A nanosecond pulse generator must do more than produce a high voltage: it must deliver a sufficiently fast, appropriately polarized trigger pulse while operating reliably at the required repetition rate. A reported design for TDI-series pseudospark switches addresses these demands with a single MOSFET, a step-up transformer and a self-break spark gap. The system produced a 7.8 kV peak trigger voltage and operated at a typical burst repetition rate of 500 Hz.
The design offers a useful case study for engineers working with pulsed power equipment, high-voltage switching and trigger circuits. Its results also show why pulse shaping, component recovery and energy-storage recharge must be considered together rather than treating peak voltage as the only design target.
Why pseudospark switches need a fast trigger
Pseudospark switches, sometimes described as cold-cathode thyratrons, are used in pulsed-power circuits that require high-current switching. The devices described in the study require a negative trigger pulse of at least 4 kV and 50 A. The trigger voltage must also rise quickly: the paper identifies a rate of change greater than 5 kV per microsecond as an important condition for reliable triggering.
Those requirements place specific demands on the driver. A circuit that reaches the required voltage but has a slow pulse front may not provide the desired trigger behavior. Conversely, a steep pulse must be generated without overstressing the switching components or compromising operation between shots.
How the nanosecond pulse generator works
The circuit stores energy in an 8 µF capacitor, which is charged from a rectified AC supply. A single power MOSFET controls discharge of that stored energy through the primary winding of a step-up transformer. The transformer raises the voltage, while a self-break spark gap on the secondary side sharply steepens the pulse front before it reaches the pseudospark igniter.
The MOSFET operates at approximately 330 V, below its 600 V rating. A gate driver controls its turn-on. The design also includes an RCD snubber across the MOSFET to limit the effects of voltage rise associated with parasitic inductance during turn-off, along with transient voltage suppression diodes for momentary overvoltage conditions. These protection elements are important because fast switching can produce high rates of current change and voltage stress.
The spark gap is made from two stainless-steel spheres separated by about 3 mm. When the voltage across the gap reaches its breakdown level, it conducts and sharpens the transformer output. This component is central to the design’s fast pulse front, but it also contributes to timing variation and limits recovery between pulses.
Measured pulse performance
With the pseudospark connected, the generator delivered a reported peak trigger voltage of 7.8 kV and peak trigger current of 63 A. The pulse had a rise time of 5 ns and a full width at half maximum (FWHM) of approximately 30 ns. These values describe the pulse delivered to the switch’s trigger pin, rather than a general performance guarantee for other loads or circuit configurations.
Tests with a 200 Ω resistive dummy load illustrate the spark gap’s pulse-shaping effect. Without the gap, the transformer output reached a peak voltage of 1.6 kV with a rise time of about 6 µs. With the gap installed, the measured pulse front rose in approximately 5 ns. Under that test condition, the gap reduced rise time by roughly three orders of magnitude. The measured peak voltage in the dummy-load test was different from the 7.8 kV reported at the pseudospark igniter, underscoring the importance of specifying the load and measurement point when comparing pulse data.
The generator also incorporated a heater supply for the pseudospark’s hydrogen reservoir. Its output was adjustable from 0 to 12 V and 0 to 3 A. A common-mode choke was included to suppress common-mode noise in the heater circuit. Combining the trigger and heater circuits made the setup an integrated driver arrangement, although their electrical functions remain distinct.
Timing jitter and repetition-rate limits
Timing jitter is the variation in when the trigger pulse begins from shot to shot. Across 35 consecutive single-shot tests, the reported jitter was approximately 50 ns. When the self-break spark gap was removed, jitter fell below 1 ns. The comparison indicates that most of the measured timing variation came from the gap’s breakdown behavior, not the MOSFET trigger circuit. Where lower jitter is necessary, the gap would therefore be a key part of the design to optimize or reconsider.
In burst operation, the generator achieved a typical repetition rate of 500 Hz with a 200 Ω load. Operation at 1 kHz was less reliable, with occasional missed pulses. Two constraints explain this limit:
- Capacitor recharge: The 8 µF energy-storage capacitor must recover its charge before the next MOSFET switching event. At higher repetition rates, less time is available between pulses.
- Spark-gap recovery: The gap must recover after breakdown. Its recovery depends in part on cooling of the hot discharge channel; the gap in this design was not specifically optimized for faster recovery.
These factors are linked. Increasing pulse frequency without allowing adequate recharge can reduce the energy available for a shot, while insufficient gap recovery can prevent consistent breakdown behavior. A repetition-rate requirement should therefore be evaluated alongside the trigger amplitude, pulse shape and jitter requirement.
Practical implications for pulsed-power design
The design demonstrates how a relatively simple switching stage can be combined with a spark gap to generate a fast, high-voltage trigger pulse. Using a single MOSFET supports a compact circuit, while the transformer and spark gap perform voltage step-up and pulse-front sharpening. Protection components and attention to the charging path remain essential parts of the implementation.
For engineers assessing a similar system, the key lesson is to define performance at the actual load and operating mode. The 7.8 kV, 63 A, 5 ns and approximately 30 ns figures describe pseudospark triggering, while the 500 Hz result is reported for burst operation with a resistive dummy load. Jitter and maximum usable repetition rate are shaped by different parts of the system: the spark gap strongly influences both timing consistency and recovery, while capacitor charging sets another frequency constraint.
In practice, selecting or developing a high-voltage pulse generator means balancing pulse-front speed, trigger current, timing repeatability, component protection and recharge time. The reported circuit provides a concrete example of those trade-offs and highlights where further optimization may be needed when application requirements differ.