Home Creators Posts Import Register Favorites Logout

Content

Hey Everyone,

Thank you so much for your feedback on the last video. Because of your feedback, we chopped out some sections at the last minute and I think it really improved the narrative.

I have one more video before this month is out and it is the follow-up to the big lightbulb circuit riddle from last year. I really want to get this one right to answer all the questions raised by the first video. So please take a look and let me know what you think.

Best,

Derek

Files

How Electricity Actually Works

Comments

-

The only critique I have is that the crux of the issue here (wish I was hoping you would address somewhere) was that your question was ambiguous, and that's what sparked (ha) the controversy. Specifically, you asked "how long would it take for the lightbulb to light up?", but that's open to interpretation, just like "how long does it take for the earthquake to reach this point?" is open to interpretation: the answer is different depending on whether you're talking about the P waves or S waves, about humans or animals or seismometer detection, about the initial jolt or or the average shake or the peak magnitude, etc.; naturally, you can't really claim people are wrong just because they interpret your question differently. I think most people interpreted your question as "how long will it take for the lightbulb to FULLY light up, and stay lit, visibly to a human?", because that's arguably the natural interpretation—if you returned a lightbulb to the manufacturer and said "it doesn't light up", you would be quite flustered if they replied "well, technically it does light up, but only at 0.1% of the normal brightness", or if they said "well, it does light up, but it then goes out 1 microsecond later". You'd feel (rightfully) annoyed at the manufacturer claiming you're wrong, because they would come across as trying to deceive you with English semantics, instead of selling you an honest product, when it would be obvious to any outsiders that your "it doesn't light up" sentence would've in fact merely been a (sloppy, yet obvious) approximation of "at best, the lightbulb only lights up dimly for a brief period". However, you meant to ask "how long would it take for a sufficiently sensitive lightbulb to show any kind of initial brightness that would be visible to a human if the human eye could detect changes on sufficiently short time scales"... at which point the answer of 1/c would've been pretty uncontroversial. (Some people may have still thought that's due to leakage current, and thus may have been wrong about the magnitude of the voltage, but it wouldn't have changed the answer to that question in particular.) But you only phrased it as "lighting up", which is not a scientifically precise question, and hence people took the simplest plain-English interpretation they could think of, and went with that, instead of treating it like a brainteaser with a non-obvious interpretation. Oh, and one other, more technical note (though I know you're already aware of this, but it might not be clear to viewers): voltage does not imply current. i.e. you can't just simulate the electric potential and then claim that means a lightbulb in that position would light up. At the absolute theoretical minimum in ideal conditions, you'd obviously need to account for the resistance in the steady state. In reality, for transient effects like this, you would also need to account for all sorts of weird effects. I recall reading that this even includes quantum tunneling (!) over the oxide layer on your bulb and/or switch. I imagine some of these effects might be transient, and thus the effective resistance might be higher initially (like a breakdown voltage of sorts), but I'm by no means an expert in this. (I love your videos by the way, and I really appreciate that you finally came back to this one. Thank you for all the awesome content!)

EEVblog

Happy that you finally did the follow-up video. I've only watched it once, but seems excellent and well thought out and adds some excellent new detail. It won't stop the nerds arguing though, the EEVblog forum thread on the video is up to 75 PAGES of debate, LOL. But I think you can sleep well at night after this one. The only thing I didn't see mentioned was the Quantum Electrodynamics (QED) hypothesis that might ultimately trump Poynting and Maxwell when it comes to energy flow inside vs outside the wire. But I can understand how that might side track this video completely. Well done.

Chris Stone

13:24 Tiny bit disappointed you didn't mention Radio. Otherwise an absolutely excellent follow-up. As another comment mentions, I think the main issue with the first video was the imprecise question.

EEVblog

You can also use transformer theory to analyse it. Many ways to skin this practical circuit cat.

Adam Berkan

It's not true that "it doesn't matter if the end of the wire is open". If the end was open, the return wire wouldn't be charged, and flipping the switch wouldn't have the same effect. (Maybe it would have half the effect since the wire would be neutral instead of charged?) I think the interesting question is: What if the switch is closed and at the same moment the end of the loop is broken. In that case it looks exactly like the loop is still closed (since information about the loop closing hasn't made it back yet). There's a bit of an interesting question if initially the end is closed, and then I open it, is the electric field still there? I think it is, except that leakage will eventually remove it. Anyway, great video! I'm back leaning towards camp Veritasium, but I'm not 100% convinced yet I'm looking forward to seeing the responses. Science YouTube at it's best!

