Talk:Passivity (engineering)

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Electronics  
WikiProject icon This article is part of WikiProject Electronics, an attempt to provide a standard approach to writing articles about electronics on Wikipedia. If you would like to participate, you can choose to edit the article attached to this page, or visit the project page, where you can join the project and see a list of open tasks. Leave messages at the project talk page
 ???  This article has not yet received a rating on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.

I had proposed merging passive component, active component, and active device several months ago. There were no objections (indeed, mostly apathy), so I just created an account and performed the merger. I have not copied any content from either of the active pages (both of which were classified as stubs, and had much less useful content than passive component). I have limited experience editing Wikipedia pages, so it may be helpful if someone more experienced can verify if I did this correctly. Also, during the merge, I removed "electrical" from 2 places, since passivity can be a property of non-electrical systems (indeed, it is commonly used in control system design for mechanical systems) -- the old active device page emphasized this. I probably should have made this two separate operations. Pmitros 00:59, 20 July 2007 (UTC)

Active and Passive[edit]

The old articles, Active component and Passive component may not have been entirely correct according to higher theory. However, they did explain the common circuit designer's usage of the terms. My understanding of this is that a passive component/circuit is one which does not contain a generator. A diode, under this thinking, is active, not because it is non-linear as it currently says in the article, but because its small-signal equivalent circuit contains a generator. This equivalent circuit may well be due to the non-linearity but it is not non-linearity per se that classifies it as active. The lists in the old articles of examples especially, made clear the division.

Another problem I have with this article is that active component and active device redirect to it but there should be a section in the article as a specific target for the link. It must be very confusing for anyone who types in "active component" to arrive here.

SpinningSpark 12:32, 4 October 2008 (UTC)

Nice article. A strange heading to find it under though. By the way, I personally define an active or passive component in my job of electronics and RF by, "If it has to have a power supply (a volatage applied to it) to make it work, it's active. If it doesn't need its own power supply, it's passive." —Preceding unsigned comment added by (talk) 21:36, 31 December 2008 (UTC)

Both of your definitions are flawed, though SpinningSpark is certainly better than the unsigned one. Transistors alone are passive devices, and that's very clear when they work by dissipating heat. It's true that the transistor will often be used with an external supply of energy (e.g., a battery), but so long as the battery is included in the diagram, the transistor is passive. Circuit elements become active when you stop drawing the power supplies. For example, an operational amplifier without power rails shown is an active device, but an operational amplifier with its supply connections drawn in is passive. Finally getting back to the diode, if you're drawing a small signal model of a diode (e.g., a battery, resistor, and ideal diode in series), within that small signal model you might call the diode an active device. However, within that small signal model, it's the battery you introduced that is "active." In reality, it's the external battery driving the large signal model (responsible for the biasing) that's the active device. In the end, the diode is entirely passive. It's best to remember where the energy is coming from. The more you drill down, the more passive devices you find. —TedPavlic (talk/contrib/@) 13:27, 7 August 2009 (UTC)

TeX is not so barbaric[edit]

Please don't write this:

 \int_0^T -< V(t),i(t)> \, dt

TeX was invented by literate people (Donald Knuth) and you can write this:

 \int_0^T -\langle V(t),i(t)\rangle \, dt

Michael Hardy (talk) 12:26, 7 August 2009 (UTC)

Yuck. You mean this:
 \int_0^T -\langle v(t),i(t)\rangle \, \mathord{\operatorname{d}}t
TedPavlic (talk/contrib/@) 13:20, 7 August 2009 (UTC)

Please explain this notation[edit]

This article says:

Given an n-port R with a state representation S, and initial state x, define available energy EA as:
E_A(x)=\sup_{x \rightarrow T>0} \int_0^T -\langle V(t),i(t)\rangle \, dt
where the notation supxT>0 indicates that the supremum is taken over all T > 0 and all admissible pairs {v(·), i(·)}.

