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

[bsfeechannel]

Relativity is one reason that Derek's proposed correct answer is clearly wrong.  If you accept his impossible magical conditions, an observer at the switch will see the light turn on at about 6.66ns, or twice as long as he says.  An observer at the light will observe both events to be nearly simultaneous.  Only an observer halfway between the two or very far away above or below the plane of the wires will see the approximately 3.33ns transition time.  So (E) None of the above--unless you are in just the right spot.

So, if I have a transmitter, 3,000 km away from you, and I send you a signal, will it take around 10 ms to reach you, or 20 ms because I need to observe you receiving the signal?

Nothing.  Look at this whole discussion!

I've looked at it and what I've found is a bunch of people nitpicking on unimportant questions and jealous because Derek managed to engage the whole internet around a theme that caught some "knowledgeable" people with their pants down.

[adx]

... I proposed no specific model for this case, others came up with the transmission line.  I think there is  more to it, as even with no resistance it should also radiate energy into space--so even an ideal transmission line is not a perfect model. But that doesn't affect the outcome of the question posed.

It's a great big terminated folded dipole. Was going to be a caveat in my scaled-down (15km cables) model.

[bdunham7]

So, if I have a transmitter, 3,000 km away from you, and I send you a signal, will it take around 10 ms to reach you, or 20 ms because I need to observe you receiving the signal?

That question can only have meaning at all because you can assume that we are in the same reference frame--not moving relative to one another.  Same for the switch and the light of course, and the answer is that you can, in fact, assume that I got the signal after 10ms and when I get it, I can assume that you sent it 10ms "ago"--if those notions are somehow comforting and help you keep things straight in your mind.  And if, after observing the light come on 6.66ns after you throw the switch, you can say that the light actually turned on 3.33ns "ago" since you know the distance.  But like the notions of suction, centrifugal force and so on--all 'fictitious' concepts you would presumably lampoon as a crutches for the weak-minded--the (incorrect) notion that there is any master clock at all will quickly prove untenable with even slightly more complex problems with relativity.  You, me and the signal itself are all on different clocks, ours just happen to be progressing at about the same rate for now.

[adx]

While I'm at it, the repeating L C model does model delay and distance (pF per length, nH per length). As shown by the delay in the signal. In Dave's simulation IIRC (and departing from calling people by their vlogger name). The same could be done with the feed lines, if they existed in the example. Looking at my schematic, you'd think they were there, and I started believing they existed in the example.

[HuronKing]

Given the choice between an EE graduate who can solder up a circuit, use an oscilloscope, and one who's able to quote verbatim Maxwell's four equations, I'm 100% sure who I'd go for.

You're free to hire anyone you want, of course, but I disagree that there is any equivalency between these two skills. In fact, I'd say the second thing isn't even a skill at all, it's a strawman.

Quoting the equations =/= being able to use the equations. If EEs come out of school unable to use the equations or conceptually apply them to engineering problems, then they never really learned them, did they? I sometimes do get those students with good memory who can recite the equation at me during a lecture - and I challenge them to actually apply the physical meaning to understanding a problem - like why reactive components have frequency response, why motors can become generators, and why harmonics can develop in motors controlled by VFDs.

All these things require EEs to understand the physics of the equations - not just quoting them.

Just like owning an oscilloscope or a soldering iron is no guarantee they know how to use it.

[Howardlong]

Given the choice between an EE graduate who can solder up a circuit, use an oscilloscope, and one who's able to quote verbatim Maxwell's four equations, I'm 100% sure who I'd go for.

You're free to hire anyone you want, of course, but I disagree that there is any equivalency between these two skills. In fact, I'd say the second thing isn't even a skill at all, it's a strawman.

….

All these things require EEs to understand the physics of the equations - not just quoting them.

Just like owning an oscilloscope or a soldering iron is no guarantee they know how to use it.

But I explicitly never made that claim, you did. And you accuse me of making a strawman?! ;-)

You’re also misrepresenting my general position for some reason.

For the last time, I am not suggesting that the general basis and awareness of these fundamentals aren’t taught, I’m questioning the level of depth imposed, particularly at the undergrad level, when it’s at the expense of other skills they’ll need from day one, like being able to measure things with an oscilloscope or read and understand a datasheet for example.

