expected performance, compared to the LHC Run2 in terms of trigger rate, rejection of anomalous signals, and selection efficiency for electrons and photons.
Summary
The electromagnetic calorimeter (ECAL) of CMS is currently operating at the LHC and its barrel part consists of 61200 lead tungstate crystals. The instantaneous luminosity during the current LHC Run2 is in the range of up to $2\times 10^{34}$ cm$^{-2}$s$^{-1}$ in routine operation. With an upgrade at the end of the decade, the High-Luminosity LHC (HL-LHC) will increase the instantaneous luminosity by about a factor 2 to 5 from current levels. The goal of the HL-LHC is to accumulate a total of at least 3 ab$^{-1}$ of data.
CMS implements a sophisticated two-level triggering system composed of the Level-1, instrumented by custom-design hardware boards, and a software High-Level-Trigger.
With the luminosity foreseen for the HL-LHC, and consequent event rate, the requirements on the CMS Level-1 trigger system are a latency of 12.5 $\mu$s, about a factor of two more than the current electronics, and a sustainable rate up to 750 kHz, about a factor of five more than the current design.
The front-end readout electronics was completely re-designed to satisfy these requirements. In contrast to the current electronics, which hosts the computation of ``trigger primitives'' (coarse information on the deposited energy in ECAL, exploited at Level-1) directly on the on-detector electronics, the upgraded design will shift their computation off-detector. This will be possible by means of 10 Gb/s high-speed optical links (lp-GBT) transmitting off-detector the single-crystal information, serialized by groups of 25 crystals. Accounting for one additional link for every 25 crystals, to transmit the system clock and control signals, an order of 10k of such links will suffice for the whole ECAL.
The processor used offline to compute trigger primitives are not required to be radiation tolerant, as they will be located in the service cavern. Therefore design choice opens up to commercial solutions, such as the CTP7/MP7 $\mu$TCA boards.
The algorithms to be implemented in this board include: conversion of digitized pulse data into transverse energy, rejection of anomalous APD signals based on topological criteria, basic clustering of localised energy. The board will also interface with the central CMS L1 trigger and DAQ systems.
The talk will discuss in detail the requirements of the new ECAL off-detector electronics and the design options and choices for its architecture. The algorithms to be implemented will be illustrated and compared to the current system, along with the first estimates of their performance in term of trigger rates, rejection of anomalous signals, selection efficiency for electrons and photons.
