Talk:Standing wave ratio

Reflection coefficient is complex

Note to ssd: I added a mention that the reflection coefficient is complex. Perhaps there is more we could say about this quantity, but I suggest that extra information goes in to the reflection coefficient article, not this one. Unless, that is, you care to explain how the phase of ρ can affect the SWR. That is beyond my understanding, but you might like to have a go. -- Heron 12:14, 24 Jul 2004 (UTC)

Probably mentioning phase angle is enough, although one of these articles (perhaps not this one) should mention complex conjugate impedance match. Actually,now that I think about it, that would best go in an article on antenna tuner which does not yet exist. As you said before, probably the equations for resistive loss and dielectric loss should go in the transmission line article where it talks about decibels per foot. --ssd 17:01, 24 Jul 2004 (UTC)

The image "Standing wave ratio for ρ = 0.5" is useful, but could someone edit out the microsoft office grammar check green squiggly underlines in the axis labels? — Preceding unsigned comment added by Jethrotull4321 (talk • contribs) 16 aug 2006 01:02‎

radiation burn

Abnormally high voltages in the antenna system increase the chance of accidental radiation burn if someone touches the antenna during transmission.

Here's my understanding:

• antennas do not produce ionizing radiation, which corresponds to UV upwards.
• radiation burn is caused by ionizing radiation.

Could someone who knows these things fix up one of these articles: Standing wave ratio, radiation burn, ionizing radiation please? njh 23:11, 2 January 2006 (UTC)

The EM Radhaz article notes that a burn can result from touching an antenna. I guess that assumes that the power/voltage is high enough. See also[1]

The sentence quoted probably isn't related to the SWR article, probably should be removed. I'm not sure what it means by abnormally high, 2 w from a 1 w system would be abnormally high, but not likely to cause a burn. Maybe it's there because if you touch an antenna it will create high VSWR? Maybe just delete it from this article. Someone can add it again later if they have a reference for it. --Dual Freq 00:08, 3 January 2006 (UTC)

Agreed rf burns have little to do with SWR, essentially a RF burn is caused by radio waved heating the body tissues (in a similar manner that a microwave heats foods) touching the antenna directly increases this effect, also has alot to do with resonant frequency of various parts of the body to calculate absorbtion and heating, ARRL has alot of good litterature on burns. Mattpalmer84 15:00, 19 May 2006 (UTC)

I think the the presence of any grounded body near the antenna will change the SWR. and the grounded body will pull most of the standing waves to it and result in microwave burns. wile causing signal interference. User:taylormayhew87 6:22 12 May 2014 (UTC) — Preceding unsigned comment added by 2600:1011:B017:E342:8412:75EE:2C85:BCE6 (talk)

Standing Wave Math Expression

It seems being a paradox. For

${\mathcal {\mid }}\Psi (x,t){\mid }^{2}=A^{2}sin^{2}(kx)*cos^{2}({\omega }t)=0{\neq }1$

which means that it does not equal to 1. Thus caused not coresponse Normalization. Known a standing wave is expressed as

${\mathcal {,}}\Psi (x,t)=Asin(kx)*cos({\omega }t),$.

Can anyone talk about your thoughts? Thanks.

• One more question that what's difference between phase velocity and group velocity? My opinions and thoughts:

By their math expression we can clearly find angular frequency of ${\mathcal {V}}_{p},$ ${\mathcal {\omega }}$ which keeps constant when a wave vibrates up and down localized. That may because of energy transports into a wave is conservative,just like a particel moves up and down in a Y axis,localizedly(which keeps energy conservative).

But for another one,it travels in an X axis,that hints its phase-angular is the function of time. By time changes,then ${\mathcal {\omega }}$ naturely changes either.

I'm a little not sure above. Could anyone discuss with me? --HydrogenSu 11:47, 7 February 2006 (UTC)

my opinion is that is something for a physicist to explain, i would rather consider the practical realitys of EE problems Mattpalmer84 15:05, 19 May 2006 (UTC)

Arcing

The article says "Arcing may occur if there is a hole, if transmitting at high power, usually 200 watts or more (Need reference for the power statement)." I doubt a reference for the power statement can be found as arcing strongly depends on the frequency and the waveguide dimensions. I frequently transmit between 2 and 3 kW through a R26 waveguide at 2.45 GHz and have never experienced any arcing. In my opinion, the sentence should be shortened to "Arcing may occur if there is a hole or if transmitting at excessive power."

