Talk:Characteristic impedance/Archive 1

Archive 1

Archive 2

Does the definition of characteristic impedance in terms of permittivity and permeability belong here?

Pozar defines the characteristic impedance of a transmission line in terms of the RLGC parameters of the transmission line, and uses the term intrinsic impedance to refer specifically to the relation between the magnitudes of the electric and magnetic fields of a plane wave traveling in an optical medium.

Should this article use the RLGC parameters consistently throughout, perhaps with a statement that the characteristic impedance of a transmission line is analogous to the intrinsic impedance of an optical medium?

Pozar, David (2004). Microwave Engineering (3rd edition ed.). {{cite book}}: |edition= has extra text (help); Unknown parameter |month= ignored (help)

IJW 01:59, 14 September 2006 (UTC)

Sounds reasonable. And if you believe this should be here, then be bold and edit the article!

Atlant 13:08, 14 September 2006 (UTC)

I have edited the article to define the characteristic impedance in terms of RLGC parameters and linked to Medium (optics) instead of defining characteristic impedance in terms of permittivity and permeability. I also removed the section on frequency dependence because it is misleading; R and G are not constant and the frequency dependence of $Z_{0}$ is not as simple as the previous version of the article stated. At AC and higher frequencies $R\sim {\sqrt {\omega }}$ and $G\sim \omega$. Only at very low frequencies (where the thickness of the conductors is comparable to the skin depth) is R relatively constant, but the frequency dependence of G remains. I will revisit this section to include these facts and give a more complete treatment.

IJW 14:41, 14 September 2006 (UTC)

Following is the section on frequency dependence that I deleted:

=== variation with frequency ===

The impedance of a real lossy transmission line is not constant, but varies with frequency. At low frequencies, when

$\omega L\ll R$ and $\omega C\ll G$,

the characteristic impedance of a transmission line is

$Z_{0}={\sqrt {R/G}}$.

At high frequencies where

$\omega L\gg R$ and $\omega C\gg G$,

then the characterstic impedance is

$Z_{0}={\sqrt {L/C}}$.

So there are two distinct characteristic impedances for every line. Usually G is very small so the low-frequency impedance is high, whereas the high-frequency impedance is low. The break points in the impedance frequency graph are at $\omega _{1}=G/C$ and $\omega _{2}=R/L$ (where $\omega =2\pi f$). If $R/G\gg L/C$, it is obvious that $\omega _{2}\gg \omega _{1}$. Between these two break frequencies the cable impedance decreases smoothly.

Example

Take the case of a 50Ω coaxial cable with polyethylene dielectric. R is about 100 mΩ/m and G < 20 pS/m (based on measurements of leakage resistance in a 1 m length). Using $L=CZ^{2}$, L can be calculated at about 250 nH/m. So,

ω₂ = R/L = 200 krad/s (f₂ = 30 kHz)

and

ω₁ = G/C = 0.2 rad/s (f₁ = 30 millihertz)

At 100 Hz the 50 ohm coaxial cable will have an impedance of about 900 ohms, only reaching 50 ohms at about 30 or 40 kHz. The phase angle of the impedance between the two break frequencies is leading (the cable looks capacitive).

IJW 17:11, 14 September 2006 (UTC)

