Talk:Ivor Catt/Archive 1

From Wikipedia, the free encyclopedia

Archive This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Contents

Location of similar discussion material

Should we move all those discussions about capacitors, TL, the meaning of life etc. here now?--Light current 14:00, 15 October 2005 (UTC)

Intend to leave things more related to capacitors at [talk:capacitor], only move Tl and cap/Tl similarity stuff here--Light current 22:16, 17 October 2005 (UTC)

Catt Anomaly

Who says the pulse travels at the speed of light? In an imperfect conductor, I think it would be slightly slower. According to the Feynman Lectures (II 24-3) the wave only travels at the speed of light if the conductors are perfect. Pfalstad 17:29, 15 October 2005 (UTC)

Speed of light in the TL dielctric medium is meant to be understood here.--Light current 17:36, 15 October 2005 (UTC)

I don't think so; I think the anomaly is stating that the electrons can't account for the additional charge because they can't move at the speed of light (in a vacuum), because nothing with mass can move that fast. They could easily move at the speed of light in a dielectric. Pfalstad 17:42, 15 October 2005 (UTC)

Thats still pretty damn fast! 0.66c in polyethelene. Can they go that fast?--Light current 21:01, 15 October 2005 (UTC)

Sure, why not? If properly accelerated. (Although I'm not saying that electrons need to go as fast as the wave to explain the anomaly. I'm just saying they could.) Pfalstad 22:15, 15 October 2005 (UTC)

I dont really know. I thought they were limited to pretty slow speeds myself. This says its very slow [1]

--Light current 22:21, 15 October 2005 (UTC)

There's nothing in there that states that electrons are limited to any particular speed, just that they typically don't move fast with small steady currents. Beta particles are fast-moving electrons, and this page says that they often move faster than 75% of c. [2] Pfalstad 21:50, 16 October 2005 (UTC)

Yeah , but beta particles (or cathode rays) are not held in a metal lattice are they?--Light current 21:53, 16 October 2005 (UTC)

In an idealized perfect conductor, electrons do not collide with the metal lattice. But this is all distracting from the fact that the electrons do not need to go fast to explain the anomaly. Pfalstad 22:13, 16 October 2005 (UTC)

Well actually I think they would have to go fast. Whats your explanation for wher the charge comes from?--Light current 22:29, 16 October 2005 (UTC)

Did you read [3]? Pfalstad 22:41, 16 October 2005 (UTC)

Yes!--Light current 23:08, 16 October 2005 (UTC)

And? Pfalstad 02:48, 17 October 2005 (UTC)

I think Dr McEwan is wrong! I dont see how you can get what is in effect a longituninal wave traveling at c if none of the particles can travel (even for a short time) at that speed. Think of sound waves in air. What speed do the molecules have to travel so that sound travels at 345 m/s?. Less than 345 m/s? or at least 345 m/s?--Light current 02:58, 17 October 2005 (UTC)

But air molecules only interact by bumping into each other. Electrons interact by means of EM waves. Bump an electron, and it will send out an EM wave which will push other electrons out of the way. This EM wave moves at c regardless of how fast the electron is moving.

Forgive me interrupting your post at this point, but I thought it essential to deal with the points you have just amde
I agree that electrons interact by means of EM waves, and that these waves travel at 'c' in the medium. What you and Dr McEwan have been arguing is that the charge can travel at 'c', whereas you have just agreed that charge cannot move at 'c'. This is the whole point of Catt's argumnet: ie charge cannot travel at 'c' but it needs to in the classical argument order to explain EM waves in a TL. So the conclusion is obvious EM waves in TLs do not require charges to move in the conductors.--Light current 22:03, 17 October 2005 (UTC)

The signal/wave velocity and the drift velocity are not the same. Also, in the case of a coax line, for example, it is the EM wave which transmits the signal. The electrons/currents have little role except to cancel the electric field at the surface of the conductor, to ensure the wave is reflected there. In order to do this, they do not need to move at c. It doesn't matter how fast the wave moves; it matters how fast the field changes at a given point.

I agree with all of this last para.--Light current 22:03, 17 October 2005 (UTC)

A straight, square edge is not realistic; it is more like a voltage ramp when the wave passes over a given point. If there are enough mobile electrons, they should easily be able to respond with a current to cancel the electric field and reflect the EM wave.

