LLNL-TR-2019311 Reference Hardware and Software Architecture for a Solenoid Pulser for Scientific Instrumentation at the National Ignition Facility S Mistyuk Jun 2026 This document was prepared as an account of work sponsored by an agency of the United States government. Neither the United States government nor Lawrence Livermore National Security, LLC, nor any of their employees makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States government or Lawrence Livermore National Security, LLC. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States government or Lawrence Livermore National Security, LLC, and shall not be used for advertising or product endorsement purposes. This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. This document was prepared as an account of work sponsored by an agency of the United States government. Neither the United States government nor Lawrence Livermore National Security, LLC, nor any of their employees makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States government or Lawrence Livermore National Security, LLC. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States government or Lawrence Livermore National Security, LLC, and shall not be used for advertising or product endorsement purposes. This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. Reference Hardware and Software Architecture for a Solenoid Pulser for Scientific Instrumentation at the National Ignition Facility 1 Surge Mistyuk 1 Department of Target Diagnostics Engineering and Science, Lawrence Livermore National Laboratory, 7000 East Avenue, Livermore, CA, USA Abstract – High current, high voltage solenoid pulser power systems for the generation of magnetic fields in a solenoid is not a radical or new concept. Despite being a cornerstone of scientific and weapons research for over 100 years, a commercial off the shelf unit easily configurable for a broad set of applications has yet to appear on the open market. Pulser systems are instead designed for specific target requirements. For gigawatt to terawatt systems, this is unavoidable. For megawatt systems however, many hundreds of thousands of dollars are wasted designing pulser systems from the ground up for clients, when it is possible to simply architect a modular system once that can have its capacitors swapped for many possible target applications. The National Ignition Facility (NIF) at the Lawrence Livermore National Laboratory (LLNL) is one such institution that would benefit from such a system, having bought such pulsers for several imaging systems. Ideally, all such a system would require is only a bachelors-level circuit analysis to find the correct capacitor values, as well as a written procedure to calibrate the software to correctly deliver the voltages required by the application. In this work, systems architecture and software engineering techniques are used to design and analyze such a system. Such a design, it is the hope of this author, will be used in scientific and engineering purposes in the future to build solenoid pulsers at the National Ignition Facility and other institutions to drastically cut development costs of scientific instrumentation. Contents 1. Introduction 4 2. Background 4 2.1. Solenoid Pulsing Basics 4 2.2. System Context 6 2.3. Systems Engineering Methodology 13 3. Body of Work 14 3.1. System Requirements 14 3.2. Behavior Modeling and Functional Architecture 17 3.3. Physical Architecture 23 3.4. Software Class Diagram 26 4. References 29 1. Introduction High current, high voltage solenoid pulser power systems for the generation of magnetic fields in a solenoid is not a radical or new concept. Despite being a cornerstone of scientific and weapons research for over 100 years, a commercial off the shelf unit has yet to appear on the open market. Pulser systems are instead designed for specific target requirements. For multi-gigawatt to terawatt systems, this is unavoidable. For megawatt systems however, many hundreds of thousands of dollars are wasted designing pulser systems from the ground up for clients, when it is possible to simply architect a modular system once that can have its capacitors swapped for many possible target applications. The National Ignition Facility (NIF) at the Lawrence Livermore National Laboratory (LLNL) is one such institution that would benefit from such a system, having bought such pulsers for several imaging systems. Ideally, all such a system would require is only a bachelors-level circuit analysis to find the correct capacitor values, as well as a written procedure to calibrate the software to correctly deliver the voltages required by the application. In this paper, systems architecture and software engineering techniques are used to design and analyze such a system. Such a design, it is the hope of this author, will be used in scientific and engineering purposes in the future to build solenoid pulsers at the National Ignition Facility and other institutions to drastically cut development costs of scientific instrumentation. 