Talk:Waveguide/Archive 1

Archive 1

"Guided Wave" Absolutely Should Not

"Guided Wave" Absolutely should not redirect here! A guided wave is not a waveguide - especially since whatever you think otherwise about a waveguide, it is technological creation made by human beings. In other words, it is artificial.
There are lots of cases of guided electromagnetic waves that use surfaces that were not created by people. Here are some of them:
We can set up a radio transmitter that transmits at a suitable frequency on one side of Lake Superior, Lake Huron, Lake Erie, or Lake Ontario (hereafter called the "Lake") and transmit radio waves over to the other side of the Lake (U.S.A. to Canada or vice-versa). If the frequency is not too high that wave will become attached to the surface of the water at its bottom end, since the water is somewhat conductive. Specifically, its electric field will be perpendicular to the surfact of the water and (for all practical purposes), its magnetic field will be perpendicular to both its magnetic field and to the direction of propagation. In this way, the radio wave will propagate across the Lake quite efficiently, with most of its power density concentrated rather close to the surface of the water, and not dispersing upwards toward the stratosphere and outer space.

A very good way to transmit these waves is with a vertical antenna: for example, one about 1/4 wavelength long, driven with one terminal at its bottom and the other one grounded. With such an antenna, the electic field of the radio wave is automatically vertical, and its magnetic field is automatically horizontal, which is just what we want.

During the 1930s and 40s, there was actually some transatlantic telephone service that did not, repeat, did not, use undersea cables. Such cables were impractical at the time because the electronics technology was not good enough. (The first undersea transaltantic telephone cable did not come into seriece until 1956.)
These systems used radio back and forth actross the Atantic. Electronics technology had not progressed enough to generate the high RF power levels necessary. What they did was manufacture high frequency alternators. The alternators had fast-spinning armatures and multiple magnetic poles to generate carrier frequencies in the range 10,000 to 100,000 hertz. (10 to 100 KHz). Amplitude modulation was impressed on the carrier wave, transmitted with a vertical antenna to get the electric field vertical. Antenna arrays were used to aim the transmitted EM waves toward the British Isles, or if on that side of the Atlantic, toward the U.S.A. and Canada. The distance was much farther, but signal propagation was aided by the fact ocean water is salty and a better conductor of electricity. Transmittter power levels could be as high as 200KW. The EM waves propagated as guided waves, with the lower end of their electric fields attached to the water. The EM waves followed the surface of the water right along the curvature of the earth, and in this way, they reached Europe, rather than propagating up to the ionosphere, to outer space, or whatever the undesirable direction was. So, these guided waves not only followed the curvature of th earth, but they also kept most of their power density at a fairly-low altitude where an antenna on the other side of the ocean could collect enough of the power.
It is not necessary to use water for the substance that guides the wave. In some other systems, soil is used - but moist soil is better, of course. This is where you find systems that use what is called the "ground wave". Those just get guided along the earth's surface, as "ground waves" rather than up into the sky . Ground waves are very important to AM radio broadcasting. How good they are depends a lot on how damp the spoil is and how flat the land is. I can remeber on a long auto trip that I took, and I passed through Kansas City on my way via Lincoln, Nebraska (to get on I-80), and then out to Cheyenne, Wyoming. It was broad daylight, and while I was in the K.C. area, I picked up the radio station KCMO. Well, I listened to KCMO all the way to the southwestern corner of Iowa, and then cut west across Iowa to get to Nebraska, and then I still listened to KCMO all the way out to Lincoln, loud and clear. There was no way that I could have done this without ground waves. I was way out of the line-of-sight from K.C., and in the AM band, sky waves are no good in the daytime. So, no line-of-sight waves, no sky waves, the only thing left was ground waves. I ate dinner in Lincoln, and then drove farther west, and I think that I was still listening to KCMO. 98.81.18.144 (talk)

IMHO, the only thing left is not just ground waves. Daytime skywaves in the AM band still exist. The layer of the atmosphere that waves bounce off of is just lower to the ground than at night, not "no good". The range would be less, of course, during the day rather than night, but it's not that strong a case that it had to be ground waves. --132.3.57.68 (talk) 16:23, 19 July 2011 (UTC)

The term "waveguide" refers to a system that guides waves. I've never heard of anyone in my field who restricts the term "waveguide" to structures created artificially. A naturally occurring waveguide is still a waveguide. — Steven G. Johnson (talk) 16:36, 19 July 2011 (UTC)

Transmission lines are not waveguides

TLs and WG work on different principles. For example: there's signal loss in TLs but no loss in waveguides. So how can they be the same thing?--Light current 00:04, 22 January 2006 (UTC)

No loss in waveguides?!! What?!! This is totally wrong. Think about the IR absorbtion bands in glass fiber which make periodic amplification necessary. Think about rectangular microwave waveguides at high power... with huge currents in the metal, problems with heat, and where adding internal silver plating greatly reduces the LOSSES. --Wjbeaty 07:07, 25 January 2006 (UTC)

Could you clarify, what is different about the operating principle of a transmission line from a waveguide? If there's a clear distinction to be made between WGs and TLs, I'd support simply removing most of the TL material from this article.