Four comments

• A definition should give the method to measure the item defined. If you define the characteristic impedance as the ratio of voltage to current in the line, you simply cannot measure it. The only place where the measure can be done is at the end of a semi-infinite line. Then, instead of talk about waves, why not define the characteristic impedance as the impedance measured at the end of a semi-infinite line?
  Of course, in the two definitions, there is always the problem of reflections if the line is finite. You can replace the "last infinite length" of the line with an impedance equal to the characteristic impedance. The proposed definition allows this.
  You can also use a finite length of line and make the measure before the arrival of reflections (a pulse generator and an oscilloscope are enough).
• In a transmission line, you cannot have a voltage wave without current or a current wave without voltage. Any wave is voltage plus current. You can write two equations, one for voltage and one for current but they form the same wave. There is not "a pair of waves".
• You can hardly talk of "transmission line" or "characteristic impedance" when the length of the line is negligible compared to the wavelength in the line. You have just a conductor with negligible inductance and stray capacitance. This is the case of IJW example of the coaxial cable at 30 kHz. It just begins to be a transmission line at a length of a few hundreds of meters and then it behaves more as a resistance than as a transmission line. The characteristic impedance depends on frequency. However, when this is the case, in low frequency, the transmission line is no more interesting. This is not the most interesting aspect of transmission lines. LPFR 12:17, 15 September 2006 (UTC)
• The vacuum characteristic impedance should be mentioned here,related to the rf domain where the carcateristical impedance is very important. —The preceding unsigned comment was added by 194.138.39.55 (talk • contribs) .

Impedance of a transmission line or impedance of an electrical circuit is the ratio of a voltage divided by a current, both of them measurable quantities. Impedance of vacuum or impedance o a substance is something ($\scriptstyle {\sqrt {\mu \over \varepsilon }}$) related to the properties of the substance in an electromagnetic field. It just happened that the units of this value are ohms and people could not avoid calling it "impedance". Impedance of a line and impedance of vacuum are very different things. The difference is still greater if you think of transmission lines built with discrete inductors and capacitors used (in the past) as delay lines. LPFR 12:05, 1 October 2006 (UTC)

Electrical and electromagnetic impedances

As Mebden himself wrote in the page "intrinsic impedance", electrical impedance and electromagnetic impedance should not be confused. The impedance of a transmission line is an electrical impedance and the impedance of a medium is an electromagnetic impedance. LPFR 08:51, 23 October 2006 (UTC)
Impedance of a transmission line or impedance of an electrical circuit is the ratio of a voltage divided by a current, both of them measurable quantities. Impedance of vacuum or impedance o a substance is something ($\scriptstyle {\sqrt {\mu \over \varepsilon }}$) related to the properties of the substance in an electromagnetic field. It just happened that the units of this value are ohms and people could not avoid calling it "impedance". Impedance of a line and impedance of vacuum are very different things. The difference is still greater if you think of transmission lines built with discrete inductors and capacitors used (in the past) as delay lines. LPFR 12:05, 1 October 2006 (UTC)

Meaning of high or low characteristic impedance

Would it be accurate to add this (e.g. to the introduction): "A high-quality (high conductance) transmission line tends to have a low characteristic impedance, and vice versa." (Or is it the other way around?) --Coppertwig 13:18, 11 January 2007 (UTC)

Electrical conductance is pretty orthogonal to impedance; you can design a transmission line in a wide variety of impedances (to suite the need) although some impedances are a lot "easier" (natural for the materials employed?) than others. So I guess I disagree with your proposed addition.

Atlant 13:41, 11 January 2007 (UTC)

My goal here is to have this page and related pages improved to the point that a person similar to myself can quickly and correctly understand the concepts being presented. So if something seems unclear or contradictory, that means it needs to be edited.

For now, I'm thinking in terms of transmission lines with zero conductance and zero inductance. The equation given for characteristic impedance in that situation seems to me to reduce to being equal to the resistance of a unit length of the transmission line.

I really like the mention of the infinitely long transmission line in the opening paragraph: it appeals to the intuition in a simple, relatively easily understandable way. However, by itself it isn't enough; I'd like at least one more sentence with similar simplicity and appeal but providing complementary information.

Problem: The first paragraph seems to be claiming that the characteristic impedance is equal to the impedance (resistance, in the case I'm considering) of an infinitely long piece of transmission line, while the equation seems to reduce (in the case I mentioned) to the resistance of a unit length of transmission line. Those can't both be true, can they? It seems to me that there's something wrong. (If they can both be true, this needs to be explained in the article.)

Also, if you're talking about a transmission line with two conductors, (such as often plug into electrical appliances), then with the infinite one you're attaching to both of the conductors, whereas when measuring the impedance of a unit length, it would seem to make sense to measure only one of the conductors at a time. This clouds the issue.