2. Background 2.1. Solenoid Pulsing Basics At their core, a pulsed solenoid is simply an LRC circuit, consisting of a solenoid inductor (L), resistor (R), and capacitor (C), in which the capacitor is separated from the resistor and inductor with a switch. This circuit can be seen in Figure 1. Figure 1 – solenoid pulser circuit schematic. L, R, and C represent the inductance, resistance, and capacitance of the circuit, respectively. S C represents the switch. V C , V R , and V L are the initial voltage of the capacitor, the voltage drops across the resistor, and the voltage drop across the inductor, respectively. While the switch is open, a capacitor is charged to an initial voltage V 0. When the generation of the pulse is desired, a trigger is sent to the switch, which closes it and allows current to flow from the capacitor to the resistor and inductor, commonly referred to as the load. A pulse then is generated, the shape and duration of which is the resultant of the interplay between the inductance, resistance, and capacitance of the circuit (the exact nature of this relationship is beyond the scope of this discussion). Suffice it to say, the problem falls into the damped undriven oscillator class of problems in physics, and types of current pulses capable of being produced can be seen in Figure 2. Figure 2 – current across the inductor as a function of time as a function of LRC parameters (ξ). ξ <1 shows underdamped oscillations, ξ =1 shows critically damped oscillations, and ξ >1 shows overdamped oscillations. Image reproduced from [1]. Current flows through the solenoid, generating a magnetic field through the Ampere-Maxwell Law, Equation (1) [2]. 𝜕𝐸⃑ ⃑ = 𝜇0 (𝐽 + ℇ0 ∇×𝐵 ) (1) 𝜕𝑡 Add to this circuit a charging power supply for the capacitor, the solenoid pulser system is ready. A block diagram of such a circuit can be seen in Figure 3. Charging Power Energy Storage Triggered Start Transmission Line Load Supply Capacitor Switch and Power Feed Trigger Pulse Generator Figure 3 – block diagram for a basic pulser system. It is important to reiterate that pulsing solenoids is only one application that such a system can be used for. Millisecond current pulser systems can drive many scientific and engineering instrumentations, such as radar [3], electromagnetic launch systems [4], some pulsed lasers [5], and plasma systems [6]. 2.2. System Context Rarely is a real system conceived from vacuum, and this case is no different. This body of work is rooted in solenoid pulsing needs at the LLNL. For decades, external vendors were contracted for the development of solenoid pulsers for diagnostic systems, including the Dilation X-ray Imager (DIXI) systems [7], (Hardened) Single Line of Sight (hSLOS) imager [8], and Multi-Frame Gated X-ray (MGXI) imager [9]. For the next generation DIXI system, the High Yield X-ray Imager (HYXI), the solenoid pulser was decided to be designed and built in-house due to available expertise of pulsed power electrical engineers. The operating principle of all these systems is the same. X-rays are generated during the NIF process due to the high energy density physics driving the process. These x-rays propagate radially outward to the target chamber wall, on which various x-ray diagnostics are mounted. The front-end of the diagnostic mentioned in this section all have a photocathode, typically a 500nm cesium iodide (CsI) thin film, a material that converts incoming x-rays to electrons through the photoelectric effect [7, 8, 9, 10]. A thin electrical conductor film coats the CsI material through which a high voltage pulse if propagated, typically gold. Because CsI is a non-conductor, contact with x-rays depletes the material of electrons, and only the portion of the photocathode through which the high voltage pulse is propagating has sufficient electrons to convert the image into a current. These electrons are then subjected to a non-uniform acceleration potential, which accelerates electrons generated early in the process faster than those generated at a later time (several picoseconds later). The non-uniformly electrons then propagate through the vacuum vessel, in which they are given time to temporally dilate. They are kept on course and focused on an imaging sensor using a magnetic field, generated by the solenoid pulser running current through the solenoids surrounding the chamber. Finally, electrons are captured by the sensor, which now can be reasonably well by a nanosecond gated silicon hCMOS sensor. Thus, by taking advantage of the physical properties of electrons, images that are otherwise too fast to capture with semiconducting technology can be captured and visually inspected by scientific staff to derive observations about the NIF process. A sample image captured by a DIXI system can be seen in Figure 4. Figure 4 – data collected from the NIF implosion with timestamps from bangtime. This led to the development of the system requirements