Ulrich — Preceding unsigned comment added by 129.69.28.66 (talk • contribs) 16 jul 2007 16:25‎

Error in claim in section "Practical implications of SWR"

The following claim mentioned in "Practical implications of SWR" is not true:

1. Higher reflected power in the transmission line is dissipated in the output stage of the RF source, in the form of heat.

This is extensively documented in the ARRL Techical Information Service at http://www.arrl.org/tis/info/reflections.html

See http://www.arrl.org/tis/info/pdf/q1106037.pdf.

Regards, LA8UU

---

—Preceding unsigned comment added by Oka (talk • contribs) 18:40, 20 February 2008 (UTC)

Well if the power is reflected, where do you figure it eventually goes? And it's a pretty well-accepted fact that running an unloaded microwave oven cooks the magnetron precisely because the magnetron ends up dissipating the power that is reflected rather than absorbed by the cooking chamber.

Atlant (talk) 00:24, 22 February 2008 (UTC)

Lede

The third paragraph of the lede makes it sound as though the VSWR is a property of the transmission line; the vast majority of the reflection will occur at the junction with the load or from within the load (barring backhoe fade). --catslash (talk) 02:03, 14 March 2012 (UTC)

Not sure if this is the intended point of the author, but it is indeed true that VSWR is used as a quality measure by cable manufacturers. I found some references on this while writing Nominal impedance#Cable quality. Of course, I agree with your statement also; cable testing requires precision adjustment of the termination first to be meaningful. SpinningSpark 08:32, 14 March 2012 (UTC)

Further analysis section

There is something wrong with the further analysis. According to the result min voltage always occurs at x=0. Clearly, this is only true for resistive loads less than Z0. The mistake seems to be ignoring the phase of Γ, but I haven't worked through it in detail. Also the diagram contradicts this by showing max voltage at x=0. Might be an idea to state where x is being measured from. SpinningSpark 13:28, 19 November 2013 (UTC)

Symbol for reflection coefficient

I don't see the advantage of changing the symbol used for the magnitude of the reflection coefficient from \rho to |\Gamma|. The former makes for cleaner expressions in my opinion, but in any case, I don't think that it is acceptable to half change the article as it now stands. It should be all or nothing, including the diagram. SpinningSpark 16:22, 27 November 2014 (UTC)

Well go back to the old version and look carefully. Rho was used for the magnitude of Gamma. It was just used in a couple locations and it didn't seem to require its own symbol for any good reason since writing |Gamma| doesn't take up much more space. I could see substituting repeated uses of a complex expression with an easier symbol, but in this case it would more likely introduce confusion (if they didn't read the definition of rho carefully) than clarity.Interferometrist (talk) 17:20, 27 November 2014 (UTC)

But you're right about the diagram though. I guess if you really want that diagram, then you should change back to rho. I didn't think the diagram was too useful to a general reader and would prefer to look for a different one. A good one, in my opinion, would be one that showed the standing wave pattern with a composite showing the net voltage at different time phases, and the SWR showing up as the envelope (of course you can only do that for one SWR in one picture). The original presentation in that section did have x as a measure of distance, but didn't show where the end of the line and load was located (so I made it explicit), so the correspondence between the figure and the text wasn't so clear. I guess, according to the figure, the line ends at 3pi, with Gamma being real positive. But how often is a line exactly 3/2 wavelengths long? (Or did it really start at x=0?). Interferometrist (talk) 17:35, 27 November 2014 (UTC)

And actually it's worse than that (possibly the reason for the section being tagged). According to the first equations, it would appear that the load was at x=0 and gamma is positive real. And with the forward wave in the +x direction (thus x is always negative along the line). So that didn't correspond to the figure at all. Of course you could reintroduce the figure stating x_0 = 3pi and spelling out rho. But as I said, the figure isn't so helpful in my opinion, and certainly shouldn't be inconsistent with the equations.Interferometrist (talk) 17:44, 27 November 2014 (UTC)

I don't agree with your argument that the user will get confused if they don't read the definitions. Of course they will, but they could get just as confused with |\Gamma| if they don't read the definition of \Gamma. SpinningSpark 19:13, 27 November 2014 (UTC)

Power standing wave ratio

I recently sourced the definition of PSWR in the article. The source was challenged with the edit summary Would like to see source <60 years old indicating current usage (also why the term is meaningful). Currently Google search for term points to wikipedia: are we to be the main source of misinformation?. First of all, the book I cited is still in print (or was until quite recently). It is from the publisher of one of the most highly prestigious electrical engineering professional societies in the world. It is hard to get more reliable or authoritative than that on the correct use of EE terminology. Secondly, even if the term could not be found in any modern source, that still remains the correct definition of it which is all that is being cited. However, a gbook search quickly establishes that the term still has some traction in books published this millenium.