But you just said its not the electrons (charge carriers) that respond but the EM field. Ill leave the bit about stright edges for the moment.--Light current 22:03, 17 October 2005 (UTC)

(If they don't, how does the EM wave get reflected at a conductor's surface?) Pfalstad 21:31, 17 October 2005 (UTC)

Just think about it for a while!  :) Pfalstad 03:19, 17 October 2005 (UTC)

You agree that electrons need to cancel the electric field at the surface of the conductor? That means they need to move. You have yet to answer the question: how does the EM wave get reflected at a conductor's surface if there is no current? Pfalstad 22:14, 17 October 2005 (UTC)

No they dont need to move, they just need to be there I think! But they're there any way in a conductor-- so no problemo!--Light current 23:18, 17 October 2005 (UTC)

How do they influence the EM wave if they don't move? All matter has electrons in it. Does all matter conduct electricity? No. Pfalstad 23:26, 17 October 2005 (UTC)

Its the em wave that moves- not the electrons as you said earlier.--Light current 23:28, 17 October 2005 (UTC)

Fine, don't answer my question. Pfalstad 00:35, 18 October 2005 (UTC)

You already answered your own question. You said its the EM wave that moves, not the charge carriers. I agree with you!--Light current 18:39, 18 October 2005 (UTC)

Actually they both move (EM wave and charge carriers). As you noted below, when I asked this same question, there are currents at the surface of a conductor when hit by a wave (see skin effect). Pfalstad 18:59, 18 October 2005 (UTC)

OK charge carriers can and do move, but slowly. They dont explain high freq effects tho', which is what im talking about. All low freq effects can be covered by 'snails pace' charge carrier movement. I suggest we eliminate slow carrier movement from the discussion to save time and confusion.

Well of course I'm not ready to eliminate currents from EM theory. BUT, high-frequency circuit behavior is probably best understood by looking at the EM wave, which is most important, and not paying much attention to charges or currents. Happy? Pfalstad 02:15, 19 October 2005 (UTC)

Another important question is: how does the slow movement of charge carriers and the fast movement of EM fields all nicely and seamlessly mesh together? For instance, if we were to charge a capacitor very slowly, would we need these eM fields (travelling waves) between the plates? Is there a difference between behavior at high speed and that at low speed. I think there must be; but they somehow fit together in the middle.--Light current 02:31, 19 October 2005 (UTC)

The pulse can move at 2/3 c while the electrons move at half a centimeter per hour. Electron movement is not the same thing as electromagnetic wave movement. — Omegatron 15:25, 22 November 2005 (UTC)

Displacement current

No idea why he thinks displacement current is not "needed" if you treat the TL as a transmission line. The telegrapher's equations, and the behavior of a TEM wave, require the displacement current term of Ampere's law for their derivation. A transmission line certainly has a changing electric field which is all that displacement current is, a quantity proportional to the derivative of the electric field. It seems to me that he has just used transmission line equations to come up with a description of capacitor behavior that doesn't mention charge or displacement current, and then said viola, charge and displacement current don't exist.

Well thats what he reckons! Have another look at his pages to see if you can find his reasoning.--Light current 20:53, 15 October 2005 (UTC)

Mystery contributor (Probably Nigel Cook) input

This led to the biggest row ever in Wireless World. After Catt's paper was published, another (and longer standing) Wireless World author, Professor David A. Bell, M.A., B.Sc., Ph.D., F.Inst.P., F.I.E.E., Professor of Electrical Engineering at Hull University, was so angered that he wrote an article which didn't mention Catt or the others by name, but was headed up No Radio Without Displacement Current. [4]

Bell first pointed out that Maxwell's physical "aether" model of displacement current had been dumped by Einstein and others, and secondly pointed out that the mathematical equation of whatever "displacement current" is required in Maxwell's equations to predict radio waves that Hertz discovered (in fact Faraday in 1846 wrote Thoughts on Ray Vibrations without Maxwell's maths, and Weber preceded Maxwell in getting light speed out from electromagnetic constants).