seen in Figure 5. To make sense of these requirements, and by extension the place this solenoid pulser in the system, a brief discussion of the HYXI instrument is required. A CAD model of the HYXI instrument, as well as a simulation of the magnetic field in the vacuum vessel, can be seen in Figure 6. Because this system was born in a complex infrastructure of services and systems within the NIF apparatus, a context diagram of the HYXI pulser is shown in Figure 7. A CAD render of the final HYXI solenoid pulser units developed for the HYXI is visible in Figure 8. Figure 5 – System requirements of the HYXI solenoid pulser system : (a) high level system requirements, (b) pulse waveform requirements, (c) physical design requirements, (d) remote control requirements, (e) triggering requirements, and (f) human safety requirements. Clearly, these requirements need further polish. Figure 6 – (left) a CAD render of the HYXI instrument, and (right) a FEMM simulation of the magnetic field lines in the vacuum vessel. HYXI Infrastructure and cabing Solenoid Pulser Global triggering Spares and Operations Offline testing configuration tracking Figure 7 – system context diagram for the HYXI solenoid pulser. NIF-specific names have been replaced with description of group functions for ease of readability. Figure 8 – CAD render of the HYXI solenoid pulser system. System is organized into five rack - mountable units, and are, top to bottom, the system controller, the photocathode pulser, the tapered region 1 pulser, the tapered region 2 pulser, and the hCMOS imager pu lser. For this analysis, infrastructure layout, triggering systems, spares, and configuration tracking was omitted. While important in the context of a greater system design, they change drastically from location to location, and their impact is largely external to system design. Only infrastructure cabling has an impact on system design in the form of capacitor selection, but this too is outside the scope of the system design. This architecture focuses on the systems surrounding those in Figure 3. 2.3. Systems Engineering Methodology If all that were required was a one-size solenoid pulser for a garage project, section 2.1 in addition to physics and circuit simulations to model the magnetic field of the system is entirely sufficient. If we wish to fulfil only the pulsing requirements of the HYXI system, the requirements presented in Section 2.3 are likely sufficient for system development. However, if we want a well-designed system that can perform its function in a multitude of situations, more effort is required. In this work we will use the model-based systems engineering methodology to analyze what works and what doesn’t work for a pulser system using the HYXI pulser system requirements and lessons learned from the DIXI and hSLOS pulsers. Using this, a robust, scalable, and safe architecture was developed to be able to generate a current pulser system as needed by working professionals that will work in a broad spectrum of contexts. For this, the systems engineering toolkit was used to first model the system behavior in Vitech’s GENESYS platform, a model-based engineering (MBE) toolkit used to analyze system behavior, manage the engineering process, and act as a single source of truth of all generated models. An iterative approach was taken in which the requirements suggested functional and physical architectures, which were then modeled with various behavioral diagrams. Analysis of these diagrams was then used to suggest changes to the requirements, and iteration continued until the architecture was considered sufficiently neutral of both hardware vendors and capacitor bank configuration. These will allow the architecture to be more rapidly adapted in situations where capacitor banks are large and require complex storage. 3. Body of Work 3.1. System Requirements The requirements from Section 2.2 were analyzed and refactored, firstly to make them more consistent with standard systems engineering practice, and later to make the architecture more fitting for an open, one-size-fits-all architecture that this work aims to achieve. The updated and reorganized system requirements can be seen in Figure 8. Table 2 lists these changes for quick reference. The new requirements structure is far more focused and easier to navigate than the initial requirements structure, yielding a cleaner system design. Figure 8 – final requirement structure . (a) top level structure, (b) controller requirements, and (c) pulser channel requirements. Some sections, such as R.1.1.2 were truncated to save space. Table 2 – requirement changes and justification. Observe the removal of all requirements pertaining to the waveform shape. This is because these requirements will change from application to application, and even from system to system in the same application. This architecture provides a platform to interface with and control the capacitor bank, not to dictate the usage of the system and restrict