Author Topic: Does Kirchhoff's Law Hold? Disagreeing with a Master  (Read 63164 times)

0 Members and 1 Guest are viewing this topic.

Offline Berni

  • Super Contributor
  • ***
  • Posts: 2473
  • Country: si
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #300 on: December 01, 2018, 10:41:47 am »
Oh and i was to add that with the using schematic circuits to calculate heat flow trough a wall, water flow trough a pipe or magnetic flow trough a core is mostly something that electrical engineers do when they have to deal with something non electrical. Its sort of just swapping out the abstraction layer underneath so they can keep using familiar analysis methods. Im pretty sure other kinds of engineers don't "misuse" schematics for this.

But this trick does show how in nature everything appears to have two intertwined quantities:
-Electricity: Volt and Amps
-Mechanics: Force and Distance
-Rotational mechanics: Torue and RPM
-Fluids: Pressure and Flowrate
-Magnetics: Magnetic density(H) and Magnetic flux(B)
-Thermal: Temperature and Thermal flux
etc...

In all of these cases if you multiply these two intertwined quantities you get power(P) in Watts. Ohms law works for all of these(Tho units are not in ohms anymore) and each of these has a inductor/capacitor equivalent that provides time dependence.

All of these act the same, just use different units because they are different physical things. Because of this cirucit mesh analysis also works on all of these as long as you know how to model it as a schematic. Once its a schematic you can apply KVL and KCL to it just fine and it still works fine, just like all the other circuit analysis tools such as Thenevnins and Nortons theorem. Circuit schematics ware never designed to do this but it just happens to work due to how nature works.

I found this realization of intertwined units quite eye opening back in university. It can help you understand how non electrical things work a lot faster.
« Last Edit: December 01, 2018, 10:43:44 am by Berni »
 

Offline bsfeechannel

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: 00
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #301 on: December 01, 2018, 03:48:59 pm »
And this is exactly why KVL is an abstraction of Maxwells equations. It makes things easier while not influencing the result (when used correctly).

Because in many of your "abstractions" you used KVL instead of Maxwell, you think that KVL is an abstraction of Maxwell. Don't say that anymore.

Saying that means that every time you have a problem solvable by Maxwell, you can immediately apply KVL "used correctly" and that's it.

It implies that you can get away without understanding the underlying phenomena. No, you can't.

That's what Cyriel Mabilde disastrously did. That's what Lewin is desperately trying to warn you about.

Quote
Don't worry i seen plenty of programmers that don't understand how a computer works and consider C being too low level. Some even stick to purely interpreted languages like javascript, php or python. I do think they should at least conceptually understand the inner workings of computers. Large software projects are often massive cobbled together messes with chunks of code that nobody understands why they are needed and how they work but if you touch them things break horribly so everyone stays away from them and works around them. Tight deadlines encouraging to just cobble on more rushed code until the project becomes unmaintainable. But that's a topic for another day.

Why you can get away with that in computing? Because those languages, including machine code, are all equivalent in computing power. So you can get a program working without knowing the details under the hood at the expense of a messy code.

Kirchhoff and Maxwell are NOT equivalent. If Kirchhoff and Maxwell were languages, Kirchhoff would describe a machine with LESS computing power than Maxwell.

Quote
Actually i was working on some phased array stuff recently and i used even more abstraction than just Kirchhoff.

I was trying to calculate the directionality of a phased array and i wanted to do it fast so that i could later use iterative optimization methods on it. I could have simulated each point in space and its interaction to the neighboring points so that i get a simulation of wave propagation trough the medium. Takes a lot of computation to do and im too lazy to program all of that. Well... instead i just did trigonometry to find the distance to each element and pretended there is an ideal delay line with its delay proportional to the distance. Worked great and it would spit out a directionality graph for all possible angles in the blink of an eye(Without even trying to optimize the code). Is this magical delay line what is actually happening on a phased array? Hell no. Does it act similar enough? For my use case plenty enough.

When the abstraction works i will definitely say "Bugger that!" to the more complicated alternatives. If the abstraction doesn't work then i will go down the long path. I don't work in an academic institution to needlessly waste time obsessing about the underlying mechanisms for cases where they are simply not important.

Again. Who told you that you could reduce the problem to KVL? Kirchhoff? Certainly not. Kirchhoff is a bird. He doesn't know anything about propagation, delay lines, fields, etc.

