"Veritasium" (YT) - "The Big Misconception About Electricity" ?

[SandyCox]

Ascribing meaning to the Poynting vector at a point leads us to the wrong conclusion, as shown by Fig. 11.3.1.

I didn't see Haus and Melcher recant what they wrote

"Even with the fields perfectly stationary in time, the power is seen to flow through the open space to be absorbed in the volume where the dissipation takes place."

Did you?

I’m not quite sure what you a trying to say by “glowing red hot”. Are you saying that energy is now transferred through thermal radiation?

No, I'm saying that by making rod and washer of very different materials you can have one glow red hot while the other stays cool, and viceversa. There still will be Poynting field lines in the space inside the can and they will account for the difference between the total power delivered by the battery and the power absorbed by the rod.
In one case you will see a lot of lines coming out of a cool washer to impinge into a red hot rod.

On the next page they say:
" we illustrate the danger of ascribing meaning to S evaluated at a point, rather than integrated over a closed surface."

I'm not sure why you are dragging thermal issues into the argument. We can also make your resistors glow, but why would we? Add a cooling system if you are worried about conductors glowing red.

The point is that there is no power being transferred from the washer to the rod. All the power is accounted for. Misinterpreting the meaning of the Poynting vector leads us to incorrectly believe that power is flowing from the washer to the rod. Please do the calculations. You have all the information you require.

[adx]

... 'Millikan tongs' ... So in this special case, can we agree that the energy is 'flowing' in the succession of 1eV-potential electrons being carried by the tongs?  It's hard for me to see it any other way.

Hard to argue with that. If you pick up a charged battery, with separated + and - ions in it, and carry them in the same direction (the conventional approach to battery transport), you are transmitting power (briefly), and inducing (or via) an EM field. It's not steady state, but is sort of DC (not in the Fourier sense, but the non-accelerating A to B sense once launched), and there's just no easy way I can wrap my head around the potential fact that this is a radio transmission, despite there being moving E and H fields.

[adx]

Enclose each resistor in a gauss surface and compute the net flux of S through that closed surface to find the power absorbed and you will see that each resistor is getting the power corresponding to V times I. ...

Do the same procedure on some space inside the loops but not where wires or resistors are. Then tell me how the result in any way supports the concept that "power doesn't flow in the wires"?

(Yes I'm aware of what will happen if you include some wire. That's not the question.)

I guess we can get to the conclusion that vacuum does not absorb energy?
(I wonder what a QFT theorist would say about that...)

Then the wires become the question. The vacuum can have E-field without charge, the conductor can have charge without E-field. If neither can absorb energy, what gives one precedence over the other to be able to transmit it? Is it because a magnetic field can't exist in the conductor? What if it is a non-ideal conductor, which permits magnetic field everywhere at DC while absorbing a tiny amount of energy similar to what a 'semi-conductive vacuum' could?

(Not saying I know the answers, just pushing for a line of reasoning which might help untangle this.)

[adx]

...
Or, if you want to fly a bit lower, Kraus

John D. Kraus
Electromagnetics 2e
section 10.20 Circuit Applications of the Poynting Vector
p. 416
on p. 418, after considering a circuit with a battery (DC) and a resistors he writes:

"In Fig. 10-19aflow lines of the Poynting vector (power flow lines) are shown. It is evident that the power flow is through the empty space surrounding the circuit, the conductors of the circuit acting as guiding elements. From the circuit point of view we usually think of the power as flowing through the wires but this is an oversimplification and does not represent the actual situation."

This (italics mine). Is an example of an academic sermonising scientific hypothesis as fact. It may seem harmless, but results in generations(s) of disciples believing stuff.

I know this was already a page ago, but can we pause for a moment to appreciate that John D. Kraus, the inventor of the helical antenna and corner reflector array, is being called a sermonizing academic?
https://en.wikipedia.org/wiki/John_D._Kraus

I mean, you can call him wrong, and you'd be wrong to call him wrong, but Kraus was hardly an ivory tower academic who didn't build anything practical... he used his exquisite knowledge of Maxwell's equations and Poynting theory to create entirely new types of antennas and waveguides - and all without numerical EM-simulators that we take for granted.