The notation EA(x) should refer to something that depends on the value of x (otherwise what's the x doing there?). The one place where x appears to the right of "=" is in "supxT>0". The notation "supxT>0" is not standard and accordingly an explanation of its meaning is given, but I don't see that the explanation succeeds. The expression following "supxT>0" does not contain any occurrences of the variable x. Such an occurrence would not be strictly necessary if somehow the subscript "xT>0" puts some constraint on T, so that the set of values of T involved somehow depends on x. But the subscript doesn't do that. The notation "x→" always makes x a bound variable in any of its uses I've ever seen, and I don't see what it would mean otherwise. If the whole thing is to depend on x as seemingly required by the notation "EA(x)", then x needs to appear as a free variable.

When you write

 \lim_{x\to\text{something}}\left(\text{something}\right) = \text{something}, \,

then the "something" at the end—the bottom-line answer—does not depend on the value of x. Likewise when you write

 \text{something} \to \text{something as }x\to\text{something}, \,

then the second "something", which is the limit that is approached—the bottom-line answer, does not depend on the value of x.

As written, it's clumsy and opaque. Michael Hardy (talk) 12:38, 7 August 2009 (UTC)

The editor who inserted the original definition was clumsy and forgot to properly define that notation. I've gone ahead and defined it as best I can without rewriting the whole section of the paper it's from. Basically, the x is an initial state that generates voltage v and current i trajectories. The integral calculates the power of those trajectories up to time T. The supremum gives a bound on that energy over all time T. So it's a little like...
\sup_{T(x) \to \infty}
but that notation makes you think of the \limsup, which would be incorrect. It's just meant to be an upper bound on a set of powers. Each initial state x has a different set of powers, and each set has a different upper bound. All upper bounds must be finite for the system to be passive. —TedPavlic (talk/contrib/@) 13:35, 7 August 2009 (UTC)

Thank you. It's comprehensible now. And I'd actually missed the infelicitous switch from capital V to lower-case v. Michael Hardy (talk) 16:10, 7 August 2009 (UTC)


I removed two factually inaccurate statements:

To be fair to this view of passivity, a non-linear device will inevitably include a generator in its small-signal equivalent circuit and so will be modelled as a source of energy. This arises because a linear approximation to a small section of the transfer function is unlikely to pass through the origin and such an offset requires a generator in the equivalent circuit to produce the value of the offset current or voltage. For instance, a forward biased diode can be modelled as a resistor in series with a DC offset as far as small signals are concerned.

This is incorrect. A diode small-signal model (for a normal diode) is just a resistor of the slope of the IV curve at the point of linearization (coincidentally, if you look closely enough, most resistors include some amount of non-linearity as well). The definition proposed by SpinningSpark is non-nonsensical (this is just the tip of the iceberg as to how badly nonsensical).

Sorry for breaking up your post but you have made two completely separate deletions here which need to be addressed separately. For a diode where the signal remains within the substantially linear portion of the slope, the transfer function is precisely identical to a d.c. generator plus a resistor. In many applications, the d.c. can be ignored because there is no d.c. component to the signal and is going to be subsequently removed with a coupling capacitor or some such. This does not invalidate the model, it just means that the full model is not needed for that application. The paragraph is not at all a proposed definition, it is merely commenting on the definition in the paragraph above it (which currently remains in the article). SpinningSpark 15:03, 9 October 2010 (UTC)
You do not understand the definition of small signal model. Please see a copy of Gray and Meyer, or any other introductory text on analog design for a definition of small signal model. You have one valid point -- the article is not very readable to beginners. To some extent, this is necessary, since it is a complex topic (largely because it is used in many engineering disciplines -- from analog design to airplane controls -- and has to be explained to readers coming from multiple fields), but also because the writing in the article could use improvement. What you are doing, however, is not cleaning up the article. You are adding statements that range from imprecise to incorrect, and they will at best confuse a beginner. I do not believe you have the knowledge or understanding to improve this article -- you very obviously do not understand the material yourself, and so are incapable of explaining it further. I do not have the time or energy to either try to educate you, or to get into revert wars with you. You are also violating the rules of Wikipedia by posting uncited "original research." Wikipedia is not the place for novel (and incorrect) explanations or rationalizations. Of course, if you really feel like vandalizing Wikipedia, that is your prerogative. (talk) 23:18, 13 February 2011 (UTC)
but the desired signal is invariably attenuated. If no resistors are used, the amount of signal loss is directly related to the quality (and the price) of the components used.

