Talk:Gyrator

Archives (Index)

1

This page is archived by ClueBot III.

Grounding

The circuit happens to be drawn connected to ground. There is no such restriction as far as I'm aware; why couldn't we float the gyrator? Oli Filth^{(talk|contribs)} 17:21, 7 April 2010 (UTC)

Of course, we may float the circuit but together with the power supply. See for example, how I have floated a VINIC in the story about negative resistance (BTW, I created it a year ago after Zen-in mutilated the Wikipedia page about negative resistance). Circuit dreamer (talk) 17:50, 7 April 2010 (UTC)

At the risk of starting a discussion about the circuit itself, why does the power supply matter? We could just as well connect R to any other voltage, and the analysis for Zin would be the same. Oli Filth^{(talk|contribs)} 17:52, 7 April 2010 (UTC)

I hope you will see the same powerful idea (modifying an initial 2-terminal real element to obtain a new artificial element) behind negative impedance converters.

I'm not sure if your clever "shifting" trick will allow us to connect the simulated inductor in series (between) two other circuit components. I need time to think about it. Circuit dreamer (talk) 18:00, 7 April 2010 (UTC)

Circuit dreamer: I object to the above statement reading in part "(BTW, I created it a year ago after Zen-in mutilated the Wikipedia page about negative resistance)". To refresh your memory, most of the edits on the negative resistance page were done by Spinningspark. The impetus for doing these same edits was the consensus reached that the article in question had too much WP:or and idiosyncratic POV. This was about the same time you were asked to change your username. So I did not "mutilate" this article any more than several other editors. Statements like this are unacceptable. An apology and a retraction on this discussion page is in order. Perhaps you need a wiki-break as well. Zen-in (talk) 00:11, 8 April 2010 (UTC)

I would like to remove the note saying that gyrator-based inductor simulations can not be used in a low-pass filter, on grounds that it is probably false. It is a claim about what can not be done, which can only be taken seriously if backed by a mathematical proof whether in the article or by reference (since otherwise one at most learns that the *author* doesn't know how to do it.) Thoughts? But of course if the impossibility of using gyrator-based inductor simulators in low-pass filters has indeed been proven (which I doubt) then a better fix would be to reference that proof. Bmord (talk) 15:55, 29 September 2010 (UTC)

It's certainly true that the op-amp circuit shown could not be used as a series element in a classic unbalanced ladder network filter, which rules out low-pass applications. But op-amps aren't the only way to do gyrators and ladders are not the only way to do filters. SpinningSpark 12:30, 1 October 2010 (UTC)

Demystifying gyrator circuits

I start this discussion here to demystify, once and for all, the basic idea, implementation and operation of these exotic circuits. I will use the text below as a base (I wrote it two days ago):

"Gyrator circuits imitate real elements by dynamic voltage sources with swapped instantaneous values of the voltage and the current (the voltage across the new virtual element is proportional to the current flowing through the initial real element and the current flowing through the virtual element is proportional to the voltage across the real element). These voltage sources (gyrator's outputs) are connected in opposite direction to the exciting input sources as they mimic voltage drops (in contrast, negative impedance converters with voltage inversion produce voltages). So gyrators are not only positive impedance inverters; they can "invert" in this manner any elements (linear, non-linear, time-dependent, sources, etc.) connected as a load..." Circuit dreamer (talk) 18:20, 11 April 2010 (UTC)

Impedance is a misleading concept here

The lede says "the gyrator is a positive impedance inverter" and this is true, especially in the case of a simulated inductor that converts a capacitive reactance into an inductive reactance (a capacitor into an inductor). But the problem is that impedance "impedes" the understanding gyrator circuits. What is the problem?

The problem is that impedance Z = V/I is defined as a ratio of the rms ("effective") values of the voltage and the current (see hyperphysics). Being some kind of averaged quantity, impedance hides the concepts behind all these odd circuits - gyrators, multipliers, negative impedance converters, etc. Impedance viewpoint does not allow us to realize what all these circuits actually do. As an example, you may see the result of this misleading "impedance approach" in the area of negative impedance if you browse these archived talk. Then I (Circuit-fantasist) wasted plenty of time to show how simple the idea behind negative impedance is but I didn't manage. Now I realize why. The reason of this failure to understand one another was simple - I was thinking in terms of instantaneous values while my opponents were thinking in terms of averaged (rms) values.