I Wikipediad him before making that comment, just to be sure I wasn't saying anything silly. I had a lecturer who may have idolised someone maybe him, he also had a friend (a different one I guess) who used to go up with a hammer and "beat the side lobes out of the antenna" such was his deep and instinctive understanding of radio and stuff (my words at the end there). It is extremely hard to refuse Poynting's vector in RF, unlike DC. But Kraus commented specifically about DC and it is profoundly unscientific in the context of Poynting theory as presented. I don't mean academic in a pejorative way, just the way that getting focused on one area of work tends to focus a person into one area of work. Add some boffinness, perhaps a bit of age and a lot of experience, and it tends to happen.

It's not lost on me that Veritasium basically committed the same faux pas. But he is intentionally asking to get called out on it, not blithely stating a fact in a textbook.

Right, better leave it there for the night and get to bed, or make a strudel with an apple I promised to use before it totally rots.

[bdunham7]

What she said is that an electron has a much higher probability of interacting with another electron a billionth of a meter apart than with one 1 m away. Just that.

I don't think any of her comments were directed at Derek's specific question regarding time.  It took me a few minutes to figure that out, as I didn't see the relevance of her points at first.

Dave says it doesn't apply to DC. He's wrong. But the issue here is not technical, as Maxwell himself pondered on.

I haven't seen Dave claim that Poynting's Theorem is incorrect in any circumstance.  If I've missed it, could you point me there? 

It's difficult to accept that an analogy that you held dear turns out in the end to be superficial and misleading.

Indeed. 

[SandyCox]

By looking closely at Example 11.3.1 of Haus and Melcher we can see that Veritasium is wrong. In this example, all the power is being transferred by the conductors. No power is transferred in the region outside the conductors. Haus and Melcher should have said "power seems to flow through the open space" instead of "power is seen to flow through the open space".

The power entering the washer from the voltage source is:
Pw = 2*pi*sigma*delta*V*V/ln(a/b)
By integrating the Poynting vector over the outer surface, we find that the power that is dissipated in the washer is:
Pw = 2*pi*sigma*delta*V*V/ln(a/b)

The power entering the rod from the voltage source is:
Pr = pi*b^2*sigma*V^2/L
By integrating the Poynting vector over the outer surface, we find that the power that is dissipated in the rod is:
Pr = pi*b^2*sigma*V^2/L

So all the power entering the washer from the voltage source is dissipated in the washer and all the power entering the rod from the voltage source is disspiated in the rod.
There is no power being transferred in the region between the washer and the rod.

Will someone please check my calulations?

[adx]

[I have a question though. Consider the following experiment. Let's take a rubidium laser, shine its beam through a beam splitter, then one of the beams goes through a thick copper plate to a detector A, and another beam goes through the air unobscured directly to a detector B (sea level, 25C air temperature, 30% relative humidity). Can I assume that the probability of a photon reaching detector A (the one behind the metal plate) would be 999999 times higher than probability of reaching detector B, considering the same Coulomb's law and QED?

PS. Simplified version of this experiment can be reproduced by every member of this forum with a flashlight and a frying pan.

Your simple experiment is so brilliant that the Poynting-haters didn't even realize it.

Strudel complete. One more.

I got A and B backwards again. The copper block is missing the transverse m return path. Try cutting a slot in it (so you can see through your screen or at least to the VESA mounting pad, use the laser, but remember to remove a few km of optical length from it or at least turn it down afterwards, and check the detectors and beamsplitter for holes and rubidium fume). While at it, focus the laser around the entry to the block, so the beam diverges quite a bit. Make that plate longer (thicker) for good measure, move the detectors away and or check for copper intrusion this time (blame the instructions).

Now measure the intensit probability of measuring a photon at A and B. Frequency is a bit high for a DC circuit, but I'll accept some energy in the space between the 'wires'.

[rfeecs]

So all the power entering the washer from the voltage source is dissipated in the washer and all the power entering the rod from the voltage source is disspiated in the rod.
There is no power being transferred in the region between the washer and the rod.

You cannot conclude that from your calculation.  It could be that power is exiting the washer and flowing into space (into the fields) and the same amount of power is flowing into the rod from the space (the fields) around it.

There is no double counting of power.  Both views are equally valid in terms of conservation of energy.

But the "alternative" approach is only valid for DC.  Using the Poynting vector is valid for all cases.

[bdunham7]

You cannot conclude that from your calculation.  It could be that power is exiting the washer and flowing into space (into the fields) and the same amount of power is flowing into the rod from the space (the fields) around it.

There is no double counting of power.  Both views are equally valid in terms of conservation of energy.

But the "alternative" approach is only valid for DC.  Using the Poynting vector is valid for all cases.

