#562 – Electroboom!

[bsfeechannel]

Dave, KVL doesn't hold under a varying magnetic field.

Repeat it until enlightened.

Mehdi now comes with yet another explanation for his lack (which he admits) of understanding of electromagnetism : the old vs new definition (i.e. what he thinks is the definition) of voltage.

As for Lewin's experiment, it was just Faraday's experiment with fancy test equipment. Faraday did what Lewin did, using an unshielded galvanometer with completely unshielded probes and using an electromagnet to generate the varying magnetic field.

If Lewin was wong in anything, Faraday was also wrong. And that's not the case for about the last two centuries.

So, give up this nonsense already. Please.

[jesuscf]

Dave, KVL doesn't hold under a varying magnetic field.

Really?  So according to the statement, AC circuits (with transformers and such) can not be solved using KVL!  You better go tell the whole electric power industry that they have been analyzing circuits under the effects of magnetic fields incorrectly for the last 160 years!!!

[bsfeechannel]

Dave, KVL doesn't hold under a varying magnetic field.

Really?

Really.

So according to the statement, AC circuits (with transformers and such) can not be solved using KVL!

AC circuits can be solved with KVL as long as you maintain all the varying magnetic fields away from the path of the analyzed circuit. This is what happens with a transformer for instance. All the varying magnetic fields are inside the transformer core, not inside the path of the circuit.

You better go tell the whole electric power industry that they have been analyzing circuits under the effects of magnetic fields incorrectly for the last 160 years!!!

I don't need to. The whole power electric industry never made the mistake of analyzing a circuit under a varying magnetic field using KVL. In fact the whole power electric industry started out because of Faraday's discoveries which was solely explained by Maxwell's equations. And that's what we use to this very day.

[thinkfat]

I think Lewins model on which he applied KVL was incorrect. The lines in a circuit diagram don't interact with a magnetic field.

If you watch the original lecture, the error creeps in when he assumes "A1=A2" and "D1=D2", but this just doesn't hold.

[bsfeechannel]

I think Lewins model on which he applied KVL was incorrect. The lines in a circuit diagram don't interact with a magnetic field.

If you watch the original lecture, the error creeps in when he assumes "A1=A2" and "D1=D2", but this just doesn't hold.

What is absolutely amusing about the people who can't accept that KVL doesn't hold under a varying magnetic field is the disparity of explanations.

For Mehdi at first, Lewin's experiment was flawed because of bad probing. When he tried to replicate the experiment and couldn't get a different result no matter how good probing he did, he tried to reinvent the theory creating what he calls the new definition of voltage.

Dave came with a Deus ex-machina called "Practical Engineering". It's a completely irrational explanation with absolutely no basis on science.

Cyriel Mabilde came with the "masked" EMF, which he states is much simpler to understand than the real thing which is Faraday's law.

Others tried to manipulate the math to make it appear that the voltages found in such a circuit actually add up to zero.

Now Lewin doesn't know how to apply Kirchhoff's law to a circuit, and, worse, doesn't know how to draw diagrams.

So summarizing, KVL always holds because Lewin didn't know how to probe the circuit, because he's adherent to the "old definition of voltage", because he's a physicist and not a practical engineer, because he doesn'1 know how to manipulate the math so that you make Maxwell's equation of magnetic inductance to always show a zero on one o the sides of the equal sign and finally because he doesn't understand how to model circuits and draw diagrams.

The truth is that "KVL always hold" is a myth, brilliantly debunked by Lewin. And up to this day no one could show that those two resistors in a loop under a varying magnetic field will not show different voltages, no matter what.

You can fool an audience. But, as Richard Feynman said, nature cannot be fooled.

[armandine2]

brilliantly debunked by Lewin.

experimentally?

[bsfeechannel]

Yes. Try it yourself.

[thinkfat]

I think Lewins model on which he applied KVL was incorrect. The lines in a circuit diagram don't interact with a magnetic field.

If you watch the original lecture, the error creeps in when he assumes "A1=A2" and "D1=D2", but this just doesn't hold.

Now Lewin doesn't know how to apply Kirchhoff's law to a circuit, and, worse, doesn't know how to draw diagrams.

