General > General Technical Chat
"Veritasium" (YT) - "The Big Misconception About Electricity" ?
<< < (184/396) > >>
Sredni:
Lets have two slides with almost no friction and with the same slope. One is at one meter above ground level (plus the minimum height to get the minimum slope to compensate for friction), the other is at 100 meters above ground. We make two one ton stones slide for 1 km to their destination: a machine that will make the block fall and turn all gravitational potential energy into heat (eat your heart out, Carnot!).

The block falling one meter generates m g (1 meter) J of energy. The block falling 100 meters generates one hundred times that energy.
Has the energy travelled along the slide?
Both slides are identical, both stone blocks are identical, and the both moved at the same speed on the slide.

Has one slope really carried 100 times the energy of the other one? Or was the function only that to bring the mass where the gravitational potential field was allowed to make the conversion of energy to heat possible?
What if we dig a hole 100 meters in the ground with a Predator's starship beam, just microseconds before the stone reaches destination? Has the energy travelled along the slide before we decided to dig the hole?

EM adds a twist to this scenario in that the charges carry a significant field and so they set up the field that allows the conversion of energy into heat on the resistor. They are basically digging their own hole as they pile up and (what's the opposite of piling up?) decrease in density at the resistors interface, but not during their journey (which is still necessary to maintain a dynamical equilibrium).
adx:

--- Quote from: bdunham7 on January 13, 2022, 10:35:03 pm ---...Which is why I kept croaking that you can't consider power to be 'in' the fields on their own without considering charges.

--- End quote ---

I get that now. I kept thinking of it in a context of trying to 'fuse' the concepts of the classical field with an aspect of physical reality, because what I learned 'must be right' on some level, and so the sum of the parts (education) has to be a better starting point than rejecting part and moving forward with another. Which of course isn't rational. What I learned at school and uni might not be wrong, but that doesn't mean it is right. There is no need to hang on to some thing that isn't right as some pillar of truth. Letting go of the concept of "field" entirely is ok. It doesn't make it disappear.

That type of reasoning also leads to the possibility that the theory isn't right for RF. It's merely not wrong.
EEVblog:

--- Quote from: bdunham7 on January 13, 2022, 11:37:02 pm ---
--- Quote from: rfeecs on January 13, 2022, 11:24:28 pm ---Shouldn't QFT give the same results as the classical model?  The classical model agrees with measurements.
Of course, if we can't actually measure "energy flow", then that could be a problem.  In that case, QFT's prediction would be just as useless as the classical model.

--- End quote ---
You've neatly summarized the issue.  Both the QFT and classical model will agree as to the movement of the charges, the dissipation of energy in the load, the surface charges and fields, the magnetic fields and so on.  So both should agree with any actual physical measurements you can take, but they arrive at their results differently.  There's no reason to assume that the two models have to agree on intermediate constructs like probability, virtual particles, S-fields or other mathematical results.

--- End quote ---

And therein lies the rub.
Many people have been screaming blue in the face at anyone who dares think that energy might actually flow inside the wire, particualy at DC. Yet along comes QTF that seems to imply exactly that with almost certain probability. In which case I doubt there is anyone on this forum with the qualifications or knowledge to actually talk on the topic of QTF. So they might have to ultimately eat humble pie and submit to the higher thoery. Because ultimately the question isn't about meadurement and transients etc, that's already conclusively modeled and answered with Ponyting or anything else classcial you want to throw at it. But does any of the energy actually flow inside the wire? Why stop at Maxwell and Poynting to expalin that?
Naej:
In this post, I will explain the true (and surprising) physics behind "The biggest misconception about electricity".

tl;dr : energy flows in wires. Or in vacuum. Whatever.

First, what is energy? The modern "definition" is "energy is conserved", namely it is a mathematical invariant.
It goes this way:
1) you select the rules of physics you want to apply (the lagrangian)
2) you check that they are invariant by time-translation
3) if they are, Noether theorem will give you A corresponding invariant https://en.wikipedia.org/wiki/Noether%27s_theorem
A *crucial* point here is that an infinite number of quantities are time-translation invariants. The simplest examples are "multiplying energies by 10" or "adding 10J" or "changing the reference frame", but there are far more complicated ones.

Maxwell's equations: Maxwell defined electromagnetism with the electric potential field V, the magnetic potential field A, the electric/magnetic field E/B (and others; I'll talk only about vacuum here).
He found that E is the gradient of V corrected by dA/dt, and B the curl of A.
He found the electrical energy as being given by E^2/2 or qV/2, and magnetic is B^2/2 or qu dot A/2 with u the speed of the charge (in the correct units)
He also interpreted qA as being the 'potential momentum' of the charge.