One problem is that Catt has to try to make his case exciting by claiming he has uniquely disproved something important (which no journalist can grasp!), instead of saying the more mundane but relevant (and obvious) fact that his work began when computers were crashing with glitches, and crosstalk failures cost people lives.

Another problem is that Catt ignores the transmission of energy between capacitor plates (which behave as radio transmission and reception aerials) while the first plate charges up:

_____||_______

______________

If you look at the March 2005 issue of Electronics World there is a letter from Cook on page 49 which points out that radio transmission is a capacitor type problem.

Even though radio is in a sense "displacement current" mathematically, the stepwise charging, and the fact that in Maxwell's theory "displacement current" goes in a direction 90 degrees to the direction of radio waves, disproves the dogma that radio proves "displacement current".

Ideally the equation for "displacement current" should be renamed to avoid confusion with the discredited physical ideas of Maxwell. The best name for it might be "radio current".

Mystery contributor-- propably Nigel Cook

Very Strange

He says that a capacitor can be treated as a TL? Well, sure, everything can, and at high frequencies, you are forced to treat everything as a TL, aren't you? But why is this a good thing? It is much easier, at lower frequencies, to use displacement current to salvage Kirchoff's laws, because Kirchoff's laws are handy. I would certainly not want to use transmission line theory on a simple low-frequency RLC circuit. I have only browsed Ivor Catt's website but it seems like many of ideas are not wrong, just overstated. It's like he's trying to make his ideas seem as cranky as possible. His praise for Heaviside might tell us something; Heaviside was brilliant, but went a little nuts in his old age, too. Pfalstad 22:08, 17 October 2005 (UTC)

Yes Im goin a little nuts in my old age too. Its working on WP thats doing it in my case ;<(--Light current 22:14, 17 October 2005 (UTC)
In circuit theory and analysis dealing with idealised components especially at low frequencies, it is of course easier to use the (erroneous) concept of displacement current to salvage KL.
THe problem is that some people tend to extend the concept of displacement current into areas where it just dont fit! (like TLs etc.) The extended concept then leads quickly to people believing in it as an actual real phenomenon (which a lot of people do). The point of Catts articles etc has been to show that the idea of diplacement current is not only unnecessary - it is wrong.
Maybe his ideas seem cranky because they are so at odds with what we were taught. At first, I not only thought they were cranky, but I couldnt be bothered to try to even understand what he was claiming. I was too busy working to think about it!--Light current 22:51, 17 October 2005 (UTC)

What does that even mean, "displacement current is erroneous"? Displacement current is simply dD/dt. What about dD/dt is erroneous? Are you saying the dD/dt doesn't belong in Maxwell's equations? You cannot get a wave equation from maxwell's equations without that term; there wouldn't be any EM waves. Are you saying the displacement current isn't a real current? Of course it isn't. I'm not aware of people misusing displacement current like you describe, but then I'm not in the industry. Kirchoff's current law, in its most general form (the del.J equation in Kirchoff's circuit laws) seems hard to argue with, since it is derived directly from Maxwell's equations. Im not sure there would be any point in using that form, though, especially in high-frequency cases, like across a TL. Pfalstad 23:13, 17 October 2005 (UTC)

Well you dont need displacement current to explian capacitors or TLs or EM waves in space. dD/dt is not current - its something else that is not needed to explain anything!--Light current 23:26, 17 October 2005 (UTC)

Fine, dD/dt is not current. But it is needed. You need dD/dt in Maxwell's equations to explain EM waves. You need dD/dt in Ampere's law, or Ampere's law is not generally valid. If you don't need it, please show us your replacement for Maxwell's equations, and show how you derive the electromagnetic wave equation from them. Pfalstad 23:28, 17 October 2005 (UTC)

Amperes law states: int (B.dl) = uoI. Where is dD/dt in this?

