RF (Frequency & Wavelength) vs Pulse (Frequency)

[chiques]

I know this:
Frequency in RF is the oscillation rate of electromagnetic radiation. The wavelength is inversely proportional to the frequency.

I don't know how to relate to this:
Frequency in a pulsed system is the amount of pulse cycles that occur per second.

***************************************************************************************
Is the wavelength also inversely proportional in a pulsed system (PWM)?? That doesn't sound right to me.

I'm trying to get clarification on this because many terms, parameters and even specifications in RF are used in high speed digital systems. That's never sat well with me so I'm just trying to understand the way the terms are used.

 

[danadakk]

A basic single frequency definition (sinewave) -

Pulse essentially same -

Freq and wvaelength same, but pulse has many harmonics in it, so the "root" freq and wavelength is
the overall composite of all harmonics and effect on wave shape.

Demo of building a square wave out of harmonics. https://pages.uoregon.edu/emi/14.php


Regards, Dana.

 

[BradtheRad]

Certain groups of parameters go with certain types of modulation.
Duty cycle is a major parameter in pulsed communication.

The notion of wavelength and electromagnetic radiation is of more interest in radio transmission where sine waves are a chief carrier. Example, wifi, bluetooth.

Broadcasting pulses over the airwaves is frowned upon. It's in the same category as EM interference.

Also when frequency is high enough a hardware circuit needs to consider the wavelength. Ten gigahertz has a wavelength of 1 inch. To generate that from a physical circuit requires that it be smaller than one wavelength.

 

[chiques]

From what I read there is no 'wavelength' on a pulsed waveform. There is a period/cycle, but nothing related to 'wavelength'.

In RF the wavelength + impedance determine what physical size a media has to have in order for the signal to propagate without power loss. For example, trying to transmit a Bluetooth signal through two jumper wires typically won't work. Coax or transmission lines with a characteristic impedance of 50 Ohm is a better way of doing it. This is due to the inherent nature of sine waves generating electromagnetic fields.

In pulsed system, ribbons are commonly used in many types of systems. Power loss is not much of an issue since the systems are detecting the peak voltages. Ethernet cables capable of transmitting very fast switch rates are simply twisted pair wires and much cheaper that RF cabling.

To measure:
RF (frequency domain) - Network Analyzers & Spectrum Analyzers are a common instrument

Pulse (time domain)- I've seen equipment that display an eye diagram. From what I understand this measures how good a pulse looks through a circuit. I don't know if there is frequency domain equipment for measuring pulses.

 

[FvM]

Spectrum of pulse signals matters the latest if you care for electromagnetic compatibility of your design, e.g. radiated emissions that violate regulations.

A rule of thumb says that circuit wiring (e.g. PCB traces or cables) should be analysed as transmission lines if the length exceeds 1/10 of signal wavelength. In case of pulse waveform, it's not the pulse period or width but the rise time that determines the frequency spectrum and respectively shortest wavelength to be considered.

 

[danadakk]

From what I read there is no 'wavelength' on a pulsed waveform. There is a period/cycle, but nothing related to 'wavelength'.

Yes and no. No in that the non sinusoidal waveform have constituent sine components, each with a
phase and amplitude and wavelength. We can talk about a wavelength associated with a non-sinusoidal
signal, such as -

Wavelength - Wikipedia

en.wikipedia.org

But for design purposes we need to think of a waveforms occupied band width in order
to allow its harmonic constituents to propagate thru circuit of interest. Or for rejecting
parts of its harmonics, eg. filtering, purposely removing harmonics not of interest.


Regards, Dana.

 

[c_mitra]

Frequency in RF is the oscillation rate of electromagnetic radiation. The wavelength is inversely proportional to the frequency.

The key concept here is a process taking place at regular interval. That is necessary to understand frequency.

Another important concept is the epoch. When it started? When it is going to end? If we are in the middle of a wave train, you can talk about frequency. Not really if you are at the beginning or near the end.

Another device is often said but poorly understood. That is a harmonic oscillator. That is the model of all models. The device lets you see all periodic processes as made up of harmonic oscillators. We are of course talking about sine waves.

For the harmonic oscillator, the frequency is a fundamental parameter that defines the oscillation. You can get away without invoking wavelength at all.

But when we talk about the propagation (e.g., EM radiation moves from point A to point B), we need to define the speed and wavelength (they are just related - the corresponding point principle).

With harmonic oscillators, we have to introduce phase- related to epoch- because that is very important if we want to couple two oscillators. Phase of any periodic process is usually defined by superimposing a sine wave.

Once you have understood these basics, wavelength can be easily defined for ANY periodic process, provided we can define a velocity.

By the way, what was the question?

--- Updated Sep 11, 2020 ---

To generate that from a physical circuit requires that it be smaller than one wavelength

This is not accurate.

The cosmic microwave background has a peak around 160GHz. That corresponds to a wavelength of less than 1mm.

Compare that to the size of the universe!

Also consider noise production. Does size matter? Wise men may tell.

 

[danadakk]

Planck units - Wikipedia

en.wikipedia.org

Electromagnetic spectrum - Wikipedia

en.wikipedia.org

The root of it all ? (not a physicist here)



Regards, Dana.

 

[c_mitra]

The root of it all ? (not a physicist here)

Let me try to put this differently.

Consider a pendulum oscillating: ignore damping for the time being. This is a repetitive process and is associated with a period and frequency.

Do we have a wavelength for this? Not really because nothing is getting propagated here. To define wavelength we need to define velocity.

Enter electromagnetic waves with their modulations! Now we have group velocity and phase velocity. They are same most of the time but not always.

The problem gets complicated if the EM waves travel in a dispersive medium. In a dispersive medium, the velocities depend on the frequency.

In essence, the frequency is a more fundamental concept compared to the wavelength. Consider, just for fun, the wave traveling in a circuit board between several components.

 

[crutschow]

The cosmic microwave background has a peak around 160GHz. That corresponds to a wavelength of less than 1mm.

Compare that to the size of the universe!

But the microwave background was created by the "Big Bang" when the universe was infinitesimally small.