TLs operate on TEM mode. WG are TE or TM modes. TEM plane waves cannot propagagate in WG.--Light current 05:55, 23 January 2006 (UTC)

How do you classify microstrip lines? They also don't support TEM modes, but are often called "transmission lines". --The Photon 03:20, 24 January 2006 (UTC)

I classify them as TLs. Predominantly TEM propagation. See here [1]--Light current 03:35, 24 January 2006 (UTC)

The definition in Ramo, Whinnery, & Van Duzer (just got this from Amazon's "search inside") is "A waveguide is a structure, or part of a structure, that causes a wave to propagate in a chosen direction with some measure of confinement in the planes transverse to the direction of propagation." (p. 398) Their first example of a waveguide is "a transmission line." I'll stand on their authority to say that the established literature considers transmission lines to be a sub-category of waveguides, not a separate concept. -- The Photon 04:12, 24 January 2006 (UTC)

I dont agree. All the books I have differentiate between the two structures.--Light current 05:02, 24 January 2006 (UTC)

I'm suspicious. Which books specifically? Kraus "Electromagnetics" treats TLs and WGs the same. Here's an important point: he puts them in two classes: "hollow" waveguides, and the non-hollow ones such as coax, helical lines, etc. Perhaps you're definition of the word "Waveguide" resembles Kraus' definition of "hollow waveguide?" Certainly the coax (etc.) can behave as hollow waveguides when wavelength is short enough to approach the line's width, so that other modes become possible. But twinlead and coax guide EM waves, so they're waveguides regardless of modes or bandwidth limitation. Even in a flashlight, the energy flows as propagating EM fields in the space outside the metal. Think in terms of Poynting vector fields, and you'll see the identity between a flashlight's wave-guiding circuitry and a rectangular microwave waveguide. --Wjbeaty 07:07, 25 January 2006 (UTC)

FE Terman 4 th ed treats WG and TLs in separate chapters. Also, Electromagnetism by Grant& Philips (Physics Dept Manchester Uni) pub John Wiley, treats them differently.--Light current 18:45, 25 January 2006 (UTC)

A transmission line can act as a waveguide if excited above its WG cutoff frequency when different propagation modes are used. But TLs are overwhelmingly used in the TEM mode. I think you need to look at that book you quoted in detail (rather than just the abstract or snippet) to see what the author is really saying.--Light current 05:08, 24 January 2006 (UTC)

I am actually very familiar with the book [R, W, & vD]-- I used it as an undergrad. I mentioned its available on Amazon in case you want to check into it. Following up, I found that Federal Standard 1037C clearly uses "rectangular closed waveguides" as an example of a type of transmission lines. Finally, another highly respected textbook, Balanis' "Advanced Engineering Electromagnetics" regularly uses "transmission line" and "waveguide" as interchangeable terms.

Based on those sources, I can accept that TL are a subgroup of WGs, or that WG are a subgroup of TLs, or that the terms are interchangeable. I didn't find any source that defines a clear distinction between TLs and WGs. If you could cite a specific reference for your claim that WG and TL work on different principles, it might help me to understand where you're coming from.

The Photon 03:38, 25 January 2006 (UTC)

Well looking at it from that point of view, I suppose you could say that any conveyor of information is a 'transmision line'. If you want to go down that path, then I would say that a waveguide is part of the larger group called 'transmission lines'.

However, you have to be careful here, because one of the primary differences (apart from the modal one) is that WGs are narrowband structures (they have cutoff and go into different, unsatisfactory, propagation modes at higher frequencies) whereas TLs (coax etc) are generally thought of as wideband 'pipes' and in fact go right down to dc from say 20 GHz. WGs dont do dc or anywhere near. So, on an octave BW basis TLs are wider band than WG and I think this is one of the main differences. The other main difference is that WGs can do much higher frequencies with low attenuation than TLs can. Does this help at all?--Light current 05:15, 25 January 2006 (UTC)

Could you provide a citation for the claim that there's no loss in waveguides? If its intended to apply to real waveguides, I simply don't believe it. Even the longest-distance waveguides I know of, undersea fiber optics, have ~0.2 dB / km losses.

No citation because WG do have losses (my exaggeration- sorry). But much lower losses than TLs because the energy is bouncing off the walls. Highly conducting polished walls will give lower losses.--Light current 06:03, 23 January 2006 (UTC)

If the claim is intended to apply to the textbook "ideal lossless waveguide", then it doesn't create a distinction from the textbook "ideal lossless transmission line." -- The Photon 04:53, 23 January 2006 (UTC)

Light current is confused. Yes, one typically requires that a waveguide confine light losslessly in the limit of perfect structures made of lossless materials, but in this limit both TLs and e.g. dielectric waveguides (i.e. index-guided) are lossless. (However, even this restriction is commonly relaxed, since one sometimes considers "leaky" waveguides that have some small intrinsic loss even for lossless materials.) There is no particular restriction to TE, TM, or non-TEM modes (and in any case any waveguide with inhomogenous dielectrics does not generally have a simple classification by polarization). Saying that waveguides are only "narrow-band" is nonsense...for lossless materials, even an optical fiber would have guided modes at all frequencies, extending arbitrarily close to DC; you get a lower-frequency cutoff for non-coaxial hollow metallic waveguides or for e.g. index-guided dielectric waveguides sitting on a substrate. —Steven G. Johnson 17:45, 25 January 2006 (UTC)

I'm sorry, while waveguides and optical fiber have "modes" down to DC, these are no longer sinusoidal, these are cutoff waves with a exp(-x) decay, not a exp(-ix) propagation. The term for below-cutoff modes is "evanescent", and the guide is not usable at all approaching cut-off, let alone below cutoff.

Continuing to add my 2 cents, 2:1 coax is typically 50 to 75 ohms while typical waveguide (2:1) is over 300. The greater impedance is quite a good reason for the significantly reduced loss -- that, and the fact that coax requires some dielectric to support the center conductor and carry away the heat, air and vacuum "dielectric" are, in fact, lossless (but finite wall resistance does account for WG loss).

BTW In general, TLs use dielectric and WGs not.

