Harmonics is one of those things that few understand especially

[psparky]

Harmonics is one of those things that few understand...especially

Harmonics is one of those things that few understand...especially me.

Does anyone have a real understanding on the subject as it pertains to electrcial engineering?

At least introduce some basic things about it that even the simplest mind can understand?

 

[berkeman]

Harmonics is one of those things that few understand...especially me.

Does anyone have a real understanding on the subject as it pertains to electrcial engineering?

At least introduce some basic things about it that even the simplest mind can understand?

The typical example is to observe how you add odd harmonics to a sine wave to gradually build up a better and better square wave. See down the page at this link, for example:

http://www.facstaff.bucknell.edu/mastascu/elessonshtml/Freq/Freq4.html

.

 

[psparky]

Well, I'm already aware that Fourier analysis is an infinite number of waves adding up to one wave.

I'm more interested in harmonics...like the 3rd harmonic...or the 5th harmonic.

If the frequency is 60 hz...then the 3rd harmonic happens at 180 hz...5th at 300 hz etc.

I think that harmonics are more like noise or disturbances in a system and ussually show up as a spike in a graph.

However, I don't know what much of this means. For example "the 3rd harmonic happens at 180 hz"...what does that mean!

And yes, I could probably pick up a textbook and learn it over days...but was just wondering if there was anyone wise enough to explain the basics of it and so forth.

Ideally, I would like to know this subject well enough to teach it...like most subjects.

Can anyone teach this subject?

 

[berkeman]

Well, I'm already aware that Fourier analysis is an infinite number of waves adding up to one wave.

I'm more interested in harmonics...like the 3rd harmonic...or the 5th harmonic.

If the frequency is 60 hz...then the 3rd harmonic happens at 180 hz...5th at 300 hz etc.

I think that harmonics are more like noise or disturbances in a system and ussually show up as a spike in a graph.

However, I don't know what much of this means. For example "the 3rd harmonic happens at 180 hz"...what does that mean!

And yes, I could probably pick up a textbook and learn it over days...but was just wondering if there was anyone wise enough to explain the basics of it and so forth.

Ideally, I would like to know this subject well enough to teach it...like most subjects.

Can anyone teach this subject?

Did you look at the waveform diagrams in the link? If you add the correct amplitude and phase of the 34d harmonic to a sine wave, you get a little squarer sine wave with some ripples at the transitions. Add in the correct amplitude and phase of the 5th harmonic, and it gets closer to a squarish wave. You don't need many harmonics added to get a pretty good square wave.

 

[psparky]

Your answers are perfectly good and all...but maybe I should ask a better questions.

Lets take a factory for example operating at 60 hz with many, many electrical motors and controlers. It's my understanding that I get harmonics...also known as noise and disturbances...that can cause problems, ripple, voltage drop...crappy signals, etc.

What exactlly is going on here?

 

[yungman]

From my pass experience, when you run electric motor, you get spikes with period of 60Hz riding on top of the 60Hz, usually at the same spot on the 60Hz. That is due to connect and disconnect of the coil in the motor. Those all can be transform into higher harmonics using Fourier Transform.

I really don't understand what your question is, you pretty much answer your own question. If you don't accept Fourier Transform, I don't know what else to say. I don't have to go through all the detail.

Maybe you should concentrate on the time domain analysis instead of frequency domain analysis( Fourier Transform). Look at how the motor work, when the center turn, coils in the middle thing( I don't even know what it call!) get connect and disconnected and voltage spike when disconnected and reflect into the power line as a spike. Of cause it is 60Hz spike. You have multiple spike because you have quite a few coil connect and disconnected every revolution. So you see spikes on the 60 Hz.

 

[jim hardy]

maybe you're mixing ideas...

study of harmonics infers to me a steady state situation
and you seem to be describing transients.
a harmonic is simply an integer multiple of a frequency.

when i was taught about Fourier transform of a function, Prof said the groundrules are:
must be continuous
must have a period, ie repeat

if it meets those two it can be expressed as a sum of a fundamental frequency, 1/the basic period, plus a series of harmonics.
One can write the Fourier transform of anything that's repetitive, from a pure sinewave monotone note which has a short period to Beethoven's Ninth which has about 55 minute period. (isn't middle C ~261.6 hz, not quite 4 msec period ? )

next we went through the exercise of graphically constructing an arbitrary waveform by adding harmonics. it is possible to derive your individual Fourier term constants graphically.
i recommend that exercise to any student of electrical engineering.
Awareness of the technique is helpful when troubleshooting with an oscilloscope.
you'll soon pick up that odd harmonics distort both peaks of a sinewave symmetrically and even ones bring about asymmetry.

