A nonlinear transmission line (NLTL) can convert an input pulse into a signal containing high-frequency oscillations, offering an alternative approach to pulsed RF generation. A laboratory study of a low-voltage, 20-section capacitive NLTL demonstrated oscillations around 230 MHz, extracted them with a Chebyshev high-pass filter, and transmitted the resulting RF signal between antennas. For automation and electronics engineers, the work illustrates how nonlinear components, filtering, simulation, and measurement can be combined to evaluate a pulsed signal chain.
How a capacitive nonlinear transmission line generates RF
A capacitive NLTL consists of repeated inductance-capacitance sections. In the tested design, inductors were paired with varactor diodes, whose capacitance changes with applied reverse voltage. This voltage-dependent capacitance makes the line nonlinear: different portions of an input pulse encounter different capacitance and therefore propagate at different speeds.
As pulse amplitude rises, varactor capacitance falls. The higher-amplitude portion of the pulse consequently travels faster than lower-amplitude portions. As these parts interact along the line, the pulse can steepen and develop oscillations, producing an RF-bearing waveform near the output. The propagation velocity depends on both inductance and capacitance, so changes in diode capacitance also affect the line’s characteristic impedance and frequency behavior.
The experiment used 20 sections with 100 nH inductors and SVC236 varactor diodes. A 13 V peak input, repeated at 5 MHz, produced oscillations around 230 MHz, with a reported voltage modulation depth of about 5 V. These results describe the tested setup; they are not a universal output specification for NLTL designs.
Measuring the output before and after filtering
Researchers first measured the NLTL output across a 50-ohm resistive load. A 30 dB attenuator connected the circuit to the oscilloscope, and the waveform was examined in both time and frequency domains. The time-domain view showed the pulse structure and repetition, while a Fast Fourier Transform (FFT) identified the RF content around 230 MHz.
To separate the higher-frequency component from lower-frequency portions of the input pulse, the team added a third-order Chebyshev high-pass filter with a nominal cutoff of 200 MHz. Its T topology used series capacitors and a shunt inductor. The filter was characterized separately using a vector network analyzer before being connected to the NLTL.
In the reported measurements, the filter had less than 1 dB insertion loss within its passband and strongly attenuated lower-frequency components, with attenuation reported at about 60 dB. The oscillation frequency remained approximately 230 MHz when measured directly at the load and after filtering. This distinction matters in signal-chain design: the filter isolates the desired RF content, while the NLTL remains the element that generates the oscillatory waveform.
Using simulation to check circuit behavior
An LT-SPICE model represented the 20-section line and its measurement configurations, including operation with a resistive load and with the high-pass filter. The model also accounted for specified circuit losses and inductor self-stray capacitance. Comparing simulated and measured waveforms showed good agreement, supporting the model as a tool for interpreting the experimental behavior.
For engineers evaluating a nonlinear circuit, combining time-domain and frequency-domain results is useful because neither view tells the whole story. Time traces reveal pulse shape and timing; FFT results show where energy appears in the spectrum. Comparing both with a circuit simulation can help assess whether the model captures the observed response and whether the filter is performing its intended separation.
Radiating and receiving the pulsed RF signal
After load and filter measurements, the RF signal was transmitted and received using a pair of Double-Ridged Guide antennas. The antennas were vertically polarized and placed two meters apart in an anechoic chamber to reduce interference. The receiving antenna was connected to a 50-ohm oscilloscope input.
The received waveform and its frequency-domain analysis showed that the RF signal could be detected over the antenna link, with the oscillation remaining near 230 MHz. This completed an important step beyond observing a voltage at the circuit output: it demonstrated transmission and reception of the generated pulsed RF signal using antennas operating across a frequency range that included the measured oscillation.
Practical considerations for NLTL evaluation
The experiment highlights several considerations for engineers assessing NLTL-based RF sources:
- Account for voltage-dependent behavior. Varactor capacitance changes with reverse voltage, influencing propagation speed, characteristic impedance, and the generated waveform.
- Measure at multiple points. A resistive load, a filter output, and an antenna receiver provide different views of signal generation, extraction, and transmission.
- Design around the signal path. The high-pass filter was selected to pass the oscillation near 230 MHz while reducing lower-frequency components; its insertion loss was checked independently.
- Use simulation alongside measurements. Agreement between LT-SPICE results and experimental observations provides a useful cross-check of circuit behavior.
The study positions low-voltage, lumped capacitive NLTLs as practical experimental platforms built from commercial components. It also notes potential areas of interest such as radar, communications, and medical applications, while the reported work focuses on laboratory signal generation and antenna transmission rather than validating a deployed system.
Conclusion
A capacitive nonlinear transmission line can produce pulsed RF through the interaction of a propagating pulse with voltage-dependent varactor capacitance. In this 20-section test, the output oscillated near 230 MHz, a Chebyshev high-pass filter separated the higher-frequency content, and Double-Ridged Guide antennas transmitted and received the signal. For engineers, the practical lesson is to evaluate the complete chain—nonlinear line, load, filter, simulation, and antenna measurement—rather than treating RF generation as an isolated circuit result.