Antennas and radiation - some questions... / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Antennas and radiation - some questions...

previous 1 2 3 4

Move Thread

LAN_403

Chris Russell

Fri Dec 28 2007, 07:39PM

... not Russel!
Registered Member #1 Joined: Thu Jan 26 2006, 12:18AM
Location: Tempe, Arizona
Posts: 1052

Marko wrote ...

Oh..
SO I must say I'm back on the beginning now, with same questions.

Why does current flow into antenna at all?

You now say that is solely due to this capacitance?

Capacitance will grow with length, but I really thought it's something completely unimportant?

ANd what resonates when I *do* resonate the antenna, simply by setting it to correct length?

I can't comment this any further because I fell overwhelmed with information that looks self contradicting to me and I'm lost.

I'm embarrassed for bothering you guys this far and still getting nowhere. As of 3 AM I need to get sleep and then re-read all of this few times again.

Sorry,

Marko

Yes, the capacitance of the antenna is what causes RF current to flow. I'm sorry if that wasn't clear from the get-go. There's nothing special about an antenna in that regard.

Capacitance will increase with length. So will inductance. An antenna is considered resonant at a frequency when inductive reactance and capacitive reactance are equal. Adding capacitance hats or loading coils changes the resonant frequency of the antenna -- this is useful for coming up with an antenna that resonates at a much longer wavelength that its physical length would normally allow.

When you set the dipole to the correct "resonant" length, the inductive and capacitive reactances are equal. They cancel out, leaving behind a purely resistive impedance. Happily, for a thin wire antenna, the correct length is very near an actual half wavelength. However, factors such as ground conductivity, nearby trees, etc. often mean that extra capacitance or inductance is coupled into the antenna, and the length must be altered slightly to obtain best performance.

The information does appear to be conflicting, probably because as Neil said above, in practice, we use a series of approximations, depending on what we want to know. In some cases we may consider a dipole to act like a series RLC circuit. In others, we will consider it a transmission line. If you want to know every minute detail of what happens in and around an antenna while transmittig, we are probably not going to be able to answer your questions satisfactorily. We can give you approximations, which do not hold up for every case, or we can present general ideas, which don't help much in the real world. A proper discussion would start, I'd imagine, with Maxwell's equations, and build from there with several hundred pages of discussion and examples.

I definitely second what Neil said above -- learn by doing. I'd even suggest taking it one step further than simulations. Get your ham radio license, and learn by doing in the real world.

Dr. Slack

Sat Dec 29 2007, 08:55AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

The clue is in the words used. "Electrical Length" therefore has something to do with length, but it's qualified by the word Electrical, which identifies it as a model. This means it captures some aspects of the situtation under certain conditions, but is not fundamental. In fact almost any "adjective - noun" pair means that "noun" is not fundamentally "noun", which is why I don't like the term Surge Impedance in this thread, because it's like impedance, and then with qualifications.

Electrical length (usually) means the length of ideal transmission line that would be required to (usually) delay the propagtion of a step wavefront or rotate the phase of a sinewave by the same amount as the structure we are describing. So 1ft of plastic insulated coax will have an electrial length of somewhere between 1.5ft and 2ft (depending on the plastic). The electrical length of the CLC ladder on Shaun's post will be given by some appropraiate formula involving SQRT(func(L,C)) * number of sections, which will be valid up to some fraction of the freqeuncy where the structure becomes a low pass filter, and stops being good model of a transmission line.

The electrical length of an antenna, if made from stright wires in air, is not much different from its physical length. The electrical length of an unloaded TC secondary, whuich is in air but is by no means straight, is usually within a factor of 2 or 3 from the physical length of the wire (not the wound length of the secondary). Various theories say the capacitive top-load loading should ideally bring the frequency down to where the physical length of the wire is exactly 1/4*freq for best efficiency, I've not thought about the fundamentals enough to know whether I subscribe to this one. But capacitive toploading does improve spark length.