I still don’t know why you’d need a working understanding at such a deep level of the details of the physics of Maxwell’s equations when no EE I know of ever thinks at that level, even when using EM design tools. As someone mentioned earlier, it’s the same as the holes and electron doping model of PN junctions, it’s an interesting sidebar for an EE, but unless you’re building semiconductors from raw materials, it has zero relevance in how an EE goes about their daily tasks. EEs are not physicists or chemists, they are about providing practical solutions built on scientific principles, but the vast majority of the time they have no need to do the scientists’ jobs for them.

Finally, in conclusion, no, I am not suggesting ceasing the teaching of Maxwell’s equations et al, I’m merely questioning the value of going any further than a general awareness to the vast majority of EEs and undergrad level when there are so many other skills they’ll actually need from day one. A chef doesn’t need to be a biologist or a chemist to make a great omelette or soufflé.

[Sredni]

The first and most important thing an engineer should learn is: know the limits of your models.
You and Mehdi think the effect of the field propagating from the switch can be modeled by a transmission line that extends in the perpendicular direction? Think again.

Which fields?  If you are talking of the E-field from the positive battery terminal, it should already have propagated to the lamp because if it can propagate over 1 meter of space it can certainly propagate over the switch contacts (and no I don't mean leakage current).  There are no new fields propagating 'from the switch' when it is closed, all of the effects in this circuit are a direct result of current flow--yes, moving charges again--down the wires and the changing fields that result from that.

The current in the wire flows when there is an E field INSIDE the wire. The E field inside the wire is created by the SURFACE CHARGE on the surface of the wire (and at any gradient in conductivity and permeability).
Before closing the switch, you have surface charge accumulated at its open terminals. When you close it, this charge starts recombining with relaxation times (veeeeeery fast) and the change in charge distribution will be reflected in a change in the electric field. While the surface charge perturbation propagates along the wire, this change in the electric field configuration propagates at the speed of light in the space around it. In particular, the perturbation in the electric field (and in the magnetic field associated with the charges set in motion inside the portion of conductor where j is nonzero) will reach the opposing load in d/c seconds.

This propagation along the transversal direction IS NOT MODELED by the transmission line model.

As far models go, the transmission line model seems to pretty closely match the experimental results for the scaled-down versions.  And those that used the lumped-element model of a transmission line have acknowledged the limitations such as the spike from the capacitor being first (Mehdi) and the additional propagation delay across the line due to the model not incorporating physical size in that dimension (Dave Jones).  It's not like they're ignorant of these things.

The transmission line model does not model the effect Derek is talking about. It models the recombination of surface charge along the cable, if you will, but on the transverse direction it is completely blind because it used lumped components that have zero dimensions. If someone had taught them the basics of EM down to that level (a level that some seem to think is useless and a waste of time), they would not waste our time with a model that does not model what needs to be modeled.

You need to fold back to EM field simulators.
So far, the only simulation I've seen that has a suitable answer is that by Ben Watson



Youtube video: https://youtu.be/aqBDFO1bEs8

[iMo]

Yea, that nice simulation supports the idea of the "path of least resistance/effort".. 

..The E and B fields follow the "path of least resistance/effort" (a fundamental law of physics, afaik) - that means directly from the battery to the bulb 1 meter apart.
Whatever shape and length of the wires you choose, the bulb X meters apart from the battery will always lit in X meters/c secs.

[SandyCox]

You need to fold back to EM field simulators.
So far, the only simulation I've seen that has a suitable answer is that by Ben Watson



Youtube video: https://youtu.be/aqBDFO1bEs8

I fully agree that a proper EM simulation and/or experimental evaluation are the only ways to study this problem.

I also looked at this interesting video. My only criticism is the fact that he integrates the magnetic field around a loop to calculate the current. We know that Ampere’s law is incomplete. He should use the full Maxwell’s 4th equation.

I can’t find anything wrong in Veritasium’s video.

[Sredni]

You need to fold back to EM field simulators.
So far, the only simulation I've seen that has a suitable answer is that by Ben Watson


Youtube video: https://youtu.be/aqBDFO1bEs8

I fully agree that a proper EM simulation and/or experimental evaluation are the only ways to study this problem.
I also looked at this interesting video. My only criticism is the fact that he integrates the magnetic field around a loop to calculate the current. We know that Ampere’s law is incomplete. He should use the full Maxwell’s 4th equation.
I can’t find anything wrong in Veritasium’s video.