I mean the Maxwell-Ampere law, Maxwell's_equations#Summary_of_the_equations, the "Ampere's law + Maxwell's extension" under "Differential form". The integral form doesn't have dD/dt but has a displacement current term. The law as you quoted is not generally valid and is not part of Maxwell's equations. Pfalstad 00:41, 18 October 2005 (UTC)

Maxwell only added this bit to Amperes law cos he couldnt explain where all the charge/current went in a capacitor. But now we know: it goes into EM energy. This is what all the fuss has been about regarding capacitors and transmission lines. Is this becoming any clearer to you now?--Light current 01:30, 18 October 2005 (UTC)

no.. So he shouldn't have added this to Ampere's law? But you just said (below) that the EM waves derivation depends on it! Pfalstad 01:33, 18 October 2005 (UTC)

I dont remember saying that. I said EM waves come out of curlB and curlE thats all. No current in sight!--Light current 01:47, 18 October 2005 (UTC)
Its the travelling electric and mag fields that make4up the wave. Any voltages, currents etc in the conductors are induced, as Heaviside said. The mag field in a n EM wave is not caused by any physical current. EM radiation is more fundamental than current, voltage or charge.--Light current 23:34, 17 October 2005 (UTC)

Currents are induced in the conductors? So there really are currents and moving charges? I'm totally lost! Pfalstad 01:29, 18 October 2005 (UTC)

If you can measure em , theyre induced. But they dont involve moving charges, thats for sure.--Light current 00:24, 19 October 2005 (UTC)

I have a simple question for you. Are Maxwell's equations correct? If not, what specifically is wrong with them, and how would you correct them? Pfalstad 00:10, 18 October 2005 (UTC)

The question is not actually that simple. I think its the interpretation of dD/dt as a current thats wrong.--Light current 00:15, 18 October 2005 (UTC)
EM waves derivation depends on 2 of Maxwells equns.
1) curl B = epsilon mu dE/dt
2) curl E = -dB/dt
NB all derivatives are partial. Now what your saying is that dE/dt constiutes a current because D =epsilonE. Why should it be a current. It isnt a current. Cant be measured as a current. Its dE/dt nothing more. Therefore, whilst EM radiation depends on the electric field varying with time - so what?. That's what you expect in a wave isnt it?? Same for the magnetic field-- varying. Who decided that dD/dt was a current in the first place? Its actually rate of change of electric flux density. This can be accounted for by accumulation of charge.Light current

Accumulation of charge? I thought we were talking about a vacuum. Pfalstad 01:14, 18 October 2005 (UTC)

No. Im obviously talking about Maxwells stupid idea about capacitors here.--Light current 01:23, 18 October 2005 (UTC)

I'm afraid to ask, but how does charge accumulate if it doesn't move?

Well its plain to see that since charge cant move fast, it cant accumulate fast (but can it be separated fast by some other means?) I dont know yet. When we know how a capacitor really charges, we will be able to answer this question. To charge a capacitor fast, one must separate lot of charge quickly. In a TL this obviously can be done quickly so it must be the EM wave that causes this fast separation. The charges cant move that fast. Interesting question!--Light current 01:43, 18 October 2005 (UTC)
E and B are symbiotic. One causes the other. No external currents or voltages are needed. That this is so only requires consideration of EM in vacuo.--Light current 00:49, 18 October 2005 (UTC)
In fact Im going off Maxwell quite rapidly now for inventing this term to describe something he couldnt expalin. Its caused a great deal of trouble in the past 100 or so years. If only hed said 'waves in space' everyone would have been happier and he would have been correct.--Light current 01:11, 18 October 2005 (UTC)

Well of course the terminology isn't great, but I thought you and Ivor Catt were objecting to the concept itself, not the terminology. Pfalstad 01:14, 18 October 2005 (UTC)

Well Im trying to keep an open mind, but the more I think about it and try to explain things, the more attractive Catts arguments become. If energy in a capacitor was stored as (counter propagating lets say) waves, this would mean we would not need to separate charge quickly-- yes? (At least we could take our time over it). This idea is therefore attractive in that it explains how a capacitor can charge fast when the charges cant move fast- neat eh? Ahhh-- its all becoming very clear to me now!!(Dave Bowman 2001/2010)--Light current 02:17, 18 October 2005 (UTC)
No. I've just seen it. When charging a TL, the charges cannot separate via the external circuit becuse they cant move that fast. What happens is that the EM wave induces equal and opposite charges on each conductor and achieves a sort of charge separation that way. This is more evidence the the energy is trapped in the EM waves not by some external movement of charges. Charges dont need to flow to 'charge' a capacitor!!!! Now Thats a controversial statement is it not?--Light current 02:27, 18 October 2005 (UTC)