Poynting's Theorem, of course, is fully falsifiable and easily proven correct for all cases.  I'm beginning to think this debate over the significance of the Poynting vector itself (as a 'true statement of what really happens' or 'proof that the energy flows in the fields through free space') is actually not resolvable because it is not falsifiable.

It can be difficult to rebut a claim that a mathematical model represents reality simply because it accurately predicts the result.  In an earlier discussion it was stated that you could consider a 9-volt battery connected to a 300R twin-lead wire (open at the end) to be a transmission line with 30mA of continuous current outbound with an equal 30mA being continuously reflected.   The math works for an ideal circuit and may be useful for solving certain problems, but I don't think anyone really thinks that the math reflects 'the real situation' or 'what is really going on'.  And lets not even discuss whether imaginary cancelling currents would have parasitic losses or not. 

So back to the Poynting vector.  What does it mean to say that an arrow on a diagram represent power flow density in W/m²?  Specifically, in the classic battery/wires/load rectangle there is a Poynting vector poynting directly away from the load on the outer side of the battery.  Many diagrams conveniently omit it, Derek's video shows it but just truncates it without explaining its meaning.  What does that arrow represent?  Is it something tangible that can be measured, used, interfered with, etc?  If I say that it is just a mathematical result (from two actual physical manifestations, E and B)  that has no corresponding manifestation of its own in reality, just like the imaginary currents in the open transmission line above, can you counter that?

[bdunham7]

Your simple experiment is so brilliant that the Poynting-haters didn't even realize it.

Yes, we're all idiots.  So please explain the brilliance of this experiment to us.  Make sure to throw in the concepts of localization and the interaction between photon and a conductor.

[SiliconWizard]

So now, let's study the same setup, but with a circular wire loop instead of a rectangular one.

I assume we are talking about two circular loops, replacing the two rectangular loops?

Uh. Sometimes, words are poor at expressing simple things. So, what I meant was essentially something like this: https://www.eevblog.com/forum/chat/veritasium-(yt)-the-big-misconception-about-electricity/msg3823973/#msg3823973

With the battery and switch (let's neglect the distance between those two again) diagonally opposite to the load, and R the radius of the circle.

So, anyone?

[SilverSolder]

So now, let's study the same setup, but with a circular wire loop instead of a rectangular one.

I assume we are talking about two circular loops, replacing the two rectangular loops?

Uh. Sometimes, words are poor at expressing simple things. So, what I meant was essentially something like this: https://www.eevblog.com/forum/chat/veritasium-(yt)-the-big-misconception-about-electricity/msg3823973/#msg3823973

With the battery and switch (let's neglect the distance between those two again) diagonally opposite to the load, and R the radius of the circle.

So, anyone?

The whole problem is easier to digest if you start with the switch closed and then think about what happens when you open it!  -  then, closing the switch is just changing to the opposite sign on all your thinking, and you're done in no time at all!

[rfeecs]

So back to the Poynting vector.  What does it mean to say that an arrow on a diagram represent power flow density in W/m²?  Specifically, in the classic battery/wires/load rectangle there is a Poynting vector poynting directly away from the load on the outer side of the battery.  Many diagrams conveniently omit it, Derek's video shows it but just truncates it without explaining its meaning.  What does that arrow represent?  Is it something tangible that can be measured, used, interfered with, etc?  If I say that it is just a mathematical result (from two actual physical manifestations, E and B)  that has no corresponding manifestation of its own in reality, just like the imaginary currents in the open transmission line above, can you counter that?

If you draw a diagram of one or two charges and draw the electric field lines, you will usually have lines that go off to infinity.  We know the field drops off as we go away from the charge, so we can say the field goes to zero at infinity and most people have no problem with that.

We can argue if something is physically real vs just a mathematical convenience.  Examples are potential (vs just E and B field), the wave function, the Poynting vector.  Then there are imaginary numbers.

I'm just going to accept that these are mathematical models that engineers can use to get hopefully the right numbers.  I leave it to philosophers to try to define their version of reality.

[bdunham7]

I'm just going to accept that these are mathematical models that engineers can use to get hopefully the right numbers.  I leave it to philosophers to try to define their version of reality.

And again, we agree.  I wrote all that in a reply to your post, but I meant to direct it more generally elsewhere...

[adx]

I think you're probably right. But that's not certain, and no amount of introspection on the mathematical details of some theory is going to answer that last part.

I get it now, and I can see the attraction. Some people just want to believe.

