|
New to Ham Radio?
My Profile
Community
Articles
Forums
News
Reviews
Friends Remembered
Speak Out
Strays
Survey Question
Operating
Contesting
DX Cluster Spots
Propagation
Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers
Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net
|
|
1-10 of 34 messages
|
Page 1 of 4
Next
|
|
75m Loading Coil Epic
|
Reply
|
|
by N3OX on October 26, 2009
|
Mail this to a friend!
|
W5DXP posts in a typically inappropriate location
http://www.eham.net/forums/Elmers/235888
He comments regarding the fact that I think he should stop harassing W8JI over this mobile antenna business that has gone on forever. The backstory is long, but in short, W5DXP has been championing a coil model laid out in the following paper which appeared in a special issue of Microwave Review, a publication of the Serbia and Montenegro chapter of the Microwave Theory and Techniques Society:
http://www.mwr.medianis.net/pdf/Vol7No2-07-JCorum.pdf
In this paper, the authors solve Maxwell's equations subject to boundary conditions on a helix and in doing so work out a continuous picture of what most of us would consider "coil self-capacitance." That is, under the assumptions in this paper they work out what I'd consider a less ad-hoc method of describing the effects of electric fields in helical coils.
This is an interesting paper and may have some applications to the easy design of things like phase reversing coils, as it describes how electromagnetic waves can slowly propagate down a coil much like they'd do in a synthetic LC delay line. Of course this is known and used in systems like microwave traveling wave tubes, but for some reason the discussion to date seems to focus solely on this paper without further literature review. However, I don't see anything wrong with the paper itself on casual inspection and it would seem that the publication is peer-reviewed.
I am not interested in taking sides in this debate largely because, at its core, the debate in the amateur community centers around the names we give to things and our mental models of various electromagnetic systems.
W5DXP calls me out on my perceived unwillingness to tell W8JI he's wrong:
"Dan, when one allows personal friendship loyalties to overshadow technical facts, one sacrifices ones ethics in the process. "
And maybe there's a slight issue. I've followed Cecil's arguments and considered his points and I think there are some MINOR problems with the page at :
http://www.w8ji.com/mobile_and_loaded_antenna.htm
To W8JI: would seem that the self capacitance associated with an isolated coil with no other surroundings involved can lead to current variations along its length. There are various current patterns you can get on a helix just based on its self capacitance. So when you mention "the outside world" on your coils page that might not be quite right.
However, to Cecil: generally speaking, most of your objections to what Tom has said and measured hinge on what English language descriptions people have in mind when they discuss these things:
"The delay through a 75m bugcatcher coil" may mean different things to different people. And mathematically it is poorly defined.
A slow wave coil is dispersive. Different frequencies travel at different speeds. A TDR measurement would result in massive pulse spreading and so you can't terminate it and measure it that way like you could with a regular transmission line.
The "electrical length" of a slow wave coil is better defined, because that can at least be described in terms of excitation by monochromatic EM fields and the resulting standing wave pattern (or with traveling waves if there's a termination). You can see this clearly in EZNEC and it's pretty neat. I think it's also responsible for the fact that you can run a TakTenna or something at near three times the basic resonant frequency. It works in pancake coils too.
However, your apparent mental model of the "electrical length" effects of slow wave coils in short mobile antennas fails to address the following points in a way that fails to convince me that a slow coil "occupies a certain number of electrical degrees" in a more profound way than a lumped inductor does. These are the issues:
1) Why I could load a mobile antenna with a lumped inductor if I had such a thing?
2) Why can I use a regular inductance calculator to build a coil that is VERY CLOSE to the right dimensions even ignoring that it's a "slow wave structure?"
The answer to #1 is an IMPORTANT PART of your antenna-as-transmission line argument. The phase shifts caused by a pure point reactance MUST cause it to "occupy" 90-N degrees, where N is the actual physical length of the antenna.