The article says what a gyrator does (inverts an impedance) but it does not say how it does this magic. But the idea behind this mystic circuit is extremely simple; we will grasp it immediately if only we forget the misleading concept impedance and begin thinking in terms of instantaneous voltage and current quantities! This is the remedy - to imagine what the particular voltage and current are at each moment, to think "instantaneously":)

What is actually a gyrator?

To understand what the particular gyrator is, it is extremely useful to understand what all related exotic circuits are. Multipliers, gyrators and negative impedance converters are used to create virtual elements that are accordingly multiplied, inverse (dual) and negative copies of actual elements. These virtual elements only mimic the original elements; they have not the same nature as the initial elements since they are implemented as dynamic voltage sources. Actual capacitors and inductors create voltage drops: a capacitor impedes the input voltage source by subtracting a voltage drop V_C from the input voltage; an inductor impedes the input voltage source by creating a back emf V_L and subtracting it from the input voltage. So, to simulate these behaviors, the "multiplied" and inverse virtual elements (voltage sources) are connected in opposite direction to the exciting input source to subtract a voltage drop while negative impedance converters (VNIC) are connected in the same direction to the input source to add a voltage.

What does a gyrator actually do?

A gyrator does only a simple donkeywork: it continuously "observes" the instantaneous values of the voltage across and the current through the original element (model, sample, pattern, load here...) and makes its own instantaneous output voltage proportional to the initial current and its current proportional to the initial voltage. Thus, the combination of the gyrator and the connected actual load acts as a dual 2-terminal virtual element. Shortly, a gyrator is a dynamic voltage source with swapped initial voltage and current.

You can guess that a multiplier does the same but without inversion - e.g., it makes its own output voltage equal to the initial voltage and its current proportional to the initial current. For example, a capacitance multiplier emulates an actual capacitor with an opposing dynamic voltage source (see AN-29, page 11, fig. 19 and try to guess how this clever circuit operates). Finally, you will probably realize that a negative impedance converter with voltage inversion (VNIC) is a dynamic voltage source emulating a negative resistor by adding a voltage that is equal to the voltage drop across the initial element (see this Chua's material and try to see the clever idea behind this circuit). But let's return to the gyrator.

Imagine how simple it is! The gyrator does not "know" what it converts; it is not "interested" in what an initial element is connected. The only thing that a gyrator can do is to observe the input (load's) voltage and current to produce an output (simulated) current and voltage.

Following this simple procedure, a gyrator can "invert" not only classical impedance elements (capacitors and inductors); it can invert elements of all kinds including non-linear ones. For example, a gyrator can convert a diode or a varistor (voltage-stable elements) into a transistor or a baretter (current-stable elements) and v.v. I suppose a gyrator can invert a negative impedance (negative capacitor into a negative inductance) and this extremely odd combination will have some fancy name (e.g., "injectoplactor":) It turns out a gyrator can convert almost everything. If even we put some wikipedian (e.g., Zen-in/Circuit dreamer) at the place of the load (this is only a joke!), the gyrator will convert the poor wikipedian into his/her inversion (Circuit dreamer/Zen-in). It is wonderful, isn't it?

How does a gyrator do this magic?

(to be continued)... Circuit dreamer (talk) 9:20 pm, Yesterday (UTC+3)

I think what you are trying to do is to apply a type of natural philosophy reasoning to electronic circuits. This way of reasoning was accepted in the 17th-19th century until it was gradually replaced by scientific methods with mathematical analysis. In those days even well-regarded scientists like Sir Isaac Newton sometimes talked in terms of metaphors, etc. However this way of thinking has to be adjusted for every new case. So while a gyrator has this interesting "inversion" property that can be spun up into all manner of interesting metaphors, they aren't useful for analyzing the circuit incrementally or for designing a different circuit. And there are all kinds of exceptions and counter-examples. One counter-example is the case of putting an active element in the feedback path of an op-amp. That results in the mathematical inversion of the active element's system function. Look at a Log amplifier. It exploits the exponential system response of a semiconductor junction. Another example is when you put a multiplier (X*Y) in the feedback path. That produces a 1/(X*Y) function. Mathematical analysis is how circuits are described, analyzed, and designed. Zen-in (talk) 20:02, 11 April 2010 (UTC)