You had to use MAXWELL, and you did it almost unconsciously, to reduce to problem to Kirchhoff. So you confirm what I said in one of my first messages on this thread about how we engineers are so used to that practice that we forget that we are in fact using Maxwell and implicitly reducing the problem to Kirchhoff.

You can abstract, but you cannot use this as an excuse to ditch the fundamentals. What people are doing is not even trying to study Maxwell, consequently not understanding what the flux is going on and criticizing Lewin for THEIR ignorance.

This is the most stupid educational move I've seen in decades.

Quote
Well i was introduced to Kirchhoffs laws before high school. Then actually had to use them in high school to manually solve circuit meshes. Inside the circuit mesh abstraction its really easy to use.

...[snip]

I found this realization of intertwined units quite eye opening back in university. It can help you understand how non electrical things work a lot faster.

TLDR.

Now that I've made you a convert, let's help others to avoid saying stupid things like "Physics has no use for Kirchhoffs law since it doesn't deal with anything physical."

I was almost using that quote as a signature. But then I decided to give you a second chance.
« Last Edit: December 01, 2018, 04:28:05 pm by bsfeechannel »
 

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 1937
  • Country: au
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #302 on: December 01, 2018, 10:17:13 pm »
The thing is.

Berni, Electroboom and Cyriel Mabilde know how to measure the voltage in the demonstration circuit.
They all demonstrated it properly.

Clearly bsfeechannel and Dr Lewin don't know how to measure the voltage.

The moment you put voltmeter on as Dr Lewin did, you are measuring a scalar voltage. He did that.
But unfortunately he did it incorrectly and drew the wrong conclusion.

Maxwell-Faraday is a model the same as any other. It has limitations, and can be used successfully if it fits within them.
It is not magic. It isn't even particularly complex.

btw bsfeechannel please update your venn diagram to fix your mistake, people will get the wrong idea.
 

Offline bsfeechannel

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: 00
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #303 on: December 02, 2018, 01:18:00 am »

Berni, Electroboom and Cyriel Mabilde know how to measure the voltage in the demonstration circuit.
They all demonstrated it properly.

They have no idea what they're measuring.

Quote
Clearly bsfeechannel and Dr Lewin don't know how to measure the voltage.

I have a series of videos where I design and build an isolation transformer. The first video starts with an explanation why a transformer works. It obviously employs Maxwell/Faraday, because transformers are a direct result of Faraday's research. After I understand what to expect, I reduce the problem to Kirchhoff and then I design, build, test and characterize it.

It was a very successful project. The transformer met all the specs and works a treat.



Quote
The moment you put voltmeter on as Dr Lewin did, you are measuring a scalar voltage. He did that.
But unfortunately he did it incorrectly and drew the wrong conclusion.

Isn't it annoying when Nature does not conform to our preconceived notions of reality?

Quote
Maxwell-Faraday is a model the same as any other. It has limitations, and can be used successfully if it fits within them.
It is not magic. It isn't even particularly complex.

btw bsfeechannel please update your venn diagram to fix your mistake, people will get the wrong idea.

As I said, I don't need to upgrade anything for circuit analysis. There is no mistake. The theory is sound. When you study it, as I, Lewin, and many others on this forum did, you'll understand it as clear as crystal. If you need any assistance in upgrading your knowledge, we are here to help.
 

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 1937
  • Country: au
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #304 on: December 02, 2018, 01:54:55 am »
Your venn diagram was wrong because you had Maxwell covering in all other cases.
It's an important point because it glosses over the limitations of modelling natural phenomena.

Dr Lewin seemingly had no idea what he was measuring, I make this judgement because he said it would blow their minds.
It certainly confused people but didn't blow minds.

Quote
Isn't it annoying when Nature does not conform to our preconceived notions of reality?
I guess it would be.
If Berni, Electroboom and Cyriel Mabilde got it wrong then let us know what you think voltages are on the loop in Dr Lewins experiment.
 

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 1937
  • Country: au
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #305 on: December 02, 2018, 02:02:29 am »
But on a brighter note we seem to agree that KVL is at lease useful for characterising transformers.

ps. I would have liked to have watched your video further but the robot voice was hurting me.
Maybe someone on EEBblog could do an English (or other language) voice over?
 

