|
Current through coils
Tom Donaly wrote:
This whole thing boils down to an engineering question, anyway, which is, is it possible to engineer a loading coil to be small enough at the lower end of the HF spectrum so that it can be modeled using network analysis? That situation can be called pure single-point inductive loading. It may not be totally practical, but it is still vital to this discussion. There is an infinite range of real-life loading coils of various shapes and sizes. Pure single-point inductive loading is the limiting case that marks one end of that range. Any successful theory has GOT to get this case right - and if it can't, it fails. Regardless of the actual method used, any correct analysis of the whole antenna MUST conclude that, for the limiting case of pure inductive loading, the voltage/current/phase relationships at the loading inductance are the SAME as those predicted by conventional circuit analysis. This limiting case is where the two kinds of analysis come together, and here they MUST agree. That means a correct analysis for the whole antenna MUST predict zero phase shift in the current (It = I0 cos wt) between the terminals of the loading inductance. This requirement only applies for pure inductance, and only at that single point where the inductor is inserted into the antenna; but for that limiting case the requirement to join up with circuit theory is real, absolute and non-negotiable. Let's be clear: in this context, "current" is the plain ordinary alternating current that we learned about in school: It = I0 cos wt. It is the simple back-and-forth movement of electrons (charge) past a given point. Nobody denies that for real-life loading coils there can be a phase shift in the current from end to end, and that it will become larger as the coil becomes longer and skinnier. That isn't the question I'm addressing here. But the question of what happens when the coil shrinks down to become a single-point loading inductance is equally important: it cannot be evaded, and it is a definitive deal-breaker. It's hard to tell for sure from the avalanche of messages, but Cecil's analysis apparently fails in the limiting case of pure inductance - or rather, he seems to deny that the test is even a valid one. In principle there is nothing wrong with attempting a traveling-wave analysis for a loaded whip. Done correctly, it will give the right results that join up seamlessly with circuit theory as well. It's just that Cecil has NOT done it correctly. I think there are several reasons, and until he corrects them all, his theory will continue to fail... and he will continue in denial of that. -- 73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB) http://www.ifwtech.co.uk/g3sek |
Current through coils
Ian White GM3SEK wrote:
. . . In principle there is nothing wrong with attempting a traveling-wave analysis for a loaded whip. Done correctly, it will give the right results that join up seamlessly with circuit theory as well. . . . One of the tests the traveling wave analysis must pass is that the results from forward current wave excitation plus the results from reverse current wave excitation must equal the results from excitation by the sum of the two, i.e., the total current. This is required by superposition, whether the network is lumped or distributed. And analysis based on a distributed model, as Ian says, must converge to the same results as a model with lumped components as the physical sizes of the components get very small. Analyses of the examples using lumped models with total current have been entirely adequate to explain the observed inductor currents. Roy Lewallen, W7EL |
Current through coils
Roy Lewallen wrote: Ian White GM3SEK wrote: . . . In principle there is nothing wrong with attempting a traveling-wave analysis for a loaded whip. Done correctly, it will give the right results that join up seamlessly with circuit theory as well. . . . One of the tests the traveling wave analysis must pass is that the results from forward current wave excitation plus the results from reverse current wave excitation must equal the results from excitation by the sum of the two, i.e., the total current. This is required by superposition, whether the network is lumped or distributed. Cecil claims I'm the ONLY one who disagrees with him! Richard C., Ian, Roy, Gene, and Tom D. also appear to not be on Cecil's team. 73 Tom |
Current through coils
wrote:
Roy Lewallen wrote: Ian White GM3SEK wrote: . . . In principle there is nothing wrong with attempting a traveling-wave analysis for a loaded whip. Done correctly, it will give the right results that join up seamlessly with circuit theory as well. . . . One of the tests the traveling wave analysis must pass is that the results from forward current wave excitation plus the results from reverse current wave excitation must equal the results from excitation by the sum of the two, i.e., the total current. This is required by superposition, whether the network is lumped or distributed. Cecil claims I'm the ONLY one who disagrees with him! Richard C., Ian, Roy, Gene, and Tom D. also appear to not be on Cecil's team. The only thing those people have in common is a desire to get the facts right... and physical reality allows no compromises. -- 73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB) http://www.ifwtech.co.uk/g3sek |