I don't know why there is any doubt about the physical meaning of the term. It is patently the ratio of the power of the forward and backward travelling waves, at least for purely real characteristic impedance. SpinningSpark 20:11, 4 December 2014 (UTC)

Well, when I searched Google, I found that all the references it turned up were either copies of Wikipedia itself (whether acknowledged or not, you can just compare the wording with old versions of this page) or copies of one other source which admitted that is was a "misleading" term (for the reason I'm about to state). So the issue isn't whether it can be found anywhere (almost anything you search for is going to be found somewhere) but whether it has any actual meaning, or more to the point, whether the people using it actually understand what they are saying rather than just repeating someone else. I'm certain it's the latter, as I'll explain in answer to your remark:

"It is patently the ratio of the power of the forward and backward travelling waves". Well no, that's definitely NOT what it is!! That ratio is the squared magnitude of the reflection coefficient. Someone had been taught that P ~ V^2 so decided that a ratio between two voltages can just be squared to get a power ratio. But no such power actually exists. Those powers that you have taken the ratio of are the powers that would be dissipated in an R ohm resistor with those voltages applied across it. But of course if you actually DID shunt a resistor at a point along the feedline you would completely change its standing wave pattern (unless it was a very large resistor)! So no, the "powers" refered to don't exist and aren't intended to exist. The actual powers are in the two waves, given by V_f^2/Z_0 and V_r^2/Z_0. Those powers have nothing to do with Vmax^2 and Vmin^2. Squaring those voltages gets you nothing.

It's as if I were to edit the page on currency exchange, so that if 1 euro equals $1.2 then the "power currency conversion" is defined as 1.2^2, or the "power temperature" of the water melting point is 273^2 etc. If I were to publish such a definition and get it mindlessly repeated (even in otherwise respectable publications by otherwise intelligent authors), does that make it valid? Interferometrist (talk) 22:53, 4 December 2014 (UTC)

You are, of course, right about the physical meaning. My bad, I had a moment of blindness there. But that is all irrelevant anyway, the article makes no claim to its meaning, only saying that it is the square of VSWR, and that is all the sources ever say as far I can ascertain. The bottom line is that it appears in reliable sources. Readers may well look it up in Wikipedia. It is perfectly reasonable that we have coverage of the term in our article. SpinningSpark 23:33, 4 December 2014 (UTC)

Alright, we basically agree, so the remaining issue is how we frame the use of a basically useless term and how much credance we give to its usage. If a useless term was incorrectly introduced into the jargon of a field and carelessly repeated by those who were concentrating on other aspects of the field (such as the antennnas themselves, which requires the mention of transmission lines and impedance matching), but never is further explained in contexts where it would be central to the discussion, then it is natural that the MOST authoritative sources (which actually do explain the phenomenon, SWR in this case) won't even mention it, and its use will fade out. Then how much should Wikipedia dignify such a term? This becomes particularly important when a great deal of the use of the term can be traced back to it appearing in Wikipedia in the first place, as appears to be the case here. If so, then Wikipedia is doing a disservice not only to its readers but to the whole discourse (given the prominence of Wikipedia as a first reference point for reliable information, which is certainly the way I use it). In those cases either the term should be avoided in Wikipedia, so as not to be reinforced, or it should come with a clear statement as to the reason for avoiding the term. So I would propose somewhat stronger language in that regard compared to the current disclaimer, if the erroneous term is to be mentioned at all.