Quickly it was realized that V,A are not uniquely defined, i.e. different V/A give the same physics, a fact now known as "gauge invariance".
The simplest example of this is that adding 420V to V does not change anything.
Which gauge you use is therefore up to you, so you use most convenient one, and often it is the Lorenz gauge (technically, almost-gauge).
The Lorenz gauge has the nice practical property that a potential difference is equal to what you measure with a voltmeter, as long as the voltmeter's wires are far from coils (this last condition is what you should learn from Lewin's KVL videos btw).
There are nice theoretical ones too: it's obvious that EM perturbations (or "light") propagate with the speed of light, i.e. it is obvious that it respects special relativity (it's covariant).

One good reason to use potentials is that you must do this if you consider a quantum mechanical electron in a field, as A contributes to its canonical momentum.
In fact it's pretty easy to derive the equation with the energy given above. ( https://en.wikipedia.org/wiki/Aharonov%E2%80%93Bohm_effect )

Are you disturbed by the gauge invariance? Well almost all fundamental physics were made with gauge invariance. https://en.wikipedia.org/wiki/Yang%E2%80%93Mills_theory

*Historical interlude* : The Maxwellians decided to "assassinate the potential", in their words, which is probably why it's usually barely mentioned in courses.
On a possibly related note, Victorian scientists believed that nature was the ether; sometimes that molecules were "vortices" of the ether.
Having an electromagnetic "proof" that energy was in the ether, and that you could stress the ether ( https://en.wikipedia.org/wiki/Electromagnetic_stress%E2%80%93energy_tensor ) was certainly a great "triumph" of the ether theory.
After Einstein's relativity, the ether was removed but the equations and concepts remained.

Heaviside-Poynting's energy flow was defined as ExB, and this is indeed correct.
Here's Poynting last sentence: "We can hardly hope, then, for any further proof of the law beyond its agreement with the experiments already known until some method is discovered of testing what goes on in the dielectric independently of the secondary circuit".
The modern take is that you apply Noether's theorem to the free-field lagrangian (so the EM equation in vacuum! or (E^2-B^2)/2), and you use a Belinfante relocation to get a symmetric stress-energy tensor.
Then you can read in it the energy, the energy flow and the symmetric stress. And you get the angular momentum flow in another tensor. https://webhome.phy.duke.edu/~rgb/Class/phy319/phy319/node143.html http://www.physics.rutgers.edu/~shapiro/613/615lects/maxwTmunu_2.pdf
The kicker is this: you *decided* that energy and energy flow have nothing to do with charge particles, and are a property of vacuum. You did not *prove* anything.
Applying it to the EM world you get the following:
- light has energy and momentum  :)
- a lightbulb is powered by light  ::)
- an (imperfect) emitter antenna is also heated by the radio waves it absorbs  ::)
- hundreds of GW are turning around every single classical electron
- you want to measure the power absorbed by a resistor? just measure E, B everywhere around it; and compute the average of ExB
- you want to compute the power radiated by a perfect antenna an AC generator? just compute E,B everywhere around it and add them

One century later, Carpenter arrives, in 1989.
If you know a bit of multivariate calculus, definitely read his papers: https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.205.4488&rep=rep1&type=pdf  http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.205.4597&rep=rep1&type=pdf
Here is the thing: you can do the exact opposite choice, namely taking the interaction lagrangian, and use the same mathematical machinery.
From it, you deduce that the energy flow is JV with J the (vector) current.
The consequences are:
- light has no energy and no momentum (this is a lie you're taught in physics!). Ackchyually, electrons wiggling over there are losing energy/momentum, and electrons over here are gaining energy/momentum.
- a lightbulb is powered by its wires
- an emitter antenna is also heated by its wires
- there's no power associated with a single classical electron at rest
- you want to measure the power absorbed by a resistor? just measure the current and potential difference (currents* at high frequency)
- you want to compute the power consumed by an AC generator a perfect antenna? just compute V,i in the wires
Here is the fun part: it *depends on the gauge chosen*.
You may think it's a problem, if you don't know that energy is arbitrary; but now that you do, it's a way to remember the fact that energy is an arbitrary invariant.
Take Gibbs' gauge V=0, and there is no energy flow, just potential energy which varies.


Now for the other problems in the video.
1) The question.
It's possible to learn 3 things from the question/answer:
a) A capacitor is not always a plate. A description of the circuit is not always the right lumped-element model that you should use. Or "capacitive coupling".
b) Current seems to go through a capacitor
c) EM disturbances move at the speed of light; and this is the only thing you need for answering the question
I admire the way the question was asked: the circuit is given with the lumped-element symbols, and you're invited to assume the wires have no resistance, as in the lumped-element model.
The "otherwise it wouldn't work" is of course wrong, you would just have a current which eventually reaches a very low value.
But it shouldn't matter because the light turns out even with a minimal current… oops we shouldn't insist on this.