As well, the article needs to have references to cutoff, impedance and wavelength equations. And TE, TM and TEM need to be defined. — Preceding unsigned comment added by Dgholstein (talk • contribs) 19:01, 28 July 2011 (UTC)

Responding to a five-year-stale discussion is seldom productive. Start a new section if there's something in the article that you are suggesting improvements to. Dicklyon (talk) 19:19, 28 July 2011 (UTC)

(My feeling is that this whole article needs to be substantially revised, in any case...it's a mishmash of information that doesn't clearly distinguish statements that are generally true from properties that are specific to particular waveguide geometries and materials. It should also be moved to Waveguide (electromagnetism) or similar, since it's perfectly possible to make waveguides for acoustic waves etc.) —Steven G. Johnson 17:45, 25 January 2006 (UTC)

Thanks, Steven and Wjbeaty for jumping in. I thought I might be kicking up dust where no-one really cared. I don't know if Light Current is confused or just biased (like all of us) toward the terminology of either his locality or his particular sub-field of expertise.

I agree the article generally needs a lot of work, but I don't think a title change is required. There are lots of examples where the most common usage of a word gets the main article of a particular title, and other less common usages have a disambiguation page to seperate them. For example, Laser and Optical. Only if someone would write an article that covers waveguides for all types of wave phenomena, then it ought to get the main article, with this article relegated to Waveguide (electromagnetism), alongside of Waveguide (optics), Waveguide (acoustics), Waveguide (fluid dynamics), etc.

The Photon 04:42, 26 January 2006 (UTC)

I will admit to being biased/ignorant about the terminology because I, like some others, have not covered the whole gamut of WG and TLs. In fact I have never bolted pieces of hollow WG simply becuase I have never been working at those frequencies. Neither have I jointed two pieces of optical cable. However, I do agree that the subject appears to be large enough that a split based on Stephens/Photons suggestions seems sensible. I will then restrict myself to editing the electromagnetic WG and TL pages.--Light current 17:49, 26 January 2006 (UTC)

My personal opinion concurs with the following extract that Stephen Johnson has found:

The term waveguide can be applied to all types of transmission lines in the sense that they are all used to guide energy from one point to another. However, usage has generally limited the term to mean a hollow metal tube or a dielectric transmission line. In this chapter, we use the term waveguide only to mean "hollow metal tube."

--Light current 23:36, 26 January 2006 (UTC)

European v US views on WGs/TLS

Before everyone gets too agitated about differences and similarities between WG and TLs, Im going to suggest that it seems to me there is is difference in view here between people on each side of the Atlantic due to the books they have been reading/studying.

It would seem that the USA view is that a TL is anything that can convey info: from a piece of wet string to an optical fibre and beyond. Over here (the old world), rightly or wrongly we have a more refined view of what constitutes a waveguide or a TL. So Im going to stick a globalise tag on the article just to remind people to be careful about representing a world view on this subject. --Light current 18:52, 25 January 2006 (UTC)

I regularly go to international conferences dealing with electromagnetic wave propagation, and I have never heard of the "refined" view of what constitutes a waveguide that you are referring to. Can you give a reference that clearly states that a TL is not a waveguide? (Historically, there were lots of analysis of TLs prior to the development of a more-general electromagnetic waveguide theory, but the existence of TL-specific terminology and analyses does not show that they are not considered "waveguides" by modern scientists.) —Steven G. Johnson 19:13, 25 January 2006 (UTC)

Are you saying that all WG are TLs, or all TLs are WG or what?--Light current 19:54, 25 January 2006 (UTC)

I mean that all TLs are waveguides. I wouldn't say the converse; in my experience the term "transmission line" seems to be mainly used for waveguides based on metallic components and operating at RF or microwave frequencies.

However, maybe I spoke too soon. It seems that there may be some differences in terminology, not so much between Europe and US, but between e.g. the optics and microwave communities. For example, I found this site (which seems to be from US Navy training material):

The term waveguide can be applied to all types of transmission lines in the sense that they are all used to guide energy from one point to another. However, usage has generally limited the term to mean a hollow metal tube or a dielectric transmission line. In this chapter, we use the term waveguide only to mean "hollow metal tube." It is interesting to note that the transmission of electromagnetic energy along a waveguide travels at a velocity somewhat slower than electromagnetic energy traveling through free space. [2]

—Steven G. Johnson 02:55, 26 January 2006 (UTC)

As a matter of encyclopedic style, it seems there's a preference for each article to take on as broad a meaning as possible. This would seem to favor waveguide being used in its broadest meaning: anything that guides waves. On the other hand, this can be taken too far. For example, see the mess that is created in Optical communications by the initial assertion that the term encompasses any communication using light as its "transmission medium". I'd like to clean that article up, but starting with the assumption that it should cover everything from aposematism to wavelength division multiplexing makes it a nightmare to contemplate.

In this case, where one local definition of the term (in the UK) has a more restrictive definition, and another (the US) has a broader definition, I'd suggest the article ought to be allowed to cover the broader field, with a note on the restrictive usage in particular regions or fields. The main thing really ought to be to actually fill out the article with some real content. Hopefully my next contribution will be on the article page, not the Talk page. The Photon 04:42, 26 January 2006 (UTC)

Fed Stan (1037C)Definition of waveguide

waveguide: A material medium that confines and guides a propagating electromagnetic wave. (188) Note 1: In the microwave regime, a waveguide normally consists of a hollow metallic conductor, usually rectangular, elliptical, or circular in cross section. This type of waveguide may, under certain conditions, contain a solid or gaseous dielectric material. Note 2: In the optical regime, a waveguide used as a long transmission line consists of a solid dielectric filament (optical fiber), usually circular in cross section. In integrated optical circuits an optical waveguide may consist of a thin dielectric film. Note 3: In the rf regime, ionized layers of the stratosphere and refractive surfaces of the troposphere may also act as a waveguide.

--Light current 23:28, 12 March 2006 (UTC)

Summary of nomenclature argument above

I wrote this out as part of another discussion, then realized it wasn't very useful there, but I didn't want to throw it away since I think it is helps to clarify the nomenclature discussion.