You'll often see a CB antenna on a car, with a centre-load coil, which is there to increase the inductance, therefore reduce the (excessive for driving with) physical length of the antenna for the same operation freqeuncy. But I'm not a ham, Chris or the others will have to expound the finer practical points of inductive base or centre loading for antennae. So physical length, capacitance to ground, capacitance between bits of the wire, loading inductance, and intrinisc inductance in the wire are all doing their bit to make the antenna resonate - that is provide a real feedpoint impedance. You could also build the same antenna longer with straight wire, or replace it with a RLC circuit of the right parameters, and the transmitter wouldn't notice any difference between them, as long as it operated over the narrow band of frequencies for whicht he three loads were equivalent.

Given the definition of antenna resonance as a real feedpioint impedance, most antennae will have several resonant frequencies. Take the model of an antenna as a bit of open circuit TX line, it will resonate at lambda/4, but also at 3*lambda/4, and every odd mulitple above that. As you increase the input frequency, the input impedance swings capactitive, then inductive, then capcitive again, crossing the real impedance axis several times. A stright wire antenna is rarely used at 3l/4 or higher, because it is much bigger than it needs to be. However, the loads on a loaded antenna are often sized to "multi-band" the antenna. As the inductance of a lumped coil changes at a different rate to the effective inductance (adjective-noun pair, beware) of the antenna wire, it's possible to shift the frequency of the next resonacne from 3lambda/4 to something else, so the ham can use one antenna to operate on several specific frequencies (a ham might provide specific examples of dual- or tri-band operation frequencies?)

Tesla secondaries also resonate at roughly 3x, 5x and odd x their fundamental frequency, but they are never used here as they are much less efficient at these higher frequncies, although they have been deomnstrated simply to prove the point.

It's really quite important to note that you start out with a mess of wires, and then drive it at various frequencies to see how it behaves, what the feedpoint impedance is. When it crosses the real axis, you define it to be resonant at that frequency. If its radiation resistance matches the feedpoint, it's called a good antenna, otherwise it's less so. The wire and its behaviour come first, then we model and approximate and describe, and get confused if we try to marry up several models, each of which are basically correct, but valid over different domains. Thin unifrom straight wires get called transmisiion lines, higher impedance ones made by coiling the wire up get called inductors, lower impedance made by shortening and widening the wires and putting them closer together get called capacitors. They are all on a continuum of behaviour. However, a really small coil of wire will behave enough like an ideal inductor for that to be a useful model for it, under certain restrictive ranges of frequency, and the same for the other components. So any bit of wire might be dewscribed as having effective inductance or cpacitance, or behaving like a transmission line. For the half-wave dipole, the bits of wire nearest the feedpoint can be regarded more or less as inductors as the voltage there is low and the current high, and the bits at the tips more or less capacitors to ground, as the voltage is high and the current low. But that LC approximation doesn't capture the 3lambda/4 behaviour, or indeed any behaviour away from the small range of frequencies around the fundamtental resonance.

Marko

Sun Dec 30 2007, 01:51AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Guys, I can't thank you enough for this!

I tend to make such a gross philosophy out of really simple things.

You guys know so much that it is very hard for you to get down to my level, and I'm very grateful for your effort.

So, fundamental resonant frequency of open transmission line, like antenna, is SAME thing as it's 1/4 wavelength frequency.

I still have one thing unclear, from my last post, (I'm sorry if I haven't figured it out from these two posts);

what is assumed for ''ideal'' dipole, with only it's length known, for it's fundamental-mode (LC) frequency to be *exactly* same as 1/2 wavelength frequency... you know what I mean?
All with vacuum E and u, assumed zero feedpoint size.

Would this just assume 1-dimensional conductors (infinite inductance and no capacitance) and thus be absurd for RLC approach?