I agree that the simulation is far from perfect in the way it finds what happens INSIDE the wire.
(also, the way the switch is simulated by two generator in opposition might create discrepancies with the actual case) but at least it answers the main question about when the perturbation arrives at the load.

What is missing in this simulation is the detail of the electric field inside the wire and load.
But it is certain that as soon as the perturbation arrives at the load (d/c seconds after the switch has been closed) the surface charge there will be perturbed and the electric field will start to build up. Of course, since it is missing the contribute of the charge recombination that is travelling on the surface of the cable to the Moon and back (so to speak) it will not be the full electric field, directed along the axis of the cable, but some abominable distribution that will result in a local current density where the load is.
So, some sort of current will flow inside the load, but I doubt it will be the orderly current we expect to see in a lumped circuit or in a distributed circuit modeled with a chain of lumped components.
(This is why in the other thread I wrote that I am not sure Derek is completely right. He sure is right about when the perturbation arrives there and that something happens in the load)

Moreover, I don't think it would be easy to measure such tiny effects without disturbing them in a non-negligible way. This would not be an easy measure to do.
Ben Watson should redo the simulation and show the electric field inside the conductor. That will give the definitive answer about what happens according to classical ED. And it might also give some insight on how to proceed to measure the 'current' in the time from d/c to 2*lenght/c.

The only sure thing here is that all models with the transmission lines placed sidewise and the reasoning about the first capacitors on the side is... the result of not knowing the limits of the model used (to say the least).

[bdunham7]

In particular, the perturbation in the electric field (and in the magnetic field associated with the charges set in motion inside the portion of conductor where j is nonzero) will reach the opposing load in d/c seconds.

I don't think anyone disagrees with that.  The questions are whether that perturbation will turn the light on and then more generally, what will the current through the lamp look like over time.

This propagation along the transversal direction IS NOT MODELED by the transmission line model.

I agree that the transmission line model has flaws which depend on the exact model used.  But a least some of those have been acknowledged.  I think there are probably additional issues in the longer-term behavior of the hypothetical circuit that will be negligible on smaller-scale versions (no I don't mean solar wind!) but so far the experiments have sort of matched a corrected transmission line model in the longer timeframes.  If you are only interested in the precise nature of the response in the first few nanoseconds, then your concerns are entirely valid and the exact geometry would need to be stipulated before you could simulate or calculate anything.  However, if the geometry is very close to an ideal transmission line, then I would expect its behavior to correspond closely to an ideal resistor with a delay in proportion to its dimensions.

However, if you can specify a reasonably scaled equivalent and then model it, and your model differs significantly from the adjusted predictions of the transmission line model, I can build it and test it experimentally--if someone hasn't beaten me to it.

The transmission line model does not model the effect Derek is talking about.

Other than the limitation of light speed and the physical dimensions, what would an accurate transmission line model lack? 

So far, the only simulation I've seen that has a suitable answer is that by Ben Watson

...who refers to the setup as "pretty much like a transmission line, which is what it is" and then shows a delayed step response quite similar to the others we've seen...

 

[HuronKing]

Given the choice between an EE graduate who can solder up a circuit, use an oscilloscope, and one who's able to quote verbatim Maxwell's four equations, I'm 100% sure who I'd go for.

You're free to hire anyone you want, of course, but I disagree that there is any equivalency between these two skills. In fact, I'd say the second thing isn't even a skill at all, it's a strawman.

….

All these things require EEs to understand the physics of the equations - not just quoting them.

Just like owning an oscilloscope or a soldering iron is no guarantee they know how to use it.

But I explicitly never made that claim, you did. And you accuse me of making a strawman?! ;-)

You’re also misrepresenting my general position for some reason.

You claimed that I, by teaching how to use Maxwell's Equations, am teaching 'unnecessary irrelevant details' and I'm better off spending more time on the intricacies of oscilloscopes or something. And, while not in direct response to me, seemed to equate the notion of understanding the Maxwell's to merely quoting them. Am I wrong?

For the last time, I am not suggesting that the general basis and awareness of these fundamentals aren’t taught, I’m questioning the level of depth imposed, particularly at the undergrad level, when it’s at the expense of other skills they’ll need from day one, like being able to measure things with an oscilloscope or read and understand a datasheet for example.