There are apparently three ways that the the categories of "waveguides" and "transmission lines" are defined:

• Waveguides are a type of transmission line. This view would come from looking at the function of the device, and this is the one used in 1037C. I don't know of anyone who uses this view, but I could imagine it being used in government contract bids or patent filings, for example.

• Transmission lines are a type of waveguide. This view comes from looking at the principle of operation: Solution of the electromagnetic wave equation with appropriate boundary conditions. This is the view used by American academics, for example in the Ramo, Whinnery, and Van Duzer book I've cited before.

• Transmission lines and waveguides are seperate non-overlapping categories. This view comes (I suspect) from practicality: it's more useful to have words that actually distinguish between the two concepts. This view seems to be used by UK academics -- for example Grant and Philips cited by LC, above.

I find the third view is used in practice by my colleagues in the fiber optics and high-speed electronics business, but when things get nasty we are likely to say "okay, so in this situation we have to consider all of the waveguide effects in this transmission line". On the other hand, the important thing here is, what gives the clearest explanation to the reader of an encyclopedia?

--The Photon 02:34, 13 March 2006 (UTC)

V useful summnary. Perhaps we could include it in the article under terminology or something!--Light current 03:30, 13 March 2006 (UTC)

Well, its hardly written in encyclopedic style... Also, this would dominate the article if inserted now, and I will stick to my proposition that an encyclopedia is about ideas, not words, so I'd find it distracting to have it stuck in there now. But it is GFDL, so I can't forbid it. -- The Photon 03:40, 13 March 2006 (UTC)

Well it would obviously need rewriting in an acceptable format. Perhaps it should go in the lede?--Light current 03:46, 13 March 2006 (UTC)

If you or someone else would like to re-write, it's GFDL. In my opinion, including the explanations it's too long for the lead. -- The Photon 04:11, 13 March 2006 (UTC)

The name "Waveguide" would suggest any structure capable of directing the propagation of electromagnetic waves (ie. anything that guides waves). A tranmission line directs the propagation of electromagnetic waves, but is specifically meant to propagate TEM waves. Therefore, isn't it logical to classify transmission lines as a subset of waveguides...? -- A Passing Student

Pleas refer to the above posts regarding this discussion.--Light current 02:14, 21 March 2006 (UTC)

other view

I came to understand transmission lines as being an idealization in circuit theory, therefore a circuit guiding waves "on paper", with the proper symbols for capacitors, coils etc. Waveguides in my opinion are the physical embodiment of such. I have no specific quote to support this, rather all references in the literature that I read conform to this view. Setreset (talk) 08:37, 24 December 2009 (UTC)

Distinguishing between Optical, RF, EM and "Dielectric" vs "Pipe" WGs

Hi all. I'm just here disambiguating "Beam" and this isn't my field but the WG article TITLEs are a little muddled. Might I suggest that:

• "Waveguide (optics)" be retitled "Waveguide (dielectric)" or similar
• "Waveguide (electromagnetism)" be retitled "Waveguide (hollow metal pipe :-)" or similar.

Reasons:

• It's all EM
• EM in fibers is often not visible
• Optical is also a general term refering to any EM bouncing around, e.g. GO, PO and BWGs (which don't even have a reference in WP!?)

I've drop this in Waveguide, WG(em) and WG(opt) as there doesn't seem to be much overlap between the "talkers". Enjoy, Saintrain 18:03, 12 February 2007 (UTC)

I also agree that the current categorization is very unsatisfactory. I don't think that the distinction should be between "dielectric" and "hollow metal", however. What about metallic waveguides with inhomogeneous dielectric interiors? Or metallic waveguides that are open? Or dielectric waveguides that use photonic band gaps rather than index-guiding (total internal reflection)? I would suggest, first there be a general article on electromagnetic waveguides, which can discuss the things that all EM waveguides have in common, regardless of the physical mechanism (e.g. give an overview of confinement mechanisms, the separable form of the eigenmodes, orthogonality relations, waveguide dispersion and group velocity, etcetera). Then separate articles divided according to the different physical mechanisms: closed metallic waveguides (waveguides completely surrounded by metal), index-guiding (a.k.a. total internal reflection in the ray-optics limit, probably with a sub-article on conventional optical fibers), surface plasmons (open metallic waveguides with imperfect metals, although this should arguably be a sub-article of index-guiding since the mechanism is similar mathematically), photonic band gaps, hybrids of gaps+index guiding, and there are a few oddball cases as well. As it stands, the Wikipedia articles reflect rather narrow perspectives. —Steven G. Johnson 21:09, 12 February 2007 (UTC)

It is a matter of terminology as evident from the long discussions of the past. My personal optinion is to use the word waveguide for any structure which confines light in 1 or 2 dimensinos not allowing it to diffract. Such confinement is always done by some "mirror". I suggest to right a general article about waveguides, distinguidhing the different geometries (slab, channel, fiber) and different mechanisms of guiding (mirrors) i.e total internal reflection, metalic, or Bragg reflection (photonic crystal waveguides).Eranus (talk) 13:47, 2 June 2009 (UTC)

Hebrew translation

I wrote the Hebrew version of waveguide which now contains lots of material that isn't found anywhere else on wikipedia, someone might want to give it a look (the Hebrew would be difficult but there are formulae and images). I'm not doing it myself since I don't want to step on any toes, all content was removed from this article, and divided into 5 different ones, for some purpose to me unknown. Setreset (talk) 08:37, 24 December 2009 (UTC)

I have begun translating. Setreset (talk) 18:07, 5 January 2010 (UTC)

Stopped translating due to contribution called "dubious" (see below) and lack of support. I would strongly suggest to change the notation of dubious to notation of "not representing a wide view of subject", based on the clarification of the user who posted the "dubious" tag. Setreset (talk) 11:11, 20 January 2010 (UTC)

Ostensive definition required

The article needs an ostensive definition in the format, "x is a waveguide because it satisfies X requirements. y is not because it does not satisfy Y requirements. LokiClock (talk) 01:10, 27 December 2009 (UTC)

Article split

I dug in the history of this article, and found that the article was split and practically deleted by User:Light current who was repeatedly blocked until permanent ban. This might be one of his early and undiscovered works. Setreset (talk) 09:09, 7 January 2010 (UTC)

Dubious additions

Hi, I marked some of the recent additions as dubious because they reflect a very limited perspective of waveguides.