Question I'd like to see answered in the first or second paragraph of the article: which has a larger characteristic impedance as a transmission line: a pair of thick copper wires, or a pair of thin copper wires (straight, separated by an insulator)? Again, I'm thinking in terms of the resistive part of the impedance. (After I understand that, I might tackle the imaginary part.) Doubling the cross-section of the copper wire cuts its resistance in half, I believe. Or do they both have the same characteristic impedance? (I don't think they do.) The answer to this question is basically the same thing as the sentence I proposed at the beginning of this discussion.

Could somebody just give a few examples? Would a coaxial cable tend to act as a capacitor at high frequencies, for example? What happens to the characteristic impedance when you double the thickness of the conductor of a coaxial cable?

I mean: if there's anybody out there who understands what characteristic impedance is, could you please explain it more fully and illustrate it with examples? Thanks. --Coppertwig 04:44, 12 January 2007 (UTC)

Characteristic Impedance? What's infinity got to do with it? ____________________________________________________________ —Preceding unsigned comment added by 92.40.34.110 (talk) 11:25, 6 September 2009 (UTC)

I wish people would avoid talking about "infinite" lines when discussing Z0. Has anybody ever seen one?

Its true, that Zin of a line is equal to "Z0" multiplying a quotient, containing the hyperbolic tangent of the product of length of the line and the propagation co-efficient (easily derived from the transfer matrix of a line). If you let the length of the line tend to infinity, then the quotient tends to unity and one is left with Zin = Z0. Which is all very well mathematically, but it has never been shown practically, because of the problem of obtaining, for example, an infinite length of 50 ohm coaxial cable!

There is a much better definition of Z0, but which requires knowledge of iterative impedance and image impedance. As follows.

One can always find an impedance which when connected to the output terminals of any two port network (including a transmission line), that will give the same impedance, measured at the input terminals. This is called the "iterative impedance" of the network Zit1. Similarly one can always find a suitable generator, whose source impedance, when placed at the input terminals of a two port network will give the same impedance, measured at the output terminals of the network. This is also an iterative impedance, Zit2.

If the network is symmetrical, i.e the determinant of the transfer matrix is unity, then Zit1 = Zit2 = Zit.

Similarly the "image impedance" of a two port network, is that input impedance (and is the complex conjugate)of the generator source impedance, due to a load at the output terminals, and causes maximum power to be transferred from the generator to the network, Zim1. Similarly if the output impedance of the network is equal to (and is the complex conjugate of) the load impedance, then maximum power will be transferred from the network to the load, Zim2. For a symmetrical network, Zim1 = Zim2 = Zim.

And now for the definition. If (and only if) for a symmetrical network, the case that the "iterative impedance" is equal to the "image impedance", this is known as the "characteristic impedance" of the network, and is given the symbol Z0. Z0 = Zim = Zit.

Note, it doesn't matter if the network is a piece of coax cable a mile long, or three resistors connected in a "T" configuration, the definition is still the same. This is charactersitic impedance, and doesn't require the mention of the word infinity.

Phil Robinson —Preceding unsigned comment added by 92.40.34.110 (talk) 11:09, 6 September 2009 (UTC)

However much you might not like it, the infinite line is a commonly accepted way of defining characteristic impedance and it is not Wikipedia's place to change the world. Do any reliable sources use your definition? I have some problems with it - if Zit is always equal to Zim then it is overcomplicated, on the other hand if you are claiming there are cases when Zit != Zim then that implies there are cases when characteristic impedance is undefined, which is nonsensical. An alternative definition which is widely found in the sources, and does not require infinite lines, is that characteristic impedance is the impedance encountered by a wave travelling in a single direction. This is the definition that the article opens with so I don't really see the problem. SpinningSpark 08:14, 1 June 2010 (UTC)

"Do any reliable sources use your definition?"