Current through coils
Ian White GM3SEK wrote:
To all: Ian is not addressing the issue which is: Can a standing wave current phase measurement be used to tell us anything about the phase shift through a loading coil? The answer is NO! W7EL's phase measurements are flawed. Therefore, they cannot be presented as evidence of anything valid. They certainly cannot be used to prove that a coil is a point inductance. The question is not whether the two models agree on the low end - they do agree. The question is whether they agree on the high end - they don't and they are not supposed to. The differences I have presented in the two models is where the lumped circuit model fails and the distributed network model is valid. There is an infinite range of real-life loading coils of various shapes and sizes. Pure single-point inductive loading is the limiting case that marks one end of that range. Any successful theory has GOT to get this case right - and if it can't, it fails. Since the distributed network model is a superset of the lumped circuit model, it does get that case right. It also gets the case right at the other end of the range where the lumped circuit model fails. 75m mobile loading coils cannot be validly modeled using the lumped circuit model. Regardless of the actual method used, any correct analysis of the whole antenna MUST conclude that, for the limiting case of pure inductive loading, the voltage/current/phase relationships at the loading inductance are the SAME as those predicted by conventional circuit analysis. This limiting case is where the two kinds of analysis come together, and here they MUST agree. And they do, no question about that. That means a correct analysis for the whole antenna MUST predict zero phase shift in the current (It = I0 cos wt) between the terminals of the loading inductance. Here you are allowing your model to dictate reality, not vice versa. A practical antenna is a large structure, usually at least an electrical 1/4 wavelength. There is no such thing as a point inductance in a real world mobile ham antenna. Let's be clear: in this context, "current" is the plain ordinary alternating current that we learned about in school: It = I0 cos wt. It is the simple back-and-forth movement of electrons (charge) past a given point. Let's be clear. That model fails in a transmission line as it does in standing wave antennas. Taking a simple-minded approach to physics is where the air, earth, fire, and water elements came from. Nobody denies that for real-life loading coils there can be a phase shift in the current from end to end, and that it will become larger as the coil becomes longer and skinnier. That isn't the question I'm addressing here. But the question of what happens when the coil shrinks down to become a single-point loading inductance is equally important: it cannot be evaded, and it is a definitive deal-breaker. In that case, both models give the same answer. But that case doesn't exist in reality in real-world antennas. In reality, the lumped circuit model fails when it is extended to large structures like transmission lines and antennas. It's hard to tell for sure from the avalanche of messages, but Cecil's analysis apparently fails in the limiting case of pure inductance - or rather, he seems to deny that the test is even a valid one. The distributed network analysis works perfectly in the limiting case since it is a superset of the lumped circuit model. There is absolutely no disagreement between the distributed network analysis and the lumped circuit analysis for point inductors. Anyone who says there is is just attempting to set up a strawman. -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
Roy Lewallen wrote:
One of the tests the traveling wave analysis must pass is that the results from forward current wave excitation plus the results from reverse current wave excitation must equal the results from excitation by the sum of the two, i.e., the total current. This is required by superposition, whether the network is lumped or distributed. It certainly does that within the bounds of the principle of superposition. But as I earlier pointed out, like two superposed PSK modem signals, phase information is lost in the superposition. You used standing wave current phase to try to measure the phase shift through a coil. Your attempt was futile since the standing wave current phase doesn't contain any phase information. And analysis based on a distributed model, as Ian says, must converge to the same results as a model with lumped components as the physical sizes of the components get very small. And it certainly does. But the distributed network model works for antennas and transmission lines where the lumped circuit model fails. The lumped circuit model is supposed to fail for transmission lines and antennas. Analyses of the examples using lumped models with total current have been entirely adequate to explain the observed inductor currents. Maybe for you, Roy, but not for me and others. Are the four elements of earth, air, water, and fire adequate to explain the physical world to your satisfaction? -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