In the "theoretical analysis" intro:

• The "rule of thumb" that "the longer the wavelength, the wider the waveguide has to be" is technically mainly for closed waveguides (although with exceptions such as coaxial waveguides). For example, a dielectric waveguide formed by a higher-index region surrounded by (infinite) lower-index regions has no cutoffs. For photonic-bandgap guidance, the guidance bandwidth is restricted by the photonic bandgap's bandwidth; no matter how big of small the waveguide is, it will never guide outside the gap (at least, not by that mechanism).

The "propagation modes and cutoff frequencies" section has multiple problems:

• A plane wave solution per se is not a waveguide mode in the usual sense, it is a solution in a homogeneous medium that fills all space, and hence has no confinement...unless you consider it as the asymptotic limit of a waveguide of infinite size. I think what the article is trying to get at is waveguide modes that can be locally described in terms of plane waves (usually more than one), but that is a very different thing. The confusion between homogeneous media and waveguides permeates this whole section.

• The "speed" of the wave that is given here is the phase velocity, which is problematic for many reasons (and is not even uniquely defined in an axially periodic waveguide).

• In general, a waveguide mode can be described by an axial propagation constant β (which is only defined modulo 2π/a in a waveguide with period a, but is well-defined in a constant-crosssection waveguide), but it has no three-dimensional wavevector k. (If you Fourier-transform the waveguide mode in the transverse directions x and y, you will get many Fourier components, not a single kx and ky. More abstractly, the existence of a well-defined propagation constant in a given direction is a consequence of translational symmetry, and there is no translational symmetry in the transverse directions.)

• There are waveguide modes with a lower-frequency cutoff, waveguide modes with an upper cutoff but no lower cutoff, and waveguide modes with both upper cutoff and lower cutoff. The section makes it sound like there are only lower cutoffs.

The "impedance-matching" section is misleading, because the ability to describe a waveguide mode by a single "impedance" number that completely characterizes the reflection at a junction between two waveguides is only true in effectively one-dimensional systems. Essentially, it requires that the waveguide mode pattern have only a single degree of freedom, e.g. be described (locally) everywhere by a plane wave at a fixed angle and hence a single ratio of the transverse E to H field magnitudes. For general waveguides, e.g. an inhomogeneous dielectric waveguide, this is not the case (except in certain limits, such as very low index contrast as applies to standard optical fibers). For more general waveguides, "impedance matching" is at best a heuristic description of what happens at a junction.

There is certainly a place for these kinds of approximations, stated carefully, but not in a general theoretical description of all possible waveguides.

— Steven G. Johnson (talk) 21:18, 7 January 2010 (UTC)

reply:

As these comments may lead to another user to simply delete all the so-called dubious additions, and I would not like that to happen, please note the following:

• Wide view does not mean that only general sentences are to be written; rather, all views should be represented. Indeed I do not have a complete view of the subject, but I do have a view of acoustic as well as EM simple waveguides, theoretically as well as experimentally, in science not in applications and systems. To limit the article only to universally true sentences will reduce it to a few poor sentences, as it was before. The special cases of rectangular and cylindrical hollow/dielectric WG are important and should be treated. Other issues which I'm not acquainted with should of course be treated.
• My contributions are sourced. the sources are noted and found full-text online. especially Balanis's book is useful in confirming them.
• If any affermation is limited in view, try to point out its limit in the text, not deleting it. For example, the wavenumber in the z direction also called beta, is right for constant cross sections, which is an important case. Maybe the whole theoretical analysis could be limited to "constant cross section theoretical analysis".
• Cutoff may be referred also as upper cutoff, but not by Balanis, who simply writes "cutoff" for lower cutoff.
• a dielectric waveguide is a good example of an exception to the rule of thumb, which should be added to the article. by definition a rule of thumb has exceptions. It is also noted in the preface, and I think it is good for giving the reader some idea of waveguide dimensions.
• A TEM mode in a coaxial cable is a plane wave, in the sense that the wavefront is plane, although it is not a free-space plane wave as the power changes with radius. It is not an approximation.
• kx and ky are useful in hollow pipe analysis. They might not be for a general case.
• Impedance matching is right I think for a single mode. Again, it gives the reader some idea of what we're about. It is a major issue in radar systems design.

Setreset (talk) 00:25, 8 January 2010 (UTC)

recent edits - removed information

Recent anonymous edits removed much information, and restricted the waveguide to its narrowest meaning. I tried in the past to improve the article by adding information, but failed. It seems that for some reason this article always will tend towards becoming a stub. I removed it from my watchlist and will not trouble you further. Setreset (talk) 08:42, 3 August 2010 (UTC)

Any Consensus?

It's a shame that the article is still plagued by a lack of consensus regarding the terminology of "transmission line" vs. "waveguide". Some are bothered by the fact that the present waveguide article uses the term "waveguide" in the narrowest (but most common) sense, namely a hollow metal pipe waveguide. They wish it to be more clear that the term waveguide is more general. On the other hand, if one goes to the "transmission line" Wikipedia webpage, it is interesting to note that this page initially defines transmission lines in a more general sense, so that waveguides would be included as part of this set! However, the article on transmisson lines then proceeds to discuss the theory for transmission lines as they are commonly thought of in the narrowest sense -- a two-conductor system supporting a TEM mode (which can be modeled using the Telegrapher's equations). It would be good for these two articles to be coordinated with the definitions.

In my own personal experience, I have always used "transmission line" to mean a two-conductor system, such as a coaxial cable or a twin-lead line, or a microstrip line. I have usually used "waveguide" to mean a hollow metal pipe waveguide, but I also realize that the term 'waveguide" can be more general. For example, one can have a "dielectric rod waveguide". One can also have a "coplanar waveguide," which in essence is simply a transmission line (albeit not with an exactly TEM mode, but a quasi-TEM mode).

I think it is relatively uncommon to use the term "transmission line" to mean something other than a two-conductor system. Presumabley a "transmimssion line" is something that obeys "transmission line theory", and this theory is all based on the theory of a two-conductor system (i.e., the Telegrapher's equations).

I would propose that in the waveguide article continue to state, as it does now, that the term waveguide is general, but that usually in common usage it means a hollow metal pipe. The article could then state that further disacussion in the article will be confined to this special class of waveguides. A more detailed discussion of other classes of waveguides could then be found in the links to other webpages. I would also propose that the article state in the beginnning that transmission lines are a special type of waveguide that consist of two parallel conductors, which is capable of supporting the TEM mode (or quasi-TEM mode if inhomogeneously filled) and which obeys (either exactly or approximately) the Telegrapher's equations. The waveguide article would then give a link to the transmisson-line webpage.

I would propose that the transmisson line article state that the term 'transmission line" can sometmes be a general one, but that normally the term is restricted to mean a two-conductor waveguiding system that can support a TEM (or quasi TEM) mode, which obeys the Telegrapher's equations. It should then state clearly that all further analysis in the article will be restricted to this type of system. —Preceding unsigned comment added by Elee1l5 (talk • contribs) 21:53, 7 October 2010 (UTC)

There's more than enough that could be said about waveguides -in the narrowest sense of the word- to make a pretty long article, and likewise for transmission lines. It should be made clear however, that each term can be used in a wider sense (even (if it is really so), that some (misguided) people count waveguides as a type of transmission line (rather than vice versa)). So yes, I agree with you proposed arrangement. The two articles do not seem to me too far from this ideal state of affairs at present though. See also this discussion. --catslash (talk) 17:56, 8 October 2010 (UTC)

Common usage in what field? In microwave, RF, and similar long-wavelength electrical engineering, maybe, but at visible and infrared wavelengths metals are very lossy, and people are usually talking about some form of dielectric waveguides when they talk about "waveguides". I continue to think that the organization of this article is fundamentally wrongheaded. It should:

• Define waveguides in general (any structures/materials that confine waves in one or more lateral directions, in any wave equation with translational symmetry in the propagation direction).
• Give a few examples showing the breadth of the concept: hollow pipes, optical fibers, acoustic waveguides, quantum wires.
• Break things up into subarticles by wave equation (electromagnetism vs. other wave equations like acoustic or Schrodinger) and by the confinement mechanism (e.g. in electromagnetism: metallic boundaries, total internal reflection / index guiding, surface plasmons, photonic band gaps, ...).

Having the main article on one specific sub-category just because that is what circuit people know best doesn't make sense to me. — Steven G. Johnson (talk) 18:38, 8 October 2010 (UTC)

I've just realized that in writing my previous comments, I was confusing this article with Waveguide (electromagnetism) (which might be better named Waveguide (microwave)). --catslash (talk) 18:56, 8 October 2010 (UTC)

Thank you for pointing out this other webpage -- I didn't realize (until I read more carefully later on in the article) that there was this other webpage for electromagnetic waveguides (I agree, it probably should be called Waveguide (microwave)). In view of this, I agree with Steven, the present article should not be restricted at all to microwave waveguides (metal-pipe waveguides) but should overview all of them. However, the article should make this aspect very clear at the beginning, and give a link to the microwave waveguide webpage early on, to minimize confusion between the two webpages. If this is the intent, however, to keep the present webpage focused on more general waveguides, then some of the later material in it should probably be moved to the other webpage (on microwave waveguides) or modified in some way -- for example, the discussion on impedance matching. I believe that this is what Steven was saying early on. Depending on how general the present webpage is intended to be, the structures do not have to even be limited to those that are translationally invariant. Certainly many waveguiding structures are periodic structures. This aspect is not discussed at all. I would also still recommend the minor clarification edit to the transmmission line webpage.Elee1l5 (talk) 00:21, 9 October 2010 (UTC)

(Note that I didn't say "translationally invariant," I said "translational symmetry". A periodic structure has discrete translational symmetry. Translational symmetry, whether discrete or continuous, is required because it is the source of the fact that there is a well-defined propagation without scattering; essentially, it gives rise to a conserved "momentum". — Steven G. Johnson (talk) 04:24, 9 October 2010 (UTC))

Okay, that's fine. A periodic structure (discrete translational symmetry) will support guided waves in general, and the definition of a waveguide can be taken as a structure with such symmetry. Maybe we shouldn't say that such symmetry is required, though, in order to have a guided wave existing on a particular structure. There could be cases, for example, where the structure is not physically periodic, but electrically it behaves as a periodic one at some frequency due to the nature of the discontinuities, and thus it supports a guided wave at this frequency. Of course there are also trivial cases where the structure is not periodic but the fields are zero in the region where it is not periodic, so a guided wave exists. Elee1l5 (talk) 19:35, 11 October 2010 (UTC)