GibbousLife

Great video Derek, as always. I can't wait for Mehdi's response.

Ram Ramabhadran

I loved this extensive circuit analysis. I love your videos, and I appreciate your breadth across many areas. I loved your video on " The Man Who Accidentally Killed The Most People In History", the one on snakes, etc.

Robert Blum

Great, this should settle it! A few comments: 3:00 - I’m struggling with this simplistic depiction, especially given the topic at hand. Since it is useful, maybe some comment about this being a simplified or traditional explanation will suffice. At 3:11, maybe the electrons/ electron sea could move even slower since they are still drifting pretty fast compared to the previously shown speeds. 12:40 - for me, the HFSS/Ansys simulations really hit home because they truly show what’s happening - no need to argue about semantics. But then you say that “it’s the electric field and not the changing magnetic field that creates the current”. To me, it’s all just an electromagnetic field. You need to consider both the changing electric and magnetic fields to arrive at the correct temporal solution. So I’d rephrase that. 14:36 - the Poynting vector really settles is again for me. Two requests: 1. Can you run the simulation without the bottom wire since it’s not part of the original problem, and 2. Can we see the steady state after the initial pulse and reflections have settled down? 18:47 - great results. I think your HFSS simulations seem to predict this pretty well, including the overshoot, so it’s a missed opportunity to not mention that. I’d also show/ say where on the x-axis your speed of light delay is (60+ ns after the green rising edge?) to emphasize that the initial rise is before that. You have about a fifth of the final voltage so that’s plenty to cause a bulb to glow - independent of whether that’s an LED or not.

katgod

I don't know if it would have been more confusing but a capacitor in series also shows the field part of electricity and you did in a way as you showed the lumped circuit which may have been more confusing to the novice. As someone who worked in the field no pun intended I am very familiar with cross talk which is that other unconected wire being lit up. Also you are really talking about AC here and I think many people are looking at it as DC. Well done as a clarification.

Tim Ludwig

Hi Derek, many thanks for the great follow up video! It clarifies a lot. That being said, I'll be a bit nit-picky in the following. 3:55 Here you present the electric field in an oversimplified way. You focus only on the electric field generated by the surface charges but disregard the electric field of the charges within the conductor. Of course, the field generated by the internal charges (electrons and ion cores) will be screened but not on such a small length scale as illustrated. 4:20 Electrons do push each other through the conductor by their repulsion; that's the source of the electric field. It's just not a short range interaction, as you correctly say at 4:30. 4:40 Misconception 2 is a bit misleading. Surface electrons (charges) are mobile as well. So, separating electrons into mobile and surface electrons is confusing. It relates to the previous point. 5:55 The gradient of the surface charges as the source of the electric field should be emphasized more, as it is the key point of your argument. You only present the steady state (DC-current) situation, which is insufficient to understand the transient effects on which you focus in the measurement. Of course, you provide details of the time-dependent simulations later on; but the connection between the two cases seems to be a bit weak. 6:50 The time the charges need to move is important, as you are focusing on the transient effect of building up the steady-state surface charge. This time-dependence is responsible for the initial signal you measure. 11:45 In your first video, you talked about the source and not the switch. You illustrate the difference nicely here. But emphasizing the difference might make the physics clearer. Your previous video was wrong in that point; or, at least, imprecise. 13:05 The results depend on the order (and potentially time delay) in which you break the circuit and open or close the switch. This can be figured out based on your previous explanation. The "initial" charge on the part of the wire (that is completely disconnected from the battery in the open-switch-broken-circuit case) depends on the way in which the system was put together; leading to different results. 16:31 You should speak of an ideal or a perfect conductor; not a superconductor. A superconductor is also an ideal diamagnet (i.e. it expels magnetic fields), which could massively affect the simulations for solving the Maxwell equations. 17:43 You seem to measure capacitance and inductance before moving the wires. Shouldn't both quantities depend on the geometry of the circuit and, thus, change when you move the wires afterwards? It doesn't really matter for your argument, of course. 10m vs. 1 light second Of course, for your main point, the difference between a 10m or a 1 light second wire doesn't make a difference. But for an AC signal with 50Hz or 60Hz it might very well make a difference: in the 10m case, 50Hz basically corresponds to a DC signal; in the 1 light second case, 50Hz means you have waves of surface charges traveling through your wires, which could interact. I hope those comments help a bit.