This is a happy example; it illustrates the unique negative feedback feature to reverse the causality in circuits. Read more about it in this wikibooks talk and in this story about the famous current mirror. IMO, it is high time to realize that understanding, explaining, inventing, analyzing and designing are different things; so, they need different means. Formal analysis will never replace human imagination (see more about the topic here). Circuit dreamer (talk) 20:23, 11 April 2010 (UTC)

I think you should desist in the POV pushing of your wiki-pages. They aren't acceptable as references in articles, as you found out in October, and so it is not a good idea to continue trying to push them in discussion pages. Zen-in (talk) 21:11, 11 April 2010 (UTC)

"The term POV pushing is primarily used in regard to the presentation of a particular POV in an article and generally does not apply to talk page discussions." Circuit dreamer (talk) 21:50, 11 April 2010 (UTC)

BTW, I have visited Log amplifier but "I haven't managed to see the forest for the trees":( The talk page was much more interesting, especially in the beginning. As you can see, our visitor is not satisfied "how the circuit is described, analyzed and designed" and wants to know "How this log amplifier works????????" Circuit dreamer (talk) 22:22, 11 April 2010 (UTC)

When you post links to your wacky-book pages on a talk page it becomes POV pushing, IMHO. Unsigned comments are usually ignored. They are often just crude attepts attempts at sock-puppetry. Some electronics articles are too advanced for the casual reader and it is not worthwhile to dumb them down. Zen-in (talk) 22:31, 11 April 2010 (UTC)

I'm not sure what you're hoping to achieve with all this text that you keep adding to talk pages? Oli Filth^{(talk|contribs)} 08:04, 12 April 2010 (UTC)

Incidentally, impedance is not defined in terms of RMS values; it is defined as the complex ratio between the voltage phasor and current phasor. Oli Filth^{(talk|contribs)}

Actually, I think I understand what you may be trying to say. Essentially: "a gyrator inverts all I-V relationships. In the case of an LTI component (resistor, capacitor, inductor), the effect of this is to invert the impedance. Therefore, the article should explain the gyrator in terms of I-V relationships, not impedances (which are a specific case)." Is that what you mean? Oli Filth^{(talk|contribs)} 07:10, 13 April 2010 (UTC)

Yes! Yes! Yes! This is exactly what I want to say! Impedance is not a suitable concept here as it is a some kind of "derivative" quantity based on the particular voltage and current quantities; impedance is a more general, global quantity that "combines" the two particular quantities into one. In the case of linear elements, it works fine, but it can't represent the circuit operation in the case of non-linear loads. But, what is worse, impedance can't explain what a gyrator does even in the case of these linear loads since it is not something real; it is some "fiction". Only the instantaneous voltages and currents are real quantities in this (and in any) circuit; at each moment, there are only a pair of voltage drop across and a current through the load and a corresponding pair of voltage across and a current through the simulating virtual element. The gyrator does not know what it actually converts - impedance or "no impedance"; it acts as a functional converter that measures "blindly" the instant input voltage and current and converts them immediately into output current and voltage. Thus, if the input element (load) has an impedance, the output virtual element will have an impedance as well and v.v. If we want to see impedance here, we will see; if we do not want, we can think in terms of special voltage and current quantities. Circuit dreamer (talk) 12:11, 13 April 2010 (UTC)

In that case, I'm inclined to agree with you. Let's wait to see what Zen-in's thoughts are (or anybody else's). However, even if we're all in agreement, we must follow the presentation of available references. What do they have to say? Oli Filth^{(talk|contribs)} 12:32, 13 April 2010 (UTC)

A few days ago, I managed to grasp "how a gyrator does this magic" and started the section above. I found a powerful viewpoint at these multiplying, gyrator and NIC circuits and prepared informative pictures revealing the clever trick. This morning, in the beginning of the lecture, I posed the problem to my students. I drew the circuit on the whiteboard and asked them to reveal the basic idea behind it (I have been using this didactic approach since 80's). One of them (the best student) immediately began thinking and proposing ideas; as a result, I didn't manage to take a cup of coffee during the break:) I hope some of them will see the brilliant idea till the next week when we will investigate these circuits in the laboratory by Microlab. Circuit dreamer (talk) 12:56, 13 April 2010 (UTC)