Since I went to the trouble of doing a Google search and glancing at the results, let me share those. These are copied from older versions of the wikipedia page itself (or perhaps our page was copied from one of them): [2][3][4]. These all contain enough common text that they have been copied from each other: [5][6][7], and to its credit it mentions that the definition is "misleading," a somewhat stronger statement than we currently have. I found one other source that mentions the definition [8], and one that is a glossary with a definition that's not even understandable out of context and which doesn't mention the context: [9]. To say that it is a term that is included in "reliable sources" is somewhat a stretch. At this point, I am more concerned about Wikipedia doing harm by propagating an invalid term than failing to mention an invalid term that may have some usage. Though I still wonder about that. Have you EVER heard someone mention a specific value of "PSWR" or seen it calibrated on an SWR meter? If not, then you can't even say the term is in usage, but just that its definition is in usage, if you see what I'm saying. Interferometrist (talk) 15:26, 5 December 2014 (UTC)

Sorry, you can't cherrypick a bunch of unreliable sources and from that conclude the term is useless, nor can you conclude that it is mostly copied from Wikipedia because unreliable sources so copy it. Let's restrict the conversation to sources normally considered reliable, like books from recognised publishers and scholarly papers. I have no idea how you have come to the conclusion that the term originated in Wikipedia after you tagged my contribution for citing it to a 1949 source, long before the internet, let alone Wikipedia, was even thought of. Just to show that that is not a one off, there are also, A Textbook of Radar (1948, CSIRO), Electric Transmission Lines (Skilling, 1951, McGraw-Hill), Handbook of Microwave Measurements (Sucher et al, Polytechnic Institute of Brooklyn, 1963), and many others. There are also many modern book sources from the last decade, for instance, Transmission Lines, Matching, and Crosstalk (Kaiser, 2005). There are also papers showing up in Scholar since 2000. So while it may not be the favourite term, it is clearly not defunct.

My understanding of the use of the term is that it is used for exactly the same purposes as VSWR (used for the same, not equal to!) and is thus no more, nor less, useless than VSWR. Most sources start off with VSWR and then say PSWR is its square. However, I note that Silver does this the other way round, starting with PSWR then declaring VSWR to be the square-root. You can't claim the term to be useless in the article when sources don't support that claim. Gridey (2010)says that PSWR is sometimes used in favour of VSWR because of the square-law nature of the crystal detectors originally used. I think it is also easy to cite that the more modern detectors using Schottky barrier diodes are also square law. SpinningSpark 17:32, 5 December 2014 (UTC)

Hey cool down! I'm just trying to get this right. But please understand one thing: I did NOT cherry pick any sources. I did a Google search, and showed you EVERY relevant result I got on the first two pages of results. Period. And I certainly did NOT say that it was mostly copied from Wikipedia, but did say that Wikipedia was responsible for 3 of the sources I found on that not-cherry-picked search, and concern for the fact that, if not careful, Wikipedia could be contributing to the problem rather than helping to clarify. No more time to write just now, but I will answer further, and actually WOULD like to see any use of the term (didn't have a chance to follow your links yet) where a justification for the term is offered. But please moderate your tone: I'm just trying to get this right. I don't have a financial stake in putting PSWR out of business and if I'm making a mistake (I've certainly made mistakes before) then I want to correct it, not justify it at all costs. Ok? Interferometrist (talk) 17:50, 5 December 2014 (UTC)

Sorry you think I need to cool down. I'm not angry, like you I'm just trying to get the article right. I suppose my essential point about your search results is that raw google searches, or analyses of them, are rarely helpful in constructing an encyclopaedia article. SpinningSpark 19:10, 5 December 2014 (UTC)

Alright, I am pleased to hear that you aren't angry. It appeared you were because you opened with an accusation that I had "cherry-picked" sources. If I believed that someone had presented me with cherry-picked sources in order to deceive me, I would be angry. Though I wouldn't have made the charge, following the dictum of assuming good faith, unless I had strong reason to believe the accusation. Starting with, in this case, doing the same Google search and seeing if there were conflicting sources that had been conspicuously omitted from those presented. I want you to know that I did an honest Google search to find any usage of the term, not looking either for reliable or unreliable sources, but for any justification of the term (of which I found none) or commentary on the term (which is always disparaging, including from your own sources). I am editing the page accordingly.

