Does electromagnetic waves are generated by dc current?

[vishvajeet]

I am beginner in physics.I have question that when dc current flow they flow with apparent drift velocity so it appears that it should not emit em waves.But in actual current flow electrons are accelerated due to applied potential and also deaccelerated during collision so even they appear to have constant velocity they are accelerated so they must emit em waves. am i right?



I have another question more basic than above why accelerated charge should emit em waves?please describe precisely

 

[Simon Bridge]

Welcome to PF;
When you first switch a DC circuit on you do get a short burst of EM waves as the charges are acceleated.
That is not normally thought of as "DC" however.

Note that DC and AC are only two possibilities - the current may vary without alternating for example.

Accelerated charges radiate due to the invarience of the speed of light.
It's a property of the Universe - as Richard Feynman would say, "them's the rules."
Basically - if you shifted a charge from one place to another, then the electric field at some distance point must also change - but it cannot do this right away or you could use it for FTL communication. So - the change in the field much travel out from the charge. The edge of the shift manifests as an electromagnetic pulse.

 

[vishvajeet]

but why electron moving in conductor which is also accelerated can't emit em waves??

Again in second explanation why only accelerating charge and not charge moving with constant velocity produce em? Also what about particle nature of em waves?

 

[Simon Bridge]

Electrons accelerating in a conductor do emit EM waves - that is how radio transmitters work.

All charges have an electric field.
When you move a charge at a constant velocity, you also get a magnetic field.

To get EM waves you need to accelerate the charges - like wiggling them back and forth or turn them in a circle ... accelerations.

Words can only provide a basic qualitative feel for how things are happening - to understand in depth you need the math. Look up "Maxwell's Equations".

What about the particle nature?
The "particle nature" of EM is in the way EM "waves" deliver energy - see "photoelectric effect" to start with.

This motivates a statistical treatment of light where we imagine it as a bunch of corpuscles traveling in space. The statistics work out to the same relations but the math is harder.

 

[vishvajeet]

sir,you are not understanding me in radio transmitter we use oscillator circuits.In which electrons doesn't move with constant drift velocity but they oscillate and so generate em.But even without oscillator circuit i thought there should be production of em. sir please read www2.warwick.ac.uk/fac/sci/physics/current/teach/module_home/px263/lectures/sefton.pdf

 

[yungman]

sir,you are not understanding me in radio transmitter we use oscillator circuits.In which electrons doesn't move with constant drift velocity but they oscillate and so generate em.But even without oscillator circuit i thought there should be production of em. sir please read www2.warwick.ac.uk/fac/sci/physics/current/teach/module_home/px263/lectures/sefton.pdf

If the voltage is oscillating, it's not DC. Of cause there will be EM wave radiating out.

Another important thing. It is the EM wave that move and oscillate, electron drift velocity is not important in this kind of circuit. AC signal travel as EM wave, not as electrons. drift velocity is way way too slow. Current you measure is only the consequence of the boundary condition of the EM wave with the conductor.

Any circuit has to comprise of a signal forward and a signal return path. This forms a transmission line guided structure where EM wave travels. You apply a signal, you just launch an EM wave down the transmission line structure. If you really want to see the electron that you inject at one end of the transmission line, you are going to have to wait a long time before the same electron comes out from the other end even if you put a voltage across the line.

 

[vishvajeet]

But even in electron model of current flow all the electrons in conductor start drifting at once due to application of electric field at one end of conductor which just applied at all points of conductor atonce(because speed of electric field propogation is 3*10^8 so light lights up so fast ie electrons at everywhere start drifiting at once

 

[yungman]

But even in electron model of current flow all the electrons in conductor start drifting at once due to application of electric field at one end of conductor which just applied at all points of conductor atonce(because speed of electric field propogation is 3*10^8 so light lights up so fast ie electrons at everywhere start drifiting at once

It really does not work this way. You need to look at the boundary condition of E and H field. Mainly it's the H field that cause surface current to flow, not the injecting of the electrons at one end. All electronic AC signals transmitted by EM, not electrons. It's all EM wave through guided structure.

You have to be very careful in a lot of the electronics theory they called EQUIVALENT circuit. It cannot be explained by physics. Sure, current and voltage traveling is the easiest way to visualize, but it really does not work this way. Books are not that good in explaining this either.

When you have high enough frequency signal travel down a long wire( where length of wire larger than $\lambda$ of the signal), different part of the wire have different voltage( consequence of E field, not current). You might even think the current is moving in opposite direction in different part of the wire. This cannot be explained by all electrons move at the same time along the wire due to the potential.