"it can't represent the circuit operation in the case of non-linear loads" I think that statement applies to the use of I-V analysis and not to transfer functions. We have all seen your tortoured explanations with multi-colored diagrams showing load lines, voltage levels, etc. There is no support from me to subject this article to that kind of obfuscation. The Transfer function is a widely accepted method of describing a circuit. It is adaptable to all types of circuits. There is no reason to introduce WP:OR. BTW: What Circuit dreamer is calling "impedance" are the transfer functions that appear in this article. Zen-in (talk) 15:30, 13 April 2010 (UTC)

Impedance implies LTI, which in turn implies the load has a defined transfer function. A diode (for instance) doesn't have a defined transfer function, and hence doesn't have a meaningful impedance (at least far as I understand it; it certainly can't be expressed in terms of a ratio of phasors). However, the gyrator still inverts its I-V characteristic. So I think Circuit-Dreamer's point stands.

However, we don't need a tortured explanation to present this. The article structure could be "Gyrator inverts I-V. In the case of LTI networks, this is equivalent to inverting impedance. etc." No need for multi-coloured diagrams. Oli Filth^{(talk|contribs)} 15:58, 13 April 2010 (UTC)

All electronic devices have transfer functions, since all are ultimately derived from incremental models. A diode's active region is modeled as an exponential function, di/dv = I_Se^qv/kT(q/kT). The incremental model takes into account junction capacitances and resistances. Impedance, as it relates to device models (transfer functions) Z = H(s)I_in/V_out. That is the same as a ratio of phasors because the incremental model and the driving force are both functions of the frequency domain. The problem with explaining an electronic circuit with a DC current and voltage approach is that the transient response is ignored. You are only looking at the steady-state response, ie: ignoring reactances. Which brings us back to the recurring thread of this discussion. Zen-in (talk) 21:26, 13 April 2010 (UTC)

A function such as I_Se^qv/kT(q/kT) may be the I-V characteristic of a diode, but it's not a transfer function in the sense that that article describes; it does not have a Laplace or frequency-domain equivalent, for instance, and cannot be described by an LTI convolution. Equivalently, it cannot be described as a ratio of phasors, as it is non-linear. Perhaps you are referring to linearised small-signal models, but that would be missing the point, which is that the gyrator (as I understand it) inverts large-signal I-V characteristics as well.

This is not a DC analysis, nor does it "ignore reactances"; by transposing I and V, we also transpose dI/dt and dV/dt in the case of linear components, which results in inverting complex impedance. Oli Filth^{(talk|contribs)} 22:02, 13 April 2010 (UTC)

Note that I'm not necessarily saying that we should adapt the article to this approach, this should only be attempted if there are sources that support it. However, what I am saying is that Circuit-Dreamer does have a valid point. Oli Filth^{(talk|contribs)} 22:36, 13 April 2010 (UTC)

That's right it isn't a transfer function. To derive the transfer function you need the incremental model. All components have incremental models. If they didn't the people who sell Spice software would be out of business. I am not sure what CD is proposing and I don't have a lot of free time right now. So I am going to watch from the sidelines while you and CD sort this out. Zen-in (talk) 22:48, 13 April 2010 (UTC)

There is no transfer function of a diode; all there is are transfer functions of linearised (i.e. approximated) small-signal models, which only hold over small regions of operation around the bias point. Whilst this may be what Spice uses to perform AC analysis, this is not what I'm referring to (and presumably, not what C-D is referring to). The small-signal approximation is a consequence of the large-signal behaviour, which is what (IMHO) is the crux here.