But first I will answer you, now that we've gone this far in discussing a matter of so little importance. I clicked on each of your sources, though in cases of Google books it often showed just one sentence with no context. I'd be particularly interested in seeing "Silver does this the other way round, starting with PSWR" which sounds unquestionably wrong. But I was able to view Gridley fully and will comment. In addition to disparaging the term, he (as you point out) attempts to explain its origin. Now, his work looks fine in all respects but I can only conclude that he wasn't thinking when he wrote that but was just coming up with what seemed plausible to him in order to make sense of non-sense. Exactly as you did when you wrote: "It is patently the ratio of the power of the forward and backward travelling waves," only because you didn't take 5 seconds to think about it. I don't fault you for that. I regularly find, from very intelligent people, that when you challenge conventional wisdom or the statement of a recognized "authority," they often will immediately come up with a plausible-sounding but often invalid (as they will admit after careful consideration) reason that the authority is right and I am wrong. I'm sure I've done it too. Our psychology is such that we try to make order of the world and resolve such tension. But Gridley's remark is wrong on two counts, and I'd bet that if you pointed this out to him he would agree right away and retract his explanation.

It is true that a diode, operated forward-biased in the small-signal regime, will perform square-law demodulation. However in an SWR meter that doesn't apply, because you are measuring substantial RF voltages with a transmitter sending down watts of power. Now (obviously) when you operate a rectifier in the large-signal regime, the rectified DC voltage (neglecting any forward voltage drop) is EQUAL to the peak RF voltage. Period.

Secondly, while SWR is the ratio of Vmax to Vmin, you never measure those particular voltages in practice. First, you wouldn't know where along the transmission line to find them. And most transmission lines are insulated so you couldn't make that measurement, and unless you were careful attaching a test lead would change the SWR (granted, you could use a 10x scope probe). The only times I've seen (on video) such a measurement is for classroom demonstration of UHF standing waves along an open transmission line setup for that purpose. The old SWR meters I've seen (haven't looked at any recently) work using directional couplers to obtain RF voltages proportional to Vf and Vr. Even if you had directly measured the powers of those waves, Pf ~ Vf^2 and Pr ~ Vr^2, the computation of PSWR still takes one more step than SWR. PSWR = ((1+sqrt(Pf/Pr)) / (1-sqrt(Pf/Pr)))^2 which cannot be further simplified to avoid the square roots for instance. I rest my case. Interferometrist (talk) 14:44, 6 December 2014 (UTC)

Errata: the above equation should have read:

PSWR = ((1+sqrt(Pr/Pf)) / (1-sqrt(Pr/Pf)))^2 Interferometrist (talk) 16:28, 7 December 2014 (UTC)

Gridley is almost certainly talking in the context of slotted lines or similar instruments, although he doesn't say so. On such instruments the extrema are directly measured and the square law issue applies. SpinningSpark 16:35, 6 December 2014 (UTC)

Well I guess I agree with you now! I hadn't heard of slotted lines and have no practical experience with waveguides and such technology. But now I see that this does involve direct measurement of Vmax and Vmin. And from reading the page on slotted lines, I see that these are low power measurements with (understandably) tiny antennas sensing the electric field, so that the diode never receives a large signal. (That article also mentions using thermistors for sensing, which of course are also sensitive to power and thus ~ V^2). So this all adds up now, and I will make only non-controversial changes in wording.

For one, you were a bit hasty in reformatting the quote, because the second sentence wasn't by Gridley but was what I had written to emphasize his point! Perhaps you're spending too much time staring at the source and not formatted text (where this would have been clear). I'm fixing it now. (BTW, should the quote be enclosed in quotation marks if it is offset in this manner? I gather not, using the template you supplied.)

Again, I'm glad we have this resolved, and regret having made an issue of it except that I was unaware of the direct measurement using slotted lines (and wish you had mentioned that earlier!). It also still bothers me that not one of the references (that I've seen so far) specifically refers to slotted lines, but now that I'm familiar with their use I have to agree that that is what Gridley must have had in mind. Interferometrist (talk) 16:28, 7 December 2014 (UTC)

P.S. I don't have the book, but was the Rollin reference over the slotted line measurement, or just over the square law of a crystal detector? If the latter, the ref needs to be moved up. Interferometrist (talk) 16:38, 7 December 2014 (UTC)

Sorry for not mentioning slotted lines earlier, as I wrote the article, I don't even have your excuse of never having heard of it before! In my day, these were quite ubiquitous and I tend to assume they are still well known. I was quite shocked to find Wikipedia did not have an article. The whole reason that VSWR is used at all is because it is the easiest thing to measure on slotted lines (and similar measuring techniques). The reflection coefficient would be the natural choice, but it can't be found directly. If it weren't for this, VSWR would probably never have been used at all. By rights, it should probably be abandoned now, but it is much too well established.