"Antenna-as-transmission-line" is a productive model that has been successfully used by many... I've flipped through Kraus now and then and he favors that. But the insertion of a point inductor has to mesh with this model. I think it probably does. And in that context, "occupation of electrical degrees" also becomes. at least partially, semantic argument.
And I'm sick of semantic arguments. They're boring and unproductive. They're a waste of time. They allow us to argue over things that have one mathematical answer that none of us want to actually work out.
And as far as I can tell, you and Tom are BOTH right on some points and probably both wrong on others and mostly you're both talking about the same system using different language.
And Cecil, this is not anywhere close to a matter of "you're right and Tom's wrong" because, as usual, we haven't even asked all the questions and settled on what we're discussing when we're discussing it.
If Tom knows all this stuff but imagines little tiny capacitors everywhere displacement current flows, he is NOT WRONG.
This is important.
Mental models of physics are not right or wrong. Mathematical models can be right or wrong and can live on a continuum of applicability, but I can't look inside anyone's brain and see how their mental model analogy jives with the solutions to boundary value problems on coils.
The answer, in the end, is the math, and I don't see anyone jumping to do the math.
73
Dan
|
| |
|
RE: 75m Loading Coil Epic
|
Reply
|
|
by N3OX on October 26, 2009
|
Mail this to a friend!
|
"Mental models of physics are not right or wrong"
This is too strong.
I should have said something like "Mental models of physics are not wrong if they don't lead to wrong predictions"
This is why I don't like discussing this stuff in words.
|
| |
|
RE: 75m Loading Coil Epic
|
Reply
|
|
by W5DXP on October 26, 2009
|
Mail this to a friend!
|
Dan, this is not a complicated problem. Either the coil has a ~3ns delay, as W8JI asserts, or it has a ~21.5 ns delay as suggested by the inductance calculator at:
http://hamwaves.com/antennas/inductance.html
Which is it? Those values are almost a magnitude apart. Who is closest? Hamwaves or W8JI?
--
73, Cecil, w5dxp.com
|
| |
|
RE: 75m Loading Coil Epic
|
Reply
|
|
by N3OX on October 26, 2009
|
Mail this to a friend!
|
Wait... I guess if we assume an infinite coil or the terminated one we can define it based on the speed of a monochromatic wave on that structure.
In that case, you're right and Tom is wrong.
You win that one. 21ns is what you'd measure if you could measure that.
Since the coil is highly dispersive, though, it doesn't mean anything for finite coils beyond the standing wave pattern it causes after steady state has set in and I think you'd have better luck convincing others if you were more careful to explain "delay" that way.
The current taper can be explained the way you explain it but you need to be careful how you put it so that you're saying something true and not something poorly defined. And I think you'd get more converts if you would explain how it replaces parasitic capacitance with a continuous self consistent version.
Apologies for the "Define Delay" snark, but you can be a real pain about this stuff which makes even me defensive.
Furthermore, if you're going to push "antenna as a transmission line" and the coil length+discontinuities as a model, you need to work lumped inductance in there as a point discontinuity. I think it can be done and will give the right answer and wraps everything up.
73
Dan
|
| |
|
RE: 75m Loading Coil Epic
|
Reply
|
|
by W5DXP on October 26, 2009
|
Mail this to a friend!
|
> N3OX wrote: Cecil, Define "Delay" <
Dan, I see you have been taking lessons from our past president, Bill Clinton. "It depends upon what the definition of 'is' is."
Why am I not surprised? Is there no length to which you guys will go to promote your myths and old wives' tales?
--
73, Cecil, w5dxp.com
|
| |
|
RE: 75m Loading Coil Epic
|
Reply
|
|
by W5DXP on October 27, 2009
|
Mail this to a friend!
|
My apologies for the tone of the above posting. I shouldn't have posted anything shortly after slicing my left pinkie finger with a butcher knife.
The delay through a coil (or through a wire) is the time it takes for a pure traveling wave to travel from one point to another point. In the case of the coil, that's from the bottom of the coil to the top.
