Talk:Impedance matching/Archive 3

MAXIMUM power
MAXIMUM power transfer is using voltage bridging. But only if you are able to choose a LOW output impedance. You will have to use some kind of fuse for NOT transferring maximum power and overload the source. If you have to use a significant output impedance, then impedance bridging is the solution. You will loose half the power or even more (not precise impedance matching).

--AK45500 (talk) 14:41, 12 July 2018 (UTC)


 * Not sure what your point is. For low power signals, such a Ethernet, you do not need a fuse for protection.  For the AC  Mains, you never load it for maximum power transfer; it would melt the wires and probably the load. Constant314 (talk) 14:55, 12 July 2018 (UTC)
 * Ha ha, it would melt the national grid as well if you actually succeeded. SpinningSpark 16:18, 12 July 2018 (UTC)
 * My point is: Real MAXIMUM power transfer occurs with voltage bridging (see: AC mains). But you need some kind of protection (fuse). Ethernet has a significant source impedance (and save against overload). You will use impedance matching to minimize reflections. Maximum Power transfer is not important. Impedance Matching can only transfer 1/4 of the power. (1/2 due to attenuation, 1/2 because you will draw double current using Impedance bridging compared to impedance matching at same load impedance).
 * My point is: Some people think MAXIMUM power transfer  is used (or has to be used) for AC Mains. They like not just to melt their ice cream. This article should state the misunderstanding explicitly. And people think whenever relexions are optimized, and perfect impulse transmission is needed it uses Maximum Power Transfer. i.e. AES3 (digital AUDIO) was optimized for perfect impulse retention and group delay. There is no need for energy transfer (no level problems).
 * --AK45500 (talk) 09:57, 11 September 2018 (UTC)


 * Still not understanding your point. Maximum power transfer is not very relevant to AC mains.  The AC mains article you refer us to does not even mention "characteristic impedance", "voltage bridging", or "impedance bridging".  I am not aware of your claimed common misunderstanding.  Rather than continuing to try to explain your perceived problem with the article, perhaps you could link to a source that discusses the issue?  Without a reliable source, nothing can go in the article anyway. SpinningSpark 14:35, 11 September 2018 (UTC)
 * In the old days (perhaps the days of Edison and DC), the concept of maximum power transfer misled some engineers, before they came to understand that mains should be voltage sources. This is in books, and may be what he's referring to. I'll try to find it. Dicklyon (talk) 14:39, 11 September 2018 (UTC)
 * The problem with Edison's DC system was that it was of necessity low-voltage. That made it highly lossy and unsuitable for distribution over long distances.  Attempting to achieve max power transfer would only have made its problems worse.  It would be interesting to hear if Edison or his contemporaries actually thought that was the way to go. SpinningSpark 17:27, 11 September 2018 (UTC)


 * Maximum power in this article is not about getting the maximum power out of a generator or distributing maximum power over the grid. It is about what you can do with the load impedance to get the maximum power in the load given that everything else is fixed.  It is not about the maximum sustained power you can get without burning anything up or down.  It is about the theoretical maximum regardless of the consequences.Constant314 (talk) 17:55, 11 September 2018 (UTC)
 * @Constant314 ; Thank you for your statement. This (explicitly) is what I am thinking of. Some People are misinterpreting the MAXIMUM. It is just the maximum power in the load given that everything else is fixed. IF I am able to change source impedance the minimum I am able to choose is the better maximum. This is back to my very first sentence. Just a statement to handle these different maximums. The better maximum is 4 times the power maximum of the maximum power transfer (same load).
 * --AK45500 (talk) 11:04, 12 September 2018 (UTC)
 * @AK45500 - careful of the assumptions you are also making. Your comment about lowering the source impedance to increase power transfer into the load is true at low frequencies (essentially at baseband/DC). At higher frequencies (eg RF and microwave) what you are suggesting will result in an impedance mismatch between the source and the load that will create reflections over the transmission line that connects the source and the load, which will actually lower the power transferred to the load. At worst case, it will cause a communication line to become unusable. — Preceding unsigned comment added by 2A00:23C8:293:5001:59DF:B7A7:1AAE:AC9F (talk) 15:31, 12 August 2020 (UTC)