it to a particular application. Instead, pulse energy requirements were moved into R.2.3.1 and the system was redesigned to make this requirement true. This is a far more relevant parameter to a general pulser system than specific pulse waveform parameters, such as rise time, flat top time, etc. 3.2. Behavior Modeling and Functional Architecture The expected use cases of the system can be seen in Figure 9, which include configuring and firing the solenoid pulses, diagnosing issues in the system, and updating system software. From these use cases, we can begin to model the system behavior, which captures both the hardware and software working together to accomplish the goal of pulse delivery. System response was modeled with sequence diagrams, the most important of which can be seen in Figure 10, Figure 11, Figure 12, and Figure 13. Finally, a system state diagram can be seen in Figure 14. All these models suggest a functional architecture that can be seen in Figure 15. Pulse Solenoid(s) User Configure pulse(s) <> Query Internal State <> Run System Diagnostic Power Cycle System Engineer Update System Software Figure 9 – system use case diagram. Magnetic NIF Microcontroller Pulsers Solenoids Pulser IBC Initialize Initialize subsystem Subsystem initialized Initialized Triggers Output pulse (4x) Collect current, magnetic field data Wait until cap voltage is 0V Return success Return success Figure 10 - sequence diagram of the solenoid pulser typical usage. Magnetic Magnetic DMM NIF Solenoid Scope Pulser IBC Pulser IBC Initialize Initialize subsystem Subsystem initialized Initialized Trigger Trigger all circuits Wait 25ms Output Pulse (4x) Request cap voltages Return cap voltages Turn on discharge circuit Monitor voltage until discharge complete Turn off discharge circuit Return failure Figure 11 – typical usage of the solenoid pulser system, but system fails output the current pulses. Magnetic Magnetic DMM NIF Pulser IBC Pulser IBC Initialize Initialization Procedure Initialized Canel Operation Clear armed thread Request cap voltages Return cap voltages Turn on discharge circuit Monitor voltage until discharge complete Turn off discharge circuit Return success Figure 12 – normal operation, but pulsing is cancelled. Magnetic Magnetic DMM NIF Solenoid Scope Pulser IBC Pulser IBC Initialize Initialize subsystem Subsystem initialized Initialized Trigger Trigger all circuits Wait 25ms Output Pulse (4x) Request cap voltages Return cap voltages Turn on discharge circuit Discharge circuit fails Turn off discharge circuit Return alarm Monitor capacitors forf 30 min Return all clear Figure 13 - normal operation, but pulsing is cancelled and discharge circuit fails. Failure Reinit Success +Relaunch lost process Check Software Integrity Error Detected +Check hardware IO process No Error Any other Command +Check control logic process No Command +Check network comms process Error State Re-Init Controller Command System Not Ready +Status LED RED Any Command Error Initialization Discharge Charging Init Pulsing Ini Cancel Command +Read default settings from XML file Update Front Board Idle Charge +Check interlocks System +Charge caps Ready +Init harware IO thread No Error Charged +Status LED GREEN +Set timeout timer Pulsers +Check output resistence Ready +Arm monitor channels +Wait for command +Init control logic thread Pulse +Check if system enabled +Prep cleanup threads +Init network comms thread Trigger Run Self Self-Diagnostics Diagnostic +Run diagnostics routines +Generate report Figure 14 – system state transition diagram demonstrates all the possible states of the system. These behaviors are universal to all pulsers, managing trigger states, internal configuration, charging, discharging, and hardware failure safely. These suggest a functional architecture for the system, which in turn suggests a physical architecture for the system. The functional architecture can be seen in Figure 15. 3.3. Physical Architecture Analysis of the system lead to the physical architecture seen in Figure 16. The system supports up to 4 pulser channels, but only one channel is modeled here. This is because all 4 channels have the same architecture, and only the channel number (via an internal jumper on C.2.2) and the number and type of capacitors are expected to change from channel to channel. Thus, only one channel is explored in Figure 16, C.2. The remaining channels, C.3, C.4, and C.5, have been omitted in this report to reduce space. For some components, such as C.2.2, the design does not change from application to application, and thus hardware can be designed or selected that can be used across all use cases. A recommended design for C.2.2 is shown in Figure 17. However, some components, such as C.2.3.1.1 (capacitor) will change from system to system, and as a result C.2.3.1.2 (the capacitor mount PCB). This produces difficulty in maintaining standardized interfaces across different applications. For example, C.2.3.1.2.1, the voltage monitor interface