Sonix

The European Spallation Source (ESS) in Lund, Sweden is just entering the construction phase with 3 neutron instruments having started in its design concept phase in 2014. As a collaboration of 17 European countries the majority of hardware devices for neutron instrumentation will be provided in-kind. This presents numerous technical and organisational challenges for the construction and the integration of the instruments into the facility wide infrastructure; notably the EPICS control network with standardised hardware interfaces and the facilities absolute timing system. Additionally the new generation of pulsed source requires a new complexity and flexibility of instrumentation to fully exploit its opportunities. In this contribution we present a strategy for the modularity of the instrument hardware with well-defined standardized functionality and control & data interfaces integrating into EPICS and the facilities timing system. It allows for in-kind contribution of dedicated modules for each instrument (horizontal approach) as well as of whole instruments (vertical approach). Key point of the strategy is the time stamping of all readings from the instruments control electronics extending the event mode data acquisition from neutron events to all metadata. This gives the control software the flexibility necessary to adapt the functionality of the instruments to the demands of each single experiment. We present the advantages of that approach for operation and diagnostics and discuss additional hardware requirements necessary.

Many of the world׳s time-of-flight spallation neutrons sources are migrating to recording individual neutron events. This provides for new opportunities in data processing, the least of which is to filter the events based on correlating them with logs of sample environment and other ancillary equipment. This paper will describe techniques for processing neutron scattering data acquired in event mode which preserve event information all the way to a final spectrum, including any necessary corrections or normalizations. This results in smaller final uncertainties compared to traditional methods, while significantly reducing processing time and memory requirements in typical experiments. Results with traditional histogramming techniques will be shown for comparison.

Imaging using scintillators is a widespread and cost efective approach in radiography. While different types of scintillator and sensor confgurations exist, it can be stated that the detection efciency and resolution of a scintillator based system strongly depend on the scintillator material and its thickness. Recently developed event driven detectors are capable of registering spots of light emitted by the scintillator after a particle interaction, allowing to reconstruct the Center of Mass of the interaction within the scintillator. This results in a more precise location of the event and therefore provides a pathway to overcome the scintillator thickness limitation and increase the efective spatial resolution of the system. Utilizing this principle, we present a detector capable of Time of Flight imaging with an adjustable feld of view, ad hoc binning and re binning of data based on the requirements of the experiment including the possibility of particle discrimination via the analysis of the event shape in space and time. It is considered that this novel concept might replace regular cameras in neutron imaging detectors as it provides superior detection capabilities with the most recent results providing an increase by a factor 3 in image resolution and an increase by up to a factor of 7.5 in signal to noise for thermal neutron imaging.

The next generation of muon spin spectrometers at the ISIS pulsed source are being developed to make efficient use of the increased source intensity. They will provide a transformational improvement in counting rates: ‘Super-MuSR’ will be the first of these instruments, capable of counting at ≈1 G·event·hr-1. Key to delivering this capability is the development of highly pixelated, high density detector arrays that cover an appreciable solid angle, with each detector element optimised for counting at very high data rates. A series of ‘firsts’ are planned to optimise individual element count rate capability, where analogue waveforms recorded from SiPMs are fully digitised and processed using digital signal processing (DSP) methods at either software or firmware level. Full raw-signal digitisation will be achieved using the Xilinx Zynq® UltraScale+TM series of ‘system on a chip’ operating with ADCs capable of 1 GHz sampling, data handling using event streaming technology, and DSP to provide novel data correction techniques. We will discuss our concept and present preliminary results. Our prototype digitising data acquisition system, which is key to implementing a ‘digital data pipeline’ (DDP) is presented.

This work provides an ideal opportunity to replace the existing ISIS streaming mechanism with a new data streaming system based on Apache Kafka, which not only provides enhanced functionality for ISIS but also functions as a prototype for developing and testing requirements of the ESS.

In the last years SINQ has undergone a major upgrade both from the hardware and the sofware point of view. The neutron flux has been increased thanks to new neutron guides, instruments have been upgraded (AMOR, DMC) and new instruments will be installed (SANS-LLB, Falcon). NICOS and EPICS replaced the old control sofware (SICS). At the same time, we are undergoing a paradigm shift in the data acquisition sector. From the old “histogramming” approach we are transitioning (where possible) to “event streaming” mode. This is the first step toward time-resolved experiments. In this talk we will present the status of data streaming at SINQ and the current limitations.

As a contribution to the European Spallation Source as part of BrightnESS, the Paul Scherrer Institut is involved in the streaming of EPICS data and the writing of NeXus compliant HDF5 files. We combine this development with the transition of the AMOR instrument at the Paul Scherrer Institut to EPICS and a streaming based data architecture. To guide our development before ESS has operational equipment, we use a detailed simulation of the instrument AMOR at SINQ to test and integrate our data streaming components. We convert EPICS data sources to Google FlatBuffers as our message format and distribute them using Apache Kafka. On the file writing side, we combine the messages from EPICS data sources as well as from neutron events to write HDF5 files at rates up to 4.8 GiB/s using Parallel HDF. This platform will also be used for testing the experiment control software on top of EPICS

As part of the UK’s in-kind contribution to the European Spallation Source [1], ISIS [2] is working alongside the ESS and other partners to develop a new data streaming system for managing and distributing neutron experiment data. The new data streaming system is based on the open-source distributed streaming platform Apache Kafka [3]. A central requirement of the system is to be able to supply live experiment data for processing and visualisation in near real-time via the Mantid data analysis framework [4]. There already exists a basic TCP socket-based data streaming system at ISIS, but it has limitations in terms of scalability, reliability and functionality. The intention is for the new Kafka-based system to replace the existing system at ISIS. This migration will not only provide enhanced functionality for ISIS but also an opportunity for developing and testing the system prior to use at the ESS.