--
73, Cecil, w5dxp.com
|
| |
|
RE: 75m Loading Coil Epic
|
Reply
|
|
by N3OX on October 28, 2009
|
Mail this to a friend!
|
"The delay through a coil (or through a wire) "
I would call it "on a coil" or "on a wire" ... that is, the wire or coil provides a boundary condition on EM energy ...
For the sake of others who may not be familiar with this:
For coil of the form of a bugcatcher type coil, several inches in diameter and wound with fairly tight pitch, the paper referenced in the first post shows that such a coil supports a mode of propagation of electromagnetic waves at a very slow pace compared to the free space speed.
This makes it something like a transmission line with very low velocity factor.
Unlike a usual transmission line, the speed depends highly on frequency (see my comments about dispersion above) but nonetheless if you set up an infinite coil of wire like this and looked at the phase speed of single-frequency waves traveling down it, you'd see it was much less than the speed of light in a vacuum.
This kind of "slow wave" behavior is commonly exploited using much larger helices, in traveling wave antennas such as the axial-mode helix and in things like traveling wave tube amplifiers.
These modes of operation are somewhat different than the slow waves on smaller coils. But nonetheless, you can get slow phase speeds for EM traveling modes at HF frequencies on what looks like an ordinary loading inductor.
And, indeed, many people would describe it as a loading inductor with a lot of "stray capacitance" and they would arrive at similar conclusions if they were well versed in properly imagining "stray capacitors" everywhere that the electric field was important on such a coil.
Since the phase speed of the traveling EM mode is low on such a coil, and wavelength = speed/frequency, you can see that the mode has a wavelength on the coil shorter than the free space wavelength by the same factor as the speed is slower. This is the same reason that coaxial cables have an "electrical length" that is shorter than the free space wavelength. If you look at a wave traveling down an infinite coax cable, the physical distance between crests will be shorter than it would be in free space.
This has consequences when devices are operated with standing waves, in particular, if you can actually fit a good fraction of a wave cycle on the physical length of a coil, there is significant current variation along its length. Indeed this is due to the action of displacement currents but is not entirely due to stray objects in the environment of the coil (though perturbations like nearby metal objects will indeed have a LARGE influence due to the strong electric fields).
It's straightforward to see this in antenna modeling software if you make a physical helix... you can see the current variation from turn to turn. If you have a helix that's very long in terms of the modal wavelength then you can get several crests and troughs of coil current.
If you build a coil such that the physical length admits one half cycle exactly, and feed that coil in the center, you've built a helically wound dipole. I believe this effect is detrimental to mobile antennas, however it would seem it would tend to be inevitable to a certain extent on the lower bands.
And this is where it comes back to the old debate: whether or not the current in and out of the ends of a loading coil is always the same.
The answer is "not necessarily," but to go beyond that requires careful definition of the exact problem and rational, non-confrontational discussion of the details of the causes of this.
And that's where it's easy to fall back into a semantic argument about delays, electrical length, the role of displacement currents...
All of that information is contained in a proper mathematical consideration of the problem. Displacement current, modal wavelengths, modal phase velocity: all of these things are describing properties of a solution to Maxwell's equations subject to the boundary conditions on the mobile antenna.
If we want to talk about *that* I'll be happy to ponder and discuss it in public. If we want to discuss the "reality" of the "electrical length" of slow coils compared to the "electrical length" of anything else... if we want to talk about the influence of the "environment" and the absolute necessity of stray objects to result in significant displacement current...
Well, with those things we're not going to get anywhere, as a look back at rec.radio.amateur.antenna on this subject is concerned.
This has been a strange and interesting phenomenon to observe and I think it can serve as a lesson when we're trying to understand electromagnetic devices. We pretty much know the answer. But in this case I feel like the poor answer quietly sat on the sidelines of this argument, timidly whispering as people shouted.