On the Rollin cite, he is discussing a student exercise using a waveguide slotted line. The relevant sentence is "Since for low level input, the crystal is approximately a square law detector, this is the power standing wave ratio (P.S.W.R.)." Sorry, it was the only source I could find that intersected square law of detectors with PSWR. Intersecting square law/detectors/slotted line are easy to find (there are more accessible sources in the slotted line article) as are sources intersecting slotted line/PSWR but getting both in the same source proved a bit more difficult. SpinningSpark 18:03, 7 December 2014 (UTC)

Well good, we've just about converged. And I trust you'll consider my latest edits. On your point, though: "It is dangerous to say the readings are power, I thought we had just established they weren't." Ha, I guess I was saying that the power delivered by the probe was being measured. It occurred to me (but I'm not the one with practical experience in this regard) that such a probe would feed a rectifier and analog meter, where the meter's scale would already be warped to take into account deviations from the square law response at higher/lower voltages (like the scales on the low AC voltage ranges of old multimeters), in which case what's important is whether the calibration is in terms of relative voltage or power. And although I'm not surprised that thermistors aren't used much anymore (of course I just got that from the slotted waveguide page), it was another perfect example of a true power (from the probe!) detector. But never mind. It's looking fine, and I think I'm done with it, unless you can find something else to disagree about ;-) Interferometrist (talk) 18:45, 7 December 2014 (UTC)

Article rating

@Steve Quinn:, I find it baffling why you have demoted the article to start class. The requirement for C class (in part) is "The article is substantial, but is still missing important content...". The page has nearly 4000 words of readable prose. It is difficult to see how that cannot be substantial. The reader experience for C class is "Useful to a casual reader, but would not provide a complete picture for even a moderately detailed study." Reader experience for Start class is "Provides some meaningful content, but most readers will need more." Demoting the article implies that either the article is not useful to a casual reader, or that most readers will require more information. So can you explain how the article fails one of those criteria? SpinningSpark 15:26, 21 March 2017 (UTC)

@Spinningspark: you have presented some really valid arguments. I agree with you on this. I am changing this back to Class C. Thanks for pointing this out. Steve Quinn (talk) 04:32, 24 March 2017 (UTC)

Gamma greater than unity

@Interferometrist: You reverted the edit by user:AE6TY that claimed that gamma can be greater than unity. The source cited is only available in snippet on gbooks, but clearly supports the claim. The source also addresses your concern that gain has been implied. I don't think your claim can be right that gamma is always specified with respect to a positive real impedance, and not the actual characteristic impedance. Do you have a source for that? SpinningSpark 16:31, 24 July 2019 (UTC)

Hi, of course it's possible to plug in a complex characteristic impedance into a formula and mechanically apply the formula. But AFAIK (and if I'm wrong I'll concede the point, but I'm not) whenever SWR or Gamma is used in practice (such as on a Smith chart, which is a plot of Gamma) it is with respect to a real impedance, usually 50 or 75 ohms. That impedance may or may not be the same as a coax used, or there may be no coax ever involved anywhere, such as when you measure s-parameters (Gamma == S_11) with an analyzer having a specified reference impedance. SWR and Gamma are only useful in this way and there is no reason anyone would use a complex reference impedance. Even when using a cable whose characteristic impedance has an imaginary component (which is normal, but tiny and uninteresting in this context) you don't redefine SWR due to that.

No, I didn't read the reference, but all I can say using my judgement based on my familiarity with these fields, is that the author decided to be cute and show that if you do allow a complex reference impedance then you get strange, but essentially nonsensical results. After all, the MEANING of SWR -- forget the definition in terms of impedances -- is the ratio of the peak voltage (or current) magnitude (phase isn't involved) to the minimum, which is obviously >=1. Long ago someone translated that into a formula involving impedances without spelling out the hidden (but obvious) assumption that the cable impedance is taken to be real. Finding such a flippant remark based on inappropriately applying a formula in one otherwise-RS doesn't mean that this would be accepted by any other equally competent authors who don't bother themselves with fringe cases (the whole point of transmission line is to transmit power, not to dissipate it which is what adds a SMALL imaginary part to its characteristic impedance). Agree? Interferometrist (talk) 18:48, 24 July 2019 (UTC)