Short description
The short description is meaningless nonsense. SpinningSpark 17:06, 13 May 2022 (UTC)


 * A more descriptive description would go past the 40 character limit for descriptions Ffffrr (talk) 17:11, 13 May 2022 (UTC)
 * 2001:16B8:2D2D:D100:F9F9:9CF9:4639:F4D5 (talk) 18:42, 20 May 2022 (UTC)
 * SpinningSpark's reply appears to be misplaced on "Impedance Matching" talk.
 * It may refer to Ffffrr's user page. 2001:16B8:2D2D:D100:F9F9:9CF9:4639:F4D5 (talk) 18:42, 20 May 2022 (UTC)
 * No, I am referring to this. <b style="background:#FAFAD2;color:#C08000">Spinning</b><b style="color:#4840A0">Spark</b> 10:03, 5 June 2022 (UTC)

Composite reflection factor
If you have been following the recent discussions, there is some confusion over exactly what reflection and reflection coefficient is being discussed. I hope we can agree on some terminology. Let's suppose a sequence of transmission lines labeled 1, 2, 3, etc. with the load attached to #1. Power is transmitted from the highest numbered toward the load. Let's consider the junction of #2 and #3. There are two reflection coefficients of interest. One, that I propose to call the local reflection coefficient, depends on the characteristic impedances of #2 and #3. The other, which I propose to call the composite reflection coefficient, depends on the characteristic impedance of #3 and everything downstream toward the load from the junction of #2 and #3.

In the case of a transmitter connected to an antenna by a coaxial cable, there is a local reflection coefficient at each end of the cable and a composite reflection coefficient at the transmitter end looking through the cable to the antenna. The matching problem is to put some network between the transmitter and the cable to minimize the composite reflection coefficient, even if that means there is a substantial local reflection coefficient at each end of the cable.

Likewise, at each junction there is a local and composite reflection.

We might find some standard terminology in the literature of multi-layer optical coatings.

I hope some of my helpful fellow editors can suggest better terminology or confirm my proposal to use local and composite. Overall might stand in for composite. Thanks. — Preceding unsigned comment added by Constant314 (talk • contribs)
 * The accepted terminology for this seems to be local and global reflection coefficient. I found usages of that in fields as diverse as ultrasonics, fluid flow, and hematology as well as electromagnetism. <b style="background:#FAFAD2;color:#C08000">Spinning</b><b style="color:#4840A0">Spark</b> 06:06, 11 June 2022 (UTC)


 * I agree that local and global are common differentiators. I've also seen 'systemic reflection coefficient' or the slightly wordier 'reflection coefficient of the system'. Any of those are fine, really. MrOllie (talk) 11:31, 11 June 2022 (UTC)

Transmission lines
Please correct:

instead of: ... "The reflection creates a standing wave if there is reflection at both ends of the transmission line, ... "

please write: ... "The reflection creates a standing wave if there is reflection at the end of the transmission line, ... "

Why? Please see the Figure "Coaxial transmission line with one source and one load" and apply it's indexing to the \Gamma_{12} equation given, inserting Z_L for Z_2 and Z_C for Z_1.

$$ \Gamma_{CL} = \frac {Z_L - Z_C} {Z_L + Z_C} $$ unless Z_L = Z_C. And once, you get standing waves, independant of Z_S at the other end.

DJ7BA (talk) 16:21, 13 June 2022 (UTC)

Unreasonable reverting
@Constant314:

Issue 1:<

Completely misunderstood. Read again. If that doesn't help:

I don't argue impedance bridging at all, but the existing context - not me - of the bad sentence did call impedance bridging "unsuitable".

The bad sentence (second sentence in subsection Transmission lines) said:

The reflection creates a standing wave if there is reflection at both ends of the transmission line

This bad sentence states in other words, that for standing waves to exist, there must always be two mismatched ends.

No!

Your assumption is, I didn't think of impedance bridging, thus making a false statement.

No!

But the sentence is not generally true, however. One example is:

A mismatched antenna is connected by a transmission line to the well matching pi filter of a tube amp transmitter. In this case there is just one single mismatched end of transmission line, but that single end's mismatch certainly creates reflections on the line, causing SWR.