Yes, there are a few counter-examples: structures where the field is excluded from the non-periodic region by a metal, structures where the non-periodicity is negligible, and even structures where the non-periodicity is perfectly "impedance-matched" (an often misleading term outside of one dimension) so that it is effectively invisible to the wave (which is only possible with certain materials, and is nearly impossible in practice for 3d high-contrast dielectric systems). However, these are the exceptions that prove the rule: the only allowable non-periodicity is "invisible" non-periodicity. Periodicity is the source of the conserved wave momentum and the lack of scattering (e.g. backwards) and is integral to the notion of a waveguide mode — Steven G. Johnson (talk) 05:20, 12 October 2010 (UTC)

Another interesting class of guided waves are "leaky waves," which radiate (leak power) as they propagate. They can exist on both uniform and periodic structures. In microwave terminology, a guided wave on an open structure (e.g. microstrip line) is "bound" if there is no radiation, while a leaky mode is fundamentally unbound. Normally when most people use the term "guided wave" I believe that they mean either a wave inside of a closed structure (e.g., rectangular waveguide) or a bound mode on an open structure (e.g., the HE11 mode on a cylindrical dielectric rod, or the quasi-TEM mode on a microstrip line). I'm not sure if we wish to broaden the discussion of guided waves to include leaky waves.

Back to the main issue -- editing the webpages. I am recommending minor edits to the Transmission Line and Waveguide (electromagnetism) pages, including changing the name to Waveguide (Microwave) as suggested by Catslash. For the Waveguide webpage, I suggest that you, Steven, might be the best candidate to make the changes, if you accept such a duty. Probably the highest priority, in my opinion, is to address the sections where the content is labeled as being in dispute. Do we need some kind of a consensus before proceeding? Elee1l5 (talk) 22:11, 12 October 2010 (UTC)

In general I agree with Steven Johnson's suggestion,

• Define waveguides in general (any structures/materials that confine waves in one or more lateral directions, in any wave equation with translational symmetry in the propagation direction).
• Give a few examples showing the breadth of the concept: hollow pipes, optical fibers, acoustic waveguides, quantum wires.
• Break things up into subarticles by wave equation (electromagnetism vs. other wave equations like acoustic or Schrodinger) and by the confinement mechanism (e.g. in electromagnetism: metallic boundaries, total internal reflection / index guiding, surface plasmons, photonic band gaps, ...).

But rather than breaking the article up by wave equation, which has little meaning to most people, does it make sense to refer to applications instead? Optical waveguides fall under EM but they deserve a separate subarticle because there's a wealth of applications and technologies. 206.47.231.164 (talk) 18:36, 26 April 2011 (UTC)

This needs to be fixed. Yes coaxial cables, and other transmission lines are waveguides. The article could explain this early on, and then mention that the WP:COMMMONAME use is for hollow waveguides. Yes squares are also rectangles, but we commonly use the word for things that are not squares. (Hopefully the article on rectangles gets this right.) Gah4 (talk) 20:31, 14 August 2020 (UTC)

Sad situation

This whole article is in very very poor shape, one of the worst I have come accross in Wikipedia. Quality is naturally in contrast to the length (not depth) of discussion here. It is such a fundemental component and intersting and important transmission line type. Why folks who have very little to contribute, marginally informed, take it upon themselves to edit articles I will never know. As someone who has actually designed and built transmission lines of all kinds, including waveguides, and who makes a living engineering such circuits, I would like to clean things up a little but I am afraid of wasting effort given the dedication of non-contributers to non-contribute here. What does impedance matching have anything to do with waveguides? How can an artcicle like this not include waveguide sizes and standards and frequency bands? Usually copper? Even a freshman EM book has a better explanation of propagation properties. How about countless waveguide components, from filters, to couplers, to magic tees, to loads, irises, transformers, etc.? I suppose all were busy dissecting "transmission lines"... —Preceding unsigned comment added by 173.77.162.40 (talk) 05:06, 24 February 2011 (UTC)

You may want to take a look at Waveguide (electromagnetism). --catslash (talk) 10:26, 24 February 2011 (UTC)

microwave ovens

a bit should be added about how the holes in the microwave oven cannot contain the frequency (a waveguide not tuned to the freq acts as a reflector), and removes most of the radiation whilst letting shorter wavelengths through like light. i do not know the term for this kind of blocking.

i wrote "most" because of the article written called "light sneaks through small holes", it can get through when its 7% of a wavelength wide. i also read that if the hole is less then 100% of a wavelength then you get attenuation. perhaps show a graph of how much gets trough depending on % of wavelength. — Preceding unsigned comment added by Charlieb000 (talk • contribs) 02:24, 9 October 2011 (UTC)

"Not tuned to the frequency" is not quite right. A waveguide is not a resonant cavity; it guides a continuous range of frequencies, not just a single one or a discrete set. However, a hollow metallic tube does not guide waves with a wavelength greater than a "cutoff" proportional to the diameter; longer wavelengths are exponentially attenuated "evanescent waves" within the tube. In a microwave oven, the operating wavelength is far greater than the cutoff wavelength of the holes, and so the exponential attenuation is very rapid within the holes; hence the escaped energy is very small (and could easily be calculated, but that calculation seems far beyond the scope of a generic article on waveguides). — Steven G. Johnson (talk) 03:48, 9 October 2011 (UTC)

This whole article is messed up. — Preceding unsigned comment added by 138.163.0.42 (talk) 00:03, 1 July 2014 (UTC)

Mathematical Waveguides Clarification

The section on mathematical waveguides references something called a "bound state", but without enough information for an inexperienced layman (me) to figure out what that means. The terminology does not appear in the 'main article' for the section, either. KE7KTO (talk) 20:58, 15 May 2017 (UTC)

August 2017 merge

Waveguide rotary joint is a single paragraph which could benefit by moving here to give it context. --Wtshymanski (talk) 20:53, 6 August 2017 (UTC)