I make students read datasheets and use meters in lab. But that's neither here nor there. Let's move this forward a bit - how much is too much depth for you? In the context of this discussion, the depth is teaching the Poynting Vector and that energy in the EM field does not flow IN wires but in the space around them, even at DC. Is that too much detail? And if it is... then how the heck are students supposed to understand how a coax works? Where do the datasheet parameters on a transmission line come from? Why is there such a thing as impedance matching? Reflection coefficient? SWR? Harmonics?

I still don’t know why you’d need a working understanding at such a deep level of the details of the physics of Maxwell’s equations when no EE I know of ever thinks at that level, even when using EM design tools. As someone mentioned earlier, it’s the same as the holes and electron doping model of PN junctions, it’s an interesting sidebar for an EE, but unless you’re building semiconductors from raw materials, it has zero relevance in how an EE goes about their daily tasks. EEs are not physicists or chemists, they are about providing practical solutions built on scientific principles, but the vast majority of the time they have no need to do the scientists’ jobs for them.

Because I'm training engineers - not technicians. How is an engineer supposed to know the difference between a PNP and a NPN transistor without understanding electron-hole and doping theory? Why do diodes have a threshold voltage? Why do MOSFETs have zero gate current? Where does the equivalent model come from? Why do transistors have active, triode, and saturation regions? What do those terms even mean?

When I was doing my MSEE I had a professor who worked for Samsung come teach a class on advanced amplifier design and he was appalled at the lack of understanding in the graduate students about all these things.

Finally, in conclusion, no, I am not suggesting ceasing the teaching of Maxwell’s equations et al, I’m merely questioning the value of going any further than a general awareness to the vast majority of EEs and undergrad level when there are so many other skills they’ll actually need from day one. A chef doesn’t need to be a biologist or a chemist to make a great omelette or soufflé.

So, a typical EE's direct educational exposure to Maxwell's Equations is:
1 course on vector calculus (Stokes' Theorem, curl, divergence, vector and scalar products, so Heaviside's expression of Maxwell is comprehensible. Fun fact - it was Heaviside and Gibbs applying vector calculus to EM that caused it to be adopted by ALL of physics)
1 course on undergraduate EM physics
1 course on Applied EM

And that's too much? 

I go a step further and ask my students to conceptually apply the equations to topics studied in lab. Earlier in this thread, you told me that's too much? 

And the part in bold is funny you should mention that. I have a friend who is a professionally trained chef who is always talking about the Maillard Reaction, sugar breakdowns, caramelizations, and the different temperature thresholds involved in making great omelettes and souffles. I suppose your mileage varies but I was impressed to learn how much physical chemistry education he received:
https://en.wikipedia.org/wiki/Maillard_reaction

There is a difference between a chef and a common line cook.

[vad]

You, me and the signal itself are all on different clocks, ours just happen to be progressing at about the same rate for now.

Speaking of signal’s clock, does photon experience proper time?

[vad]

I am wondering, what would be the correlation between the group of electrical engineers who strongly argue against Veritasium, and a group of engineers who often struggle with passing EMI/EMC tests.

[bsfeechannel]

I am wondering, what would be the correlation between the group of electrical engineers who strongly argue against Veritasium, and a group of engineers who often struggle with passing EMI/EMC tests.

I bet on 100%.

[Sredni]

In particular, the perturbation in the electric field (and in the magnetic field associated with the charges set in motion inside the portion of conductor where j is nonzero) will reach the opposing load in d/c seconds.

I don't think anyone disagrees with that.

Well, the transmission line model does not model that transverse propagation of the perturbation.

The questions are whether that perturbation will turn the light on and then more generally, what will the current through the lamp look like over time.

Agreed.

This propagation along the transversal direction IS NOT MODELED by the transmission line model.

I agree that the transmission line model has flaws

My point is that it's not flaws. It just can't model that. The propagation of voltage and current in the direction of the transmission line is modeled thanks to the exchange of energy between capacitors and inductors that represent capacitance and inductance per unit length along the direction of the line. But transversely it's basically zero dimensional.

If you are only interested in the precise nature of the response in the first few nanoseconds, then your concerns are entirely valid and the exact geometry would need to be stipulated before you could simulate or calculate anything.  However, if the geometry is very close to an ideal transmission line, then I would expect its behavior to correspond closely to an ideal resistor with a delay in proportion to its dimensions.
...
Other than the limitation of light speed and the physical dimensions, what would an accurate transmission line model lack? 