Be that often or not, suitable or not is not the question for the absolute statement the bad sentence claims. You don't need any match or mismatch situation at the source end for that antenna end mismatch caused SWR to happen. It is a totally source mismatch independant, load mismatch caused, SWR.

And even: You didn't revert the thing you are complaining about: Often impedance bridging must be considered, but the existing context - not me - called it unsuitable. You kept the part you complained about, but reverted my reasonable contribution.

Issue 2:

You didn't even care to explain why you reverted my second contribution with one blow.

I am not amused.
 * I see your point. I'll fix it.  However, I think that maybe impedance bridging is being overused in this article, as it only seems to be a term used in audio engineering. Constant<b style="color: #4400bb;">314</b> (talk) 20:02, 14 June 2022 (UTC)


 * Thank you, I appreciate it.
 * Overuse of impedance bridging? I agree, but it deserves treatment in RF electronics as well. Wiki doesn't demand any well proportionate coverage, or? And if: What exactly will be the adequate proportion? That would be a door opener to endless reverting and discussions. If one aspect is, however, obviously all too much overused, it is perhaps possible to reduce some of it to a nutshell instead of being too lengthy, and improve the page that way.
 * Both impedance matching aspects, conjugate and impedance bridging, are precious contributions, I think.
 * Some examples:
 * Almost all high power transmitters today make use of it because of energy cost. Battery operated transmitters need best efficiency, too.
 * On the other hand, as far as conjugate matching is concerned, perhaps a billion people use it, but are unaware of it: In the LNB of a sattelite disk the antenna is built into the low noise amplifier input with no line in between. Here you need to make the best of the extremely low reception signal. Another such example is radio astronomy: See the liquid helium cooled LNA with antenna in the focus of the huge parabolic reflector. What an extreme effort. So why loose any portion of the precious small signal at all. Reception antennas for a wide enough frequency band cannot be standard constructed to be 50 +j0 Ohm, as standard transmitters can. Instead, they use conjugate impedance, power maximizing match. DJ7BA (talk) 07:22, 15 June 2022 (UTC)

Lead section
In electronics, impedance matching is the practice of designing or adjusting the input impedance or output impedance of an electrical device for a desired value. Often, the desired value is selected to maximize power transfer or minimize signal reflection.

Excellent. Couldn't be better.

In the next sentence "same impedance" is not 100% correct:

"For example, a radio transmitter can most efficiently transfer power to an antenna if the antenna and interconnecting transmission line have the same impedance as the transmitter."

Based on Stutzman-Thiele, for a radio transmitter to most efficiently transfer power to the antenna, two matching networks are needed: The network to the right matches the antenna impedance to the transmission line's characteristic impedance. It adjusts for identical impedance match. The one to the left matches the transmitter impedance to the impedance, the transmitter sees at the left end of the transmission line. It adjusts for conjugate impedance match.

In many cases, for the sake of simplicity, there is only the (easy to reach and tune) transmittter end matching network, as shown by Stutzman-Thiele fig. 30a. In this common case, conjugate match only is used, not identical impedance match. In many cases this will yield good overall power transfer, but not full achievable optimum.

Let's change the sentence to:

"For example, a radio transmitter can most efficiently transfer power to an antenna if the antenna and interconnecting transmission line are matched for same impedances, and the transmitter and the transmission line are matched to have conjugate impedances." [1]

If that is a bit too complicated in a lead, let's better say alternatively:

"For example, a radio transmitter can most efficiently transfer power to an antenna if the antenna is matched to the interconnecting transmission line and the transmitter is matched to the line, too." [1]

This will postpone the different types of match to a more detailed, later section.

[1] Warren L. Stutzman, Gary A. Thiele, Antenna Theory and Design, – 3rd ed. ISBN 978-0-470-57664-9 (hardback) Antennas (Electronics) I. Thiele, Gary A. ch. 6.4.1 transmission lines. p. 176, eq. 6-30b

DJ7BA (talk) 09:33, 14 June 2022 (UTC)


