MicroTCA implementation of synchronous Ethernet-Based DAQ systems for large scale experiments

摘要

Large LAr TPCs are among the most powerful detectors to address open problemsin particle and astro-particle physics, such as CP violation in leptonicsector, neutrino properties and their astrophysical implications, proton decaysearch etc. The scale of such detector implies severe constraints on theirreadout and DAQ system. In this article we describe a data acquisition schemefor this new generation of large detectors. The main challenge is to propose ascalable and easy to use solution able to manage a large number of channels atthe lowest cost. It is interesting to note that these constraints are verysimilar to those existing in Network Telecommunication Industry. We propose tostudy how emerging technologies like ATCA and $\mu$TCA could be used inneutrino experiments. We describe the design of an Advanced Mezzanine Board(AMC) including 32 ADC channels. This board receives 32 analogical channels atthe front panel and sends the formatted data through the $\mu$TCA backplaneusing a Gigabit Ethernet link. The gigabit switch of the MCH is used tocentralize and to send the data to the event building computer. The core ofthis card is a FPGA (ARIA-GX from ALTERA) including the whole system except thememories. A hardware accelerator has been implemented using a NIOS II $\mu$Pand a Gigabit MAC IP. Obviously, in order to be able to reconstruct the tracksfrom the events a time synchronisation system is mandatory. We decided toimplement the IEEE1588 standard also called Precision Timing Protocol, anotheremerging and promising technology in Telecommunication Industry. In thisarticle we describe a Gigabit PTP implementation using the recovered clock ofthe gigabit link. By doing so the drift is directly cancelled and the PTP willbe used only to evaluate and to correct the offset.