Offline bsfeechannel

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: 00
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #306 on: December 02, 2018, 02:50:22 am »
But on a brighter note we seem to agree that KVL is at lease useful for characterising transformers.

ps. I would have liked to have watched your video further but the robot voice was hurting me.
Maybe someone on EEBblog could do an English (or other language) voice over?

I wrongly thought that since my videos are directed to a technical audience, my viewers would not be bothered by a computer-generated voice. If someone, especially a native speaker of English, could do the voice-over that would be wonderful. It would be a voluntary work, though, since my channel is not monetized. I was mustering the courage to do it myself. But I am not sure how the viewers would react.

Anyway, thanks for watching.
 

Online GeorgeOfTheJungle

  • Super Contributor
  • ***
  • Posts: 1967
  • Country: pl
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #307 on: December 02, 2018, 09:36:57 am »
Hey, bsfeechannel, where do you come from?
Even when the experts all agree, they may well be mistaken.
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 2473
  • Country: si
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #308 on: December 02, 2018, 10:25:01 am »
And this is exactly why KVL is an abstraction of Maxwells equations. It makes things easier while not influencing the result (when used correctly).

Because in many of your "abstractions" you used KVL instead of Maxwell, you think that KVL is an abstraction of Maxwell. Don't say that anymore.

Saying that means that every time you have a problem solvable by Maxwell, you can immediately apply KVL "used correctly" and that's it.

It implies that you can get away without understanding the underlying phenomena. No, you can't.

That's what Cyriel Mabilde disastrously did. That's what Lewin is desperately trying to warn you about.

That's because abstractions also have there limits. So do Maxwells equations when you get down to really small scales where quantum effects take over. You have to know about those limits and simply make sure you don't use that particular abstraction when outside of them.




Why you can get away with that in computing? Because those languages, including machine code, are all equivalent in computing power. So you can get a program working without knowing the details under the hood at the expense of a messy code.

Kirchhoff and Maxwell are NOT equivalent. If Kirchhoff and Maxwell were languages, Kirchhoff would describe a machine with LESS computing power than Maxwell.

In theory yes any truing complete language can do anything.

In practice the capabilities of programing languages vary a lot. Some languages are simply faster to compute a given task, no matter how well you optimize your program (While others complete it with less program code). But there is also certain low level functionality that many languages are simply not capable of. Modern CPUs often have special instructions that most compilers don't know how to use so they don't. There are special locked registers in CPUs that require a sequence of instructions that a lot of high level languages can't reproduce. These special things are usually not something that most programs need to do, but operating systems or hypervisors and such really need it, as they have to do things like set up the MMU, manage execution privilege levels, perform context switching, switch the CPU from 8086 compatibility mode into the full instruction set on boot etc... Normal programs running under a OS also make use of some special features such as JIT compiling where the whole program is basically self modifying code that compiles itself on the fly as it runs by jumping back into the compiler whenever needed. All of this is simply not possible in high level languages like python (Well apart from loading raw binary data into memory and then crashing the program in just the right way that the CPU ends up executing that area by mistake, but that's basically bypassing the language and using a hex editor to program)

This is much how Kirchhoff and Maxwell are NOT equivalent. Kirchhoff can do most things one would normally need to do, but not everything. The things it can do it usually does in a way that is move convenient than the alternative, for the things it can't do then you have no choice but to use the alternative.


Again. Who told you that you could reduce the problem to KVL? Kirchhoff? Certainly not. Kirchhoff is a bird. He doesn't know anything about propagation, delay lines, fields, etc.

You had to use MAXWELL, and you did it almost unconsciously, to reduce to problem to Kirchhoff. So you confirm what I said in one of my first messages on this thread about how we engineers are so used to that practice that we forget that we are in fact using Maxwell and implicitly reducing the problem to Kirchhoff.

You can abstract, but you cannot use this as an excuse to ditch the fundamentals. What people are doing is not even trying to study Maxwell, consequently not understanding what the flux is going on and criticizing Lewin for THEIR ignorance.

This is the most stupid educational move I've seen in decades.

Kirchhoff was not involved in in that phased array. I was making an example why wave propagation doesn't automatically make Maxwell necessary, or even make physics necessary.

I was making use of physics of wave propagation to abstract the problem down to just geometry. Waves traveling in a constant uniform medium always travel at the same speed, this leads to a conclusion that the time delay from the transmitter to the receiver is only a function of distance. With that i can craft simple factor to multiply with in order to translate distance into time delay. With this number in hand i can then predict the waveform this element will receive and feed that on into the phased array beam steering math. No physics involves what so ever, only geometry.