 * I agree that "For example, a radio transmitter can most efficiently transfer power to an antenna if the antenna and interconnecting transmission line have the same impedance as the transmitter" is incorrect. It is not abut efficiency.  Most antennae have significant reactance.  You can put a matching network at the antenna, but that is not usually done.  Usually, a mismatch at the antenna is accepted.  A matching network at the transmitter makes it work for the transmitter.  There are two cases that I am aware of where matching the cable at both ends provide advantages.  First, when there is two-way communication, it is desirable to minimize reflections ate both ends of the cable.  The second advantage is that when the cable is matched at both ends, changing the length of the cable does not disturb the matching.  This can be a great convenience in broadband, instrumentation, and any situation where you do not want to tune the system after the cable is installed. Constant<b style="color: #4400bb;">314</b> (talk) 15:45, 14 June 2022 (UTC)


 * All you said is correct. Thanks for good cooperation.
 * We are in a lead section. So things should not be overly detailed at this early place. We cannot include all your well explained reasons here in talk.
 * Currently, you deleted the not 100% correct sentence. That's an improvement, but the good part's info of the sentence would be nice.
 * How about this version?
 * "For example, impedance matching typically is used to improve power transfer from a radio transmitter via the interconnecting transmission line to the antenna."
 * I'll now so insert it. It may be helpful as a first, coarse info in the lead section.
 * Don't hesitate to change or delete it.
 * DJ7BA (talk) 15:48, 15 June 2022 (UTC)

"... either the source or load has a reactive component" doesn't necessarily make conjugate matching and reflectionless matching look different.
Instead of:

Maximum power transfer matching
Complex conjugate matching is used when maximum power transfer is required, namely



where a superscript * indicates the complex conjugate. A conjugate match is different from a reflection-less match when either the source or load has a reactive component.

It should read:

Maximum power transfer matching
Complex conjugate matching is used when maximum power transfer is required, namely



where a superscript * indicates the complex conjugate of the source impedance

Reflectionless matching
Reflectionless matching is used when reflection suppression on a transmission line is required, namely



where is the characteristic impedance of the transmission line.

This will make the disputed equations visible. No convincing reason for earlier deletion was given.

The bad sentence to be removed sais:
 * “A conjugate match is different from a reflection-less match when either the source or load has a reactive component.”

This is bad because: 1.	Comparing apples with pears is misleading: A conjugate match is always different from a reflection-less match by matching goal, physics, equation, quantities given, and character of real part of impedance type: The real part of source impedance is dissipative, while the real part of the characteristic line impedance is not (or just partially in case of a lossy line) dissipative.

2.	If anything is just looking alike (though being different anyway), it is the equation in the case of a real only source impedance, when

3.	 means that is real only, independant of the load impedance, that does not matter for the look-alike.

4.	If “either the source or the load impedance” can include the case that only the load impedance is reactive, and the source impedance is not, both equations do look alike. For some English (slang?) language backgrounds, however, the word “either” may be ambiguous and could mean “both”, but not for all backgrounds. This is why it is a bad choice of words even in that look alike case.

5. If, however, only the source impedance (or the characteristic line impedance, respectively) is reactive, but he load impedance is not, The equations d o  n o t  look alike. In that case the logic of the bad sentence fails.

5.	A reason, why this apple-pear comparison often goes undisputed is: It is quite common practice to use a common misnomer and call reflection, what rigorously speaking is just conjugate impedance mismatch. Reflection of waves has the necessary prerequisite of a medium for the waves to travel and to be reflected, i.e. a transmission line. But many are inclined to think that both types are indeed reflections in some undefined, never proven way of meaning. The page should not endorse this misleading ambigouity, that probably had influenced the bad sentence.

6. The second bulleted quote in impedance matching devices is not saying the same thing as the bad sentence. As far as I can see, it is 100% correct. If it was the same, but it isn't, the bad sentence would mean doubling, which would not be excellent wiki style.