I had good fortune to be in on some early work developing handheld FFT analyzers now widely used for diagnosing rotating equipment. It is astonishing to observe what an expert can find in industrial instrument sensor signals using a FFT analyzer. We saw several pump speeds, natural mechanical frequencies of reactor core internals, mechanical resonance of stuctures supporting sensors, you name it it was there.
what a blast!



A practical book to have is "Vibration Primer" published by Bentley-Nevada company.

old jim

 

[paulfr]

Yes you are mixing concepts here both [many] of which share the need for
the concept of a harmonic frequency.

When a system is excited by an impulse, it causes the system to resonate
at its characteristic frequency, also called the Fundamental frequency.
But it also can oscillate at multiples of this fundamental frequency.
Why ?
Think musical instruments. The guitar or piano string resonates at a frequency determined by its physical parameters [size, length, material, etc.] But other higher order frequencies can also resonate in this physical structure and they do in response to a disturbance like plucking the string.

So harmonics are a common phenomena in nature due to the compatible relationship
between the mathematics and the physical structure.

 

[Studiot]

Here is a real world, non mathematical answer.

There are many forms of interference or in other words unwanted signals.

As Old Jim says, transients may or may not have something to do with harmonics. You can analyse harmonically the momentary dimming of the lights when you switch on some heavy machinery or you can simply attribute the phenomenon (transient) to the switch on.

True noise has zero harmonic content, its structure is totally random. Noise analysis recognises many types and forms of noise, eg shot noise and flicker noise to name but a few.

Hum (pickup) in electronic equipment is susceptible to harmonic analysis however.
If the frequency of the hum is the same as the mains (60Hz) it is probably due to direct pickup in the wiring.
If, however the frequency is twice the fundmantal (120Hz) then it is due to inadequate filtering of the rectifier output in the power supply.

Distortion is another unwanted signal.
Distortion can be analysed as an addition of harmonics and can be reduced by using the fact that certain circuit configurations cannot pass even order harmonics.

http://en.wikipedia.org/wiki/Push–pull_output

go well

 

[paulfr]

Again one needs to define what type of environment you are dealing with.
For an open loop system like a musical instrument, harmonics arise because
many frequencies can be sustained by the resonant structure.
But on a factory floor, one now has systems problems which means feedback
loops, multiple grounds and the oscillations that result from this complexity.
Harmonics here are created by loop instabilities.

I think you are making this harder than it needs to be.
Or your foundation in Physics is weak.

But the subject is fascinating so keep at it.
You'll get it eventually.

 

[Studiot]

Good morning, Paul.

I'm not sure where you are coming from or to whom you are addressing your comments but be aware that this is the electrical engineering forum and in particular this was the question from psparky

Does anyone have a real understanding on the subject as it pertains to electrcial engineering?

go well

 

[psparky]

Interesting comments...I believe I was referring to steady state analysis.

Perhaps, I am so weak on the subject that I don't even know the correct question to ask!

I have emailed this question straight to the top...so I shall suceed eventually.

This is a difficult subject and the bell isn't quite ringing yet, but it will.

Keep the comments coming.

And yes, I was referring to electrical engineering, but the musical connect with the subject is certainly fascinating as well.

 

[paulfr]

Electrical Engineering is a subset of Physics
If you want to understand Harmonics, you need to understand the
definition first and the common sense of the basic Physics behind it.
In my opinion anyway.

To answer Studiot ...
I am coming from 35 years of experience working in Nuclear Physics
at several Particle Beam Accelerators in Berkeley.

 

[psparky]

I hear you paulfr...

Electricity in physics and in electrical engineering are two different beasts.

When taking physics II a ways back, I did have a difficult time with the way physics approaches electricity. Ironically, I am very strong in electrical engineering (sans harmonics and some other things).

However, I am strong in other physics categores such as statics, dynamics, structrual...etc.

Weird huh?

 

[Studiot]

So what did you make of my comments in post#9?

 

[psparky]

I read it...interestingly, I do play guitar, electric or acoustic.

I am aware of all the notes and how you can tune your guitar by using the harmonics on double dot fret and 5 frets to the left on one string lower without actually pushing down on the string...just touching it on top of the fret. If you push down on the string you get a different tone I believe.

The 12th fret is obviously one octave higher and twice the frequency.

Not sure where I'm going with this...just throwin it out there.

I am seeing that somehow when frequencies double or land on top of each other the tend to resonate or something to that degree.

 

[Studiot]

I read it......

Was this an answer to me or someone else since on my screen post#9 looks like this

Here is a real world, non mathematical answer.