The results it gave matched up with other tools and with experimental results.

You only have to understand enough of the underlying physics to determine what abstraction is appropriate (if any). No need to calculate the whole thing using fundamental physics first. The understanding is more valuable than being able to blindly put numbers into famous equations. Sticking numbers into equations blindly without trying to understand what they are is ignorance. Applying understanding of the subject to form a simpler abstraction to make things easier and faster is instead called "getting stuff done".


TLDR.

Now that I've made you a convert, let's help others to avoid saying stupid things like "Physics has no use for Kirchhoffs law since it doesn't deal with anything physical."

I was almost using that quote as a signature. But then I decided to give you a second chance.

Made me convert to what?

Kirchhoffs cirucit laws still are not some fundamental law of the universe or something. Its just one of the sets of laws that make circuit meshes work. I have yet to see Kirchhoffs laws be wrong when they are used as intended. Its a great abstraction that helps you make sense of physical things.

You have demonstrated in the transformer video how useful the circuit mesh abstraction is. Transformers don't have additional winding that make leakage inductance, they don't have a physical resistor inside them that causes core losses. Yet it acts pretty much like that was the case so that's why the real transformer model uses it. It makes things much simpler to work with while acting close enough. You even use such simplifications before you get to the equivalent circuit model. For example you consider two turns in a coil as simply being 2 times a single turn, while showing segments of wire that go up diagonally to connect the two and a set of wires coming out and then showing a voltage across the two wires without closing them into a loop. I'm not saying its wrong to do this, it makes perfect sense to do it, but for these same reasons is why other people had issues with my lumped model of Dr. Lewins experimental circuit. For some reason i was not allowed to insert inductors into the equivalent model and not allowed to have a voltage on a non closed loop wire segment.

Absolutely nothing wrong in that video (Okay maybe apart from the voice)
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 2667
  • Country: lv
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #309 on: December 02, 2018, 11:10:28 am »
For some reason i was not allowed to insert inductors into the equivalent model and not allowed to have a voltage on a non closed loop wire segment.

It's privilege of physics [self-acclaimed] gurus only. As you did not worship Dr.Lewin - you are not worthy to use inductor models.

Quote
Absolutely nothing wrong in that video (Okay maybe apart from the voice)

Gloves :D

Hey, bsfeechannel, where do you come from?

My guess would be: Brazil?
« Last Edit: December 03, 2018, 03:11:15 pm by ogden »
 

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 1937
  • Country: au
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #310 on: December 03, 2018, 12:47:06 am »
Sorry you still don't get it.
The only thing the oscilloscope reads is the voltage at its terminals.
It displays a representation of that scalar voltage.
It doesn't show the voltage at at A and D even though the probes are connected to A and D.

If you want the oscilloscope to read the voltage at A and D you must not add extra flux into the loop.
Cyriel Mabilde demonstrated how to do this.
It is not magic.
It is measurable.
He even knows the value.

I think 15 year olds would understand this better than some older people seem to be able to do.
 

Offline bsfeechannel

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: 00
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #311 on: December 03, 2018, 02:26:58 am »
You have demonstrated in the transformer video how useful the circuit mesh abstraction is. Transformers don't have additional winding that make leakage inductance, they don't have a physical resistor inside them that causes core losses. Yet it acts pretty much like that was the case so that's why the real transformer model uses it. It makes things much simpler to work with while acting close enough.

I had to make a lot of implicit assumptions to REDUCE the model. That is stated in the video. These assumptions were only possible because Maxwell told me what is going on. Kirchhoff can't explain them in any way. For instance, where does the leakage inductance come from? What exactly produces it? How can it be modeled as a lumped component? Why do we have a magnetizing current on the primary even without a load on the secondary. Is that a "parasitic" current? Can we reduce it? If we can, how? How will this affect the other parameters?

While I think about that and derive the calculations to model the device, Kirchhoff sits in the corner like Jack Horner. I only call him to the party in the last minute, when all the components of my model are figured out.

So Kirchhoff is just a fancy interface in the end to hide a complicated mechanism. It serves no purpose in the design of the transformer.

Quote
You even use such simplifications before you get to the equivalent circuit model. For example you consider two turns in a coil as simply being 2 times a single turn, while showing segments of wire that go up diagonally to connect the two and a set of wires coming out and then showing a voltage across the two wires without closing them into a loop.