Yes see "Advanced Electrical Engineering" by AH Morton Phil Robinson — Preceding unsigned comment added by 94.72.252.35 (talk) 11:24, 20 October 2016 (UTC)

Another way of looking at/understanding Characteristic impedance

This impedance, remains the same, no matter how long the line is, because the ratio of voltage applied to the current, remains the same, but their actual values reduce, along the length, of a lossy line.

Is the above statement false?, please tell me how to correct it, is the one below better?

This impedance, remains the same, no matter how far along a uniform line, you measure it, because the ratio of voltage applied to the current, remains the same, but their actual values reduce, along the length of a lossy line.

or

This impedance, remains the same, no matter how long the line is, even if it is infinite, because the ratio of voltage applied to the current, remains the same, but their actual values reduce, along the length, of a lossy line.

Bookbuddi (talk) 17:05, 15 April 2012 (UTC)

At the place you tried to add this the article defines characteristic impedance as the input impedance of an infinitely long line (also uniform line is meant but not stated). It makes no sense at all to talk about measuring the characteristic impedance of other lenghts of line if we are taking that as the definition - by definition it is constant since it is defined in terms of an infinite line for any length.

If you were to measure the impedance some way along an infinite line you would not measure Z0; if the line extended to infinity in both directions you would measure Z0/2, otherwise some other impedance. If you were to cut the line some distance from the source and then measure the impedance looking to the right you would indeed measure Z0, but that is still the input impedance looking into an infinite line. The impedance looking to the left (the finite portion of line) would not be Z0 unless the line was terminated in Z0 - that is, you would need to know the quantity you were trying to measure prior to measuring it.

I think what you are trying to say is something like what is already in the lede "Z0, is the ratio of the amplitudes of a single pair of voltage and current waves propagating along the line in the absence of reflections." The ratio remains constant because Z0 is constant, not the other way around. The limitation to waves travelling in one direction only is essential. Applied voltages and currents will generally not be in the ratio Z0 because of reflections. SpinningSpark 19:01, 15 April 2012 (UTC)

I am trying to give a practical example of what happens with real non infinite lines (to make the concept easier to understand, as we don't have any infinite lines to measure). so can I say:

On a real, non-infinite terminated line, we can see that the impedance remains the same, no matter how far along the line you measure it, because the ratio of voltage applied to the current remains the same, but their actual values reduce, along the length of a lossy line. Bookbuddi (talk) 19:32, 15 April 2012 (UTC)

Edits by Zolot

Zolot, the characteristic impedance of a lossless transmission line is real, i.e., resistive. For example:

The impedance is determined by the speed of the signal and the capacitance per length of the pair of conductors, both intrinsic properties of the line. This intrinsic impedance is termed the characteristic impedance of the line (Z0).

If a measurement is made at one end of the line in a short time compared to the round trip time delay, the line behaves like a resistor with a resistance equal to the characteristic impedance of the line. — Preceding unsigned comment added by Alfred Centauri (talk • contribs)

I just reverted this edit, which appears to be aimed at clarifying Zolot's misunderstanding. My edit summary was truncated by Twinkle because it was too long. For the record, the full summary was "It is not essential to introduce complex numbers to summarize this phenomenon and doing so requires more prior knowledge from the reader. Thinking that Z0 is connected to line resistance is mistaken and "resistive" helps the reader understand this mistake." SpinningSpark 18:05, 23 November 2013 (UTC)

Would like to see considerations on the characteristic impedance of transmission lines connected in parallel. The article does not touch the subject nor calculations. Above, there is an edit/comment "Zolot, the characteristic impedance of a lossless transmission line is real, i.e., resistive." and "the line behaves like a resistor with a resistance equal to the characteristic impedance of the line" Then, two equal lenght, equal 50 ohm impedance transmission lines would exhibit 25 ohms ?. Then what about equal 50 ohm impedance transmission lines in 'series' ? 75.249.27.246 (talk) 18:17, 4 February 2015 (UTC)MB