A final excuse: Because of multiple editors editing at the same time, my edit summary explanation got lost and the deletion looked like an edit war attempt. Sorry for that. I certainly didn't mean edit warring. DJ7BA (talk) 10:20, 17 June 2022 (UTC)


 * I agree that there is scope for confusion here, but I think adjusting the statement is preferable to removing altogether. The article is beginning to lose sight of there being two different things here.  Prior to this edit that was clearly stated in the lead with expressions for both cases.  In fact the entire article needs a rework.  The "Maximum power transfer matching" section leads one to believe that it should be followed by a "Reflectionless matching section", but it isn't.  There is no logical organisation to the headings.  We then go up one heading level to "Power transfer" which has no business being a separate section to "Maximum power transfer matching" and the preceding introduction to the "Theory" section contains no actual theory.  I know that the "Transmission lines" section is essentially about reflectionless matching but that is far from clear from the structure of the article. <b style="background:#FAFAD2;color:#C08000">Spinning</b><b style="color:#4840A0">Spark</b> 18:04, 17 June 2022 (UTC)
 * I cannot argue with that. I have been thinking along the same lines, especially the theory section.  I think some of issues that  is having is caused by the assumption that there is a transmission line between the source and the load.  Except for the transmission line section, where there is an explicit transmission line, Z_source and Z_load are the global impedances seen at the point of interface.  If there are transmiFig. 1 ssion lines, their effects are absorbed into Z_source and Z_load.  We need to make that clear.  I an working on a diagram, but haven't made much progress. Constant<b style="color: #4400bb;">314</b> (talk) 20:17, 17 June 2022 (UTC)

Yes, I agree with both replies. The problem was that by over-reverting to some too early date, and large scale deletions much reflectionless matching subject was co-deleted. Orphaned remainder now doesn't make sense without the lost context. Good contributions simply disappeared, too, in one blow, without explaination or discussion in talk by RF experts. So the article currently is partially truncated.

Yes, rework is needed. This means much more work collecting and arranging good puzzle pieces than quick one-blow deletions. But there is also a good chance for improvement. Yes, we should again have sections for reflectionless matching and for maximum power transfer matching. After these sections we can use an adjusted, distinguishing, comparative statement. Yes, the bad sentence probably was based on three impedances: source, transmission line and a load - as is common practice. An earlier figure had these, but also was victim of co-deleting.

Thanks for understanding. Good cooperation. DJ7BA (talk) 08:29, 18 June 2022 (UTC)

Suggestion for a new section:

Placeholder for S. Roberts Fig. 1 - Equivalent circuit of generator and load

(Naming and index Question: Is it ok by wiki rules, to use our existing, electrically identical circuit - almost a copy of the very similar Roberts circuit - and adjust the equation quantities and indices in the derivation to conform with our existing circuit? Or must we be so narrow-minded and are indeed forced to show two almost identical circuits? Copying the original might perhaps infringe copyrights. Help me, please. As newcomer and I don't know it. And I don't want to get avoidable reverts either.)


 * Yes, you can change indexes to match the figure already in the article. Constant<b style="color: #4400bb;">314</b> (talk) 15:39, 21 June 2022 (UTC)

Reduced power transfer by source and load impedance mismatch
A Thévenin's equivalent circuit of a generator with voltage source E_0 and fixed, complex source impedance Z_0 is terminated by an adjustable complex load impedance Z_1. Maximum real power transfer to the resistive part R_1 of the load occurs, if impedances Z_0 and Z_1 are conjugate complex to each other. S. Roberts in 1946 published a derivation of how real power P in the load resistance R_1 will be affected if the impedances differ from the optimum conjugate values. In the following the asterisk * denotes conjugate complex: $$ P = ii^* R_1 = \frac {e_0} {(Z_1 + Z_0)} \cdot \frac {{e_0}^*} {({Z_1}^* + {Z_0}^*)} \cdot \frac {(Z_1 + {Z_1}^*)} {2} $$

$$ = \frac {e_0{e_0}^*} {2 \cdot (Z_0 + {Z_0}^*)} \cdot \frac {(Z_0 + {Z_0}^*) \cdot (Z_1 + {Z_1}^*)} {(Z_1 + Z_0)({Z_1}^* + {Z_0}^*)} $$

$$ = \frac { e_0{e_0}^*} {4R_0} \cdot \left [1 - \frac {(Z_1 - {Z_0}^*) \cdot ({Z_1}^* - Z_0)} {(Z_1 + Z_0) \cdot ({Z_1}^* + {Z_0}^*)} \right ]$$

where R_0 and R_1 are the real parts of Z_0 and Z_1. This can also be written $$ \frac {P} {P_0} = 1 - \alpha_{10} {\alpha_{10}}^* $$ where $$ \alpha_{10} = \frac {Z_1 - {Z_0}^*} {Z_1 + Z_0} $$ and $$ P_0 = \frac {e_0{e_0}^*} {4 R_0} $$

By analogy with transmission lines, \alpha_{10} is called the "reflection coefficient" of the load as viewed from the generator. It differs from the corresponding reflection coefficient on the ordinary image basis in that the numerator contains the complex conjugate of Z_0. (Emphasis because of reflection coefficient ambiguaty.)

