LEAPS-INNOV D5.7 - Installation of the synchronization protocol and demonstrator experiments at several facilities

Deliverable no. D5.7 Page 1 of 7 Project information Project full title LEAPS pilot to foster open innovation for accelerator-based light sources in Europe Project acronym LEAPS-INNOV Grant agreement no. 101004728 Instrument Research and Innovation Action (RIA) Duration 01/04/2021 – 30/09/2025 Website www.leaps-innov.eu Deliverable information Deliverable no. D5.7 Deliverable title Installation of the synchronization protocol and demonstrator experiments at several facilities Deliverable responsible DESY Related WorkPackage/Task WP5 “POSIT” / Task 5.3 “Synchronization” Type (e.g. Report; other) Report Author(s) Jose Avila Abellan, Oliver Seeck Dissemination level public Document Version 1.0 Date 10.09.2025 Download page Document information Version no. Date Author(s) Comment 1.0 10.09.2025 Oliver Seek, Jose Avila Table of Contents Introduction ............................................................................................................................................. 2 General remarks to the synchronization protocol and the demonstrator experiment .......................... 2 Feedback from the experiment at P08 at PETRA III ................................................................................ 3 Revision of the protocol by experts from ALBA Synchrotron and Diamond Light Source ...................... 5 PandABox2 and Next Steps ..................................................................................................................... 7 Deliverable no. D5.7 Page 2 of 7 Introduction Synchrotron radiation sources are pulsed X-ray sources with typical pulse lengths of 100 ps and interpulse distance of some nanoseconds up to about 200 ns. This pulsed source can be accounted as continuous for standard applications with acquisition times of more than 0.01 s and relatively slow change of actuators or other devices, with “slow” meaning a speed of one X-ray beam diameter per 0.01 seconds. Should an experiment require much higher speeds, e.g. for pump-probe measurements, the granularity of the X-ray beam has to be considered or is even required to collect meaningful data. For this a synchronization of beamline devices, such as detectors, the accelerator clock and external fields is realized by means of fast electronics. In the future, with detectors operating at acquisition rates of more than 100 kHz and with voice coil and piezo stages reaching speeds of 1m/s, synchronization will be required even for non-pump-probe measurements. By offering a human and machine-readable protocol, which describes the synchronization tasks for an experiment, the installation of experiments becomes straight forward at the different synchrotron radiation sources. The protocol should be useful for users and for the staff of these X-ray sources. This protocol as presented in deliverable D5.3 was rolled out at beamline P08 at PETRA III for testing synchronization with the defined demonstrator experiment from the company TXProducts. In addition, it has been evaluated by a synchronization expert and by a beamline and a laser scientist at ALBA and Diamond Light Source. During the testing at P08 it turns out that the actual implementation of the hardware is straight forward and the more complex task is to actually use the protocol. Due to limited time for testing the protocol with hardware at several synchrotron radiation facilities we could unfortunately not achieve more than this beamtime. However, as the more critical task is to actually use the protocol, we asked several experts at other facilities to review the protocol in terms of functionality and usability for the beamlines. General remarks to the synchronization protocol and the demonstrator experiment The experimentalist comes to the facility with a certain equipment/hardware to run the experiment. The synchronization of this equipment with the beamline and the accelerator is key for the success of the experiment. He/she must describe the synchronization that this equipment needs by submitting a form. The output of this form is a JSON data file. This data file is processed to 1) check that it fits the beamline capabilities 2) configure the beamline synchronization equipment for that experiment. Each required signal will be defined independently. The format of the protocol is strictly defined. As demonstrator experiment a pulse picker from the company TXProducts was chosen. A pulse picker is a device which with a speed of 10MHz can switch the X-ray beam on and off. This is particularly useful for pump-probe experiments with decay times of milli seconds where several subsequent X-ray probe pulses are created to detect the decay. Similarly, the pulse picker is useful if the measurement id of stroboscopic type, where the experiment is repeated multiple time with exact same conditions and time delay (see Figure 1a). Deliverable no. D5.7 Page 3 of 7 Feedback from the experiment at P08 at PETRA III The test experiment has been implemented at beamline P08 at PETRA III. Beamline P08 offers a laser for pump-probe experiments at sample surfaces. For this test experiment (see Figure 1a) the laser was not connected but an X-ray detector (Pilatus 300k). The function of the pulse picker has been tested by scanning the temporal gate of the pulse picker versus the bunch clock from the accelerator with a pulse of 9 ns width. The setting of the protocol below resembles the definition of the timing parameters as seen in Figure 1b. The experiment has been done in PETRA III 480 bunch mode with bunch separation of 16 ns. Furthermore, the Pilatus 300k detector has been described using the protocol. Here are the both devices described within the protocol: 1) A Synchronization signal for the WaveGate Pulse Picker: with external trigger, no inhibit, variable divider, no holdoff and variable delay Signal: Name: WaveGate Trigger Description: WaveGate Pulse Picker, input signal is an integer division of bunch clock Direction: Input FormFactor: BNC Type: Trigger Reference: None