 

[yungman]

Here is the diagram I copy and modified from Cheng's electromagnetic book:

[PLAIN]http://i40.tinypic.com/30hx9wj.jpg[/PLAIN]

This is a drawing of EM wave propagate in z direction ( from left to right) through a parallel plate transmission line. The orange arrow is the E field that is from one plate to the other. The green circle with a dot is the H field coming towards you, the green circle with a cross is H going away from you. The dark purple arrow inside the top and bottom plate is the direction of the current flow on the surface of the conductor plates. The red + and - are charges on the surface of the plate.

the current is what you can measure, but it is absolutely not from the electron flow as you think. It is the consequence of the magnetic boundary condition.

$$\int_c \vec H\cdot d\vec l=\vec J+\epsilon \frac {\partial \vec E}{\partial t}$$

You see the direction of the current density change direction along the tx line?

The voltage along the line is integrating the E field in the tx line structure.

$$V=-\int_c \vec E\cdot d\vec l$$

 

[Simon Bridge]

sir,you are not understanding me in radio transmitter we use oscillator circuits

Indeed - when I spoke of a radio transmitter, I was answering your question about charges accelerating in a wire where you said:

but why electron moving in conductor which is also accelerated can't emit em waves??

...
the answer is that charges accelerated in a conductor do emit EM radiation.
But perhaps you meant to ask about the charges in a conductor, where the conductor is accelerating ... and the answer is the same: they do ... but, the conductor contains positive and negative charges so the total field is zero.

In physics we understand the phenomena we investigate in terms of theoretical models.

It is important not to mix up the models from different disciplines - they are all expected to be related but the relationships are not always straight forward. It does no good to insist that should apply equally when they don't.

The theory of electric circuits is a subset of the theory of electrodynamics, which is a subset of quantum theory.

You can understand drift velocity of charges in a solid in terms of random accelerations and decelerations.
However, these accelerations and decelerations are random - so each of their contributions to the overall EM field sums to zero. What is left over is the general continuous motion - which happens on average - which produces the field we see.

But what is important about the fields generated by the charge carriers in a wire, say, is the disturbance of the electric field rather than the individual charges. This is especially important in AC.

To understand this sort of model you need to read about Maxwell's equations.

However, your own reference (the pdf linked in post #5) cautions you against thinking about the moving charges carrying the energy in a circuit. It is the very first thing the author talks about. The author continues to take great pains to talk about the fields throughout the rest of the paper.

Was there are particular part of the paper you wanted to draw to my attention with regard to your questions?

 

[yungman]

It is important not to mix up the models from different disciplines - they are all expected to be related but the relationships are not always straight forward. It does no good to insist that should apply equally when they don't.

The theory of electric circuits is a subset of the theory of electrodynamics, which is a subset of quantum theory.

This is so true. I am a long time EE and in my retire years digging into electromagnetics. I learn so much and realize there are a lot of circuit theory that can be challenged by classical physics. Ohm's law, Kirchhoff loop etc. fall apart under certain condition. It was very hard for me to swallow at the beginning also. As I study more, I start to understand more ( not nearly enough!). A lot of the circuit theory can be very misleading. People need to take those circuit theory with a grain of salt and really need to know their limitations. A lot of them are nothing more than an equivalent, not a theory.

Particular now a days when speed and frequency of the circuit goes higher and higher, a lot of the normal circuit theory falls apart and not useful in predicting the behavior of the circuit. I think the reason EE books still try to use voltage and current is because it is more familiar and more important, easier to measure with instruments. E and H are not as easy to measure!

That's the reason most RF and microwave theory books start with EM theory as the ordinary circuit theories fall apart at that frequency and a lot of them fail to predict the behavior of the circuit.

 

[vanhees71]

The circuit theory used by electrical engineers is of course completely in line with classical electromagnetics. The only assumption made is that the typical frequencies in the Fourier transform of the time variations of the fields, charge and current distributions are such that the corresponding wavelengths $\lambda=2\pi c/\omega$ are large compared to the extension of the circuit and that thus one can lump the geometry and consituent laws into paramameters like resistance, capacitance, and inductivities, because one can necglect the discplacement currents. Physically this means also to neglect retardation and thus you use the quasistationary approximation, and this is very accurate for any practical purposes. For a thorough discussion, see Jackson, Classical Electrodynamics.

 

[yungman]

I don't think circuit theory necessary in line with electromagnetics theory. Thevenin theorem is not a physical theorem, it's only an equivalent. The Thevenin voltage source and resistor do not exist, only the equivalent in given condition. Neither is Superposition and some other ones.
There was a big discussion of the MIT professor proofing Kirchhoff voltage loop doesn't hold.

Signal do not travel as electron or charge movement, it would be way too slow to be useful as electrons move very slow in good conductor. The reason signal travel in fraction of light speed is because it's the EM wave that travel. A simple proof is the speed of the signal travels through a conductor depends on the dielectric surrounding it. If it is a trace on a pcb with $\epsilon=4$, speed is only half the light speed. It's the EM wave that travels, not current or voltage.