"excuse me, gentlemen... there's no need... fellows... there's erm.. no need to fight over me, I'm right here..."
And all the while people squinted down and pointed to pieces of the answer, sometimes correctly, sometimes incorrectly, but we're still not looking it in the eye.
I think it's important to remember that people can be various shades of right and wrong. That they can be totally right about one thing while being wrong about another. And working together to reconcile and reinforce the correct things instead of fighting over the ways in which the "opposition" is wrong is, I think, ultimately much more productive. In the end, the wrong ideas still get ironed out, but in the interim the discussion is fun, interesting, and educational instead of bitter and distracting. This fight has had a lot of people sticking to what they know and their own further hypotheses based on that (some of which are off base). It's had a lot of people doing experiments designed to reinforce their own argument instead of designed to TRY to get the system to do what the other guy says it would do.
Some have cried persecution at the hands of orthodoxy, some have labeled others crackpots, trolls...
If you look past all that, mix everyone's past arguments up in a blender and strain off the base bulls*** at the bottom, there's something pretty interesting going on. But it's easy to add more bulls*** when you're convinced that the other guy is fundamentally wrong and somehow dangerous because of it.
73
Dan
|
| |
|
RE: 75m Loading Coil Epic
|
Reply
|
|
by W5DXP on October 29, 2009
|
Mail this to a friend!
|
> N3OX wrote: And this is where it comes back to the old debate: whether or not the current in and out of the ends of a loading coil is always the same. <
With the lumped-circuit model, that is a presumption since the lumped inductor is assumed to exist at a point. But the lumped circuit model is known not to work for helical antennas, e.g. where the current at the bottom of the coil is maximum and the current at the top of the coil is zero. The question is: At what point does the lumped inductor model fail? Dr. Corum says that when phase is important, the lumped inductor model starts to fail when the coil is electrically around 15 degrees long. Of course, when one presumes a lumped inductor, one presumes that it is zero degrees long, by definition. With that presumption, one cannot know when the 15 degree electrical length is reached. To compound the problem, there is no appreciable phase shift in the total current on an electrical 1/4WL standing wave antenna either in the coil or in the wire (stinger).
One thing is for sure, however. An air core 75m mobile loading coil, e.g. a 75m Texas Bugcatcher coil is virtually always electrically longer than 15 degrees.
One interesting tidbit. One can assume that the forward current and reflected current are equal with lossless conditions existing including zero radiation and no stray capacitance. The forward current has the same amplitude at both ends of the coil. The reflected current has the same amplitude at both ends of the coil. Interestingly, the resulting total current waveform is very much like the current waveform in a real-world loading coil. This indicates that radiation, stray capacitance, and I^2*R losses are secondary effects. The primary effect is the phase difference between the forward wave and the reflected wave on the standing-wave antenna. The same thing is true of a transmission line.
--
73, Cecil, w5dxp.com
|
| |
|
RE: 75m Loading Coil Epic
|
Reply
|
|
by N3OX on October 29, 2009
|
Mail this to a friend!
|
"The question is: At what point does the lumped inductor model fail?"
It fails continuously starting at the point that the inductor has zero length :-)
All inductors have parasitic capacitance. Again, I'm accepting of that term even in the case where quantitatively you need to calculate the full electric field distribution. This may seem wrong if you've strong feelings about the physical reality of conceptual models, but I personally think that "parasitic capacitance" is still a useful term if it's not taken too literally. I think that's the objection that the Corum paper raises to the concept, that a coil is NOT well approximated by some finite network of lumped inductors and capacitors. That is a bad electrical engineering model. It is well approximated by some continuous network of infinitesimal capacitors and infinitesimal inductors... this is one place where semantic arguments get very heated. Someone can imagine little tiny capacitors and still understand what the distributed network does.
"One thing is for sure, however. An air core 75m mobile loading coil, e.g. a 75m Texas Bugcatcher coil is virtually always electrically longer than 15 degrees."