The electrical length of an antenna, if made from stright wires in air, is not much different from its physical length. The electrical length of an unloaded TC secondary, whuich is in air but is by no means straight, is usually within a factor of 2 or 3 from the physical length of the wire (not the wound length of the secondary). Various theories say the capacitive top-load loading should ideally bring the frequency down to where the physical length of the wire is exactly 1/4*freq for best efficiency, I've not thought about the fundamentals enough to know whether I subscribe to this one. But capacitive toploading does improve spark length.

I may have unjustly dismissed this for TC's, maybe without figuring it's importance.

Secondary that would have it's quarter wave frequency equal to it's resonant frequency would be pretty stubby; still, toroid can set the frequency down to it.

So could I, if resonator really behaves like a transmission line, really have voltage peak lower that the top of resonator?

Finally it comes to famous Steve Conner's paradox, if I unwind the wire from a tipycal tesla resonator, and straight it out, it will have *lower* resonant frequency than the resonator, despite the wire appears to have much more capacitance and inductance (turns being closer together)?

How could this actually be resolved?

Does the increased capacitance in imagined 'parallel' LC components actually lower the overall resonant frequency?

I need to thank you guys once again for this... Reading all your long posts did take me some time but I enjoyed it.

Marko

CT2

Sun Dec 30 2007, 07:11AM

Registered Member #180 Joined: Thu Feb 16 2006, 02:12AM
Location: Ontario, Canada
Posts: 187

This is a very interesting thread, I'm going to have to read it all over to get all the information out of it.

So could I, if resonator really behaves like a transmission line, really have voltage peak lower that the top of resonator?

When I was playing around with my SSTC I burned the secondary up, so I was just messing around with it. I lashed up a DRSSTC with it, and I noticed it didn't make sparks from the toroid, but about half way down the secondary it made like 6" sparks right off the wire (not sharp points or broken pieces). No idea how it happened or what I did, I was kind of annoyed at it but I figured it was because I just threw it together. So I guess it is possible.

Chris Russell

Sun Dec 30 2007, 08:36PM

... not Russel!
Registered Member #1 Joined: Thu Jan 26 2006, 12:18AM
Location: Tempe, Arizona
Posts: 1052

wrote ...

I still have one thing unclear, from my last post, (I'm sorry if I haven't figured it out from these two posts);

what is assumed for ''ideal'' dipole, with only it's length known, for it's fundamental-mode (LC) frequency to be *exactly* same as 1/2 wavelength frequency... you know what I mean?
All with vacuum E and u, assumed zero feedpoint size.

Would this just assume 1-dimensional conductors (infinite inductance and no capacitance) and thus be absurd for RLC approach?

No, it just assumes a low-resistance wire whose thickness is insignificant with regards to wavelength, that the antenna is more than half a wavelength above ground, and that the area is free of conductors, wires, or other things the dipole might couple to. There wouldn't be a huge length difference for a 10MHz (30m wavelength) antenna between 18awg and 12awg wire, and both would be pretty close to the actual half-wave length. However, if you changed to 5-inch metal drain pipe, you would probably start seeing a bigger deflection from that length. Similarly, if you were building a wifi antenna, 2.45GHz, there would be a much more noticeable difference between the 18awg and 12awg wire.

Finally it comes to famous Steve Conner's paradox, if I unwind the wire from a tipycal tesla resonator, and straight it out, it will have *lower* resonant frequency than the resonator, despite the wire appears to have much more capacitance and inductance (turns being closer together)?

How could this actually be resolved?

Does the increased capacitance in imagined 'parallel' LC components actually lower the overall resonant frequency?

Just a guess here, as I am not much of a coiler, but I would assume that the capacitance of the topload and inter-turn capacitance is pretty insiginficant compared to what the capacitance of the antenna would be if all that wire were stretched out. I know from experience that removing 10 feet from an antenna and then baseloading it generally requires the use of significantly more than 10 feet of wire in the base loading coil. I would not be surprised to see the same applies to Tesla coils.

previous 1 2 3 4

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob