Talk:Near and far field

From Wikipedia, the free encyclopedia

This article contains material from the Federal Standard 1037C (in support of MIL-STD-188), which, as a work of the United States Government, is in the public domain.

[edit] Misleading

The text claims that the 1/r³ term is the electrostatic field term, and the 1/r² term is the induction term.

1. These should be called "magnetic" and "electric" terms; not electrostatic. Near and far field have no meaning in electrostatics
2. The terms switch meanings when you talk about a current loop. The assignments referred to in the text are only about dipoles

This article is confusing, otherwise, and there's no reason it couldn't be made accessible to laymen. — Omegatron 02:22, 25 November 2006 (UTC)

I have a BS in Physics and I have a copy of the Jackson text somewhere... I'll put this on my todo list. I'm not sure that this can be accessible to a true "layman", but it could be somewhat simpler. The main point of the article should be that the "near field" (in classical theory at least) is the part of an EM field (or also sound waves, etc) that does not propagate (though does theoretically extend "infinitely") and concentrates its energy nearby, whereas the "far" field allows the energy to propagate arbitrarily far, falling off as an inverse square of the distance (unless perfectly collimated, which I suppose is theoretically possible). - JustinWick 23:11, 7 December 2006 (UTC)

Also I would dispute your claim that near/far fields have no meaning in electrostatics. Consider an electrostatic dipole, or quadrapole, etc. The net "flux" through a sphere centered on a monopole is finite and constant, no matter how large or small the sphere is. Any higher order moment is necessarily neutral, rendering its net flux zero - this is, IMHO, a fundamental difference (see plasma physics for an example of this), as it allows electrostatic monopoles to feel interactions from arbitrarily far away, as the amoun tof available monopoles at any given distance would typically go up as the square of the distance, balancing out the weakening of the E field. Higher order moments only interact at short range (as they fall off too quickly to be offset by the greater abundance of sources as distance goes up). But if you can find something in Jackson that states right out that near/far field are purely electrodynamic concepts, I won't argue :) Gosh, it's been a long time since I thought about any of this - hope I make sense! - JustinWick 00:17, 8 December 2006 (UTC)

Hope you will but could you please read a bit of the text in Planck vs Stirling talk (just finished). Maybe too long but I guess it's for people who both like far/near fields and don't mind donuts (at least statistical donuts). To sum up the idea is that if you have:

a straight wire of 5 wavelengths grounded on both ends then you should have n=10 antinodes and with a high standing wave ratio this may be treated as an isolated system for "bench" (statistics). This could be one of the versions of the earliest statistical system Planck originally proposed for the black body radiation. Maybe the wire should be round the table though (read the talk there). If you suddenly disconnect one end from ground, but just to add one antinode, so that you have 11 antinodes, you will spoil the standing wave and some field should be radiated (like a droplet or a tornado in the examples there ) . Now the problem is that Planck wrote something about Stirling that nobody seems to have been able to understand so far. It is suggested that some higher Stirling terms should be somehow visible especially in the near field. Have you ever heard of anyone ever tried something like that? How to combine that with the near/far field? Would you mind leaving your comments here or there.--C. Trifle 09:57, 15 December 2006 (UTC)

I'm not sure that this can be accessible to a true "layman"

Anything can be accessible to anyone. Where "accessible" just means "they can figure out the context, and if they can't understand the details, they at least understand the basic idea, and know what prerequisites they'll need to research to fully understand it". Explaining practical applications helps make it accessible, too.

The main point of the article should be that the "near field" (in classical theory at least) is the part of an EM field (or also sound waves, etc) that does not propagate (though does theoretically extend "infinitely") and concentrates its energy nearby, whereas the "far" field allows the energy to propagate arbitrarily far, falling off as an inverse square of the distance

That's not very clear to me. Both "parts" of the EM field exist everywhere; from the source to infinity. It's just that one is stronger in one region and the other dominates in the other region, right? I wonder if a log-log plot would make this clearer... If it would make the differently powered terms into straight lines that intersect at the boundary between near and far, like the (unrelated except visually) frequency response plots I am more familiar with.

Consider an electrostatic dipole, or quadrapole, etc.

Hmmm... I'm not sure I understand that example. An electrostatic dipole would feel interactions from arbitrarily far away, too. — Omegatron 16:53, 15 December 2006 (UTC)

I wouldn't like to distract you but I am not quite sure if you see my point. It's about the near/far field and the series named after Stirling. There's the quantum theory which was introduced to science using the so-called "first order Stirling approximation" which was then used for at least 25 years by different writers.