75.205.89.30 (talk) 18:09, 4 February 2015 (UTC)MB

There is no good reason for the article to cover this subject, it is not about connections of transmission lines. The characteristic impedance of a line will not change regardless of how it is connected, it is an intrinsic property of the line. However, the impedance presented to the source may well change with connection arrangement. See reflections on copper lines and stub (electronics) for more details. And to answer your question, yes, two 50 ohm lines in parallel would exhibit 25 ohms and in series 100 ohms. SpinningSpark 19:25, 4 February 2015 (UTC)

Surge Impedance a subset (barely) of Characteristic Impedance

I know this is splitting hairs, but this article probably needs to address the differences between Characteristic Impedance and Surge Impedance. I have numerous references, but the basic reference to trigger some thought is the IEEE-STD-100-1984, IEEE Standard Dictionary of Electrical and Electronics Terms. On page 136 we find...

Characteristic Impedance

(1)Data transmission

(1.A) Two-conductor transmission line for a traveling transverse electromagnetic wave

The ratio of the complex voltage between the conductors to the complex current on the conductors in the same transverse plane with the sign so chosen that the real part is positive.

(1.B) Coaxial transmission line

The driving impedance of the forward traveling transverse electromagnetic wave.

[. . .]

(6)Surge Impedance

The driving-point impedance that the line would have if it were of infinite length. Note: It is recommended that this term be applied only to lines having approximate electric uniformity. For other lines or structures the corresponding term is iterative impedance.

Then on page 904 we have the standalone definition of surge impedance...

Surge impedance (self-surge impedance)

The ratio between voltage and current of a wave that travels on a line of infinite length and of the same characteristics of the relevant line. See: characteristic impedance.

The first sentence in this article suggests characteristic impedance and surge impedance are the same, but they have separate definitions for a reason. Few in the RF industry would be confused, but if the IEEE makes a distinction so should we I propose. Crcwiki (talk) 18:06, 28 September 2015 (UTC)

It is easy to see why a distinction has been made between surge impedance and iterative impedance. The latter is more usually applied to lumped component networks. That is, networks that are finite in extent. It does not make mcuh sense to talk of the characteristic impedance of an infinite line that is not uniform. For instance, take a line that has a steadily increasing characteristic impedance. Let's say that it starts off at 50 ohm and uniformly increases to 75 ohm at the other end. The correct terminating impedance is 75 ohm, not 50 ohm. This is the same as terminating the line with an identical line of the same length but reversed in direction and that line also so terminated and so on. This is the very definition of iterative impedance.

The distinction between conductor pairs and coaxial is a little more intriguing. I am guessing that this is to do with the propagation in twisted pair not being entirely in TEM mode due to the slight circular mode imposed by the twist. The characteristic impedance of non-TEM waves is mode dependent and there is also not one unique definition. These are the modes that exist in waveguides, but as the hatnote makes clear, waveguide modes are not within the scope of this article. It has to be said, though, that waveguide modes can exist in coaxial cable - they just ignore the central conductor. SpinningSpark 20:51, 28 September 2015 (UTC)

I’m just joining the conversation. I would point out that a dictionary is a catalog of the way a word is used rather than being an authority on what the word means. The multiple definitions come from different sources. They don’t necessarily mean different things. I have reproduced part of the definitions from the 1997 edition.

characteristic impedance (1) (A) (data transmission) (two-conductor transmission line for a traveling transverse electromagnetic wave). The ratio of the complex voltage between the conductors to the complex current on the conductors in the same transverse plane with the sign so chosen that the real part is positive. (B) (data transmission) (coaxial transmission line). The driving impedance of the forward-traveling transverse electromagnetic wave. (PE) 599-1985w

(4) (two-conductor transmission line) (for a traveling transverse electromagnetic wave). The ratio of the complex voltage between the conductors to the complex current on the conductors in the same transverse plane with the sign so chosen that the real part is positive. (Std100)

(5) (coaxial transmission line) The driving impedance of the forward-traveling transverse electromagnetic wave. (MTT) 146-1980w

my comment: a coaxial line is also a two conductor line

I don’t know, but surge impedance seems to be a term used by the electric power transmission and distribution community. Its most common usage seems to be the same as the most common usage of characteristic impedance.

surge impedance (1) (self-surge impedance) The ratio between the voltage and current of a wave that travels on a line of infinite length and of the same characteristics as the relevant line. See also: characteristic impedance. (PE) [8], [84]

(2) The impedance of an electrical circuit under surge conditions (which may differ significantly from the impedance of a circuit under steady state conditions). (PE) 1143-1994

(3) The ratio between the voltage and current of a wave that travels on a conductor. (PE/SUB) 998-1996

driving-point impedance (networks) At a pair of terminals the ratio of an applied potential difference to the resultant current at these terminals, all terminals being terminated in any specified manner. See also: self-impedance. (EEC/PE) [119]

Categories:

PE= Power engineering

EEC = Electrical Equipment and Components

MTT = Microwave Theory and Techniques

Std100 = older terms for which no category could be found.

SUB= Substations

Standards:

599-1985w This standard includes definitions of those terms deemed necessary of performing or understanding measurements on channels used for data communications. terms that are applicable to audio interfaces of wireline, leased lines, and multi-channel equipments and baseband interfaces of multichannel equipment.

1143-1994 = Guide on Shielding Practice for Low Voltage Cables

998-1996 = IEEE Guide for Direct Lightning Stroke Shielding of Substations

The fact that there are multiple definitions don’t require anything to be done to the article, but an article or section on surge impedance including surge impedance under surge condition might be useful. Constant314 (talk) 14:31, 10 October 2015 (UTC)

References

Lossless line

I thought that the section on the lossless line could use some context for why lossless lines are considered during transmission line analysis. I added some motivation behind the lossless line model, and discussed some implications of analyzing lossless lines. Prayerfortheworld (talk) 08:57, 8 December 2015 (UTC)

Cumbersome sentence

" It can be shown that an equivalent definition is: the characteristic impedance of a line is that impedance which when terminating an arbitrary length of line at its output will produce an input impedance equal to the characteristic impedance."

I'm not sure the sentence's logic can be followed - it seems circular, in the end using "characteristic impedance" itself to define "characteristic impedance". If there is logic (that I cannot see), perhaps it can be reworded?

Mike (talk) 17:44, 1 February 2016 (UTC)

It is not circular logic, which is a logical fallacy, but a circular definiton, which is not only acceptable in engineering, but also highly useful. It is even accepable in mathematics: if I define x as

$x={x^{2}+1 \over 2}$

then I have defined x in terms of itself, but nevertheless, x has a well-defined value (x=1). In the case of the transmission line, defining it in this way gives a practical method of measuring the characteristic impedance. We can't use the definition involving an infinite line because infinite lines are not possible to manufacture. This alternate definition of characteristic impedance requires only a finite length of line. It also provides a means of calculating the characteristic impedance in theory from the primary line constants of an infinitesimally short piece of line. Much simpler than trying to integrate over an infinite line. SpinningSpark 19:19, 1 February 2016 (UTC)

Assessment comment

The comment(s) below were originally left at Talk:Characteristic impedance/Comments, and are posted here for posterity. Following several discussions in past years, these subpages are now deprecated. The comments may be irrelevant or outdated; if so, please feel free to remove this section.

I don't see a reason why this page cannot be unified with the page dealing with "wave impedance" Characteristic impedence in electrical transmission lines is an important topic in electrical engineering. Though it is a particular case of wave impedence, it is an important one from a practical standpoint. I think it is worthy of it's own page, where the details can be discussed. Merging it with wave impedence would not leave sufficient space to explain (to the general audience) how electrical signal transmission relates to with wave impedence, perhaps with links to the more detailed wave impedence page. This page provides the opportunity to describe the composition and nature of the transmission line as a wave medium. Zolot (talk) 19:35, 16 November 2012 (UTC)

Last edited at 19:35, 16 November 2012 (UTC). Substituted at 11:16, 29 April 2016 (UTC)