Precisely. You cannot on one hand assume those "wires" in the circuit be "ideal" and effectively form one "node", and at the same time expect them to span up a area for the magnetic flux to pass through. It doesn't work that way. These circuit lines are dimensionless. They have zero length. The area is zero, therefore, and no magnetic flux passing through them. Besides, KVL is a mere calculation tool for network analysis. If your result doesn't match the measurements, clearly something is wrong with the model. But not with the method as such. That's a bit like blaming spice for not matching reality when you forgot to put an element.

[bsfeechannel]

Precisely. You cannot on one hand assume those "wires" in the circuit be "ideal" and effectively form one "node", and at the same time expect them to span up a area for the magnetic flux to pass through. It doesn't work that way. These circuit lines are dimensionless. They have zero length. The area is zero, therefore, and no magnetic flux passing through them. Besides, KVL is a mere calculation tool for network analysis. If your result doesn't match the measurements, clearly something is wrong with the model. But not with the method as such. That's a bit like blaming spice for not matching reality when you forgot to put an element.

The myth says that KVL ALWAYS holds even when a circuit is under the effect of a varying magnetic field. Lewin showed at least one circuit where KVL doesn't hold. Everyone repeated his experiment and obtained the same result. Conclusion: KVL does NOT ALWAYS hold. End of story.

If Lewin didn't wear the appropriate hairstyle or used a chalk of the wrong color, that's absolutely irrelevant.

[thinkfat]

Do you have an actual argument or are you just forever going to repeat yourself?

There is nothing wrong with the experiment showing Faradays Law in action. There is also nothing wrong with the probing or the instruments. That is undeniable. But if you then incorrectly translate that into a circuit diagram and blindly apply KVL, making wrong assumptions, then of course you will not get the correct result.

Do you understand that a circuit diagram assumes lumped elements? So where's the element that corresponds to the "coil", the area the magnetic flux permeates? At the very least you'd have to model that with an ideal current source that represents the current induced by the magnetic flux. But lo and behold, the internal resistance of an ideal current source in infinite! Does Lewins' "A1=A2" and "D1=D2" still hold?

[bsfeechannel]

That's precisely your, Mehdi's, Dave's and everybody-else-that-insist-that-KVL-always-hold's mistake.

KVL says that the voltages measured along the path of a mesh add up to zero.

But Lewin showed at least one circuit where the voltages do not add up to zero.

Therefore KVL doesn't always hold. Period.

There's nothing you can do to remedy this. It is not about Lewin. It is the way nature works. Go to your lab, set up the same experiment, measure the voltages around the circuit and see for yourself. We are glad that those who repeated the experiment were precisely the guys who thought that Lewin had somehow cheated or blundered the experiment.

They're all dead inside now. All of them.

You cannot solve such circuit using KVL. It is impossible. You'll have to resort to the full monty and use Maxwell's equations at least to calculate the EMF produced by the varying magnetic field.

After that, you can of course devise an equivalent circuit where instead of a varying magnetic field producing the extra EMF upon the entire circuit, you have a battery, a generator, a transformer, or any other equivalent lumped (i.e. localized) component to produce the same EMF and get the exact same effect on the other components. In that case, you can solve the equivalent circuit using KVL because you theoretically removed the varying field from the circuit and stashed it away in the equivalent component.

But that is just a theoretical trick that has a lot of caveats.

One of them is that you will not find this extra EMF on the circuit. Not even if you fart your way through the Bohemian Rapsody.

So how's that possible? How can we have, so to speak, a spooky "component" that produces an EMF in a circuit, but is not present there? This seems to violate the principle of conservation of energy, doesn't it? Those are legit questions. But to answer them you need to abandon KVL, which these people are not prepared to accept. To reconcile their cognitive bias with the real phenomenon that contradicts it, they create all the irrational arguments I listed above and more.

[jesuscf]

You better go tell the whole electric power industry that they have been analyzing circuits under the effects of magnetic fields incorrectly for the last 160 years!!!

I don't need to. The whole power electric industry never made the mistake of analyzing a circuit under a varying magnetic field using KVL. In fact the whole power electric industry started out because of Faraday's discoveries which was solely explained by Maxwell's equations. And that's what we use to this very day.

Right there is proof that you have no idea of what you are talking about!  If (and that is a big IF) you are an electrical/electronics engineer, you are embarrassing yourself.  If your background is in physics, then I know were you are coming from, and your ignorance is understandable.

[thinkfat]

That's precisely your, Mehdi's, Dave's and everybody-else-that-insist-that-KVL-always-hold's mistake.

KVL says that the voltages measured along the path of a mesh add up to zero.

But Lewin showed at least one circuit where the voltages do not add up to zero.

Only because he omitted a critical element of the circuit. His diagram was simply incomplete: he drew a line in the diagram and then promptly neglected that it's a real, physical wire and not an ideal "net" you would find in a schematic.

You cannot solve such circuit using KVL. It is impossible. You'll have to resort to the full monty and use Maxwell's equations at least to calculate the EMF produced by the varying magnetic field.

After that, you can of course devise an equivalent circuit where instead of a varying magnetic field producing the extra EMF upon the entire circuit, you have a battery, a generator, a transformer, or any other equivalent lumped (i.e. localized) component to produce the same EMF and get the exact same effect on the other components. In that case, you can solve the equivalent circuit using KVL because you theoretically removed the varying field from the circuit and stashed it away in the equivalent component.

But that is just a theoretical trick that has a lot of caveats.

And that is exactly the error that Dr. Lewin made: He drew up a circuit, but it was not the equivalent of his experiment and then tried to solve that with KVL, and of course it failed.
Solving an equivalent circuit is not a theoretical trick, it is exactly how network analysis works. How do you calculate a circuit that has a BJT in it? With an equivalent circuit using e.g. the Ebers-Moll transistor model.

So how's that possible? How can we have, so to speak, a spooky "component" that produces an EMF in a circuit, but is not present there? This seems to violate the principle of conservation of energy, doesn't it? Those are legit questions. But to answer them you need to abandon KVL, which these people are not prepared to accept. To reconcile their cognitive bias with the real phenomenon that contradicts it, they create all the irrational arguments I listed above and more.

That's exactly your problem. There cannot be a "spooky component" that is not present but causes an effect. You have to have it in your model, otherwise all calculations are nonsense. Yes, you cannot calculate current induced by magnetic flux with KVL, yes you need to solve the Maxwell-Faraday equation to calculate the EMF, but after you've done that, you cannot simply forget about it and omit if from your circuit and postulate: "current flows in this circuit but without anything that causes it".

[jesuscf]

So how's that possible? How can we have, so to speak, a spooky "component" that produces an EMF in a circuit, but is not present there? This seems to violate the principle of conservation of energy, doesn't it? Those are legit questions. But to answer them you need to abandon KVL, which these people are not prepared to accept. To reconcile their cognitive bias with the real phenomenon that contradicts it, they create all the irrational arguments I listed above and more.

That's exactly your problem. There cannot be a "spooky component" that is not present but causes an effect. You have to have it in your model, otherwise all calculations are nonsense. Yes, you cannot calculate current induced by magnetic flux with KVL, yes you need to solve the Maxwell-Faraday equation to calculate the EMF, but after you've done that, you cannot simply forget about it and omit if from your circuit and postulate: "current flows in this circuit but without anything that causes it".

bsfeechannel understanding of KVL is stuck in 1845.

[bsfeechannel]

Right there is proof that you have no idea of what you are talking about!  If (and that is a big IF) you are an electrical/electronics engineer, you are embarrassing yourself.  If your background is in physics, then I know were you are coming from, and your ignorance is understandable.

Ah the ad hominem argument. Let's add it to the list above.

KVL always holds because in the eyes of EEVBlog forum member jesuscf, EEVBlog member bsfeechannel is perhaps not an engineer, but probably a physicist.

[bsfeechannel]

bsfeechannel understanding of KVL is stuck in 1845.

Ooh! I didn't see that one.

KVL always holds because bsfeechannel understanding of KVL is stuck in 1845.

[jesuscf]

Right there is proof that you have no idea of what you are talking about!  If (and that is a big IF) you are an electrical/electronics engineer, you are embarrassing yourself.  If your background is in physics, then I know were you are coming from, and your ignorance is understandable.

Ah the ad hominem argument. Let's add it to the list above.

KVL always holds because in the eyes of EEVBlog forum member jesuscf, EEVBlog member bsfeechannel is perhaps not an engineer, but probably a physicist.

Hey bsfeechannel, I am honestly curious about your credentials.  Are you an electrical/electronics engineer with a degree from a  university (or similar) or not?

I'll need you to prove it by solving "problem #24" from your hero Dr. Lewin, but as an electrical/electronics engineer will do:

[bsfeechannel]

But Lewin showed at least one circuit where the voltages do not add up to zero.

Only because he omitted a critical element of the circuit.

He, and everyone who repeated the experiment, measured the voltages and they didn't add up to zero. That's how he showed it.

His diagram was simply incomplete: he drew a line in the diagram and then promptly neglected that it's a real, physical wire and not an ideal "net" you would find in a schematic.

That didn't change the outcome of the experiment. Neither on his bench, nor on the bench of everyone else's. I. e., the voltages didn't add up to zero.

And that is exactly the error that Dr. Lewin made: He drew up a circuit, but it was not the equivalent of his experiment and then tried to solve that with KVL, and of course it failed.

The circuit had just two resistors and nothing else. And his model predicted exactly what happened in practice. KVL failed on the board. And then failed on the bench.

Solving an equivalent circuit is not a theoretical trick, it is exactly how network analysis works. How do you calculate a circuit that has a BJT in it? With an equivalent circuit using e.g. the Ebers-Moll transistor model.

You do not expect to find two diodes and two current sources inside a transistor, do you?

That's exactly your problem. There cannot be a "spooky component" that is not present but causes an effect.

Exactly. So what is going on?

You have to have it in your model, otherwise all calculations are nonsense.

Perhaps, what you need to have in your model is not a component. It is something else.

Yes, you cannot calculate current induced by magnetic flux with KVL, yes you need to solve the Maxwell-Faraday equation to calculate the EMF,

Perfect.

but after you've done that, you cannot simply forget about it and omit if from your circuit and postulate: "current flows in this circuit but without anything that causes it".

Who said that?

[bsfeechannel]

Hey bsfeechannel, I am honestly curious about your credentials.

Thank you for your interest in my credentials.

Are you an electrical/electronics engineer with a degree from a  university (or similar) or not?

I think that's irrelevant for the present discussion.

I'll need you to prove it by solving "problem #24" from your hero Dr. Lewin, but as an electrical/electronics engineer will do:

Lewin is not my "hero". I don't subscribe to his channel, nor follow him anywhere on the social networks.

He just happened to show a physical phenomenon whose understanding is very important for electronics engineering.

[jesuscf]

You do not expect to find two diodes and two current sources inside a transistor, do you?

Actually you do! As a matter of fact you can use the BE or BC junction of a BJT as a diode as represented in the Ebers-Moll model!  As for the current sources, that is the whole point of BJTs!!! So I don't see what you are trying to imply here.  Oh, wait, I know what were you may be going: because Kirchhoff didn't know about the physics of semiconductors, KVL can not be applied when you have semiconductors in a circuit???  As ridiculous at it sounds it is the same argument you are using with Faraday's law and KVL after all.

[bdunham7]

The circuit had just two resistors and nothing else. And his model predicted exactly what happened in practice. KVL failed on the board. And then failed on the bench.

I really don't want to wade into this mess, but the glaring issue for me in the original was the fact that Dr. Lewin made a dramatic point of saying (and drawing) that the two voltage measuring devices were connected to the same points and that they would demonstrate the the voltage from one point to another was 'path dependent' by displaying different values.  As a practical matter, if there are two different voltages between the two points depending on which branch you follow, how do the oscilloscopes 'know' which branch they are measuring?  Is that determined by where they are physically placed?  Are they briefed beforehand?  If you want to debate, correct or wrangle about anything I've said, please answer this question first as I have no desire to argue this issue until that is cleared up.

And as a theoretical matter, the voltage between two points can never be 'path dependent', that's ridiculous.  Voltage is at its core an absolute value, we just typically use and measure relative values because, well, circuits and current.

Now as for KVL and magnetic fields, you can all have at it, but there's one thing I haven't seen mentioned, so someone point it out if I've missed it:  Change in flux through a loop causes EMF, EMF causes current to flow, that current then causes......counter EMF?  No?  Anyway, connecting two test instruments to different points on a wire with current flowing in it and in a changing magnetic field and then claiming they are connected to the 'same' point makes me not want to try to solve the problem.

And for whoever posted the 5-resistor video, maybe I've misunderstood, but I seem to see another glaring error regarding the supposed symmetry of the current between the L-lower and R-upper resistors regardless of the resistance.  Try 0R for the right upper one and see if that holds! 

So no need for a pissing match, if someone can explain where I'm wrong here I'll listen.   

[jesuscf]

Hey bsfeechannel, I am honestly curious about your credentials.

Thank you for your interest in my credentials.