My take is that no, you can't use a transmission line model. You get something similar but it certainly misses in toto the behavior object of the discussion. That initial 'spherical explosion' around the region of the switch+battery and the effect it has on the electric field inside the load is what we're after. In a transmission line model there is no delay between the modification of the charge on the lower plate of the capacitor and its upper plate. It's instantaneous. Faster than light (for convergence reason certain simulator might add some smoothing here and there - trapezoidal instead of exactly rising signals, parasitic losses to avoid infinities and so on...), therefore the model that uses a chain of lumped components is completely oblivious of the fact that your line conductors are separated by a distance d. And if you think that the capacitance per unit length encodes this information, no. You only know the capacitance per unit length. You can change that, for example, by making the cables flat and ten times the surface but at the same distance d. So C does not give you information on d. 

EDIT: off the top of my head, it's possible that the transmission lines model is modeling what would happen if the separation between the cables was infinitesimally small (but still magically keeping the same values for L' and C'). In that case, instead of a principal 'red glow' going strong on the lower branch with just a faint ghost on the upper branch that we see in Ben Watson's simulation, we would see an almost equally strong red glow on both branches bouncing back and forth. Or maybe it's the beer.

However, if you can specify a reasonably scaled equivalent and then model it, and your model differs significantly from the adjusted predictions of the transmission line model, I can build it and test it experimentally--if someone hasn't beaten me to it.

My problem here is that I do not actually know how delicate this effect is. When you attach the probes and the scope, you add a lot of stuff. And if this current is just flowing temporarily in the load but not in the rest of the circuit (WHAT??? Yes.) how can you measure it with an external instrument?
I don't know, maybe ditching instruments and using a load that will spit photons with the slightest current? Oh, but engineers should know how to use oscilloscopes and solder PCBs, not know this exotic and useless physics... :-)

(I am just gently pulling some legs, relax people... ;-P )

[bdunham7]

Speaking of signal’s clock, does photon experience proper time?

I don't know.  Using the most basic concepts of relativity, it would appear that at least in a vacuum, photons exist for zero time, or perhaps Planck time.  Not that I'm going to be able to tell the difference.  But even if it is zero, that just means its clock is stopped. 

I am wondering, what would be the correlation between the group of electrical engineers who strongly argue against Veritasium, and a group of engineers who often struggle with passing EMI/EMC tests.

Engineers pass EMI tests by using appropriate design tools, known design methods and models, experience and EMI pretesting. I think the objections to Veritasium's video are being mischaracterized.  Nobody seriously doubts that there will be some effect at the load within a few nanoseconds of closing the switch.  The rest of the story is in details, which the video itself sorely lacked.  My criticism of it is that by gearing it to the 'wide audience', it causes some who won't comprehend the details to have a grossly erroneous idea of how power distribution and signals work in practice.  I posted an example of this earlier in the thread.

[Someone]

Lets roll the preface again, you even quoted it in full:

lol, still going strong. Yet only one mention of Litz!

The "argument" is framed around such a nonsense assumption/premise, its intentionally trying to break peoples heads by equating radically different modes of propagation.

Back in the real world, with the sort of power people actually use to power lighting devices (I'll send you a $1 if you actually have an installed and in use microwave power system) is so low in frequency that all these "clever" ways of moving power are irrelevant and lost in the noise.

Litz wire gives the practical example that people can play with at reasonable frequencies and see the effects, at some frequency the energy does concentrate to the surface/outside, and below that it appears the same as a solid of the same cross section...

Litz wire and the skin effect has NOTHING to do with the transfer of power at steady state in DC.
One might argue that the proximity effect can change a hair here or there but that is again missing the point.
But I am sure that a lot of people who did not study EM and just heard Dave - uncorrected - talking about the Poynting vector pointing the other way around in DC (LOL) and the supposed role of the  skin effect (in what is not clear), well they will think you have a point.
This is how disinformation works on Youtube. It's self-feeding.

Noting to do with DC? its an excellent example of just that point! At DC all that matters is the cross sectional area (for conductors of same material). Yet one has much higher surface area, if surface area had something to do with DC power then we'd expect it to be different. ergo DC power delivery is not related to surface area, proximity effect, external fields etc. A practical and tangible way for people to consider and experiment with where currents are flowing in wires, that shows quite easily measurable results many hobbyists would be able to reproduce (much easier than the speed of light scale, measuring propagation delay in "long" wires).