Well I have no problem with the idea of the EM fields playing a major part in the charging of a capacitor. It makes a lot of sense. In fact, you can model the charging of a capacitor very well by ignoring the currents and movement of charge and concentrating on the fields. I've done it, in a simulation I wrote. Once you have the fields, you can easily calculate the surface currents/charge from them. Very nice. But it only works for perfect conductors, or high frequencies. Once you introduce resistance, or face the fact that the current is flowing through the whole wire at low frequencies, you're out of luck. To take a nice mathematical trick like that, and to take it too far and and totally throw out the idea of current, seems totally unjustified, and introduces a host of problems, in my opinion. And I've already told you what I think of the "undetectable waves that cancel out to DC" theory. Pfalstad 03:38, 18 October 2005 (UTC)

There may be one way of indicating the existence of these waves experimentally, and Im in the process of designing an experiment. The only trouble is, its going to take some rather large sheets of metal or copper clad PCB to do it! Im also working on theoretical energy considerations to prove the concept.--Light current 05:05, 18 October 2005 (UTC)
Your simulation sounds very interseting. Can you give more details of what software package you used etc? I would like to try stimulation myself.--Light current 04:34, 18 October 2005 (UTC)

It's here.. [5] select "Setup: Capacitor" from the upper right menu. It's a capacitor inside a resonant cavity. Arrows indicate E field, green/red indicate B field. You will be happy to know that current and charge play little role in the simulation; it's purely an EM field simulation, and the charge/current are calculated from the fields. Pfalstad 15:26, 18 October 2005 (UTC)

DISPLACEMENT CURRENT IS NOT A CURRENT

There, satisfied? I've said it many times, but apparently you didn't hear me. dD/dt is not a current. Happy now? Nobody says it is a current. But the dD/dt term is needed. Displacement current is simply what the dD/dt term is called. I don't know why, but it is. In a vacuum, this is proportional to dE/dt as you quoted. It is needed to get waves. Can we agree on that?

Did you say displacement current is not a current? OKAY!

Sorry, but I couldn't resist a little theatrics. Pfalstad 01:44, 18 October 2005 (UTC)

No its not needed to get waves (see above).--Light current 01:07, 18 October 2005 (UTC)

Well I give up, because I have no idea what you're saying. You just said: "EM waves derivation depends on 2 of Maxwells equns. 1) curl B = epsilon mu dE/dt". So that means that the curl B equation (Ampere-Maxwell) is needed to get waves doesn't it? dE/dt is the displacement current term in a vacuum. Right? It's not a current, I know. But dE/dt is needed to get waves. Pfalstad 01:10, 18 October 2005 (UTC)

dE/dt along with -dB/dt actually describes the wave when transformed into the cartesian coordinates and leads quite simply to the traveling wave equations with which we are all familiar. So rather than being needed to get the waves - it is the wave (well the electric part anyway)--Light current 01:18, 18 October 2005 (UTC)

Now as to whether currents are needed... How does a wave get reflected by a conductor but not by other matter (which contains electrons)? Pfalstad 00:58, 18 October 2005 (UTC)

I dont know at the moment. But I'll think about it Time for bed! goodnight--Light current 01:32, 18 October 2005 (UTC)
Small localised currents flow in the surface of (see skin effect) conductors but do not involve any major movement of charge carriers along the line. Is this answer OK?--Light current 02:32, 18 October 2005 (UTC)

Localized how.. In terms of thickness from the surface? Well the skin effect article predicts that for low-frequencies, the skin depth can be pretty wide, and can go through the whole conductor if the conductor is small enough. Why would there not be any major movement of charge carriers along the line.. What would prevent it?

Low frequencies are not the problem- any theory (almost) works at LF--Light current 04:30, 18 October 2005 (UTC)

I have no idea what it is about Catt's theory that you find attractive. To me it seems completely nonsensical. How do you explain resistance in terms of EM waves? The resistance of a wire depends inversely on its cross-sectional area. If there's no current in the wires (not just the skin but the entire wire), how is this possible?