The arrow indicates where you are going to place your voltmeter, and then the loop will be closed. If the diagonal wire bothers you, just rotate the top loop clockwise a little until the connecting wire becomes perpendicular.

Quote

For some reason i was not allowed to insert inductors into the equivalent model and not allowed to have a voltage on a non closed loop wire segment.

Because my transformer and Lewin's experiment have a crucial difference. While the transformer has a fixed topology, Lewin's experiment doesn't. And Maxwell showed that topology is everything in electromagnetism.

If you move the voltmeter an isty bitsy tiny little femtometer, it will measure a different voltage. Your voltmeter may not be sensitive enough to catch very small variations, but they will be there. There's simply no right way to measure voltages on Lewin's experiment. It is undefined.

Since Kirchhoff knows nothing about how your circuit is arranged in space, if you want to reduce Lewin's experiment to Kirchhoff you have to define the topology, either explicitly, or implicitly like in the case of a transformer, but once you do that, it will not be Lewin's experiment anymore.

For more details, revisit this post:

https://www.eevblog.com/forum/chat/does-kirchhoffs-law-hold-disagreeing-with-a-master/msg1962869/#msg1962869

Quote
Absolutely nothing wrong in that video (Okay maybe apart from the voice)

Thanks.
 

Offline bsfeechannel

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: 00
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #312 on: December 03, 2018, 03:55:29 am »
If Berni, Electroboom and Cyriel Mabilde got it wrong then let us know what you think voltages are on the loop in Dr Lewins experiment.

After all these posts, you still don't get it.

We've been discussing this topic for a month now and, although apparently the vast majority of EEs on this forum understands what is at stake, there are still a few people who think that Mehdi is a hero, Lewin is a charlatan, I am a troll, and Maxwell is just a complicated version of Kirchhoff.

The problem seems to be that Lewin's "Kirchhoff is for the birds" rant is being perceived as an attack on those who learned to rely on Kirchhoff. But Lewin's rant is directed at some of those who TEACH circuit analysis.

What he is exposing is not new. Somewhere on the thread I recommended the book "Electromagnetic Waves and Radiating Systems" by Jordan and Balmain. Its first edition is from 1950. The very first section of the first chapter approaches the problem using very well chosen words and taking care not to offend susceptibilities.

But the issue is the same. Circuit analysis (a.k.a Kirchhoff) is being taught by some as an equivalent theory to Maxwell. Worse than that: it is being taught as a theory based on different postulates, as if Maxwell had nothing to do with Kirchhoff. Worse still: Maxwell is an extension of Kirchhoff for very limited cases where Kirchhoff, surprisingly, doesn't work. Even worse still: Kirchhoff is a series of tricks to solve circuits, it has nothing to do with physics. Maxwell has nothing to do with practical engineering: it is a thing of interest to physicists disconnected from the reality of practical engineering.

All of this is absolutely far from the truth. And the attempts at remedying it are timid.

We must admit that Lewin had a lot of balls to defy the educational establishment and maintain his integrity even after having retired from teaching.
 
The following users thanked this post: boB

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 1937
  • Country: au
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #313 on: December 03, 2018, 04:48:39 am »
Quote
you must not add extra flux into the loop
Quote
So, let me ask you again: what is 'extra flux'?

Not sure why you cant understand me.
I mean any non zero net flux that is intercepted by the act of probing the points A and D.
Watch Mabilde he knows how to probe a voltage correctly.
The reason why I call it a voltage and not an emf is because the oscilloscope can only probe voltages.

Quote
You do realize you are saying that the reading does not depends on the endpoints only but also on the path?
Yes
Quote
So, you are saying Lewin is right.
No
Quote
And yet, you still talk about "voltage on the loop" as if it were a property of the loop.
?? I talked about measuring the voltages at A and D  but only because I think that  is what Dr Lewin is trying to do.

Anyway I think I have said my piece.
Thanks for the discussion all.

ps . I like Dr Lewins lectures. I have only seen a few though.
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 2473
  • Country: si
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #314 on: December 03, 2018, 07:11:04 am »
Nobody here was trying to say that Maxwells equations or Faradays law are wrong or even just an extension of Kirchhoffs laws.

