A Marx generator creates a high-voltage pulse by charging capacitors in parallel and then connecting them in series. In a solid-state Marx generator, semiconductor switches such as MOSFETs perform the switching that conventional designs assign to spark gaps. This approach offers engineers and educators a way to study staged voltage multiplication and impulse generation in a compact laboratory setup.
A reported 11-stage design combines a boost converter, capacitor stages, diodes, MOSFETs, and pulse-control circuitry. Its simulated and experimental configurations illustrate the operating principle, while also showing why component selection, switching control, and insulation are central to high-voltage design.
How a Marx generator multiplies voltage
Each stage stores energy in a capacitor. During the charging interval, the capacitors are charged in parallel from a supply, so each holds approximately the same voltage. During discharge, the switching arrangement connects the charged capacitors in series. Their voltages then add together, producing an output that can approach the supply voltage multiplied by the number of stages.
This multiplication is an idealized result. Practical output is generally lower because of losses and circuit constraints. The generator’s purpose is typically to produce a brief high-voltage impulse rather than to provide a continuous, steady DC supply.
In a conventional Marx generator, spark gaps act as switches. Charging resistors limit current while the capacitors charge, and the gaps break down to initiate the series discharge. The design described in the project instead uses MOSFETs and diodes. The electronic switches control the transition between charging and discharge, while diodes provide the charging path and help isolate parts of the circuit.
Solid-state Marx generator operating sequence
Charging the stages
The project’s system begins with a low-voltage DC source and a boost converter. In the reported simulation, a 24 V input and a 94% MOSFET duty cycle produce a boost-converter output of about 390 V DC. That elevated voltage feeds the Marx stages. A 555 timer configured in astable mode supplies pulses for the switching sequence.
During the charging state, the stage capacitors are charged in parallel through the diode paths. The timing circuit and switching arrangement determine when the circuit changes state. A design must allow the capacitors to reach their intended charge before initiating the discharge sequence.
Connecting capacitors in series
When the switching state changes, the MOSFETs connect the capacitors in series so their stored voltages add at the output. In the ideal case, an 11-stage arrangement supplied at 400 V would produce approximately 11 times that voltage. The project reports an output above 4 kV for a 400 V input, which is below the ideal 4.4 kV multiplication figure.
The paper also describes a separate five-stage prototype driven by an Arduino-based pulse circuit, with a reported 48 V input and a spark at the output gap. These results refer to different configurations; they should not be treated as measurements from the same setup. The distinction matters when comparing simulated performance, an 11-stage circuit, and a physical prototype.
Design considerations for high-voltage pulses
Stage count alone does not determine performance. Capacitor sizing, charging time, discharge timing, switching behavior, load conditions, and insulation all influence the delivered pulse. The source’s design calculations select capacitance based on output voltage, load resistance, and a specified time interval, then relate the equivalent capacitance to the number of stages. Such calculations are part of matching the generator to its intended load and pulse requirements.
Charging paths also involve trade-offs. In conventional circuits, charging resistors limit current but can contribute to longer charging times and lower efficiency. Semiconductor-based designs replace spark-gap switching with MOSFETs and use diodes in the charging and isolation paths. The paper presents these changes as a way to improve controllability and support a more compact implementation; actual results depend on the complete circuit and its components.
High voltage also makes insulation a design priority. The paper notes that conventional Marx generators may use transformer oil or high-pressure electronegative gas for insulation. Any implementation must account for the voltage present at each stage and at the output, as well as the layout and separation of energized components.
Where this design can be useful
The reported generator is intended for small-scale industrial use and academic laboratories. Potential uses identified in the paper include demonstrating impulse voltage generation and carrying out lower-rating tests on insulators and transformers. For system integrators and plant engineering teams, the design is relevant as a teaching and evaluation platform for staged energy storage and switching—not as evidence that a particular configuration is suitable for every production test requirement.
Simulation can help examine circuit behavior before hardware is assembled. The project uses Proteus for the boost-converter and Marx-circuit simulations, then compares simulated and experimental output magnitudes. Keeping the stages of verification distinct is useful: a simulated waveform, a measured output, and a prototype demonstration each describe different evidence about circuit behavior.
Practical implications
A solid-state Marx generator demonstrates how parallel charging and series discharge can turn a moderate DC input into a brief, much higher-voltage output. MOSFET switching provides an alternative to spark gaps and enables electronic control of the discharge sequence, while the boost converter supplies the elevated charging voltage needed by the stages.
For engineers assessing this architecture, the key takeaway is to evaluate the complete pulse-generation chain: input conversion, capacitor charging, switch timing, series connection, load, and insulation. Reported stage count and ideal voltage multiplication are useful starting points, but simulation and hardware results must be interpreted in the context of the specific configuration and test conditions.