### Author Topic: Does Kirchhoff's Law Hold? Disagreeing with a Master  (Read 52015 times)

0 Members and 3 Guests are viewing this topic.

#### bsfeechannel

• Frequent Contributor
• Posts: 611
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1025 on: April 25, 2019, 01:29:25 pm »

Quote
And I always thought, the EMF is defined as the tangential force per unit charge in the wire integrated over length, once around the complete circuit/loop! But not at all, it was hiding in the resistor and my voltmeter!

You clearly thought wrong, I regret. EMFs will never "hide" inside (static) wires (considered as ideal, that is).
Sorry, my wording was wrong, I should have said ‘I always believed, the EMF...’.!
Because the clearly wrong ‘thought’ is from Feynman Volume II chapter 16 Induced Currents.
As you say, electromag is so unintuitive, so much bullshit around.... you can’t even trust books anymore.

Thank you for bringing that up. This is the source of much confusion.

In chapter 16, Feynman is talking about REAL wires, that admit the presence of electric fields inside them. Lewin does the same in his famous Lecture 16 and also generates the same kind of confusion. In the middle of the lecture he talks about EMFs in the path of the wires.

Real wires behave just like a resistor. In fact they ARE resistors.

At the end of his lecture, Lewin uses real wires, but his setup allows the approximation to ideal wires because the resistance of the wires are much less than the resistors he uses (100 and 900 ohms). Since he doesn't make that explicitly clear, there are people up to this day looking for the missing EMF in the wires.

In chapter 22, Feynman treat the wires as ideal wires, that he calls "perfect conductor". The electric field inside those wires are ideally (duh!) zero. And that's the kind of wires we've been using in this discussion, unless noted differently.

So, to make it clear, I'll repeat: EMFs will never "hide" inside (static) wires (considered as ideal, that is).

#### bsfeechannel

• Frequent Contributor
• Posts: 611
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1026 on: April 25, 2019, 01:45:56 pm »
This phenomena is well known in literature as the ‘One-Wire-Single Current-Sophism”.

Sophism or not there is only one wire there. That's all that counts for KCL.

Quote
For sure, can only be solved with Maxwell equations

And I did it in from of your eyes: ∇·J = - ∂ρ/∂t

Quote
Maybe you missed the point, the ‘connection point’ (Berührungspunkt)?
Hence, a single wire hitting the connection point is a dead end, hence I1 = 0 .

Nope. If I is zero, then charge conservation is violated. The volume inside the dashed line is having a variation in the amount of charge, and there's no current crossing the border to account for this variation. It is like a bank account having its balance increasing or decreasing without money being deposited or withdrawn.

#### bsfeechannel

• Frequent Contributor
• Posts: 611
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1027 on: April 25, 2019, 03:02:19 pm »
Yes exactly my point that a voltage source is not an impedance.

In that chapter 22 that you mention it clearly shows that induced voltage acts like a voltage source. So including the induced voltage when calculating impedance of a component is just as nonsense as directly calculating the impedance of a voltage source using the voltage across its terminals.

If the voltage source is NOT an impedance there's no sense in talking about direct or indirect calculation of its "impedance" or the inclusion or not of said "induced voltage". It is not an impedance, period.

Quote
2 terminal components can't have multiple impedance at the same moment in time.

For that to happen you have to comply to at least the following conditions: you can't have external varying fields; the field generated by the component has to be confined and away from its terminals.

Quote
is there any reason why you replaced the world field with underscores?

Because I wanted to check if you really understand that the answers to explain how circuits work are not always found in the path of circuits themselves.

Quote
Its not exactly a secret that magnetic fields are what makes inductors work.

Well, fields are what makes everything work in electronics, even when you are not thinking about them.

Quote
Exactly it doesn't behave like an inductor anymore, but instead behaves like a coupled inductor.

Coupled inductors cannot be considered 2 terminal components anymore.

Quote
But it doesn't mean its inductance simply disappeared. The same magnetizing current is still needed to hold up the field, but because the two inductors are sharing the same flux means they can also share the magnetizing current. Since the resistor on the secondary can't provide a source of reactive current means that all the magnetizing current comes from the voltage source on the primary. The resistive load does however cause a in phase current that makes a opposing field in the secondary that affects the field in the primary, where it induces voltage according to Faradays law. Since there is a voltage source forcing a well defined voltage across the primary, this instead draws extra current to correct this field. This extra  in phase current is what drags the total current back away from lagging 90 degrees. As for the secondary, it has no need for out of phase current, because all of its magnetizing current is provided by the primary.

So the inductance didn't go anywhere. The effect of inductance is simply buried under the other currents.