I don't have anything against Dr. Lewin and i think he makes some amazing lectures. Its just his explanation of Kirchhoffs law that i don't agree with. Not that i have some secret love relationship with Kirchhoff, but from my experience with the electronics engineering field i never saw this way of using Kirchhoffs voltage law to be the correct use for it, hence why it does not work in his case.



I had to make a lot of implicit assumptions to REDUCE the model. That is stated in the video. These assumptions were only possible because Maxwell told me what is going on. Kirchhoff can't explain them in any way. For instance, where does the leakage inductance come from? What exactly produces it? How can it be modeled as a lumped component? Why do we have a magnetizing current on the primary even without a load on the secondary. Is that a "parasitic" current? Can we reduce it? If we can, how? How will this affect the other parameters?

While I think about that and derive the calculations to model the device, Kirchhoff sits in the corner like Jack Horner. I only call him to the party in the last minute, when all the components of my model are figured out.

So Kirchhoff is just a fancy interface in the end to hide a complicated mechanism. It serves no purpose in the design of the transformer.

Leakage inductance are simply magnetic field lines that don't pass trough both coils (Including connecting wires to the coils). The field has to be simulated in space to find this inductance computationally since it heavily depends on the coil placement and core design. It behaves like extra uncoupled inductance so can be modeled as a inductor. Magnetizing current is determined by the core properties(or in an abstract way by inductance and frequency). It can be reduced by changing the number of turns or swapping out the core for a higher permeability one. Alternatively the reactive part of this current can be completely removed using a compensation capacitor while the real current can only be reduced using lower resistance coils and lower loss material in the core. It affects other parts of the transformer depending on what you do but in most cases makes the transformer larger and heavier or more expensive to produce(Ferrite core).

All of this is indeed all Maxwell. I never said Maxwell was useless. Just that its not always the best choice if there are simpler alternatives.

Never said that Kirchhoff could do everything. It also can't calculate how many miles per gallon my car can do. Its simply not what Kirchhoff is for. Its only job is to describe how currents and voltages work inside ideal circuit meshes. That's all it does, it doesn't do anything more than that.


The arrow indicates where you are going to place your voltmeter, and then the loop will be closed. If the diagonal wire bothers you, just rotate the top loop clockwise a little until the connecting wire becomes perpendicular.

Yes but the path you take when connecting your voltmeter matters. So the sensible way to connect it is in a way that makes no EMF on the wires, as i have did before. Nothing wrong with that method, yet it was somehow wrong when i used it.


Because my transformer and Lewin's experiment have a crucial difference. While the transformer has a fixed topology, Lewin's experiment doesn't. And Maxwell showed that topology is everything in electromagnetism.

If you move the voltmeter an isty bitsy tiny little femtometer, it will measure a different voltage. Your voltmeter may not be sensitive enough to catch very small variations, but they will be there. There's simply no right way to measure voltages on Lewin's experiment. It is undefined.

Since Kirchhoff knows nothing about how your circuit is arranged in space, if you want to reduce Lewin's experiment to Kirchhoff you have to define the topology, either explicitly, or implicitly like in the case of a transformer, but once you do that, it will not be Lewin's experiment anymore.

For more details, revisit this post:

https://www.eevblog.com/forum/chat/does-kirchhoffs-law-hold-disagreeing-with-a-master/msg1962869/#msg1962869

Lewins experiment had no moving parts, the wires to the resistor and wires to both oscilloscopes remained stationary throughout the experiment.

Maxwell also would need to know the exact femtometer accurate path the wire takes to make an accurate prediction. Such measurements ware not provided so i assumed the wires to run perfectly in the same path as does his math. In my case Maxwell was used to obtain the inductance of the wire segments. After all of it i had a circuit that acts within the margin of error like Dr. Lewins experiment.

I fully agree that there are two voltages over points A and B in Dr. Lewins circuit. The math does work out that way.

I just don't agree in his application of Kirchhoffs law to prove its wrong. It needs a lumped circuit mesh to work right and i was able to turn his circuit into a valid and accurately behaving circuit mesh in 5 minutes where it shows that Kirchhoff indeed works.

So let me ask you then what is the correct use of Kirchhoffs laws?
 

Offline Sredni

  • Regular Contributor
  • *
  • Posts: 186
  • Country: aq
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #315 on: December 03, 2018, 06:18:06 pm »
I just don't agree in his application of Kirchhoffs law to prove its wrong. It needs a lumped circuit mesh to work right

? If you are referring to the mesh shown in "Science and believing..." (or what is the name), you are mistaken.
It might look like a lumped circuit mesh, but it's not. In fact, the emf is not lumped anywhere.
This is why we keep saying he's using Faraday.