The idea should be testable (and perhaps was tested ^{[citation needed]} I don't know) in antenna radiation. For example, if you look from the layman's point of view, there should be some phenomena that only depend on the number of quanta for big numbers n as well (not only for single photons). Quanta should also exist in UHF/VHF range if they do exist at all. So if you set a transmitter frequency to ν and the power level to P, say 110 MHz and 1W, you will have n=P/hν quanta per second, which is a certain very big number. For the chosen set you should have a certain radiation lobe pattern for the near field and usually a different pattern for the far field. Now if you divide the frequency by half and reduce the power level by half, and use an antenna twice as big, being the exact copy of the one you used before, you will have exactly the same number of quanta n and nothing should change in the far/near field lobe pattern. But if you change the power level for the same frequency, for example 1 W to 0.1 mW, there should be some effect related purely to the number. If there is no effect of any kind on the radiation field pattern depending on number n, one should conclude that the Stirling series is not related to radiation at all. Maybe the effect is only visible for very big differences in number n though, such as ten orders of magnitude or more, I don't know. Have you ever come across anything like that or should it rather be treated as a research topic to be proposed to some research institution? --C. Trifle 11:40, 18 December 2006 (UTC)

I looked at some graphs, one as old as 70 years (Terman), eg. Length of Wire in Wave Lengths , and some more recent. They give various radiation patterns but they were all practical measurements, accuracy is not stated. Here the problem is rather an exercise of different kind, a bit of philosophical (is h really a constant for radio frequency range and below, and are there separate quanta in that range. One can't say using those graphs.) This is last time from me about Stirling and antenna on this page (unless someone finds something really interesting).--C. Trifle 18:13, 18 December 2006 (UTC)

Sorry if I didn't really go into detail on which modes are "propagating" and which modes are "non propagating." Propagating modes are those in which the total energy of the field does not decrease with distance. Because we live in three regular spatial dimensions, a 1/(r^2) field will not decrease in total strength as the sphere of measurement gets larger (i.e. flux is conserved). So, waves resulting from the dipole effect allow their energy to extend arbitrarily far *without* the energy being reduced. The density is reduces, but the region it inhabits increases to counter this precisely. With other higher order moments, this is not true - the total energy of the field at a given distance goes down drastically with distance. That's why it's considered "bound" even though it does technically extend "infinitely" - JustinWick 18:46, 11 January 2007 (UTC)

Re the discussion of whether the terms near-field and far-field are purely dynamic, the issue needs more detailed explanation. The key distinction between near field and far field is that in the near field, the wavelengths are long compared to the structures of interest (say antennas, or other objects interacting with the field). Observe that Maxwell's equations have no lengthscale in them, so the definition of near- and far-field has to be relative to a particular physical structure of interest. So, the near field is defined as the region in which the electromagnetic wavelength is long compared to the lengthscale of the objects in the field. The near field regime can also be descibed as the "electro-quasi-static" or "magneto-quasi-static" regime. What this means is that even though we are in fact talking about dynamic fields & AC circuits, the behavior you get is the same as in the static case, in the following sense. In the near field regime, you can predict the values of current and voltage throughout the system using "lumped component" values of capacitance and inductance, which are determined by geometry and have the same values whether in the DC or quasi-static case. For example, suppose the circuit includes two plates, both much smaller than the wavelength. In the near field case, we know that the AC current through the pair of plates is given by I = 2 pi f C V. (We would use this same value of C to compute stored charge from voltage in the electrostatic case using the expression Q=CV.) If we were to cut the frequency (f) by a factor of 2, the current would be reduced by the same amount. On the other hand, if we are not in the near field case, antenna-type effects must be considered. It would be possible to cut the frequency by 2 and have the current out of the "transmit" plate go up, if the wavelength happened to come into the proper relationship with the plate size (so that it became a more efficient radiator). You can find a great discussion of near-field, far-field, and the quasi-static / AC Circuits regime in a book by Fano, Chu, and Adler. 192.52.57.33 18:58, 7 February 2007 (UTC)JoshSmith

Adhering to the J.D. Jackson text is a good idea. Two articles that exemplify what this page could communicate are still availible in Google's cash. http://216.239.51.104/search?q=cache:RB7gPXONO9oJ:journals.iranscience.net:800/www.conformity.com/www.conformity.com/0102reflections.html+%22near+and+far+fields%22&hl=en&ct=clnk&cd=16&gl=us

http://216.239.51.104/search?q=cache:BCTF_Wny7BwJ:www.sm.luth.se/~urban/master/Theory/3.html+near+far+fields+urban&hl=en&ct=clnk&cd=3&gl=us

It might be possible to get rights to the material through Urban Lundgren or Isadore Strauss or http://www.conformity.com which formerly hosted the Strauss article. Sue... suzysewnshow@yahoo.com.au 65.41.254.175 14:52, 20 June 2007 (UTC)

I have to agree with Omegatron that the terms Near-field and Far-field have no meaning in electrostatics, and as far as I am concerned, when dealing with magnetic fields, they are useless too. The unit of measurement for near and far fields is the wave length; since there are no wavelengths in static fields, how can you apply this unit of measurement. Even when we are dealing with quasi-static fields, who is to say that energy transfer occurs at the speed of light. Has anyone ever measured it? If someone reading this is aware of it, please let me know. The experts usually grow silent when asked this question. Steinhauer 22:21, 22 September 2007 (UTC)