The secondary is a generator, not an inductor. Although what is going on between the wires is the phenomenon of induction, for the circuit connected to the terminals of those wires they are not. Always remember that lumped components are only seen at their terminals. So if a piece of wire looks like a generator and smells like a generator at its terminals it is because it is a generator. The primary acts as an impedance that can change depending on the load connected to the secondary.

Quote
So yeah, sorry i still don't see the exact circumstance when a piece of wire stops being an inductor. Or does becoming a coupled inductor not count as being inductive? If so please explain why.

I think it is clear to you now that a piece of wire can be many things, an inductor, a generator, or whatever impedance depending on whether or not it's under a varying magnetic field.

If you consider it always an inductor, this will lead you to try to write things like L(di/dt) - R*I = 0, when you should have written R*I = -d/dt (∬ΣB·dS).

#### Berni

• Super Contributor
• Posts: 2141
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1028 on: April 25, 2019, 04:02:46 pm »
So why do we still use Maxwells equations if they are wrong according to Quantum electrodynamics?  Its much like the reason why we still use Kirchhoffs circuit laws despite being wrong according to Maxwell.

Kirchhoff's laws are not wrong according to Maxwell. Maxwell's equations do not only confirm that KVL and KCL are right, but also state in WHAT CONDITION they are right.

The source of the errors is elsewhere.

Yes the source of the error is people like Dr. Lewin applying it directly to a circuit without consideration if its applicable in those CONDITIONS. Fix those conditions by properly modeling the thing as a circuit mesh, if not then forget about Kirchhoffs circuit laws and stick to Maxwell.

Quote
On a macroscopic level and under certain conditions these laws still work just fine, while being much more convenient to work with when you want to actually calculate them with actual numbers. The 3 sets of laws are basically just different abstraction layers for electricity. Just chose the desired abstraction level and be aware of its known limitations.

A cat is a special case of mammal, and the description of a mammal is an abstraction of a cat (and every other mammal, as a matter of fact).

QED is an abstraction of Maxwell's equations, as Maxwell's equations are abstractions of KVL/KCL.

KVL/KCL are a special case of Maxwell's equations, as Maxwell's equations are special cases of QED.

KVL/KCL are NOT abstractions of Maxwell's equations by the simple fact that they don't work for all the cases for which Maxwell's equations do.

This is another illusion. Because we ram KVL/KCL down the throats of our circuits and close our eyes to the errors due to this approximation (and we manage to get away with it in many cases) we have the impression that KVL/KCL are just simplified (hence abstracted) cases of Maxwell's equations. But this mindset will bite you in the butt sooner or later. You need to be always aware of the limitations of KCL/KVL and, if the error due to approximating Maxwell to Kirchhoff is gross enough, resort to what will give you the most accurate prediction.

Okay yes i was using more the software engineering definition of abstraction layers rather than the mathematics definition of abstraction. With the mathematics definition it is indeed the other way around.

And i fully agree, all of the levels of explaining electricity work fine as long as you use them within those limitations. Hence why i never really had any problems applying KVL to Dr. Lewins experimental circuit from his lecture. Any paradox between KVL and Maxwell with that circuit is simply down to using them wrong.

Quote
Tho to be honest there is very very little reason to go any deeper than Maxwells level of abstraction for engineering use.
I think the inventors of the transistor would not agree with you, but anyway...

I'm pretty sure QED had nothing to do with the development of the first transistors.

There are patents for field effect transistors from before QED was a thing, tho nobody tried to make them at that point. Then the first working BJT device popped up by accident from a bunch of people trying to build a better solid state RF mixer.

The usual transistors can still be explained rather well with Maxwells fields pushing charged particles around. Mostly just a case of controlling where the charge carriers are rather than making use of any quantum mechanical effect. Tho the worlds finest lithography processes used for digital chips are now getting to making transistors small enough for quantum effects to start becoming a problem.

I'm pretty sure very few forum members here work in a semiconductor fab, let alone one that works with such fine feature capabilities or work on building quantum computers. Hence why very very few engineers would have a good reason to dig deeper than Maxwell, heck for 95% of cases even Kirchhoff is close enough(As long as you know about the other 5%).

#### ogden

• Super Contributor
• Posts: 2043
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1029 on: April 25, 2019, 05:32:54 pm »
In case you insist that there is just one wire and just one current meaning no path for return current - then I say that you don't even have circuit, thus Kirchoff's Circuit Law do not apply. There is huge difference between "do not apply" and "do not hold".

Nope. Kirchhoff says clearly that the wires can be connected in an entirely arbitrary way. So there is no requirement for them to form a circuit or to provide a path for a return current.