Next step from there is real word things like current crowding, where finite element analysis comes in to play as its hard to come up with practical examples large enough to observe/measure.

Back to the real world where copper has a finite conductivity. Yes, now size matters. Still, the power is not carried inside the wires

Baloney! If the power were outside or on the surface then the above example you immediately dismissed as being irrelevant at DC (changing the surface area and relative positions of multiple conductors) would make a difference. You cant have it both ways.

Who really cares that a superconductor has a singularity in its EM equations producing a relative permeability of 0, so that infinitely thin shells can replace solid conductors. Pretty much no-one, and "real" superconductors don't actually reach that point of perfection either. Its just as stupid a thing as this original "thought experiment". Adding very little value to anyone.

DC is not AC, AC mains so far from GHz voodoo that they are not treated the same, EM field propagation is irrelevant for DC battery (or mains) examples of light bulbs connected with wires.

[adx]

Quoting myself again...

While I'm at it, the repeating L C model does model delay and distance (pF per length, nH per length). As shown by the delay in the signal. In Dave's simulation IIRC (and departing from calling people by their vlogger name). The same could be done with the feed lines, if they existed in the example. Looking at my schematic, you'd think they were there, and I started believing they existed in the example.

Looks like I misinterpreted / misremembered what someone was saying (probably Sredni). I can't find any claim that physical delay isn't present in the longitudinal direction of a SPICE themed transmission line model. I forgot that the original question is the transverse delay, another one of the strings that Derek's video tugs at which are usually left untugged. If it came up in work (like that SFP camera I mentioned, getting a 10GHz+ signal out in any way sideways from the connector is going to cause nervousness about differential delays), then it's EM field solver for you. Unless you're rich, that means doing it in your head, or grabbing a VNA (cheap compared to a good EM field solver or even ps scope) and a fistful of SMA connectors to see what you got wrong.

As as a sort of aside, no one (not even a student at a university building such a camera) is going to symbolically derive formulas to describe the SFP connector + cage + board, plot stuff for days to get a feel for what they do, discuss the solvability of it all, reluctantly throw it in Matlab and kiss their supervisor's chances of a research paper goodbye, so they can hit a button and get one number out to say oh 1.23ps, um I guess that doesn't matter. Perhaps I was overly hard on educators and academics in recent posts because it's not about the intention (as I said), but (as I see it) a steadfast and intractable refusal to accept the reality (and imprecision) of how it's done in industry. Not all, of course, but as a system.

What I was getting at with the feed lines is I'm sure I've seen 2D grid arrangements of lumped elements used to implement a kind of field solver in SPICE for RF work.

[Sredni]

What I was getting at with the feed lines is I'm sure I've seen 2D grid arrangements of lumped elements used to implement a kind of field solver in SPICE for RF work.

Now, this is interesting (and potentially useful). Please, should you find it post it here.

[SandyCox]

What I was getting at with the feed lines is I'm sure I've seen 2D grid arrangements of lumped elements used to implement a kind of field solver in SPICE for RF work.

I always thought Spice uses the ABCD parameters of the transmission line, like I did in my theoretical analysis on page 19.
Back to the original problem:
My view is that our understanding of the physical world is like the onion with many layers that Feynman spoke of:
1.   The top layer is Kirchhoff’s laws and electron flow. I use them with great confidence in circuit analysis even though I know that they do not tell the full story.
2.   Then come Maxwell’s equations, of which Kirchhoff’s laws are a special case. Now we can explain electromagnetic waves.
3.   Beyond that I’m lost, but I know that there are more layers of understanding.
I’ve heard that everything in Electromagnetism can be explained in terms of either the electric field or the magnetic field. We don’t need both concepts.

Maybe this is all a special case of the Dunning Kruger effect?

[EEVblog]

... I proposed no specific model for this case, others came up with the transmission line.  I think there is  more to it, as even with no resistance it should also radiate energy into space--so even an ideal transmission line is not a perfect model. But that doesn't affect the outcome of the question posed.

It's a great big terminated folded dipole. Was going to be a caveat in my scaled-down (15km cables) model.