I'm not suggesting that we change an awful lot, merely that we start in terms of I-V relationships, rather than impedances (if sources permit us). I may attempt this tomorrow. Oli Filth^{(talk|contribs)} 23:07, 13 April 2010 (UTC)

(outdent) Right, I've made some edits to the article. I still need to update the image, but I'll get that done soon. Oli Filth^{(talk|contribs)} 22:26, 14 April 2010 (UTC)

Possible sources

• [1] "The voltage and current on the right hand side of the gyrator...are transformed into dual quantities on the other side..."
• [2] Defines TF and GY in a similar way to ref below.
• [3] "current...voltage...treated in a symmetric, dual way". "TF and GY are each others partial dual"
• [4] "The potential difference sources depend on a flow through the source in the other sub-system" and vice versa
• [5]
• [6] "yielding a relation between current, voltage or power at the input and output"

SpinningSpark 22:31, 14 April 2010 (UTC)

Mythical beast

ADDED: Someone is trying to create a fake legend here. I direct people to read a copy of the ACTUAL paper from Dr. Bernard D. H. Tellegen in 1948. He did not say he actually created such a device. He said that it was conceivable (first sentence). ref: http://theeestory.com/files/article-tellegen-gyrator.pdf

Other than fancy wording in the below article there are no actual details for a new network element. Notice further, there are no pictures of this supposed "gyrator" (Telleger called it an "ideal gyrator"). The schematic referenced below shows it to be, not a new circuit element, but a circuit with resistors, inductors capacitors and an opamp. Remember that Op-amps did not exist in 1948. In the below article, the first sentence shows it was hpothetical, but the rest of the article says it is real and actually in use in POTS (plain old telephone systems) but the telephone company says they have never heard of a "gyrator". To see the real article, check the address above. You will see it was only a theoretical proposal (that never panned out). --Dan Mickle, PhD. 21MAY2012.

Hi Dan Mickle, PhD, Please use this talk page for comments, rather than prefixing them to the article. Thanks. Taking your points one by one:

• The article already links to Tellegan's original paper.
• Although Tellegan did not build such a device, he did propose a number of ways in which one might be made.
• It's true that the article offers no details on the construction of a passive gyrator - perhaps you would like to remedy this deficiency yourself?
• We haven't got any pictures of stand-alone passive gyrators. However these isolators File:AisladorG.JPG, File:AisladorC.JPG are essentially gyrators connected like this File:circulator-from-gyrator-1.svg, with a matched termination on the third port.
• Ideal here is used in the same sense as in ideal transformer; it does not preclude the existence of real (though imperfect) transformers.
• A gyrator is defined by its behaviour, therefore an op-amp circuit which exhibits that behaviour is considered to be a gyrator. Whether or not we agree with this, the fact is that such circuits are commonly called gyrators.
• Although 741s did not exist in 1948, op-amps did.
• You don't state which telephone company you contacted, nor who you spoke to, nor whether you also asked about circulators and isolators.
• The proposal did "pan out". Lester Hogan built the first low-loss passive gyrator in 1952. Here is his paper [7]. See also the isolators and circulators in the above-mentioned pictures.

Please do feel free to clarify and expand the article for the benefit of other readers. --catslash (talk) 11:59, 22 May 2012 (UTC)

Hi. My point about the gyrator centers on Bernard D. H. Tellegen's description that his theoretical gyrator was a new element that would render coils/inductors and capacitors/condensers "redundant". He maintained that his gyrator was neither coil nor capacitor but a new element. This is one reason some of the description showing a "gyrator" as a resistor-capacitor network or any other sort of network using resistors, inductors or capacitors are NOT in keeping with Tellegen's description. There are several different circuits or circuit elements called or nicknamed 'gyrators', but I do not see that any of these satisfy Tellegen's proposal. There is indeed something called a gyrator used in microwave technology. It is a section of waveguide with a graphite element shaped in such a way that microwaves passing through one direction pass through normally, while in the other direction, they are delayed for a half cycle. Tellegen was not working with microwaves and I fail to see how this can apply to non-microwave signals. As to Op-amps (operational amplifiers) my real point, I believe was missed. Perhaps I did not state it clearly. The use of any circuit made up of op-amps, inductors, capacitors and resistors does not qualify as Tellegen's 5th circuit element. It is not a new element if it is made up of other already existing elements. Dr. Tellegen insisted that his new element was none of these but something totally new and different. (Please read his description to see what I mean.) What I would like to see is a drawing or patent application showing what Tellegen actually proposed.

Thank you, —Dan Mickle, PhD.  — Preceding unsigned comment added by 99.18.45.0 (talk) 08:43, 25 September 2012 (UTC) 

History moves on, and whatever Tellegen said in his original paper, engineers have since devised circuits which approximate the behaviour of an ideal gyrator. These circuits are called gyrators and whether or not you consider that an "impure" usage, it is Wikipedia's job to describe what is out there in the world, not what we would ideally like there to be. Resistors are called resistor despite there not being a real resistor that perfectly behaves like the ideal element. There is no such thing as an ideal voltage source, yet we continue to call real power sources voltage source, and so on and so on.