What is funny is that the Poynting theorem was independently studied by Oliver Heaviside, the same guy who developed the transmission line model, that you attempted, without success, to use to try to "debunk" Derek. His study of energy flow through fields is what made it possible for him to come up with the modern version of Maxwell's equations, without which we wouldn't have the transmission line model, nor the high frequency electronics we have today, from computers to radio.

An 'old' Apple M1 die is about 11mm on the side. It goes say 5GHz ooh looking it up (not an Apple fanboi but it's a nice chip) 3.2GHz, wavelength "Have aliens found you?" no click through that ad, hmm Find out by watching our video on exoplanets, no, clickbait, be strong, https://www.omnicalculator.com/physics/wavelength = 93.6851 mm. Not far off the fabled 1/10th wavelength where thumbrulers say you can ignore wave effects and treat it all as Ls and and Cs. But even if that weren't notionally so, somewhere on that schematic, perhaps sheet 47926, will be a note next to say Q15469947265 saying "layout designer - keep close to Q1 per ECO56789". It's not all about transmission lines and RF.

But ~all trolling aside, for anyone reading this thread in the future or even coming in fresh right now, very few of us have tried to "debunk" Derek's video, all basically agreed with the result and its unavoidable conclusion that transient energy must travel through space. Some question his statement on the physical reality that all power flows outside of the wires at DC (or 50/60Hz in the main rendering with the LED filament bulb). Some are unsure what electricity "is", despite many years of education (and some are too sure). And some want to know what the waveforms look like.

[SandyCox]

So all the power entering the washer from the voltage source is dissipated in the washer and all the power entering the rod from the voltage source is disspiated in the rod.
There is no power being transferred in the region between the washer and the rod.

You cannot conclude that from your calculation.  It could be that power is exiting the washer and flowing into space (into the fields) and the same amount of power is flowing into the rod from the space (the fields) around it.

There is no double counting of power.  Both views are equally valid in terms of conservation of energy.

But the "alternative" approach is only valid for DC.  Using the Poynting vector is valid for all cases.

Let's look at the problem without making use of the Poynting vector. We can simply calculate the integral of sigma E.E over the volume of each conductor. This gives us the total power dissipated (in the form of thermal energy) in each conductor. The power delivered by the voltage source to the rod is equal to the power dissipated in the rod. The same is true for the washer. So there is no net flow of power from the washer to the rod, as the Poynting vector seems to suggest.

There is nothing wrong with the Poynting vector or with Poynting’s theorem. The problem is that people misinterpret the meaning of the Poynting vector. It has no meaning unless it is integrated over the surface of a closed volume in space. This simple example makes this misinterpretation painfully clear.

[adx]

I can't get my head around the washer and rod exercise without leading to a circular argument, at which point my brain shuts down and spits out the most recent result with an error flag set. The results are the same either way. I cannot see any point in considering the washer and rod simultaneously if I know they are independent. The power that's stored in any magnetic field may come out of it and go back in somehow, but it has no known effect. (Plus we have the benefit these days of knowing what makes the magnetic field, and being able to test the 'moral' reasonableness of any weird situations.)

So roll with that idea, and don't just consider the washer and rod separately, but split them into infinitesimally small threads, or in the case of that example, sectors. This alters the electric field, but has no effect on the potential difference that drives energy transfer (work function in a conservative field). Compute the Poynting vector of that and superpose. Which should be possible if the energies add up. The Poynting vector should then show something quite different.

Now apply that to Veritasium's example (at DC), which is something I did think of earlier but thought it mightn't work (as in, be a reasonable partitioning of the current flow).

[snarkysparky]

In the linked section from Haus and Melcher about the S vector.
Eq 23

S = phi( J + part_D / part_t)

In the free space surrounding the wires for DC current both  J and part_D / part_t   are zero.   

 S is zero

[snarkysparky]

[SandyCox]

In the linked section from Haus and Melcher about the S vector.
Eq 23

S = phi( J + part_D / part_t)

In the free space surrounding the wires for DC current both  J and part_D / part_t   are zero.   

 S is zero

Eq. (23) is an alternative formulation. By comparing Figs. 11.3.1. and 11.3.2 it is clear that the new electroquasitatic flux in not the same as the original Poynting vector, even in the static (DC) case. The point is that their alternative formulation is not prone to the same misinterpretation as the Poynting vector.

The physics’ professors are also still arguing about the interpretation of the Poynting vector: 
https://www.researchgate.net/publication/321440917_Energy_in_Electromagnetism_The_Poynting_Vector_Historical_Corner

My personal viewpoint is not to attach any meaning to the Poynting vector without integrating it over the surface of a volume in space.