John Cox

@8:40 "now drift velocity is directly proportional to [the] electric field so the electric field must be largest inside the resistor" Verbal typo: Missing "the" @11:13 "simultaneously the new electric field from the modified surface charges radiates outwards at essentially the speed of light" (animation shows a radial electric flux emanating from the switch) The radial electric field shown in the animation differs from the 2D field shown @6:24 and from the 3D field shown @7:18. The lack of consistency is confusing. @12:00+ (simulations) These simulations are very interesting. I thought that the electric field dropped off much faster. But now I have a question: Why does this still work if the switch and the bulb are off-center from each other? If the switch and the bulb are off-center from each other, then you have a situation where the electric field on one side of the switch influences the wires on _both_ sides of the bulb. I gathered from the simulations that, given some time, the radial electric fields from the negative and positive wires will be straight out _except_ for very close to the switch, which is when the electric field changes direction. Aside from that location, the electric fields appear to be straight across. Suppose the bulb was offset a large amount such that it is opposite the negative wire and quite a distance from the switch so that there is no positive electric field in sight. By summing up the electric field vectors across a sufficiently long piece of wire from the negative power terminal, we would see a negative electric field radiating straight out from the negative wire towards the bulb opposite it (no tilting towards the positive terminal because the switch is a long ways away). According to these simulations and the explanation presented so far, the bulb would then be entirely under the influence of this negative electric field, and so there there would be no electric field _across_ the bulb and therefore no electricity would flow and the bulb would not turn on. But the bulb _will_ turn on if you wait a bit (specifically "distance/c" seconds). Why? What changed such that the negative electric field that was radiating straight out from the negative wire is now overridden such that there is a net electric field across the bulb, causing it to glow?

Corey Gemme

Good morning oh and thank you by the way for the shade ball :-). one thing I noticed and in all videos in general it is a little confusing or at least hard to concentrate or follow when the audio/narration does not match the text that that is displayed. You lose concentration on one when you’re trying to pay attention to the other :-)

Bahman Engheta

The statement @ 13m14s that "the bulb goes on regardless of whether the circuit is complete or not" will undoubtedly raise some eyebrows. I immediately wondered whether you were able to experimentally verify that with your set-up. You do justify that claim later @ 15m30s by bringing up the Lumped Element Model and revising the circuit diagram. Stating what the Lumped Element Model is, why it is used, and what the drawbacks of that way of thinking are– that was one of my most favorite parts of the video.

Chas Evans

Extremely good video, probably one of my favorites you've done recently! I love that you saw all the feedback and, frankly, confusion generated by the first video and came back with a deep dive into the aspects that generated the most controversy. Speaking as someone who's done some RF engineering (from the materials side of things), I found this video very informative, well explained, and well arranged. I especially liked the discussion the different ways we model circuits and how they simplify Maxwell's equations. As an aide, the thing that really solidified for me the idea that the energy is in the fields and not the wires was when an old boss of mine pulled a stripline bandpass filter from a pile of interesting circuit boards he'd collected over the years. If you've never seen one, you should google it: it's a circuit that is "disconnected," in that the copper traces on the board are not physically connected together. They're spaced such that a narrow range of frequencies can actually pass through them efficiently, but if you analyzed it as a lumped element circuit you would say that it was open. Seeing that circuit really convinced me that it was the fields that were of the utmost importance, not the electrons themselves.

Brad Hallisey

I'm glad you did a follow up. I'll admit I was not a fan of the original video. After a couple viewings I got what you were saying but you usually do a much better job at getting your point across. This video absolves those sins. I'm glad to see the comments by other YouTubers joining. I agree, it was a fun debate to have. Learning often takes some amount of disagreeing!

David Chatenay

I watched one of these response videos and I found the negativity was too much for me. It didn't make me want to subscribe to their channel, and it also made me realize that your approach -positive, open, humble- is the reason I enjoy your videos so much, and this new one doesn't disappoint. So thank you! Personally I found the "third wire" to be a bit of a distraction later in the video, which is pretty dense But it also made me think about the trick where neon tubes will light up if you put them under high-voltage transmission lines (which I guess is due to the intense fields from the lines above). Maybe for another video?

Chris Rogan

One of my favorite videos. Thought experiment that has been vexing me. If you had a perfectly taught cable in this experiment and tugged on it on one end, would it instantaneously be pulled on the other end and therefore a simple message received on the other end instantaneously (e.g. a binary one or on bit)? Or do we have to assume some elasticity in any material that would create a lag from one end to the other that is more than the speed of light?