[time-nuts] Types of noise (was: Phase Station 53100A Questions)

[Attila Kinali]

On Fri, 11 Feb 2022 18:25:05 -0500
Joseph Gwinn <joegwinn at comcast.net> wrote:

> May not realize that thermal noise (additive) and phase 
> noise (multiplicative) are not the same, and do not behave the same.

It seems like you are mixing up here quite a few different concepts:
Phase noise vs amplitude noise, additive vs multiplicative noise,
thermal vs other noise sources, white noise vs 1/f^a-noise.

All these are orthogonal to each other and you can pick and match them.
I.e. Phase noise can be additive, 1/f^2-noise and thermal.

Amplitude and phase noise are looking at noise from two different
perspective. One is how large the variation of the peak of a sine
wave is, the other is how much the zero crossing varies in time.
Note that all natural noise sources will be both amplitude and
phase noise. What makes them one or the other is how we treat them
in our system. E.g., amplitude noise is often relative easy to supress
with some AGC system. Or, passing through amplifiers amplitude noise
and phase noise gets different "treatment" and thus the noise figure
amplifier for each of them will end up being different.

Whether noise is additive or multiplicative is not a property of the 
noise itself, but a property where in the system it appears. E.g. while
in school we always treat all noise sources in an amplifiers additive,
once we get out and start engineering we add up noise figure values in dB.
Which means we treat them as multiplicative noise. In reality most systems
have both additive and multiplicative behaviour, but often one is dominant
over the other. In RF systems, due to having multiple non-linear stages,
noise is almost always dominantly multiplicative. 

Last but not least, the power spectral density or whether noise is white
or 1/f^a is depends which of the noise sources is dominant at which
frequency. Obviously, 1/f^a is the dominant one for low frequencies.
While for white noise we have good models that explain them (either
thermal/Johnson noise or shot noise, usually), we do not have a proper
explanation for 1/f^a noise. Most seem to agree that at least 1/f^1 noise
in semiconductors, the most likely cause is trapping of electrons (though
that explanation has quite some problem). Please note that for very low
frequencies what we declare to be noise in noise spectra is usually related
to environmental effects, which I personally would not bunch together with
"normal" noise processes, as their treatment/mitigation has to be done
differently.

			Attila Kinali

-- 
The driving force behind research is the question: "Why?"
There are things we don't understand and things we always 
wonder about. And that's why we do research.
		-- Kobayashi Makoto