Quote
and i was able to turn his circuit into a valid and accurately behaving circuit mesh in 5 minutes where it shows that Kirchhoff indeed works.

I will get to the shortcomings of the lumped representation later, but for now, let me ask you: how did you dimension the probe's inductors? Would those values change if the probes' cables were twice as long, and covered a different area?
All instruments lie. Usually on the bench.
 

Offline MiDi

  • Frequent Contributor
  • **
  • Posts: 281
  • Country: de
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 2473
  • Country: si
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #317 on: December 03, 2018, 06:32:22 pm »
Yes that video does indeed clear up a lot of things. Its also worth taking a look at the document that Dr. John W. Belcher wrote on the topic of his videos.

I will get to the shortcomings of the lumped representation later, but for now, let me ask you: how did you dimension the probe's inductors? Would those values change if the probes' cables were twice as long, and covered a different area?

I got it from seeing how his experiment is set up and applied Maxwells concepts to give wire segments a realistic inductance. Since all of these wire segments take the same path 1/4 turn around the circle means they all have to have  the same inductance value (All of it being coupled inductance rather than self inductance)

Of course they would change if you move the wires. Its the inductor values that are capturing the physical and magnetic properties of a circuit. So changing the wire paths changes the circuits behavior so its a new different circuit. Update the inductance values to match it and the equivalent lumped circuit will again behave the same as the real one.

You have to put new updated numbers into Maxwells equations too if you change the path, right?
 

Online dr.diesel

  • Super Contributor
  • ***
  • Posts: 2132
  • Country: us
  • Cramming the magic smoke back in...
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #318 on: December 03, 2018, 07:24:24 pm »
It's sad to hear Dr. Lewin attack Mehdi with the education jab, just so unnecessary. 

Mehdi's approach and professional comments on this topic clearly show he is of the highest moral character and only wishes to constructively discuss the topic.  A very well done to him.

Anxious to hear Lewin's response to the last video.

Offline bsfeechannel

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: 00
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #319 on: December 03, 2018, 11:57:13 pm »
It's sad to hear Dr. Lewin attack Mehdi with the education jab, just so unnecessary. 

The title of his first video is "Disagreeing with a master". Science has no word of authority. It seems that Mehdi wanted to provoke that kind of reaction and Lewin fell prey beautifully.

If you go to the comment section of Mehdi latest video, the most popular comments are all bashing the octogenarian professor, with hearts conceded by Mehdi. It gives us the impression that that's what Mehdi was after the whole time. But it is just an impression and I may be wrong.

Quote
Mehdi's approach and professional comments on this topic clearly show he is of the highest moral character and only wishes to constructively discuss the topic.  A very well done to him.

I read Belcher's paper. Nowhere he says KVL always holds or that Lewin is wrong.

He says: In this sense [i.e. not always], KVL holds, as argued by Mehdi Sadaghdar, but one must always remember that the voltage difference across the inductor is defined in a very different way compared to the voltage difference across the other three elements.

What is this different way Belcher refers to? It is described in The Feynman Lectures on Physics Vol II, p22‐2.

Feynman says:  Suppose that we have a coil like an inductance except that it has very few turns, so that we may neglect the magnetic field of its own current. This coil, however, sits in a changing magnetic field such as might be produced by a rotating magnet, as sketched in Fig. 22-5. (We have seen earlier that such a rotating magnetic field can also be produced by a suitable set of coils with alternating currents). Again we must make several simplifying assumptions. The assumptions we will make are all the ones that we described for the case of the inductance. In particular, we assume that the varying magnetic field is restricted to a definite region in the vicinity of the coil and does not appear outside the generator in the space between the terminals.



It is still true, however, that the line integral of E around a complete loop, including the return from b to a outside the generator, must be zero, because there are no changing magnetic fields.


One thing that Mehdi omits is that Feynman actually derives Kirchhoff's rules (not laws)--by the way, Lewin also calls them rules, not laws--from Maxwell [Section 22-3]. And doesn't say anywhere that Kirchhoff always holds. What he says is: With those two rules it is possible to solve for the currents and voltages in any network.

Network of LUMPED elements, i.e., no changing fields outside of the components and this is the abc of Kirchhoff.