Are you an electrical/electronics engineer with a degree from a  university (or similar) or not?

I think that's irrelevant for the present discussion.

I'll need you to prove it by solving "problem #24" from your hero Dr. Lewin, but as an electrical/electronics engineer will do:

Lewin is not my "hero". I don't subscribe to his channel, nor follow him anywhere on the social networks.

He just happened to show a physical phenomenon whose understanding is very important for electronics engineering.

Then, solve the f***ing circuit!!!  Show me that you know what you are talking about.  It really only takes a couple of minutes.

[jesuscf]

And for whoever posted the 5-resistor video, maybe I've misunderstood, but I seem to see another glaring error regarding the supposed symmetry of the current between the L-lower and R-upper resistors regardless of the resistance.  Try 0R for the right upper one and see if that holds! 

It is the way Lewin solves the circuit and the pomposity of "his" solution as he follows up with three more videos more that the circuit and explanation itself!  It really takes around 2 minutes to properly solve such a trivial circuit without any trickery whatsoever.  I'll post the solution here tomorrow if bsfeechannel doesn't do it first.

EDIT: the solution is attached.  A day has passed and bsfeechannel is still talking BS but it is not capable of solving a very simple circuit.

[bsfeechannel]

As a practical matter, if there are two different voltages between the two points depending on which branch you follow, how do the oscilloscopes 'know' which branch they are measuring?

Good question. For your answer see below.

Is that determined by where they are physically placed?

Yes. That's why we study vector calculus at any engineering graduation course before we study electromagnetism. Because to understand this bleep you need to think "fourth-dimentionally". I.e. you need to understand that electricity and magnetism are not phenomena confined to electronic components, and how this thing behave in space.

So, frame of references, relative positions, relative velocities, paths, rates of change. All of that counts.

Are they briefed beforehand?

Nope. That would break causality.

If you want to debate, correct or wrangle about anything I've said, please answer this question first as I have no desire to argue this issue until that is cleared up.

The oscilloscopes "know" what branch they are measuring because they form a loop with each resistor. If you pause Lewin's presentation on Youtube ( /watch?v=nGQbA2jwkWI ) at 41:54, you'll see that the scope on the right forms a loop with R2. In that loop, there's no varying magnetic field. So, all the voltages will add up to zero according to Faraday's law and, in this case, to KVL, which is nothing more than a special case of Faraday's law when you have no varying magnetic field inside the path of the circuit. So the voltage on the right scope will have to be exactly the voltage on R2.

The same thing is happening to the left scope and R1.

And as a theoretical matter, the voltage between two points can never be 'path dependent', that's ridiculous.  Voltage is at its core an absolute value, we just typically use and measure relative values because, well, circuits and current.

Voltage can be path dependent if you are dealing with a non conservative electric field (i.e. one generated by a varying magnetic field).

Now as for KVL and magnetic fields, you can all have at it, but there's one thing I haven't seen mentioned, so someone point it out if I've missed it:  Change in flux through a loop causes EMF, EMF causes current to flow, that current then causes......counter EMF?  No?

That current will produce a magnetic field. While this current, and the corresponding magnetic field are varying, they will produce a counter EMF. The amplitude  will depend on the intensity of this magnetic field and its rate of change. But if you conveniently choose resistors that will allow a relatively low current, this counter EMF will be very low and can be neglected.

Anyway, connecting two test instruments to different points on a wire with current flowing in it and in a changing magnetic field and then claiming they are connected to the 'same' point makes me not want to try to solve the problem.

Since the wires have very low resistance compared to the resistors, they can be considered practically dead shorts.

[thinkfat]

I just have to ask: do you fundamentally understand the Maxwell-Faraday Equation? I have my doubts. Otherwise you would not claim that the circuit consists of only two resistors connected by wires that can be seen as "dead shorts" and will have no voltage across them.

Do you agree that the Maxwell-Faraday equation relates a time varying magnetic flux to an electric field? Do you agree that this changing flux _causes_ an electric field in the wire? Do you agree that an electric field is a difference in potential? Do you agree that a potential difference is measurable as a voltage between the wire ends? This is very much how a transformer works.

If you agree on all this, then you should be able to see that Dr. Lewins circuit is not just two resistors. It is in fact a transformer (or maybe generator) with the single secondary winding cut open in two places to insert resistors. And it is the voltage across those two wire that you need to take into account when you add up all the voltages in the loop.