 

[Dale]

The circuit theory used by electrical engineers is of course completely in line with classical electromagnetics. The only assumption made is that the typical frequencies in the Fourier transform of the time variations of the fields, charge and current distributions are such that the corresponding wavelengths $\lambda=2\pi c/\omega$ are large compared to the extension of the circuit and that thus one can lump the geometry and consituent laws into paramameters like resistance, capacitance, and inductivities, because one can necglect the discplacement currents.

Yes, that is often called the "small circuit" approximation. There are two other assumptions: that none of the lumped elements have a net charge, and that none of the lumped elements have any mutual inductance. I.e. all of the E and B fields are confined to the various lumped elements.

I don't think circuit theory necessary in line with electromagnetics theory.

This is incorrect. Circuit theory is a simplification of classical EM under the above assumptions. Any disagreement with EM can be traced to a violation of one or more of those assumptions.

 

[yungman]

This is incorrect. Circuit theory is a simplification of classical EM under the above assumptions. Any disagreement with EM can be traced to a violation of one or more of those assumptions.

Problem, sometimes is too simplified that it can be misleading. I am EE, not a physicist, I usually don't join in for discussion in this part. I only join in as this thread about electrons drift or display current and circuit. Electronic theory do everything to make it a voltage and current signal as if they are really the ones that is doing the job. But in fact, the electron and charge moving has little to do with AC circuit. It's the EM wave that is in action. Voltage and current is only from integration of E and consequence of boundary condition of H field. The AC signal characteristic can only be explained by EM theory unless it is so so slow.

BTW, it is not correct to say circuit only behave differently when length of line is approaching $\lambda$. Even low frequency in less than 1MHz on pcb trace has to follow tx line theory that is EM. This is a major part of the field of Signal Integrity Engineering in the last 15 years...the way to predict signal return path to avoid EM emission...which is EM, it cannot be explained by just Ohm's Law and charge moving.

Like you said, the lump element is an approximation...it's just that. Electronic theory use a lot of approximation that you have to know their limitation. It is not necessary the law of the land. I know there is one level up...Quantum theory. But EM theory is closer than the electronic theory in a lot of cases. Problem is it's very inconvenient to use EM in circuit all the time, so again, it is simplified with limitations.

I just join in only because people really think voltage and current traveling in circuits...and it is The end of the means. It is only the Means to an end. I shouldn't get into any deeper discussion as I am still studying EM. I just feel it's so important to me to finally realize this and it really help me to move up one level of appreciation in EE after all these years in the field. And I think it's very important for EE people to realize this.

 

[Dale]

Electronic theory use a lot of approximation that you have to know their limitation.

Yes, it is essential when using any approximation that you understand what the assumptions of the approximation are and that you check if those assumptions are valid for the situation you are modeling.

If in your particular scenario the assumptions of circuit theory are valid then the simplified calculations will give you the same results as the full EM theory. Therefore, within its domain of applicability, circuit theory is "in line with electromagnetics theory".

BTW, it is not correct to say circuit only behave differently when length of line is approaching . Even low frequency in less than 1MHz on pcb trace has to follow tx line theory that is EM. This is a major part of the field of Signal Integrity Engineering in the last 15 years...the way to predict signal return path to avoid EM emission...which is EM, it cannot be explained by just Ohm's Law and charge moving.

That is why I mentioned that there are three assumptions for circuit theory, not just one. Which one do you think is violated in this scenario?

 

[yungman]

Yes, it is essential when using any approximation that you understand what the assumptions of the approximation are and that you check if those assumptions are valid for the situation you are modeling.

If in your particular scenario the assumptions of circuit theory are valid then the simplified calculations will give you the same results as the full EM theory. Therefore, within its domain of applicability, circuit theory is "in line with electromagnetics theory".

That is why I mentioned that there are three assumptions for circuit theory, not just one. Which one do you think is violated in this scenario?

I am not trying to get into the details, just a general statement. Yes, I agree with you 100% about knowing the limitation and use the circuit theory. It would not be very useful to pull out EM theory every time for circuits.

I go into detail ONLY to draw OP's attention that there are limitations on the assumption and don't use it as if it is the absolute truth.

Ha ha, I learn my lesson here 3 years ago, I was one of those that held onto the electronic theory by heart. This is one threat that really make me start to take a second look. I don't even know whether I agree with the professor and still not 100% conceded yet on that thread, I just ran out of EM theory...Arguing against the MIT professor and some PHDs here! But this really open my eyes to take everything a second look:

https://www.physicsforums.com/showthread.php?t=453575&highlight=yungman

Believe me, I spent the whole Christmas holiday on this, setting up experiment, theorized it and present my theory and finding according to equivalent circuits and theory.

 

[Dale]

I go into detail ONLY to draw OP's attention that there are limitations on the assumption and don't use it as if it is the absolute truth.

I completely agree with that.