I'm really not trying to be cute. I'm trying to stamp out a misconception rampant in the field. There is almost universal (and passionate) belief that for any passive network, the |Gamma| cannot be larger than 1, regardless of the reference impedance. This is patently false. I understand that the overwhelming majority of applications can safely assume the reference impedance is purely real but there ARE applications where this is not the case. I am not an expert but have been told that power transmission lines and POTS phone lines need to take the complex reference into account. Papers by Korukawa and Awakawa dealing with optimal power exchange also indicate the use of a complex reference at higher frequencies. I don't want to complicate things needlessly but I do think that a footnote is a good compromise. For those trying to get started, nothing is obvious. AE6TY (talk) 21:38, 24 July 2019 (UTC)

Interferometrist, I don't agree that transmission line impedance (always) has an insignificant imaginary part. I agree with AE6TY on this; twisted pair at audio frequencies is anything but a constant real number. I spent a good many years in my career dealing with just that issue. Yes, one uses a reference impedance on a Smith chart plotting the effect of a complex load, but there are also times when we want to know the effect on a fixed resistive load of a non-ideal transmission line. Having said that, I agree with you that the definition of SWR is the ratio and must be greater than unity. I also agree that the impedance formula becomes nonsense with a greater than unity gamma. That whole issue seems to me to be something more appropriate to discuss in the reflection coefficient article than here. SpinningSpark 22:30, 24 July 2019 (UTC)

I agree that a better place is probably the article on reflection coefficient... I'm only trying to get a footnote in here. It is amazing to me how passionate people get about this issue. I think it would improve wikipedia's image to at least provide a footnote and references for further reading. Please note that the second reference above is capitalized incorrectly. It should be 'Kurokawa'. Too much Java programming makes my fingers do weird things. — Preceding unsigned comment added by AE6TY (talk • contribs) 00:34, 25 July 2019 (UTC)

I would also point out that I ran across this when someone asked ME why my SWR report was negative. It occurred when the user was driving a highly reactive antenna with a highly reactive source. There was no transmission line present. In researching the topic I found that the overwhelming majority of sources quietly assume (or are unaware that they are assuming) that the reference impedance must be purely real for these equation to apply. Understanding that SWR is a transmission line thing I went looking at transmission line authorities such as my cited Chipman. Other in depth papers discuss the (perhaps esoteric) issues involved with complex reference impedances.

So how does Wikipedia expect us to proceed? AE6TY (talk) 15:34, 25 July 2019 (UTC)

In the first instance, say what it is you want to do to this article and see if other editors agree with you. Then you can do it. But I thought you agreed with me earlier that reflection coefficient was the right article to add this. In which case you should add it there, and discuss it there if it is challenged. SpinningSpark 18:45, 25 July 2019 (UTC)

Hi, thanks for your thoughtful inputs. I basically stand by what I wrote but will refine my views. @AE6TY:, editors "feel passionate" when text is added to a page (including within a footnote) which is incorrect (including the incorrect usage of a term according to its general understanding within a field, regardless of outlier definitions) or which is misleading (including placing excessive weight on a fringe definition). I did not look at your two references, but if there are only two such papers over 50 years, that proves my point. If there are many more, OR if you can write some CONTENT on how a complex reference impedance is used in practice to actually solve a problem (hopefully a method used by more than one individual, though even that's not essential), then I would concede the point and tell you to go ahead and put in such content. If you do, it would be responsible to point out that this liberty is almost never taken in practice which is why |Gamma|<=1 is normally asserted etc.

Yes @Spinningspark: I agree that at some frequencies (especially as freq-> 0) the imaginary part of a cable's characteristic impedance can become important or even large. However what I haven't seen (and which would change my judgement) is how using that cable's complex Z_0 in an equation for Gamma would help solve a problem. You seem to be well-versed in these matters, so I will let you answer that and edit accordingly. If that adds some new content then I will be happy to learn it for myself. But just placing a footnote to say that at least two authors over 50 years have used a generally disallowed notation and not being able to say why it is useful -- no.

Also, I think the SWR and Reflection coefficient pages should be merged because they are about the same thing. But I know how hard that is to get such a thing done, so never mind! Interferometrist (talk) 13:30, 1 August 2019 (UTC)

References