All projects

Project · 2026

RF · Oscillators · Measurement

SDR Local Oscillator

A discrete quadrature ring oscillator characterized across frequency, supply noise, interference, and FLL-control experiments.

Overview

This local-oscillator experiment used a direct-quadrature ring topology built from discrete components. I characterized its tuning range and jitter, compared its spectrum with an Si5351 clock generator, and explored a frequency-locked-loop control path.

Completed SDR local-oscillator board with controls and two OLED displays.

The assembled SDR local-oscillator board.

Hand-built discrete ring oscillator on copper-clad board.

The discrete ring-oscillator prototype.

Schematic of the direct-quadrature ring-oscillator topology.

The direct-quadrature topology used for the oscillator.

Frequency range and spectral behavior

The oscillator operated from roughly 4 kHz at the low end to about 22.5 MHz at the high end.

Oscilloscope capture of the oscillator running at approximately 4 kHz.

Low-frequency operation at approximately 4 kHz.

Oscilloscope capture of the oscillator running at approximately 22.5 MHz.

High-frequency operation at approximately 22.5 MHz.

The discrete oscillator showed visible jitter. For comparison, the Si5351 produced a much cleaner spectrum.

Spectrum-analyzer view showing jitter and sidebands around the oscillator carrier.

Jitter observed on the discrete oscillator.

Spectrum-analyzer view of the Si5351 output with a cleaner carrier.

Si5351 spectrum used as a comparison.

Interference and control experiments

OLED activity coupled noise into the oscillator supply. With the displays enabled the output became erratic; disabling them removed much of the disturbance.

Spectrum and spectrogram comparing oscillator behavior with the OLED displays enabled and disabled.

The effect of display-related power-supply noise.

Clean oscillator waveform before interference is introduced.

Waveform without the interfering activity.

Distorted oscillator waveform while interference is present.

Waveform with interference present.

I also tested an FLL control voltage, but the oscillator’s jitter made stable control difficult.

Oscilloscope trace of the FLL control voltage varying over time.

FLL control-voltage behavior during the experiment.

PWM filtering

The likely cause of the jitter was the ESL of the filter capacitors on the PWM, which would cause spikes to remain on the drive voltage of the VCO. The effect of this is the pulses frequency modulate the oscillator, which also explains the jitter of the frequency, having stable periods and limited excursion.

Oscilloscope capture showing a PWM waveform and its filtered counterpart.

PWM and filtered waveforms.

Oscilloscope capture of a short ESL-related pulse on the filtered PWM output.

An ESL-related pulse on the filtered PWM waveform.