Both active and passive filters will include gain errors due to imperfect components. In passive filters, these will typically show up as attenuation. In active filters, these may show up as attenuation or gain, but attenuation is also more common. There isn't a fundamental difference here. There are also corner cases (e.g. superconductors) where there is no attenuation. This is misleading in the typical case, and incorrect in the corner case. —Preceding unsigned comment added by (talk) 08:20, 9 October 2010 (UTC)

I agree that this is badly worded and the quality/price thing is at least irrelevant and possibly wrong. However, there is a valid point buried in here, the fundamental difference with an active filter is that losses can be compensated with a gain adjustment which is impossible for a passive filter. SpinningSpark 15:08, 9 October 2010 (UTC)
When you figure out your buried point, feel free to repost it, with a proper citation. For now, you are posting incorrect, uncited nonsense. Transfer function is described in terms of voltage gain (or, in rare cases, current gain), not power gain. Passive circuits can have voltage gain, or current gain. They just cannot have power gain. It is perfectly straightforward to compensate for voltage attenuation in a passive filter with voltage gain. In most cases, this is a bad idea for both passive and active filters, since it increases sensitivity to component variations, but there are cases where it makes sense. (talk) 23:18, 13 February 2011 (UTC)

Classic diode and transistor models are misleading[edit]

To present a non-linear resistor having an orthogonal IV curve as a source (the classic wide-spread approach) is a completely misleading concept. Typical examples of this misconception are the forward-biased diode presented as a voltage source and the collector-emitter junction of a transistor driven with constant base voltage presented as a current source. They are not sources at all; they only resemble sources in certain parts of their IV curves, and sources acting only as loads at that. A diode resembles a voltage source driven by a current source and a transistor resembles a current source driven by a voltage source. In this role, the diode keeps constant voltage drop across itself while the voltage source keeps constant voltage (the two are different). These elements are passive resistors - only not "static" (ohmic) but dynamic (differential) resistors.

Diodes and transistors are modelled by voltage and current sources because they consider there are not generic elements with such characteristics. Actually there are such generic elements but they are forgotten. These elements are (or can be) named respectively "voltage-stable resistors" and "current-stable resistors" (IMO "voltage-stabilizing" and "current-stabilizing" would be better). So, a diode (ordinary, zener, LED, varistor, etc.) is a passive voltage-stable nonlinear resistor and a transistor (PPTC, etc.) is a passive current-stable nonlinear resistor

The sentence "...a non-linear device will inevitably include a generator in its small-signal equivalent circuit..." (especially this "inevitably") is misleading. The next sentences - " will be modelled as a source of energy..." and "...such an offset requires a generator in the equivalent circuit to produce the value of the offset current or voltage...", strengthen this misconception. Of course, "...a forward biased diode can be modelled as a resistor in series with a DC offset..." but there is no need to do it if there is a generic voltage-stable element. Circuit dreamer (talk, contribs, email) 18:58, 16 November 2010 (UTC)

Passive and active adapters in computing and audio[edit]

In computing, audio, and some other fields where similar-but-different things are often connected to each other, it is useful to make a distinction between passive and active adapters. Passive adapters are usually simple connector and pinout adapters, while active ones usually contain logic circuits and/or one or more analog gain stages. For example...

  • For example, older "USB and PS/2 compatible" mice usually were equipped with a USB connector, but came with a passive adapter—a pinout adapter with no logic in it. The device's onboard firmware could determine whether it was connected to a USB or PS/2 port and so would "speak" the appropriate protocol on the appropriate pins. Whereas older PS/2 peripherals require an "active" adapter (one that includes logic so that it looks like a USB HID to a host computer) to connect to a USB port.
  • The ubiquitous connector that allows connecting a VGA monitor to a DVI-I port on a graphics card is a passive adapter (i.e. only a pair of connectors with wires in between).
  • In audio work there are passive mixers and active mixers, the former including only pots and resistors, the latter including gain stages.
  • In audio work one can also find both passive and active converters between balanced and unbalanced lines. (A passive bal/unbal converter is generally just a transformer along with, of course, the requisite connectors.)

These applications seem to me to be in the realm of this article but are not explicitly described here. I feel they should be. Comments? Jeh (talk) 10:08, 31 January 2013 (UTC)