was selected to interface a capacitor to C.2.2. In the HYXI case a 50Ω BNC connector was selected as the interface, and if the system designer wishes to utilize the recommended C.2.2 PCB to control and monitor the capacitor they must use the same component on the capacitor mount PCB. The alternative is to redesign C.2.2, an unnecessary and expensive task. To solve this problem, components where the system designer has latitude in selection have been marked with an <> stereotype, a concept borrowed from UML class diagrams, which is used to specify a standard interface or structure for a component in software engineering, but reminds the engineer that the component should not be used as is, that it is missing critical pieces that make it an incomplete component [11]. This allows system designers to easily identify what components they have freedom to change, and what components should remain unchanged. Figure 16 – physical architecture of the pulser system. (a) top level architecture, (b) controller architecture, and (c) pulser channel architecture. Some parts , such as all children of C.1.1 .1, most of C.1.1.2 , C.2.1, and C.2.2 have been omitted to save space. 3.4. Software Class Diagram Using the rich diagrammatic modeling approach of GENESYS allows for the analysis and development of the software structure to support hardware functions. To this end a class diagram, the software equivalent of the physical architecture diagram, was developed to fully flesh out the software system architecture and behavior. The initial software architecture was simplified from a complex central messaging architecture to a Model-View-ViewModel (MVVM) architecture for a better separation between the external control of the system and the internal representation (model) of the hardware, a block diagram can be seen in Figure 18. Figure 19 has the class diagram of each component of the class diagram. DAQ Control Object Hardware Driver Hardware Emulator Adapter Hardware Controller (Model) Core Logic SCPI Server (ViewModel (View) Figure 18 – block diagram of software for the system major components, including a system managing physical pulser hardware, and a system emulator system. Figure 19 – (a) software class diagram , and (b) the PowerDNA hardware object as a reference hardware management system . Because of a very large number of SCPI command objects, only a representative number was selected to be shown here to save on space. The MVVM architecture has been determined as the ideal architecture as it most elegantly encapsulates the necessary components into one of the three subcomponents, and for allowing for future expansion and growth of the system. For example, one of the desired future expansions for the system is an HTTP server for controlling the pulser easily on an online browser for laboratory use. In some centralized hardware control megastructures, simple network control protocol (SNMP) is utilized for quick integration of new hardware into the megastructure. This architecture allows for the simple, if not easy, implementation of such features in the future. The decoupling of the SCPI server and core logic object was chosen for reuse in other software systems. 4. References 1. Lehr, J., Ron, P. Foundations of Pulsed Power Technology. 1st ed. Hoboken: Wiley, 2013 2. Griffiths, D. J. Introduction to Electrodynamics. 4th ed. Boston: Pearson, 2013 3. Ryszard B., Wolfgang H., Modern phased array radar systems in Germany. In: Proceedings of 2016 IEEE international symposium on phased array systems and technology (PAST). 2016, p. 1-7. 4. Tamer M. A., Ahmed L. E., Amr A. A., Osama A. M., Performance analysis of coil-gun electromagnetic launcher using a finite element coupled model. In: Proceedings of international middle east power systems conference. 2016, p. 506-511. 5. Tsibidis G. D., Stratakis E., Ripple formation on silver after irradiation with radially polarised ultrashort-pulsed lasers. J Appl Phys, 121.16 (2017) (2017), Article 163106 6. Zabihi S., Zare F., Ledwich G., Ghosh A., Akiyama H., A novel high-voltage pulsed-power supply based on low-voltage switch–capacitor units. IEEE Trans Plasma Sci, 38.10 (2010), pp. 2877-2887 7. Nagel S. R. et. al., Dilation x-ray imager a new/faster gated x-ray imager for the NIF. Rev. Sci. Instrum. 83, 10E116 (2012) 8. Engelhorn K. et. al., Sub-nanosecond single line-of-sight (SLOS) x-ray imagers (invited), Rev. Sci. Instrum. 89, 10G123 (2018) 9. Nagel S. R. et. al., The dilation aided single–line–of–sight x–ray camera for the National Ignition Facility: Characterization and fielding. Rev. Sci. Instrum. 89, 10G125 (2018) 10. Almeida J. et. al., Review of the development of cesium iodide photocathodes for application to large RICH detectors. Nucl. Instrum. Methods Phys. Res. A, Volume 367, Issues 1–3, Pages 332-336 (1995) 11. Lehr, J., Ron, P. Foundations of Pulsed Power Technology. 1st ed. Hoboken: Wiley, 2013 12. Pilone, D., Pitman, N. UML 2.0 in a Nutshell. 1st ed. Sebastopol: O’Reilly, 2005