Don't forget that it takes time for a signal to propagate along a dipole antenna too.
Many people seem to forget that the 1m/c answer only applies when it's 1m away. But practically any physical wire that is more than 1m away is not going to give you the 1m/c answer.
We are being trolled.

[bsfeechannel]

I always thought Spice uses the ABCD parameters of the transmission line, like I did in my theoretical analysis on page 19.
Back to the original problem:
My view is that our understanding of the physical world is like the onion with many layers that Feynman spoke of:
1.   The top layer is Kirchhoff’s laws and electron flow. I use them with great confidence in circuit analysis even though I know that they do not tell the full story.
2.   Then come Maxwell’s equations, of which Kirchhoff’s laws are a special case. Now we can explain electromagnetic waves.
3.   Beyond that I’m lost, but I know that there are more layers of understanding.
I’ve heard that everything in Electromagnetism can be explained in terms of either the electric field or the magnetic field. We don’t need both concepts.

Maybe this is all a special case of the Dunning Kruger effect?

The Dunning-Kruger "effect" says in short that ignorance can't recognize itself.

Since you recognized your own ignorance with item number 3, you're not under its "effect".

As for the electric and magnetic fields, yes, they are two manifestations of the same phenomenon. That's another thing that Maxwell's equations tell us.

[bsfeechannel]

Many people seem to forget that the 1m/c answer only applies when it's 1m away. But practically any physical wire that is more than 1m away is not going to give you the 1m/c answer.
We are being trolled.

Dave, come on man, we're not being trolled. We're being challenged. Our "engineering 101", whatever that be, is not cutting the mustard anymore.

Let's take this opportunity to upgrade the most important piece of equipment in any engineering lab: our brains.

[adx]

... I proposed no specific model for this case, others came up with the transmission line.  I think there is  more to it, as even with no resistance it should also radiate energy into space--so even an ideal transmission line is not a perfect model. But that doesn't affect the outcome of the question posed.

It's a great big terminated folded dipole. Was going to be a caveat in my scaled-down (15km cables) model.

Don't forget that it takes time for a signal to propagate along a dipole antenna too.
Many people seem to forget that the 1m/c answer only applies when it's 1m away. But practically any physical wire that is more than 1m away is not going to give you the 1m/c answer.
We are being trolled.

Yes, he is being intentionally tricky. He doesn't have to be right, just has to have a way of showing he had a good explanation in his next video(s). That way he can prove he's right, rather than just be right. And say stuff like "I used 1/c to show that not all parts of the circuit are 1m from the bulb, you see..." and proceed to show an EM simulation done on one of Sandia Labs' supercomputers (after poo-pooing all those engineers arguing over simple SPICE transmission line models). Despite the original question being "after I close this switch, how long would it take for ...?". Not complaining, just thinking out loud what I would do!

Someone here (can't find easily) said it'd work without the wires, which is correct in the sense an EM transmission would occur in the blip as the charged switch is closed (the light wouldn't "come on"). That has no option but to be 1m/c.

Warning really bad explanation follows:

The switch is a dipole antenna which fills (or really defills) the electric field in the space on the right differently from the space on the left, as the voltage appears (or disappears) over the switch's length. While that is happening, the current from it being shorted out and flowing from left to right, creates (or does it?!) a magnetic field around the axis of the switch. The fields travel outward in time and hits the lamp, inducing a tiny fraction but the same kind of effects in its terminals (still arranged left to right), because it is sampling a bit of the space to taste the electric field between left to right and magnetic field with a wire parallel to the transmitter (it will be mainly an E-field receiver because there is no current loop on the terminals).

Depending on frequency the structures might be a good match to the impedance of space, that is the voltage and current travelling down the arms of the dipole and maybe reflecting back is timed in such a way (by shaping the antenna) can both focus the radiation pattern and ensure V and I are in the correct mix to get straight into the fabric of the universe. In any event, in the far field, the E and H fields 'convert' (somehow, I have no physical understanding of how this works beyond what Maxwell says) to self-propelling EM radiation (both E and H fields radiating out in the same form as described above, but self-sustaining, able to contain multiple cycles of alternating E and H field in an expanding shell). But the antenna doesn't have to be a good match for this to work, it will sort itself out.

That ended up a lot more messy than I expected. If it weren't such a mess, it'd be all anyone really needs to know about radio stuff!