It seems to me the article goes out of its way already to emphasis that gyrator is an ideal circuit element. If you feel it necessary to further clarify the difference between practical implementations and the abstract ideal then feel free to edit the article. However, we really do not like contradiction and argument in our articles - everything should be written as a finished encyclopedia page. What we do like is that new material is accompanied by up-to-date references to reliable sources. SpinningSpark 15:30, 25 September 2012 (UTC)

Archiving size

User:Spinningspark wants a smaller archive size 17500 bytes. I think that the maximum size 17500 Byte of one archive is very small. Sawol (talk) 07:18, 26 March 2022 (UTC)

Let's not have this discussion repeated on multiple pages. See Talk:Fractal antenna#Archiving size. SpinningSpark 09:13, 27 March 2022 (UTC)

Vagueness

User:Alexander Davronov, is it possible to say what is vague in the following?

Gyrators permit network realizations of two-(or-more)-port devices which cannot be realized with just the conventional four elements. In particular, gyrators make possible network realizations of isolators and circulators.

Of course it may be hard to say what is lacking when it isn't there. Taking a statement of the same form, but substituting protractor for gyrator:

Protractors permit geometric constructions which cannot be achieved with straightedge and compass alone. In particular, protractors make possible trisection of any given angle.

...is that equally vague? or is the vagueness specific to the gyrator version? catslash (talk) 00:48, 21 June 2022 (UTC)

The lede should just summarize the body. If there is vagueness, those tags ought to be in the body. Constant314 (talk) 00:55, 21 June 2022 (UTC)

True; this matter should be mentioned in the body of the article, and it is not. A little while ago, I drew circuit diagrams of gyrator-based circulators (probably based on the book cited in the lede), File:Circulator-from-gyrator-1.svg, File:Circulator-from-gyrator-2.svg, but failed to get around to using them. catslash (talk) 01:28, 21 June 2022 (UTC)

@Catslash: Thanks for reaching me out. I think the word "realized" should be clarified. Best. AXONOV (talk) ⚑ 01:12, 21 June 2022 (UTC)

"Realization" is not a vague term, it is a well recognised term of art in network analysis. It means a given response function can be achieved with a construction of real components or elements. There is something on this at Network synthesis#Realisation. That could serve as a link target, but it is not totally general. SpinningSpark 13:51, 21 June 2022 (UTC)

wikt:realization meaning #2 has "The act of making real", which shows that this is a perfectly normal usage of the word. SpinningSpark 13:57, 21 June 2022 (UTC)

@Spinning🧹Spark🧹: Well... I've undone my last edit: [Jun 21, 2022, 17:15]. In this context the term is rather technical, than vague. I'm going to link the word to the appropriate article if you don't mind. My best. AXONOV (talk) ⚑ 17:18, 21 June 2022 (UTC)

Passive lumped circuit

The gyrator article under the "Passive gyrators" heading states this: "Numerous passive circuits exist in theory for a gyrator function. However, when constructed of lumped elements there are always negative elements present." I can't seem to find such a circuit, does anyone know what this is referring to? 100.16.222.85 (talk) 19:56, 8 May 2023 (UTC)

I have no clue. I tagged it with needing a citation. Constant314 (talk) 20:13, 8 May 2023 (UTC)

Unfortunately, the editor that added that information is no longer active. Constant314 (talk) 21:35, 8 May 2023 (UTC)

If our esteemed friend was referring to circuits consisting of only transformers and linear impedances (R, L & C), then he may have been mistaken. It is not obvious how the admission of negative impedances would allow such circuits to evade the reciprocity theorem. I had intended to raise this point with him years ago, but failed to do so. If nobody objects in the next week or so, then I shall delete this paragraph. catslash (talk) 00:11, 12 May 2023 (UTC)

I think that if we cannot get clarification or a reliable source, then it should probably be removed. Constant314 (talk) 00:24, 12 May 2023 (UTC)

Done. catslash (talk) 16:17, 18 May 2023 (UTC)