|
Current through coils
Roy, W7EL wrote:
"Just a couple of hours ago you said the current would have to wind its way atound each turn, following the wire from one end to the other, and it would take nearly the wire length divided by the speed of light." Yes, and I`m still convinced that is the case in an air cored r-f coil that is long because the coupling between ends of the coil isn`t enough to bypass the delay of the coil. I posted speculations on bypassing the delay in the coil. Capacitance between turns is too small over the length of the coil, said to be about one loot, and about 100 turns. Tom, W8JI had said that magnetic coupling between the start and finish of the coil bypassed the time delay of following the path of the wire. Well, nothing happens instantly when voltage is applied across a coil. 90-degrees after the voltage has crossed the zero axis on its way up, the current does the same. It lags the voltage by 90-degrees. It`s the current which induces a voltage in the coil and this is delayed by the forces predicted by Lenz`s law. 90-degrees at 4 MHz equates to about the time required for a radio wave to traverse about 60 feet of thin wire. 100 turns of wire on a 2-inch form requires about 52 feet of wire. The current travels from start to finish on the coil before the current reaches its maximum in the coil and before energy could be effectively induced from one end of the coil to the other. The wave velocity is about 984 feet per microsecond. These are just musings aloud and confirm my speculation that signal progress is through conduction on the surface of the wire of the coil. Best regards, Richard Harrison, KB5WZI |
Current through coils
|
Current through coils
Heck, Cecil, I don't see Ian saying that a proper distributed model is
incorrect at all. I see him saying that YOUR model is incorrect! -- For the record, YOU and YOUR below translate to Cecil and Cecil's. YOUR model doesn't have any capacitance to the outside world, as YOU have posted and repeated till you are perhaps blue in the face. (No charge storage translates very directly to no capacitance, in case anyone could have missed that.) Because of that YOUR model has, guess what, ZERO time delay along the coil. It is ONLY through a combination of both inductance AND capacitance that you get propagation velocities equal to or less than the speed of light. NOTE that in freespace, the speed of light, c, is exactly equal to 1/sqrt(epsilon_naught * mu_naught) -- and the units of epsilon_naught (freespace permittivity) are FARADS per meter, and mu_naught (freespace permeability) are HENRIES per meter. A coil increases the henries per meter, but somehow we seem to have gotten in YOUR model to some space around the coil in which the permittivity is _zero_ so that we don't have any capacitance in YOUR model. That would be a good trick, but it's not one that Ian and I are buying. I suppose that W8JI and Tom D and Roy and Wes and Reg and Gene and John P and probably the two Richards and some others are ALSO not buying. Why, even the detailed Tesla coil calculations I've seen consider the distributed capacitance to the outside world, in great detail. I'd bet that the authors of those calculations would ALSO not buy your model. Putting it another way, propagation of an EM wave requires an interchange of energy between electric and magnetic fields. With zero permittivity, there would be no energy stored in the electric field, and no EM wave. (In a TEM transmission line, it's often said that the energy is stored in capacitance and inductance along the line, but that's no different than saying the energy is stored in electric and magnetic fields.) It amazes me that you fought so hard for a distributed model in which that capacitance to the outside world is missing, but insist that the resulting model allows a non-zero time delay. That's YOUR model; it's all in the stuff YOU have posted for anyone that wants to go look at it. By denying the capacitance in the model, YOU are the one who doesn't accept what's been know since Maxwell and Faraday and Tesla and... But YOUR model isn't any use to me, and it seems that it's no use to Ian. We'd prefer a model that actually accounts for all the currents correctly, and actually allows for a delay along a transmission-line structure. Then, knowing it's an ACCURATE model when we can verify through specific measurements that it agrees to an acceptable level with those cases we measured, we can look at ways to use that model as-is, or to use a model which makes life easier for us which matches the very accurate one closely enough for our purposes. (Not only does YOUR model with no capacitance to the outside world have zero propagation delay, but it ALSO leads to a line with infinite impedance, which I'm ALSO not buying, though with zero delay, the line impedance really doesn't matter. Also notice that with no capacitance to the outside world in that area, the model collapses to exactly the one YOU are arguing AGAINST, except that the straight sections of antenna are apparently not in the zero-permittivity area, so we need to keep them separated...it leads to a very strange model, indeed!) And please note that in the paragraphs above, the only models specifically mentioned are DISTRIBUTED ones, so don't go giving me any bull**** about the other kind. And you can save any bull**** about accepting the fact of capacitance to the outside world, because your postings repeatedly say otherwise. |
Current through coils
wrote:
It seems to me Cecil now agrees the system can be modeled as a lumped components and loads and we do not need to use standing waves. A 100% false statement but we are accustomed to such from W8JI. Since the lumped circuit model is a subset of the distributed network model, if there is any disagreement between the two models, the distributed network model wins every time. They are both right under certain conditions and the lumped circuit model is wrong under certain conditions. Quoting from: http://www.ttr.com/corum/index.htm "There are no standing waves [allowed] on a lumped element circuit component. ... for coils whose WIRE LENGTH exceeds 1/6WL", the distributed network model is required. Quoting from: http://www.ttr.com/TELSIKS2001-MASTER-1.pdf Concerning the *impedance only* of a loading coil: "The formula will NOT (and, being lumped, can not) give the voltage magnification by VSWR dur to physically true current standing waves on the structure ... If impedance is the only item of interest, the empirical Medhurst approximation is acceptable out to about 60 degrees." But we haven't been arguing about impedance. We have been arguing about phase. Here's what the above paper says about phase shift through a loading coil. "Further, the voltage distribution passes from the loop of a sinusoid (at 90 degrees) to the linear portion of the sinusoid (for heights less than 15 degrees)." It is necessary to use the distributed network model if the phase shift through the coil is greater than 15 degrees. Continuing the quote: "Lumped elements 'have no physical dimensions and no preferred orientation in space; they can be moved around and rotated at will.' Not so for real world coils. ... The concept of coil 'self capacitance' is an attempt to circumvent transmission line effects on small coils when the current distributions begins to depart from its DC behavior." -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
Gene Fuller wrote:
You issued a "challenge" to design and report on a loading coil for 4 MHz, with a whip of 8 feet. I responded with a solution that used a whip length of 10 feet. No, you didn't! You responded with a antenna length of 11.775 feet, 3.775 feet longer than the agreed upon 8 foot antenna. It wasn't the whip that was to be 8 feet, it was the entire antenna. I made that perfectly clear early on so it would match the mobile antennas in the ARRL Antenna Handbook. You did not acknowledge the changes you made until I complained. Of course I did. Go back and read my postings about such. It wasn't until you complained that I sent you the corrected EZNEC files. I made a severe blunder in the traveling wave model and you copied my blunder. -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
Richard Harrison wrote:
100 turns of wire on a 2-inch form requires about 52 feet of wire. 52 feet of wire on 4 MHz is 0.21 WL. Dr. Corum says anything over 0.17 WL requires the distributed network model. The 3 nS delay measured by W8JI through that coil is simply technically impossible except in his mind. -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
Richard Clark wrote:
"The apparent capacitance based on reported resonances and modeled reactance is on the order of 12 -14 pF." Have you calculated the self-capacitance of a 2in x 12in single-layer coil for yourself? The length to diameter ratio is 6. H = .92 D = 5 cm HD = 4.6 pF by the formula on page 451 of the "Radiotron Designer`s Handbook". Course, formulas are a dime a dozen and disputed. Best regards, Richard Harrison, KB5WZI |
Current through coils
Why, yes, in fact it is true that I don't agree in every case with the
unknown guru you've placed on a pedestal above all reproach. So there. It is also true that I do not agree with your (mis)interpretation of what your unknown guru has written. I REGULARLY model transmission lines as "lumped elements" and do NOT "presuppose that the speed of light" through them is infinite. I REGULARLY model op amps as "lumped elements" and do NOT presuppose that the phase shift (and therefore propagation time) through them is infinite. I REGULARLY model analog to digital converters as "lumped elements" and do NOT presuppose that the result comes out at the same time as the signal that goes in. I REGULARLY model inductors as "lumped elements", and do not presuppose that they have no resistances and capacitances parasitic to their inductANCE. I find that my models very reliably predict the behaviour I actually observe in the circuits I build. I am served very well by the models I use. By the way, what's EE203? It's very likely that I missed not only that day but all such days. You yourself may well be presupposing something that isn't true. Cheers, Tom |
Current through coils
K7ITM wrote:
YOUR model doesn't have any capacitance to the outside world, as YOU have posted and repeated till you are perhaps blue in the face. (No charge storage translates very directly to no capacitance, in case anyone could have missed that.) I do wish you guys would argue in good faith. ***STRAWMAN ALERT*** I didn't say there was no capacitance to the outside world. I said such is a secondary effect, not a primary effect, and for the sake of the present argument, can be ignored as secondary effects often are ignored. Because of that YOUR model has, guess what, ZERO time delay along the coil. No, transmission lines have negligible capacitance to the outside world and their time delays are NOT zero. You straw man is just not believable. It is ONLY through a combination of both inductance AND capacitance that you get propagation velocities equal to or less than the speed of light. Yes, and that capacitance can be either internal or external. I'm ignoring the rest of your posting because it is based on the false premises of your straw man. But you get an 'A' in Obfuscation 101. It amazes me that you fought so hard for a distributed model in which that capacitance to the outside world is missing, ... It is *NOT* missing. That is just your straw man. It is just secondary to the addition of the forward and reflected current phasors. At a point where the the forward and reflected current phasors add up to zero, it's hard for anything else to contribute much of an effect. -- 73, Cecil http://www.qsl.net/w5dxp |
| All times are GMT +1. The time now is 06:01 AM. |
Powered by vBulletin® Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
RadioBanter.com