There are many forms of interference or in other words unwanted signals.

As Old Jim says, transients may or may not have something to do with harmonics. You can analyse harmonically the momentary dimming of the lights when you switch on some heavy machinery or you can simply attribute the phenomenon (transient) to the switch on.

True noise has zero harmonic content, its structure is totally random. Noise analysis recognises many types and forms of noise, eg shot noise and flicker noise to name but a few.

Hum (pickup) in electronic equipment is susceptible to harmonic analysis however.
If the frequency of the hum is the same as the mains (60Hz) it is probably due to direct pickup in the wiring.
If, however the frequency is twice the fundmantal (120Hz) then it is due to inadequate filtering of the rectifier output in the power supply.

Distortion is another unwanted signal.
Distortion can be analysed as an addition of harmonics and can be reduced by using the fact that certain circuit configurations cannot pass even order harmonics.

http://en.wikipedia.org/wiki/Push–pull_output

go well

 

[psparky]

Ok...I got the response from the email I requested. I consider this person to be one of the more intelligent people on Earth. He will explain in small doses the deal...not that any of the previous answers were wrong. But when George talks...people listen.

From the beginning:

Remember in DC circuits the word harmonics never came up because there were no time
varying signals. In AC circuits its assumed that you are dealing with linear circuits. That is,
any time varying signal varies at exactly the same frequency as the input signal. Resistors,
inductors and capacitors cannot change the frequency of the signal applied to them and
therefore they cannot introduce harmonics.

At the end of AC circuits when you considered Fourier series you introduced harmonics in
order to model one time varying signal by a sum of sines and cosines. This is just a mathematical
trick.

In communications you multiplied one sine wave by another sine wave and generated harnonics!
Multiplicaion is not a linear operation!

In electronics when ever you passed an ac signal through a non-linear device you generated
harmonics. Any form of switching device will chop up the ac signal and generate harmonics.
The switching circuits in the motor controller will generate harmonics that need to be filtered out.

Any device that draws current in pulses will generate harmonics. Equipment that operates on
duty cycles for example.

Furthermore...any electrical or electronic device that has a switch mode power
supply in it is going to generate harmonics because of the high speed switching going
on.

More to follow

George.

 

[psparky]

Some further explanations of the above...

A harmonic is a whole-number multiple of a given frequency. A 60 Hz sine wave has harmonics at 120 Hz,
180 Hz, 240 Hz...i.e. 2nd, 3rd and 4th harmonics of 60 Hz. The higher the frequency of a harmonic the
lower its amplitude.

Remember when you plotted Ohm's law you had straight line (linear relation). When you plotted the i-v characteristics
of a diode you had that weird looking curve (non-linear). Non-linear means it doesn't have a simple
y=ax+b relation between voltage and current.

Semi-conductor-controlled switches and drives are fancy off/on switches. Switching power supplies
(used in most tthings) are more efficient because of the high speed semi-conductor (transistor) switching that occurs.
But, the on/off switching generates harmonics. Turning voltage or current off and on generates harmonics.

 

[yungman]

So which part are you confused? The email is about the most fundamental stuff!

 

[psparky]

Yes...I want to start from the bottom and go from there.

There is more to come...

 

[psparky]

George, don't all these harmonics cause problems in factories??

Psparky,

Yes they cause problems because of how they affect impedance.

Think about it in terms of inductors and capacitors.

If you chop up an ac signal and generate a lot of high frequency harmonics look
at what happens to the impedance. The inductor's impedance goes up while the
capacitor's impedance goes down.

The parallel power factor correction capacitor's capacitive reactance goes down
with high frequency harmonics. Power factor correct stops working!

The most serious problems will occur if the harmonics introduce resonance in
the system.

For what you need to know get a copy of "Operating, Testing, and Preventive
Maintenance of Electrical Power Apparatus," by Charles. I. Hubert, Prentice Hall,
Chapter 7, "Resonance, Harmonics, and their Harmful Effects in Electrical Power
Systems." It's an easy read and will give you all you need.

George

 

[jim hardy]

in three phase, third harmonics all return through the neutral.
When computers became prevalent this overloaded neutral wires in older office buildings, caused troubles.
That's because computer power supplies draw current not smoothly like a lightbulb
but in big gulps near sinewave peak
and that's rich in harmonics especially third

nowadays neutral wires are bigger
and electronic stuff is equipped with power conditioning circuitry to make it draw smoother current, not just at peaks
there's hundreds of IC's for that purpose .

interesting but complex. you younger engineers live in a marvelous time - looks like a wonderland to this old dawg.

old jim