... so far for now, to be continued. DJ7BA (talk) 14:42, 21 June 2022 (UTC)


 * The math looks OK. I could not verify the last step by inspection, but I suspect that it will hold up.  It is OK to define $$ \alpha_{10} = \frac {Z_1 - {Z_0}^*} {Z_1 + Z_0} $$, but the notability of calling it reflection coefficient of the load as viewed from the generator  has not been established.  We would need multiple secondary sources for that.  Johnson is a primary source.  He could simply be coining a term for his own paper.  We need evidence that the term was accepted and became established.  We also need a reliable source for the interpretation for $$ \alpha_{10} $$.  Right now all we have is that it is a symbol useful for simplifying a more complicated expression.  As I say, the math looks fine.  It is just another derivation of the maximum power condition.  It could go in the theory section as a justification of $$ Z_L = {Z_s}^*  $$ for maximum power transfer.  Constant<b style="color: #4400bb;">314</b> (talk) 18:08, 21 June 2022 (UTC)
 * Thank you. Notability is a good point. Authors always slightly differ, though. Nobody wants to be a copy-cat, as narrowly thinking wiki editors might perhaps want them to be. Author's freedom is to be respected. That makes their work not less valuable for being 1st class notable, reliable sources for the important substance (as compared to less important indexing etc. differences).
 * Whom do you mean by Johnson? How did (what) Johnson name it? Any reference? Or did you mean Shepard Roberts, not Johnson perhaps?
 * For S. Roberts, see the .pdf, paragraph right after the equation 3b, second sentence.
 * The correct name used by Roberts is "reflection coefficient" by analogy with transmission lines. The rest "of the load as viewed fom the generator" just simply distinguishes direction, but is not part of the general name.
 * It plays a role later, when you will get (as promised by "to be continued"):
 * Likewise, by analogy, too, $$ \alpha_{01} = \frac {Z_0 - {Z_1}^*} {Z_0 + Z_1} $$ is the "reflection coefficient" of the generator as viewed from the load.
 * Others have done the same sort of power mismatch derivation, but used their own text, referring to what they wrote in their chapters before. Same basic stuff, different naming or indexing - of course.
 * See for example
 * A good source is Cuthbert. He doesn't present the derivation, but uses our already existing figure's (source and load impedance circuit) indexing.
 * He calls it $$ \alpha = \frac {Z_L - {Z_S}^*} {Z_L + Z_S} $$.
 * Is that enough notability?
 * There are more sources mentioning this type of coefficient, especially if you include so called "power waves" on wave guides, where the authors use for derivation the same lumped element serial circuit, too.
 * These are ok for "See also".
 * DJ7BA (talk) 21:16, 21 June 2022 (UTC) DJ7BA (talk) 21:16, 21 June 2022 (UTC)
 * "I could not verify the last step":
 * This comes easy, if you insert in the numerator before this last step $$ Z_0 = R_0 +j X_0 $$ and $$ Z_1 = R_1 +j X_1 $$, then multiply the terms in the numerator, rearrange things and use the fact that $$ (a+b)^2 = a^2 + 2ab + b^2 $$, and that $$ (a-b)^2 = a^2 - 2ab + b^2 $$. The difference between these yields $$ 4ab $$. DJ7BA (talk) 07:10, 22 June 2022 (UTC)
 * This comes easy, if you insert in the numerator before this last step $$ Z_0 = R_0 +j X_0 $$ and $$ Z_1 = R_1 +j X_1 $$, then multiply the terms in the numerator, rearrange things and use the fact that $$ (a+b)^2 = a^2 + 2ab + b^2 $$, and that $$ (a-b)^2 = a^2 - 2ab + b^2 $$. The difference between these yields $$ 4ab $$. DJ7BA (talk) 07:10, 22 June 2022 (UTC)