External links modified

Hello fellow Wikipedians,

I have just modified one external link on Characteristic impedance. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20150906040834/http://www.ewh.ieee.org:80/r5/denver/rockymountainemc/archive/2000/diffimp.pdf to https://www.ewh.ieee.org/r5/denver/rockymountainemc/archive/2000/diffimp.pdf

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 17:04, 19 November 2016 (UTC)

Recent undo

Constant314, regarding your last undo you may be right, so I just want to understand. Consider the following scenario: you have a transmission line opened at one end, and you apply a single pulse at the other extremity. Before the pulse has reached the opened end, you measure the current I and voltage V at some point. The ratio V/I is the characteristic impedance. So far so good. Now, you repeat the same experience but with a constant DC voltage of 1V applied together with the pulse (meaning you have to apply the DC voltage before the pulse and wait some time the voltage stabilises). Since the line is opened, the DC voltage causes no DC current to flow, and the characteristic impedance you measure is now (V+1)/I. Isn't it a contradiction?maimonid (talk) 11:34, 22 January 2018 (UTC)

Since this is a discussion of characteristic impedance instead of a discussion about the article, I have responded on your talk page. Constant314 (talk) 12:57, 22 January 2018 (UTC)

Constant314: OK for the maths but this was not the point. I now understand what has confused you in my previous contribution. I hope you will also understand my point of view. Let me express it in the following way. I suggest to replace the sentence:

"The characteristic impedance of a transmission line is the ratio of the voltage and current of a wave travelling along the line." by

"The characteristic impedance $Z(\omega )$ of a transmission line at a given angular frequency $\omega$ is the ratio of the voltage and current of a pure sinusoidal wave of the same frequency travelling along the line. This definition extends to DC by letting $\omega$ tend to 0." Do we agree? maimonid (talk) 13:54, 22 January 2018 (UTC)

That is pretty good. My main concern is that there be no implication that the definitions do not include DC. But since were changing it, lets make it clear that it is the ratio of the voltage of the forward wave to the current in the forward wave, and/or the ratio of the voltage of the reverse wave to the current in the reverse wave, but not the ratio of the combined voltage to the combined current unless there is only one wave. You can get a reverse traveling wave if you have reflections at the load, or if there is a source at the load end (bi-directional use of the transmission line). Constant314 (talk) 14:56, 22 January 2018 (UTC)

OK I will take that into account. Check what I write in the article and eventually improve it. maimonid (talk) 16:42, 22 January 2018 (UTC)

It is satisfactory. Constant314 (talk) 08:21, 23 January 2018 (UTC)

Table of Practical Examples

The table of practical examples cites an NXP app note which references an Intel motherboard reference for designing the impedance of traces.

These are inconsistent with the specification. For example, HDMI according to spec v1.3a (which is freely available to download but you have to register) says the "cable area" should have 100 ohms +/- 10%, not 95 ohm +/-15%.

The subtlety here is that Intel has published some papers which indicate that emperically, designing a motherboard with a slightly lower impedance than specification can lead to improved performance (see "Improve Storage IO Performance by Using 85Ohm Package and Motherboard Routing, https://ieeexplore.ieee.org/document/5642794). However, the table is unclear on the fact that the specification actually calls for 100 ohms.

i would suggest revising the table to have an extra column: one for spec number, one for recommended PCB/package design. It's helpful to have both on hand. But as the table is, if it's meant to portray the actual committee-approved spec for impedance of the standards, it's wrong. — Preceding unsigned comment added by 132.147.66.42 (talk) 07:48, 5 October 2018 (UTC)

If you are are sure, then go ahead and change it to 100 ohms +/- 10%. You can cite the HDMI spec even if it is behind a paywall.Constant314 (talk) 07:54, 5 October 2018 (UTC)

Opening sentence

I reviewed the previous relevant talk.

