Get A Handle On Oscillator Phase Noise

Oct. 20, 2011
As electronic applications proliferate and their frequencies rise, designers are increasingly calling for time bases that are more stable with lower noise. For oscillator designers, the challenge is lowering phase noise. Every oscillator is ...

As electronic applications proliferate and their frequencies rise, designers are increasingly calling for time bases that are more stable with lower noise. For oscillator designers, the challenge is lowering phase noise. Every oscillator is inherently unstable, and this instability manifests itself as a spectrum of noise around the oscillator's resonant or tuned frequency. This noise band, which is generally measured from the carrier to 1 MHz away from the carrier, is depicted as a graph of dBc/Hz versus offset frequency from the carrier, f(Hz). in an application note titled, "Oscillator Phase Noise: Theory vs. Practicality," Greenray Industries explains why phase noise is actually a manageable problem.

The 12-page document begins by looking at different time bases and their properties. Compared to cesium and rubidium, quartz is generally chosen as a resonant material for consumer electronics because it is cheaper, smaller, and has a high quality factor (Q). A look at the different types of crystal oscillators follows, with oven-controlled crystal oscillators (OCXOs) and double-oven-controlled crystal oscillators (DOCXOs) promising the highest stability. Despite their high Q, however, real-world oscillators suffer from some amplitude and phase fluctuations. Their frequency is affected by temperature, long-term drift (aging), and short-term instability. The oscillator's frequency drift will diminish over time, though, which bodes well for long-term performance.

Jitter also is discussed in the application note, as it represents a change in the time of the waveform edge from the ideal nominal-frequency edge. Because most applications will operate within a certain band of frequencies, the jitter effect only needs to be measured in that band. This step calls for conversion to the frequency domain, which results in the phase-noise measurement. The note explains one way to measure jitter and then calculate phase noise at a given frequency. Equations are provided as well.

The Leeson equation also is included for a scenario that accounts for real-world components and the noise that is generated in those circuits. A graph shows how this equation fits into the phase-noise plot. Utilizing the plot, it is easier to come up with guidelines for minimizing phase noise in oscillator designs. Devices with low-flicker noise should be used. Because bipolar junction transistors (BJTs) have a much lower flicker noise than field-effect transistors (FETs), they are more suitable for low-phase-noise applications. Higher drive power also is desirable, as it impacts the phase-noise floor. The note ends with an in-depth section on making phase-noise measurements and the types of instruments required.

Sponsored Recommendations

Forging the Future of Defense

Oct. 11, 2024
Raytheon’s Advanced Technology team incubates capabilities that fuel the future of defense. Together with leading research and development organizations, def...

Phase-Matched Cable Assemblies

Oct. 8, 2024
Phase-matched cable assemblies are ubiquitous, and growing in popularity. Electrical length matching requirements continue to tighten and the mechanical precision of cable construction...

3 New Wideband MMIC LNAs Cover 5.5 to 20 GHz

Oct. 8, 2024
Mini-Circuits’ expanded PMA3-series of wideband, ultra-low NF MMIC amplifiers operates in ranges between 5.5 and 20 GHz.

Wideband Amplifiers Variable and Temperature-Compensated Gain

Oct. 8, 2024
Many types of RF systems and applications that span from the upper end of microwave frequencies to the lower end of mmWave have arisen in recent years. Meeting system requirements...