I'm not so sure about that. That seems like excessive detail to be putting in this top level article. Also, don't assume that because the article is currently a single paragraph it will always be so. You might think there is not much scope for expansion, but take a look at what user:Catslash has done with waveguide flange. That's turned out a decent article on an equally low-level waveguide part. SpinningSpark 13:37, 7 August 2017 (UTC)

There is plenty that could be written about waveguide rotary junctions (though I'm not volunteering), so on the basis of WP:MERGEREASON:Merging-should-be-avoided-if (2), I would oppose merger. (Thank you for your kind comments SpinningSpark! - these didn't sway my opinion, as I'd already made up my mind.) --catslash (talk) 18:48, 7 August 2017 (UTC)

Untagged. Looking forward to expansion. --Wtshymanski (talk) 03:14, 19 September 2017 (UTC)

Short description

Yet another SD that has been reduced from something that makes sense to something entirely useless. Blind rule-following is so much more important than helping the reader. SpinningSpark 10:44, 8 June 2022 (UTC)

Is there any use for short descriptions, aside from keeping us editors happily busy? If the title isn't clear, then the short description is already in trouble because it has to fit in.. 40?...32?.. 20?.. characters, unlike a title. Hey, let's just replace all the short descriptions with "Thing". --Wtshymanski (talk) 03:41, 15 June 2022 (UTC)

It's something that shows up when searching in mobile view (which the majority of users are on nowadays). For instance if you click here and type "rabbit" in the search box (without pressing enter) you will get a list of article titles along with their SD and lead image. The SD is meant to tell you what the context of the article is (rather than an exact definition). In the "rabbit" case that is entirely self evident, but something like "Kirchhoffs law" is not, except to the initiated (and on top of that it is ambiguous). I get the point of SDs, but people are trying to abbreviate them to the point where they are telling you nothing at all.

SDs show up in other views in mobile, but I don't know how to get to them because I don't use it much. The 40 character limit is apparently due to the inability of the devs to make it wrap in these views. Something people are using SDs for which does show up in desktop view is a template that automatically adds the SD to "See also" entries. I don't approve of that practice – the relevance of a see also is highly dependent on the specific article so any comment, if needed at all, will need to be hand-crafted. The ghastly Wikidata also makes use of SDs. SpinningSpark 08:46, 15 June 2022 (UTC)

Oh. Not a mobile phone person, so have no exposure to this problem. I wonder what happens when article titles ( like Compagnie française du télégraphe de Paris à New-York ) exceed 40 characters? You know, I remember the days when we were told "don't worry about the technology" and were told just to write the best we could. Paper encyclopedias don't give any context and just assume the title describes the article. How much hand-holding and second-guessing does a user need ..."when you typed in "rabbit", did you really mean (insert list of wrong things here)...."? --Wtshymanski (talk) 22:57, 15 June 2022 (UTC)

Article as displayed on IOS

Some time ago, I was offered this image as an example of where exceeding 40 characters might cause a problem. However, this is an IOS mobile view and my phone is Android. It's laughable that this view is entirely different to my Android phone view and to the view I get when switching to mobile view in desktop. Why have the devs made it so different and doubled the amount of work they have to do to maintain it? The SD is displayed immediately below the article title in this view. I'm told the SD won't wrap, so is truncated at a point depending on the width of the screen. But since I don't have an IOS device I can't confirm that and I can't say whether or not your long article title will wrap or not. It's certainly not a problem on Android which wraps just fine and does not display the SD within the article in any case. IMO, if there is a problem with IOS not wrapping, the devs should fix it rather having hundreds of editors running around butchering perfectly good text to make it fit into a view they can't actually see. SpinningSpark 09:39, 16 June 2022 (UTC)

I've started a central discussion at Wikipedia_talk:Short_description#Too-short_descriptions ~Kvng (talk) 15:44, 20 June 2022 (UTC)

Per WP:SDCONTENT, Editors should bear in mind that short descriptions are not intended to define the subject of the article. Rather, they provide a very brief indication of the field that is covered, a short descriptive annotation, and a disambiguation in searches. The SD for this article, for example, could be "Type of telecommunications equipment". – Jonesey95 (talk) 16:15, 20 June 2022 (UTC)

But many waveguides are not used for telecommunication, but RF power transport. Gah4 (talk) 18:09, 20 June 2022 (UTC)

Yes, quite easy to think of numerous examples that are not telecomms. The light pipes used to transport the light from LEDs on a PCB to a display panel for instance. Or the guides used in loudspeaker cabinets. SpinningSpark 09:39, 22 June 2022 (UTC)

Jonesey, your suggestion (besides being inaccurate) is just the kind of vague SD I am railing against. To say to me "telecommunications equipment" without context I would immediately think of a telecom tower like this one or this one. Admittedly, they have waveguides inside them but that is beside the point. The SD should help the reader decide if they have found the article they are looking for. In that respect they serve a similar purpose to disambiguation pages and should have similar text. I cannot imagine that I would ever write your suggested text on a dab page. We can be more helpful than that. Yes, SDs are not meant to be definitions, but they are not meant to be top-level categories either. That's the job of the category system. SpinningSpark 09:52, 22 June 2022 (UTC)

elongate

What is elongate? The waveguide is longer than it is wide? They don't have to be, though that would be usual. Gah4 (talk) 04:28, 22 June 2022 (UTC)

Yes, that would be usual, as you say. There are of course exceptions, but bear in mind that the short description is not a definition and is not intended to capture each and every aspect of the subject. Feel free to revert if you want. MichaelMaggs (talk) 08:05, 22 June 2022 (UTC)

"Elongate" is a grammatical error. I've corrected it to "elongated". SpinningSpark 09:40, 22 June 2022 (UTC)

Have you never come across "elongate" as an adjective? Any decent dictionary includes it. MichaelMaggs (talk) 10:58, 22 June 2022 (UTC)