Maintaining strong frequency stability in electronic RF circuits by eliminating phase noise is important in many high-end communication applications. This is especially true for precise targeting in radar systems and spectral purity in other communication systems.
Let's take a deep dive into exactly what the terms phase noise and jitter mean. This will help give you a better idea why reducing a system's phase noise is so important.
What Is Phase Noise?
Phase noise is represented in the
frequency domain of a waveform and consists of rapid, short-term, random fluctuations in the phase (frequency). This is caused by time domain instabilities (jitter).
Be sure not to confuse phase noise with jitter. Jitter is a method of describing the stability of an oscillator in the
time domain. It combines all the noise sources together and shows their effect with respect to time.
In the simplest terms, phase noise describes the stability of an oscillator in the frequency domain, while jitter describes stability in the time domain.
A Simple 5-Step Process to Understanding Phase Noise
To build an in-depth understanding of phase noise, try using this simple 5 step process:
Step 1: Spectral Density
Spectral Density is a measure of a signal's power intensity in the frequency domain. The spectral density provides a useful way to characterize the amplitude versus frequency content of a random signal.
When plotting each spectral density point at varied frequency intervals of your choosing (In this case every 1Hz), you're left with a graph that looks like this:
Step 2: Signal Power Spectral Density
You are now looking at what is known as the signal power spectral density of the noise.
Now focus only on the upper sideband of the graph from fstart to fstop; this is called the "single sideband."
Step 3: Noise Power Density
We can now refer to the plotted part of the single sideband as noise. Anything above the nominal oscillator frequency (Fosc) and not harmonically related can be considered phase noise. The technical term for this part of our graph is noise power density. We measure noise power density in dBW (LOG(Watts)) at this point because of the large range at which we are looking.
Step 4: SSB Noise Density
When we combine the single sideband and noise power density, we are actually measuring what's called SSB (single sideband) noise density.
Step 5: Phase Noise
Finally, when we look at this in the time domain and see a "jittery" waveform, we are looking at "jitter". Because the jitter is much smaller than one complete period, we can say it is caused by "phase fluctuations" (instead of frequency fluctuations). Since these fluctuations are noise, it's actually phase noise.
In other words:
SSB noise density = phase noise
...and that's where phase noise comes from! Easy, right?
What Causes Phase Noise?
Phase noise in higher-end applications (such as radar communications, military communications, and space and satcom communications) is typically caused by:
- High vibrations
- Micro vibrations
- G-forces and acceleration sensitivity
Using an
anti-vibration, g-sensitivity crystal oscillator is the best way to eliminate phase noise from all of these potential sources.