Ian White GM3SEK wrote:
The only thing those people have in common is a desire to get the facts right... I have asked you guys to explain the technical facts behind about a dozen technical questions of mine. The silence has been deafening. Many readers of this newsgroup have noticed the same thing. Here are the technical facts about the two models. At one end of the spectrum, we have lumped inductance. At the other end we have physically huge coils. The crossover point where the lumped circuit model becomes invalid is about 0.04 wavelength. A mobile antenna is a lot longer than that. At lengths above, 0.04 wavelength, standing waves have to be taken into account. The lumped inductance model cannot take standing waves into account. It assumes instantaneous faster-than-light propagation of current. DC|-------------distributed network model valid----------------| DC|---lumped circuit model valid---| 0.04WL cutoff Here are the two main technical points: 1. Nothing valid is proven by using standing wave current phase to measure anything. Gene F. and Tom D. seem to realize that. 2. The taper of the standing wave current through a coil depends upon where the coil is placed in the standing wave environment. You guys have looked only at mobile antennas. Try looking at longer antennas like the one at: http://www.qsl.net/w5dxp/current.htm You will find that the current taper through a coil can be zero, positive, negative, or even reverse phase. The phase reversal can be considered to be current flowing into both ends of the coil at the same time. Please come out of the deep dark lumped circuit cave and see what the rest of the world is like. You guys have been seduced by your model. You have assumed the presuppositions of the model are valid without technical proof. Here's an analogy: Ian W. says: I believe water is one of the basic elements and as proof, I offer a quart of water. Roy L. says: I believe earth is one of the basic elements and as proof, I offer a bucket of earth. Tom R. says: I believe fire is one of the basic elements and as proof, I offer this burning torch. Richard C. says: I believe air is one of the basic elements and as proof, I offer this balloon full of it. Cecil says: I believe there are over 100 elements and as proof, I offer this periodic chart of those elements. I will be on spring break until Monday so don't think I have given up on getting the technical facts out there for all to see. -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
Cecil,
Sorry, Ian had it exactly correct. I cannot speak for Roy or others, but I am quite sure I did not take any "side" in this topic. I believe in lumped circuit models, and I believe in network models. My aim was to try to correct some of the basic math and physics flaws, not to argue over the exact regimes of applicability for lumped and non-lumped models. If you choose to use my postings as support for your "side", so be it. At the same time, if proving your "side" includes mobile antennas with 48 foot whips and giant bugcatcher coils working at three times their self-resonant frequency, then please leave me off of your "side". 73, Gene W4SZ Cecil Moore wrote: wrote: Cecil claims I'm the ONLY one who disagrees with him! False! Gene Fuller, Tom Donaly, and Roy Lewallen have all made postings that support my side of the argument against yours. That doesn't mean they agree with me 100%. They agree with me on some technical things that are not in dispute except by you. |
Current through coils
Cecil Moore wrote:
To all: Ian is not addressing the issue which is: Can a standing wave current phase measurement be used to tell us anything about the phase shift through a loading coil? The answer is NO! I flatly do not accept your notion of a special kind of "standing wave current" that has its own special kind of phase properties. The current that the loading coil experiences is plain old ordinary alternating current flowing in the wire - the simple movement of electrons back and forth past a point, according to It = I0 cos wt. That is also the current that an RF ammeter reports, and the current that a computer simulation reports too. Any special kind of current that requires electronic components to behave in some different way from normal is simply not real. You have a fundamental misconception of what a standing wave of current really is. You repeat all the words about "standing waves", "cos kz", "scientific logic", "laws of physics" etc; but you don't actually let any of it into your mind. All the questions you ask other people are rooted in your own misconceptions. In other words, the questions are rigged so that they cannot be answered except by agreeing with you. And if someone very sensibly does not answer - why, you "win" that way too. I don't believe you realise how regularly you do this. -- 73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB) http://www.ifwtech.co.uk/g3sek |
| All times are GMT +1. The time now is 05:44 PM. |
Powered by vBulletin® Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
RadioBanter.com