[adx]

The Poynting vector should then show something quite different.

Not for that 2D example - might be best to quit while I'm behind .

[snarkysparky]

More beating of the dead horse.

https://phys.libretexts.org/Bookshelves/Electricity_and_Magnetism/Book%3A_Electromagnetics_II_(Ellingson)/03%3A_Wave_Propagation_in_General_Media/3.01%3A_Poynting’s_Theorem

""The quantity  E×H  is the Poynting vector, which quantifies the spatial power density (SI base units of W/m 2 ) of an electromagnetic wave and the direction in which it propagates. The reader has likely already encountered this concept. Regardless, we’ll confirm this interpretation of the quantity  E×H  in Section 3.2. For now, observe that integration of the Poynting vector over  S  as indicated in Equation  3.1.20  yields the total power flowing out of  V  through  S . The negative sign in Equation  3.1.20  indicates that the combined quantity represents power flow in to  V  through  S . Finally, note the use of a single quantity  Pnet,in  does not imply that power is entirely inward-directed or outward-directed. Rather,  Pnet,in  represents the net flux; i.e., the sum of the inward- and outward-flowing power.""

bold text is my $.02

[adx]

... of an electromagnetic wave and the direction in which it propagates. ...

bold text is my $.02

italic text is mine

... at which point my brain shuts down and spits out the most recent result with an error flag set.

That result is something like a standing wave of order 0, even though though that does not make sense (since the theory holds that the resistor(s) is dissipating the power), but neither does the first iteration of the endless loop; how can a travelling wave "propagate" in DC? It implies continuous conduction, moving charges, static fields.

[SandyCox]

... of an electromagnetic wave and the direction in which it propagates. ...

bold text is my $.02

italic text is mine

... at which point my brain shuts down and spits out the most recent result with an error flag set.

That result is something like a standing wave of order 0, even though though that does not make sense (since the theory holds that the resistor(s) is dissipating the power), but neither does the first iteration of the endless loop; how can a travelling wave "propagate" in DC? It implies continuous conduction, moving charges, static fields.

I am only aware of two mechanisms by which electromagnetic energy is transferred:
1.   Flow of charge
2.   Time-varying electric and/or magnetic fields
At DC, energy is only transferred by means of the flow of charge through an electric field. In this case the Poynting vector is the result of the flow of charge.
For AC, in the far field, only time-varying electric and magnetic fields.
But this is only the picture I have in my mind. Maybe there are more mechanisms.
I’m not sure if Maxwell’s equations can answer this question. Maybe we need to peel another layer of Feynman’s onion.

[adx]

... of an electromagnetic wave and the direction in which it propagates. ...

bold text is my $.02

italic text is mine

... at which point my brain shuts down and spits out the most recent result with an error flag set.

That result is something like a standing wave of order 0, even though though that does not make sense (since the theory holds that the resistor(s) is dissipating the power), but neither does the first iteration of the endless loop; how can a travelling wave "propagate" in DC? It implies continuous conduction, moving charges, static fields.

I am only aware of two mechanisms by which electromagnetic energy is transferred:
1.   Flow of charge
2.   Time-varying electric and/or magnetic fields
At DC, energy is only transferred by means of the flow of charge through an electric field. In this case the Poynting vector is the result of the flow of charge.
For AC, in the far field, only time-varying electric and magnetic fields.
But this is only the picture I have in my mind. Maybe there are more mechanisms.
I’m not sure if Maxwell’s equations can answer this question. Maybe we need to peel another layer of Feynman’s onion.

I suppose time-varying electric fields will be produced by flowing charge. If you consider that a circuit must have a return path, then an electron going forward will push on the electron coming back, taking work from it and putting it into the field, which is returned after it passes. I guess topped up by the relativistic effect which produces magnetism, but that will be unequal and store energy in the electric field while current flows. Maybe something like that is responsible for the Poynting vector behaving as it does here.

There must be an explanation for the mathematical behaviour even if it doesn't make physical sense, the physical is magic (in the sense that nobody knows everything about how it works) but the maths shouldn't be. If it is simply a way of quantifying one's thinking.

I don't see anything (or any way) for QED or QFT to alter the Poynting vector beyond the different insights it brings to those people skilled in that field, but I don't know (and barely have a handle on Maxwell's equations themselves). I was looking at Feynman lecture #1 earlier tonight and wondering if I should finally bite the bullet and try to properly learn what I never did at school or uni! And field solvers again, but only as a way to avoid the maths.