This:

"...the ratio of the amplitudes of voltage and current of a single wave propagating along the line; that is, a wave travelling in one direction in the absence of reflections in the other direction. "

Is not correct because it is true regardless of reflections. By saying it is the ratio of a wave traveling in one direction automatically removes the effect of a reverse wave. The presence of a reverse wave has no effect on the forward wave. The net voltage and current, however, are another story as I assume everyone understands. In fact, a wave can ONLY travel in one direction, so it even seems redundant to say so; other than to say "traveling in either direction". It would be in full:

The characteristic impedance or surge impedance (usually written Z0) of a uniform transmission line is the ratio of the amplitude of voltage to current of a single wave propagating along the line; that is, a wave travelling in either direction."

Comments? -- Steve -- (talk) 01:10, 17 June 2019 (UTC)

I think the intent was to to make clear the meaning of a single propagating wave rather than a condition on the meaning of characteristic impedance.Constant314 (talk) 02:38, 17 June 2019 (UTC)

That's right, it's what the amplitudes would be in the absence of reflections, not that there must be no reflections. I wouldn't object to that clarification. However, I think we should avoid any phrasing using "either direction" because that becomes troubling in an anisotropic medium. SpinningSpark 13:28, 17 June 2019 (UTC)

A) If that is the intent, I suggest it is misplaced.

B) It may be the (only) ratio amplitudes SEEN in the absence of reflections, but that isn't good to add at the beginning.

I must admit that I felt a bit uncomfortable with the "either direction" phrase as I tried to figure out what to put there, but the presence of a reflected wave does NOT alter the V/I ratio of the forward wave. Adding the part about "in the absence of a reflected wave" implies the V/I ratio of the forward wave IS (or may be) modified by the reflected wave and it isn't. I saw the discussions above and still think that adding too many things is confusing. A T-line is a one dimensional medium; can have waves in two directions and each wave 'sees' the Z0. This is an important concept that must be made clear from the very start.

It is important to get the fundamental principles solidly understood before introducing the advanced complexities. I see the propagation of waves in a T-line as part of an article explaining the broader subject T-Lines, not just Z0.

The net, or feed-point impedance seen due to reflections is another much more complex subject.

I could see that being more explicit and explaining that in the absence of a reflected wave this ratio is the ONLY and, therefore, the NET or gross ratio, but still feel that should be left until later. That is another complexity not actually part of the fundamental concept of Z0.

Perhaps a side issue, but related, I also found the distinction between Characteristic and Surge impedances above to be very strange because they are not different things, even acknowledging the seemingly authoritative references quoted.

Regards, -- Steve -- (talk) 15:22, 17 June 2019 (UTC)

Steve, you don't need to keep explaining the principles with all caps emphasis. I'm sure we all understand the basics here on this page. I would guess that the editor who put in the bit about reflections (@PerceptualChaos:) had in mind that the quantities directly measurable on the line are the total voltage and current. The voltages and currents associated with each wave are to some extent notional and not directly measureable. I kind of agree with you that this is a distraction. In my opinion the simplest thing to do is just remove the statement after the semicolon. Or at least explain it separately. SpinningSpark 19:58, 17 June 2019 (UTC)

Spinning, Just trying to make the emphasis as I would talk and don't know how to do bold or italics.

The individual waves are indeed directly measurable. Devices called "SWR bridges" do it all the time using two common methods and sampling directional couplers are another distinct common method. I have some of each type, myself.

I agree with removing that part. Regards, -- Steve -- (talk) 20:51, 17 June 2019 (UTC)

Well then we have different concepts of what constitutes direct measurement. In any case, I was only attempting to divine what the original editor was thinking. SpinningSpark 15:47, 19 June 2019 (UTC)

You need to keep the typical reader in mind, especially in the lede. You cannot just say wave voltage and wave current because that doesn't mean anything to the typical reader. He barely understands the idea of a wave and the ideal that there could be two waves with different voltages and current at the same point is mysterious. However, if there is one wave, then the ordinary concept of voltage and current is good enough. One way to ensure that there is one wave is if the transmission line has a reflectionless termination. It is important to give the reader a concrete example. But, you can word smith it in such a way that it doesn't exclude the other notions of wave impedance. Constant314 (talk) 16:03, 19 June 2019 (UTC)