Inhibit: None DigitalParameters: # just describing the input parameters Signaling: TTL Polarity: ActiveHigh ActiveTime: 9000 [ps] DeadTime: 1000 [ps] RiseTime: 1000 [ps] FallTime: 1000 [ps] MaxJitter: 1000 [ps] ModifierParameters: Divider: 1 - 130 Holdoff: 0 Delay: 0 – Divider/(130 kHz) In terms of this device it became evident that the Modifier Parameters lack the possibility to choose different Dividers depending on the experiment and even more important, that the Delay may depend on the Divider, or in other cases, the delay may be determined during the alignment of the experiment. So, both parameters require more flexible input than just an integer, as originally foreseen. This is definitely to be implemented in further protocols. Figure 1: a) Scheme of the WaveGate Pulse Picker. b) Definition of timing parameters Deliverable no. D5.7 Page 4 of 7 2) An X-ray detector with external trigger, no inhibit, no divider, no holdoff and no delay Signal: Name: Detector Description: Pilatus 300k, gets input from Bunch Clock Direction: Input FormFactor: BNC Type: Trigger Reference: None Inhibit: None DigitalParameters: # just describing the input parameters Signaling: TTL Polarity: ActiveHigh ActiveTime: 100000 [ps] DeadTime: 1000 [ps] RiseTime: 1000 [ps] FallTime: 1000 [ps] MaxJitter: 1000 [ps] ModifierParameters: Divider: 1-130 Holdoff: 0 Delay: 0 – Divider/(130 kHz) Again, the scientist states that a limitation to constant values is not adequate when a time window is scanned (see Divider and Delay). We immediately agree. Figure 2 shows the result of the experiment, a scan of the temporal gate of the pulse picker vs. the bunch clock signal. The spikes are pulses emitted from individual bunches in the storage ring. The amplitude envelope is a property of the external gating of the detector with 100 ns gating. The spikes visible are the bunches from the 480 bunch mode of PETRA III with 16 ns separation. Figure 2: Scan of the temporal gate of the pulse picker vs. the bunch clock signal. Deliverable no. D5.7 Page 5 of 7 Overall, he thinks, that this protocol might be handy for planning experiments, if experiments can be described with variable parameters, which depend on each other. Revision of the protocol by experts from ALBA Synchrotron and Diamond Light Source We first display the original reports of the reviewing persons: Principal Software Systems Engineer (Diamond Lights Source): “From my perspective, I guess I would be reading one of these files in software to configure the triggering systems on a beamline. Some of the information that would be needed for this is present (e.g. divider, delay times), but the nature of these experiments means that we would need a lot of beamline specific knowledge (e.g. what trigger signal type a detector is capable of accepting) in order to validate the scheme. It also seems to be very fixed in its nature, how does the experiment planner know they need an exposure time of 1000 ps for instance?” Beamline Scientist (Laser) (Diamond Lights Source): “I have gone through the document, and I think I understood the scope and many of its capabilities/functions. Please find below my feedback in the form of a narrative comprising a summary of my thoughts/understanding and a couple of critical points might be missing from the protocol. The protocol aims to provide the framework to standardise the approach in performing experiments which require temporal resolution from milliseconds (ms) to picoseconds (ps) and/or conditional validation ensuring measurements are conducted, prior and whilst certain criteria are met. The deterministic coordination of the available electrical and mechanical devices allows the preparation and execution of sophisticated experiments as well as the exploration of large experimental parameter spaces in a reproducible manner which in turn enhances the data quality and fidelity. There are various schemes that one can consider requiring complex workflows to be employed prior, during and after data collection. For example, “experiments at high pressures and temperatures (HPHT) require precise knowledge of the conditions inside high-pressure devices, such as the diamond anvil cell (DAC)” 1 prior collecting e.g. the diffraction image of the sample under study. On the other hand, several halide perovskites are prone to radiation damage, hence sample longevity would be greatly prolonged if the samples would only be irradiated/studied during data collection. 2 The above schemes also apply when the 4th dimension, time, is considered and temporal resolution is one of the experimental parameters which needs to be deduced/controlled. For applications requiring resolution at the ms and 100s of microseconds (μs) timescales the coordinated function of the detector (e.g. gate function) or the beamline equipment (e.g. choppers) with the rest of the apparatus is necessary. For applications where picoseconds (ps), nanoseconds (ns), or a few μs temporal resolution is required, the above scheme is not enough and synchronisation, of potentially multiple, devices with the accelerator needs to be employed. For example, “the first direct observation of the transient spinstate in a disordered magnetic system with time-resolved XAFS” 3 required the synchronisation of detectors with an external laser source, with the latter being synchronised with the X-ray pulses in two stages, initially with the radiofrequency cavity of the storage ring and then with the revolution frequency of the electron bunches. The latter class of experiments push the synchrotron facilities close to their limits when it comes to their conventional use and in terms of the timing accuracy they can provide. Yet, their success, often, necessitates the use of special filing patterns where a single or a few electron bunches (camshaft) are isolated from the rest and being detected. 