I dont understand the question here. Do we have to define resistance in terms of EM waves? In an imperfect conductor some current will be induced below the surface.--Light current 04:30, 18 October 2005 (UTC)

I thought that Catt was claiming that there is no current in a conductor, ever. [6] There he says "electric current does not exist." Pfalstad 15:21, 18 October 2005 (UTC)

How about semiconductor devices? How can diodes be explained in terms of EM waves on the surface of a semiconductor? What about the fact that solid-state physics specifically predicts that electrons are mobile in conductors? How about vacuum tubes? How about the photoelectric effect?

EM waves have trouble in semiconductors - thats why they are slow (unless you count the plasma devices where em waves can travel). Otherwise its the very small distances involved in semis that give the speed. What about tubes? Are photoelectric devices fast?--Light current 04:30, 18 October 2005 (UTC)

And why bother? What do you get out of this theory, other than a complete mess where an elegant, consistent, and widely accepted theory used to be? It's a mystery to me, but if you like it, you can have it. Pfalstad 03:31, 18 October 2005 (UTC)

What we get is a more complete understanding of how things work and how to design fast circuits as Catt has demonstrated. Its certainly not a complete mess. Its far simpler and more coherent than conventional theory and answers questions conventional theory cant. However, no ones asking you to accept it. Were just writing an article-- remember? at least I am!--Light current 04:30, 18 October 2005 (UTC)

Proposed experiment or simulation to prove/disprove counter-propagating wave theory

Now that I have seen the excellent simulation for EM fields, Im thinking I may be able to simulate what I had thought I would have to do in hardware.

The experimental idea is this:

If the waves are counter-propaging adding to DC, there is no way of detecting them until they are disturbed. This is easily done in one way by connecting a load at either one end or the other and observing the pulse output. However, I think you AC and others could object to the validity of that expt becuase a TL charged with plain dc would exhibit the same effect.

Outline

My alternative expt involves the use of a parallel plate capacitor made from thin square plates (say 1m sq). The energy in the form of a travelling wave is introduced from a step generator along one edge of the square by means of an exponential matching transformer to go from the generator Z to the Zo of the plates. Interposed between the matching section and the plate edge are a number of electronic switches, all capable of being operated electronically at the same time. When the energy has been deposited in the TL/capacitor (or whatever you want to call it), the switches are opened. Also, at the adjacent edge (not the opposite edge) of the square, we have a similar arrangement of switches and a matching section terminated in a load and a detector of waves (O'Scope). The output waves (if any) therefore must travel at right angles to the direction of the input wave.

Procedure

1.Charge the device with energy and isolate the source.

2.Connect the detector by operating the switch array and observe the delay and shape of the detected signal on the scope


Interpretation of results

A.If the detected signal is observed to have zero delay from switch closure and risetime only dependent on the switch performance, then it is likely that no counter propagating waves were present, becuase the system acts as tho' it was charged with plain dc.

B If, however, there is appreciable delay between switch closure and arrival of the signal at the detector (allowing for the delay in the matching section), this would tend to indicate that the counter-propagating waves did exist and they take a finite time to alter their direction of propagation. If no energy whatsoever is detected (unlikely) that would tend to prove the theory without doubt!!.


I would be very interested in Paul's (or any physicist's) comments as to the validity of such an experiment/ simulation. --Light current 18:44, 18 October 2005 (UTC)

Comments on experiment proposal

If the counter-propagating waves did exist, there would be one going in each direction, right? So the one going in the right direction would be observed immediately. How much delay would you expect to see, and why? Pfalstad 02:41, 21 October 2005 (UTC)

No. Im proposing to detect any signals at right angles to the initial direction of the counter propagating waves. As to how much delay--- I have no idea, I thought you may have!--Light current 02:46, 21 October 2005 (UTC)

Ok, so you have two parallel plates, with a generator attached to both plates at one edge, and a detector and load attached to both plates at the adjoining (not opposite) edge? Ok, that is an interesting setup. In standard physics, there will be no delay. Even if you treat the incoming voltage as a wave, it will still diffract downwards instantly when the switches are closed. Pfalstad 16:11, 21 October 2005 (UTC)