I think that's probably right. However, how much that matters to antenna design and performance is easy to overestimate or underestimate if you're not careful about the details. And I think it's been much too easy for many people to get wrapped up in their own point of view.
I still don't think the fact that one of these loading coils has a separate, independent "electrical length" as defined in isolation from an antenna makes that coil have any special properties in its contribution to the "electrical length" of a resonant mobile antenna other than being an interesting observation.
"Electrically Long" coils *will* have current taper. But at the same time I think it's important not to overestimate the importance of this model in predicting antenna performance or design, especially in a conceptual sense.
Conceptually, if you need a 100uH lumped inductor to load an antenna by inserting it at a point, you can use a wide range of physical realizations, some of which are VERY close to lumped (like toroids) and some of which are very far from lumped and have a long "electrical length" as set out here.
In all cases as viewed from the feedpoint, the coil MUST end up nearly being the equivalent of putting a lumped inductor having a reactance in the vicinity of +j*2*pi*f*100uH somewhere around the middle of the physical coil. This has important consequences to prediction of loading inductance. In particular, I think you get awfully close to the right answer for the coil needed in practical mobile antenna design even if you totally ignore what we're talking about here and use a conventional inductance calculator!
The current taper will modify the radiation resistance by a small amount, but nothing huge is going to happen. The coil radiates just like as if it's axial current profile was enforced on a straight wire of the same physical length as the coil. As you approach and cross self resonance, the current taper will get severe and then cause a phase reversal. That would be terrible for a mobile antenna, for sure.
When the "coil electrical length" is long, the current taper is severe, the coil will be extremely sensitive to external perturbation and it will be easy to make the current taper more severe in any real world measurement. Some people would probably attribute this entirely to external influence, and that would be off base. However, it would ALSO be off base to say that none of the influence was external... in other words, suggestions about hat placement, keeping the antenna away from the vehicle body, etc are all the same no matter which model you use. N7WS seems to have come to some of the same conclusions in the article you link from your website: the distributed coil model isn't really necessary even if it's somehow the "truth"
I think all these factors contribute to the duration and severity of this argument, because in the end, except for the current taper itself, the consequences of this model for actual 75m mobile antennas are pretty minimal, and the amount of current taper depends on the "electrical length" of the coil AND on external influence. And for better or worse, the electrical length of the coil is not very much related to the electrical length of the antenna. Yes, it can be considered "part" of the electrical length of the antenna, but so can a lumped load, and those two things can be chosen in whatever proportion you wish.
This aspect wasn't handled properly in this argument. You had people measuring and modeling and hypothesizing and you guys never sat down together and actually worked on exactly the same system.
I'm still uneasy about a lot of thoughts and interpretations of the models and data. I'm quite convinced that you can get lots of current taper in a loading coil that's a pretty good loading coil. I believe a well converged NEC-2 model. I wish I had software more suited for modeling such a thing than NEC-2 based software is.
But it does seem that it is possible to get *some* well converged models of these things in EZNEC. Maybe someday when I'm bored I'll build a physical realization of an antenna with lots of taper through the coil, set it up over tons of radials in a field, and take the hours and hours necessary to make a detailed current map using a W8JI design low capacitance current meter viewed through binoculars to avoid external perturbation.
It's too bad that no one did that yet, but I suppose that, in itself, is interesting.
A lot of the reason why I have persistent interest in this problem is that I know there's **one answer** and that humankind knows that answer already, but models, egos, and emotions have managed to keep our particular subset of people from agreeing on the answer.
Classical electromagnetism is a solved problem, yet we can still have a decade long argument that ends with hurt feelings, people that won't talk to each other anymore, and persistent disagreement over the problem? Oops.
73
Dan
|
| |
|
Email Subscription
You are not subscribed to this topic.
Subscribe!
My Subscriptions
Subscriptions Help
Check our help page for help using
Forum, or send questions, comments, or suggestions to the
Forum Manager.
|
|
|