Feynman uses the line integral around a circuit just like Lewin does.

Quote
Anxious to hear Lewin's response to the last video.

Mehdi is not completely honest with his audience and the literature he recommends doesn't disprove in any way Lewin's experiment.

And that is what is really sad.
« Last Edit: December 04, 2018, 12:19:58 am by bsfeechannel »
 
The following users thanked this post: boB, mhz

Offline ogden

  • Super Contributor
  • ***
  • Posts: 2667
  • Country: lv
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #320 on: December 04, 2018, 12:15:34 am »
I read Belcher's paper. Nowhere he says KVL always holds or that Lewin is wrong.

You can troll twist it as you like, but Belcher says that KVL holds for Lewin's circuit (page 16), at the same time agreeing to Mehdi.
« Last Edit: December 04, 2018, 12:19:55 am by ogden »
 

Offline Sredni

  • Regular Contributor
  • *
  • Posts: 186
  • Country: aq
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #321 on: December 04, 2018, 12:19:58 am »
We've been discussing this topic for a month now and, although apparently the vast majority of EEs on this forum understands what is at stake, there are still a few people who think that Mehdi is a hero, Lewin is a charlatan, I am a troll, and Maxwell is just a complicated version of Kirchhoff.

The problem seems to be that Lewin's "Kirchhoff is for the birds" rant is being perceived as an attack on those who learned to rely on Kirchhoff. But Lewin's rant is directed at some of those who TEACH circuit analysis.

I am afraid the problem is deeper than that.
I had a look at Mehdi's latest video (the "conclusion") and the ensuing comments and it's beyond sad. It's scary.
December the third should mark the birth of scientific populism.
And he even links an essay by a professor telling him what the field is inside the conductors! (who's gonna break it to Mabilde, now?)
All he did was to pretend the beef was about what they decided to call KVL. Then showed what it intended with 'extended KVL' and showed that it works (duh!) for lumped circuits. And even Feynman says so!
And when he tries to apply to a non-lumped circuit he tries to locate v/2 on the conductor as if potential difference had a meaning and says... "now that's not wrong but it's a bit misleading...".

"A bit misleading". "Not wrong".

And all the fanboys chanting "lock Lewin up!" (metaphorically speaking - some say he's so old he lost his wits, other point out he has an ego problem, others highlight why he was let go from MIT, others expect this to be called ElectroBoom's law...).
And he linked a document saying "if the inductor wires are perfectly conducting, this integral is zero because there is no electric field in the wires" and showing how the superposition of the coulomb and induced field is such to give (almost) zero field in the conductors and charge accumulation on the resistors (just as Lewin always said).

Science validated by upvoting.
This is worrisome.
All instruments lie. Usually on the bench.
 

Offline bsfeechannel

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: 00
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #322 on: December 04, 2018, 12:44:49 am »
Mehdi fought the old professor and lost. But made a video to appear that he won. He can fool his supporters (after all he needs their money), but not those who are experienced and versed in the theory.

Although his comment section is a despicable scene, and he has nothing really to contribute, at least he makes me laugh with his antics.
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 2667
  • Country: lv
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #323 on: December 04, 2018, 01:14:38 am »
Science validated by upvoting.
This is worrisome.

Come on. When there's paper with real arguments you suddenly [conveniently] talk about youtube comments??

Prof. Belcher's document (link below) shall be considered as peer review of Lewin's experiment.

Mehdi fought the old professor and lost. But made a video to appear that he won.

You really shall read Prof. Belcher's document which says that Mehdi won his argument.

http://www.electroboom.com/share/FaradaysLaw_Mehdi.pdf
« Last Edit: December 04, 2018, 01:27:53 am by ogden »
 

Offline Sredni

  • Regular Contributor
  • *
  • Posts: 186
  • Country: aq
Re: Does Kirchhoff's Law Hold? Disagreeing with a Master
« Reply #324 on: December 04, 2018, 01:30:29 am »
When there's paper with real arguments you talk about youtube comments??

I promised not to interact with you on this matter, but... are you asking me where is the paper stating what I put into quotes in my previous post?
Well, it's Belcher's document.

Quote
You really shall read Prof. Belcher's document

Yup. You should indeed.

Quote
which says that Mehdi won his argument.

Of course, reading is necessary but not sufficient condition to understand what it says.
All instruments lie. Usually on the bench.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf