XFEL. EU, Hamburg, Germany
STFC/RAL, Chilton, Didcot, Oxon, United Kingdom
UCL, London, United Kingdom
DESY, Hamburg, Germany
The European XFEL DAQ system has to acquire and process data in short bursts every 100ms. Bursts lasts for 600us and contain a maximum of 2700 x-ray pulses with a repetition rate of 4.5MHz which have to be captured and processed before the next burst starts. This time structure defines the boundary conditions for almost all diagnostic and detector related DAQ electronics required and currently being developed for start of operation in fall 2015. Standards used in the electronics developments are: MicroTCA.4 and AdvancedTCA crates, use of FPGAs for data processing, transfer to backend systems via 10Gbps (SFP+) links, and feedback information transfer using 3.125Gbps (SFP) links. Electronics being developed in-house or in collaboration with external institutes and companies include: a Train Builder ATCA blade for assembling and processing data of large-area image detectors, a VETO MTCA.4 development for evaluating pulse information and distributing a trigger decision to detector front-end ASICs and FPGAs with low-latency, a MTCA.4 digitizer module, interface boards for timing and similar synchronization information, etc.
XFEL. EU, Hamburg, Germany
Modern diagnostic and detector related data acquisition and processing hardware are increasingly being implemented with Field Programmable Gate Array (FPGA) technology. The level of flexibility allows for simpler hardware solutions together with the ability to implement functions during the firmware programming phase. The technology is also becoming more relevant in data processing, allowing for reduction and filtering to be done at the hardware level together with implementation of low-latency feedback systems. However, this flexibility and possibilities require a significant amount of design, programming, simulation and testing work usually done by FPGA experts. A high-level FPGA programming framework is currently under development at the European XFEL in collaboration with the Oxford University within the EU CRISP project. This framework allows for people unfamiliar with FPGA programming to develop and simulate complete algorithms and programs within the MathWorks Simulink graphical tool with real FPGA precision. Modules within the framework allow for simple code reuse by compiling them into libraries, which can be deployed to other boards or FPGAs.