Exactly. Kirchhoff says "wires", thus more than single wire you desperately insist on. When you have single wire - you do not have conditions to apply KCL. That's why I looked for return current which I BTW found in the "spatially distributed capacitor", yet you managed to shift goalposts and change your mantra into talk about dotted area, not whole circuit. It is like showing one body of mass to happily conclude that Newton's law of universal gravitation do not work. It's not even unscientific. It's utterly stupid.
« Last Edit: April 25, 2019, 05:49:55 pm by ogden »

The following users thanked this post: Berni

#### bsfeechannel

• Frequent Contributor
• Posts: 611
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1030 on: April 26, 2019, 01:26:27 am »
Yes the source of the error is people like Dr. Lewin applying it directly to a circuit without consideration if its applicable in those CONDITIONS.

Nope. Lewin was absolutely impeccable in the "application" of the laws. He used Kirchhoff to show that it works when you do not have varying magnetic fields.

He substituted the varying magnetic field for the battery and then showed how wrong it is to apply Kirchhoff to a circuit with varying magnetic fields. You will measure different voltages, because they are now path-dependent.

And left the problem for his students to solve. The solution is Maxwell and only Maxwell.

Quote
Fix those conditions by properly modeling the thing as a circuit mesh,

We've already proved in this thread that Lewin's circuit can't be modeled by lumped components.

Quote
if not then forget about Kirchhoffs circuit laws and stick to Maxwell.

The problem with your reasoning is that you think that if things are connected like a circuit, then Kirchhoff MUST hold. You think that if you find a single circuit where it can't be employed to explain its behavior then the whole theory is cactus.

This is BULLSHIT.

There are circuits for which Kirchhoff holds and others for which it doesn't. By now you should have known the difference. And the difference is that if the circuit is drenched with a varying magnetic field, KVL is out. If not, then you can merrily use your KVL and your precious LTSPICE to model it.

Lewin's circuit, if you have not paid the due attention to it yet, is just a sophisticated version of Faraday's original demonstration of induction. Lewin is just reenacting the same experiment with two fancy oscilloscopes and two resistors of relative high value. But the experiment is essentially the same.

Saying Lewin is wrong is the same as denying Faraday's law.

Quote
Okay yes i was using more the software engineering definition of abstraction layers rather than the mathematics definition of abstraction. With the mathematics definition it is indeed the other way around.

Abstraction in software is commonly thought of from the point of view of the user, but it conforms with the general definition of abstraction in which the "abstracted" software is in fact more inclusive than the special case.

Quote
And i fully agree, all of the levels of explaining electricity work fine as long as you use them within those limitations. Hence why i never really had any problems applying KVL to Dr. Lewins experimental circuit from his lecture. Any paradox between KVL and Maxwell with that circuit is simply down to using them wrong.

We've already showed that your "modelling" just don't apply to Lewin's circuit. It introduces gross errors. If I connect the lead of MY voltmeter (not yours with 250mV in series with the probes) from one resistor to the other where a wire should be in your "model" I will measure 0.5V, whereas in Lewin's circuit this voltage is zero, and everyone measured exactly zero volts.

There can't be any paradox between KVL and Maxwell. Lewin changed the condition for the validity of KVL and invited his students to explain the new behavior of that circuit. Those that think that Kirchhoff applies to just about any circuit are trying to find each one a different explanation. While those who understand Faraday's law could explain what is going on immediately.

Quote
I'm pretty sure QED had nothing to do with the development of the first transistors.
Quote
Then the first working BJT device popped up by ]accident from a bunch of people trying to build a better solid state RF mixer.

Wow! Shockley, Bardeen and Bratain won the Nobel Prize by accident! Lucky bastards!

Quote
The usual transistors can still be explained rather well with Maxwells fields pushing charged particles around.

Isn't this what essentially QED is: electrodynamics applied to subatomic particles like the electron?

Quote
Mostly just a case of controlling where the charge carriers are rather than making use of any quantum mechanical effect.

Strange. I had the impression that electrons received and emitted photons when they change their quantic levels of energy in the electronic band structure.

Quote
I'm pretty sure very few forum members here work in a semiconductor fab, let alone one that works with such fine feature capabilities or work on building quantum computers. Hence why very very few engineers would have a good reason to dig deeper than Maxwell, heck for 95% of cases even Kirchhoff is close enough(As long as you know about the other 5%).

I also had the impression that every electronics engineer around the world learns at least some rudimentary concepts of QED as part of their regular course which is required to understand solid state electronics. But I may be wrong.

#### bsfeechannel

• Frequent Contributor
• Posts: 611
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1031 on: April 26, 2019, 01:53:51 am »

Exactly. Kirchhoff says "wires", thus more than single wire you desperately insist on. When you have single wire - you do not have conditions to apply KCL.

Alright. I created an ogden version of KCL fail for you. I hope it now can make its way into your everything-must be-a-circuit head.

Quote
That's why I looked for return current which I BTW found in the "spatially distributed capacitor",

Cool! Can you connect an ammeter and measure this current flowing through a... wire?
« Last Edit: April 26, 2019, 01:55:53 am by bsfeechannel »

#### ogden

• Super Contributor
• Posts: 2043
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1032 on: April 26, 2019, 02:56:19 am »
Alright. I created an ogden version of KCL fail for you.

Brilliant. Exactly what I was looking for. Could you sign it and put high resolution file on google drive? - So I can show what stupid ingenuity looks like.
« Last Edit: May 19, 2019, 09:39:33 am by ogden »

The following users thanked this post: seagreh

#### Berni

• Super Contributor
• Posts: 2141
• Country:
##### Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #1033 on: April 26, 2019, 03:48:22 am »
Quote
if not then forget about Kirchhoffs circuit laws and stick to Maxwell.
The problem with your reasoning is that you think that if things are connected like a circuit, then Kirchhoff MUST hold. You think that if you find a single circuit where it can't be employed to explain its behavior then the whole theory is cactus.
This is BULLSHIT.
There are circuits for which Kirchhoff holds and others for which it doesn't. By now you should have known the difference. And the difference is that if the circuit is drenched with a varying magnetic field, KVL is out. If not, then you can merrily use your KVL and your precious LTSPICE to model it.
Lewin's circuit, if you have not paid the due attention to it yet, is just a sophisticated version of Faraday's original demonstration of induction. Lewin is just reenacting the same experiment with two fancy oscilloscopes and two resistors of relative high value. But the experiment is essentially the same.
Saying Lewin is wrong is the same as denying Faraday's law.

No i said that KVL holds in every circuit mesh model. Not every physical circuit, this is an important distinction so don't carelessly throw them in the same bag.

I'm not saying Lewin is wrong. Kirchoffs circuit laws indeed don't work in such a use case. All i did was show a method of using KVL in a way that does work in that circuit, useful in cases where you would want to apply other circuit analysis tools.

Quote
And i fully agree, all of the levels of explaining electricity work fine as long as you use them within those limitations. Hence why i never really had any problems applying KVL to Dr. Lewins experimental circuit from his lecture. Any paradox between KVL and Maxwell with that circuit is simply down to using them wrong.
We've already showed that your "modelling" just don't apply to Lewin's circuit. It introduces gross errors. If I connect the lead of MY voltmeter (not yours with 250mV in series with the probes) from one resistor to the other where a wire should be in your "model" I will measure 0.5V, whereas in Lewin's circuit this voltage is zero, and everyone measured exactly zero volts.
There can't be any paradox between KVL and Maxwell. Lewin changed the condition for the validity of KVL and invited his students to explain the new behavior of that circuit. Those that think that Kirchhoff applies to just about any circuit are trying to find each one a different explanation. While those who understand Faraday's law could explain what is going on immediately.

I did say there is no paradox in my own post so why are you arguing for the same thing.

Oh and if you do have a voltmeter that can integrate the work needed to move an electron along any chosen path id love to see a write up on its operation. Would give you a pretty good chance at a Nobel prize even.

If you are so good at it how about showing me how to correctly analyze this simple circuit:
https://www.eevblog.com/forum/chat/does-kirchhoffs-law-hold-disagreeing-with-a-master/msg2189216/#msg2189216

Feel free to use any method you want, as long as it shows what the circuit will do once given power.

...
Wow! Shockley, Bardeen and Bratain won the Nobel Prize by accident! Lucky bastards!
...
Isn't this what essentially QED is: electrodynamics applied to subatomic particles like the electron?
...
Strange. I had the impression that electrons received and emitted photons when they change their quantic levels of energy in the electronic band structure.
...
I also had the impression that every electronics engineer around the world learns at least some rudimentary concepts of QED as part of their regular course which is required to understand solid state electronics. But I may be wrong.

The inventors of the transistor certainly did amazing work in the field of semiconductors, but just saying that making a transistor was not what they ware trying to do when they made one. Its not the only major discovery that had a little bit of luck in it.

I was certainly ever shown any QED in lectures about semiconductors, tho to be honest those ware pretty dull lectures so i mostly just memorized enough stuff to pass the test, rather than show much interest. Its all mostly just electrostatic fields with a bit of electron physics thrown in. Nice to know about, but not terribly useful to know in deep detail. Much like teaching software engineers some assembler, not really practical for the majority of cases but good to know the basics.

Those electrons do certainly interact trough photons according to QED, but the overall behavior is more sensible to explain with Maxwell since it still works just fine in there.

The following users thanked this post: ogden

Smf