1 https://journals.iucr.org/s/issues/2023/04/00/vl5008/index.html 2 https://www.oaepublish.com/articles/energymater.2023.114 3 https://pubs.acs.org/doi/10.1021/ja907460b Deliverable no. D5.7 Page 6 of 7 An important note at this stage is that the protocol assumes that the hardware will have the required characteristics to allow implementation of the required conditions or timing protocols. It’s not clear though whether there is a validation step/layer which will clarify whether for example the required jitter values are physically possible or relevant/necessary, or whether the requested parameters fall within the limits of the device. On a side note, I found very interesting the example with the WaveGate pulse picker, particularly because we have been working with a few others on a different scheme which provides similar functionality. “ Beamline Scientist (BL11 NCD-SWEET) (ALBA Synchrotron) “After reading the document I understand that this is the definition of a new synchronization system between the X-ray pulsed source and the beamline components. This protocol is separated in two main parts: one for the source and one for the beamline (experimental) instrumentation. These comments are more focused in the experimental part. From the beamline point of view, and for a correct synchronization, the different beamline components must be fully characterized to ensure intrinsic signal delays that will be taken into account by the synchronizer. However, major issues can be encountered with commercial systems that act as a black box and that need to be fully characterized (e.g. a 2D detector, delay between the input signal arrives until the image is recorded), even if ideally should be provided by the company and evaluated during the FAT and SAT. Large efforts can be done in the electronic system to implement this synchronization, but a larger overview on how to work with current and new instrumentation should be considered (specially the requirements but also its characterization and validation). Being able to synchronize photon pulses with the sample environment and data collection is of high interest. Not only pump-prove experiments, but also time-resolved experiments that require to synchronize different hardware. Often, these experiments relay on high-repetition rates, but sometimes it is a dynamic one-way irreversible process that has to be time-resolved and that requires several systems to be synchronized to ensure that data collection is reliable and reproducible. This entails that from the user point of view, some parameters of the configuration files that could affect the experiment (e.g. delay between pump and probe, delay of triggering a device...) can be adapted on the flight, to vary a parameter (e.g. a given delay, wait for another input...) for its optimization without stopping data collection, just in a flexible and adaptable manner. Of course, any variation and change must be registered by the system and saved as a metadata, to ensure that all the exact applied conditions are well known (in value but in time). This dynamic behavior for some components will facilitate the development of top-level science. Besides, this new synchronization protocol must accept signals to trigger other actions, ensuring that a beamline component (e.g. a software doing data analysis on the flight or a hardware like a motor position) triggers an event or send and wait protocol, where a signal is sent and the second one waits until a given input. In brief, different inputs and outputs should be accessible to be programed in advance to interact with the beamline synchronization and experimental setup. The pico-switch, by its part, uses a material to shorten the X-ray photon pulse using a laser. This is system seems promising, but in view of the 4th generation synchrotrons, and for some of the beamlines, maintaining the front-wave is critical, and this technology should be evaluated and not standardized, since the beam parameters can be different for the differ rent beamlines.” In summary, the reviewers acknowledge that experiments with synchronization are very important and might become even more prominent in future synchrotron radiation sources. They also state, that if a protocol should be useful, it has to be flexible enough to implement all possible cases and in particular dynamic behavior. The latter argument is similar to what the scientist at the experiment at Deliverable no. D5.7 Page 7 of 7 PETRA III mentioned about possibility to flexibly set Delay and Divider, so this is definitely a thing to work on. PandABox2 and Next Steps During the initial periodic meetings of this Task 5.3 where representatives of the different participating facilities were attending, the needs and requirements for a future synchronization protocol standardization were collected and presented in Deliverable D5.1. From the very beginning it came to light that as a base of the synchronization protocol, it was very interesting to have a hardware solution for the beamline synchronization common among facilities. This would foster the mobility of experimentalists between facilities while facilitating the integration and experiment setup, because up to that moment each facility had obviously adapted his own solution. The idea of having a common synchronization hardware was presented in the 19th International Conference on Accelerator and Large Experimental Physics Control 4 . As a result of it, the LEAPS facilities SOLEIL, DESY, ALBA, DIAMOND and MAXIV are in strong collaboration to develop a synchronization hardware (PandABox2, the evolution of PandABox) that will cover both actual needs and future needs of the incoming years where pump-probe and time-resolved experiments will become notorious due to the upgrades to 4th generation synchrotrons. The PandABox2 collaboration is outside the scope of Task 5.3, but we can assure that it was only possible thanks to it. This high technological hardware development will continue during next couple of years. The firmware development will necessarily take into account the feedback from our protocol review regarding increasing the flexibility, in order to implement all possible cases, and guarantee the success of implementing at several facilities a common synchronization hardware with a common protocol. 4 https://proceedings.jacow.org/icalepcs2023/papers/thmbcmo22.pdf