2) "electrons dissipate their energy in the device" "my claim is that it is false"
Resistance, at high temperature (above -200°C), is due in most metals by electron-phonon scattering.
That is, electrons transfer their energy to the metal ions. If Derek has a new theory, what is it?
http://large.stanford.edu/courses/2007/ap273/rogers1/

3) The animation at 6:17 of the transmission line is correct only if the battery has the same impedance as the lightbulb and transmission line.
Clearly, this is not a video about transmission lines.

4) "You might think this is just an academic discussion, but that is not the case, and people learned this the hard way"
Indeed, it is not an academic discussion. It's a choice of invariant you can make, and the consequence is the amount of time needed to do the computation/measure.
Everything else (what invariant is the "real" one, what is "merely a mathematical concept") is a philosophical statement.

5) At 8:20 we see no Poynting's vector coming out. But Poynting's vector is supposed to tell the direction of light!
It's not wrong if you assume the light bulb is a resistor, with a very low efficiency in emitting light. But it must be very confusing!

6) "Thomson thought signals moved like water in a tube"
Well… the analogy he made was with heat in a bar. The result is known as the famous "law of squares" because the time it takes for signal/heat to travel is proportional to the square of the length.
The problem is that he didn't account for inductance, nor the leakage.
Had he taken account of inductance and capacitance, then he… would have the same equations as water (pressure waves) going through a pipe… https://en.wikipedia.org/wiki/Acoustic_ohm

7) "Electrons in the power lines are just wiggling back and forth, they never actually go anywhere" "Note that this drift velocity is extremely slow, around 0.1mm/s" "How far do the electrons go? They barely move, they probably don't move at all."
Suppose you assume electricity is like water going through a pipe. Let's take water vapor, or air if you prefer.
If you turn a pump on then you will get a pressure wave through the pipe, going at the speed of sound.
Now the speed of sound in a gas is, more or less, the average speed of molecules.
So the question we should ask is: what's the average speed of electrons.
If you answer 0.1mm/s, then you are a distracted viewer, this is the norm of the average velocity, or drift velocity. The "electron wind speed" if you prefer.
The speed of electron in a metal is around 1000 km/s, so "wiggling back and forth" "don't move at all" is widely incorrect and the "argument" is void. https://duckduckgo.com/?t=ffab&q=fermi+velocity&ia=web
The difference with air, is the air molecules are not charged, so they need to "collide" to interact while EM disturbances go at the speed of light.

0.1 mm/s is (more or less) the amount of copper you would get in an electrolysis, or consume in a battery. That's the best way you can use this information.

8 ) "Others argued that the field around the wire was carrying the energy, and ultimately this view proved correct."
Lorenz, and then Heaviside derived the telegrapher's equations, which is what "proved correct" refers to.
The big controversy at the time with Heaviside was the following:
a) Heaviside remarked that when the isolation of cables was deteriorating, the signal was improving. He proposed to degrade the cables, and his idea was rejected.
b) Heaviside derived, and then used the equations, and showed that if you artificially increase inductance (Pupin coil) and degrade the cables, you would get less losses and no distortion.
He was right, but the editor Preece had already tried the inductance part with a very poor result. Taking Heaviside for a mathematician out of touch with reality, he decided to censor Heaviside everywhere he could, until a debate, which he lost and was then fired.
So what proved correct is Lorenz equations. Lorenz who didn't believe in "energy in vacuum", but he did believe in inductance.

9) "the light bulb is 1m away and that is the limiting factor" with Derek's hands showing the distance between the battery and the bulb
Well no. The limiting factor is the distance between the switch and the bulb.
Which led to a lot of confusion (can you communicate FTL with a switch far away? no).
But if you say that the distance between the switch and the bulb is the answer, how does it appear to prove that "energy moves through vacuum" between the battery and the bulb?
This is not a convincing argument, and for a good reason: there is no valid argument possible.



Overall:
a) Derek insisted you use a counterintuitive (but correct) understanding of energy flow in a context where it is poyntless, claiming other ones are wrong, without proving they are wrong.
b) everything else is incorrect, except possibly the engineering history of the first transatlantic cables/the existence of transformers in the grid.

PS: Feynman claimed that the energies given are only correct in statics, it's not the case. Of course, the amount of electric energy depends on the gauge, so you won't get the same one.
Feynman claimed you could know where is energy using general relativity. You don't.
Einstein-Hilbert's version is defined with respect to the Einstein-Hilbert stress-energy tensor, which is the one given above. But tetrad/vierbein's version can use any stress-energy tensor.
So there is no argument here.
adx:

--- Quote from: Sredni on January 14, 2022, 01:10:01 am ---Has one slope really carried 100 times the energy of the other one? ...

--- End quote ---

I was going to say that the best I can come up with is that energy isn't real (in the sense of my previous post, being that it obviously exists in some form(s), but is not to be trusted), maybe something about what goes into a transmission line. But after Naej's great post, I might go off and think for a while. Maybe forever!
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod