Harmonics of a fundamental frequency

[0xFFF0]

Hi all, I don't have a good spectrum analyzer and have a question: Do the harmonics of a fundamental frequency actually have more jitter or is it all 100% phase-locked?

[ejeffrey]

Mostly harmonics will be synchronous with the fundamental.  However that corresponds to more phase noise because 1 picosecond of jitter is more radians of a 3 GHz third harmonic than the 1 GHz fundamental. 

[gf]

Hi all, I don't have a good spectrum analyzer and have a question: Do the harmonics of a fundamental frequency actually have more jitter or is it all 100% phase-locked?

The terms "fundamental" and "harmonics" are defined for periodic waveforms (see https://en.wikipedia.org/wiki/Fundamental_frequency).
An ideal periodic waveform does not have jitter, and a jittery waveform is strictly no longer periodic, but rather has a spread spectrum in the frequency domain.

[RoGeorge]

Hi all, I don't have a good spectrum analyzer and have a question: Do the harmonics of a fundamental frequency actually have more jitter or is it all 100% phase-locked?

You are asking about jitter, which is about the time domain, yet talking about a spectrum analyzer and asking about phase, thought a SA usually measures the magnitude of the harmonics in the frequency domain, and does not measure the phase.  An SA is to display the power distribution of the given signal into its harmonics, and phase doesn't matter.

Time jitter and phase noise are not the same.  In general, the longer you wait to measure time deviations, the larger the observed time jitter (relative to where you would expect an edge of a signal to be).

Harmonics are "phase-locked" with the fundamental, or else (if the phase of harmonics wouldn't be locked) that would mean that the fundamental changes its shape in time, so it wouldn't be a periodic function any more.



What exactly do you want to measure, or to find out, or in what context is the question you were asking?

[radiolistener]

Do the harmonics of a fundamental frequency actually have more jitter or is it all 100% phase-locked?

For what reason do you need it? If you want to use it for frequency multiplier, the output phase noise will depends on both - phase noise of the the source clock and phase noise of multiplier.

[coppercone2]

You need to use the math formula for harmonics

If you see a square wave edge, the phase has to be correct for it to be a good square wave edge.

So if you see a square wave, and what the square wave is doing is having a rise time that is slightly off each time you graph it, but each time you graph it, its still a strait line on your scope per cycle, you know that the harmonics are in phase. Think about one-shot on your scope. Do you expect to see anything other then a square wave EVER if you keep one-shotting a functional circuit? I expect noise to effect things, but there is always gonna be a vertical line  (leading edge) .. which is mathematically composed of a series of sine waves in a precise orientation for summation in your mind to a 'infiniteish rise time"

So you can decompose an ideal square wave and move around phase in the series to see what the 'ideal vertical line' looks like when you mess with the phase relationship required for a strait vertical line (which is diagnol if you scope can zoom in fast enough but I assume ~5ns pulse for this.

What is it.. comparator noise. The hysterisis/noise margin designed into the device is a design choice like heat dissipation or voltage level or switching speed is.. or at least a variable related to these things, so you end up with tradeoffs in circuits.