Talk:Active filter

Comments
It seems to me that the diagram shown here is less an active filter than an ac coupled inverting amplifier because the filter component is inactive. That is, it doesn't rely on the amplifier for its function in any way except for the virtual ground the amplifier provides. Its filtering function could be duplicated by a passive RC filter except for the buffering effect of the amplifier. The more common active filter would be in the form of a Sallen-Key filter (non-inverting) or a multiple feedback filter (inverting) in which the amplifier provides positive feedback to shape the filter's response and make it roll off more quickly into the stop-band. If I can figure out how to upload a diagram, I'll substitute a better one. Anoneditor 23:01, 11 July 2006 (UTC)


 * But it's also an impedance buffer, which is a defining characteristic of active filters. — Omegatron 23:21, 12 July 2006 (UTC)


 * Maybe, but the same result (minus the signal inversion) would occur using a voltage follower after a simple RC filter, and that combination would hardly be characterized as an active filter. I think the defining characteristic of an active filter is the use of amplifier feedback to control the shape of the filter's response.  To me, the buffering action is more ancillary (though certainly useful) than defining.  In any event, don't you agree that the original figure wasn't ideal for the purpose of showing an example of an active filter? Anoneditor 05:41, 13 July 2006 (UTC)


 * You seem to forget the R1 resistor which is connected to the output ... it is a feedback. Majorkell 15:33, 24 August 2007 (UTC)


 * No, I haven't forgotten it. The discussion with Omegatron was over the diagram that pre-existed the current one, i.e., the one that existed prior to July 11, 2006.Anoneditor 20:14, 24 August 2007 (UTC)

Categories of filter
There is the beginnings of showing the "family tree" (taxonomy?) of filter types on the main filters page, but (quite rightly) it doesn't go into details of all the types of active filter available. I think something should be here; a diagram of some sort, but there are all sorts of ways to look at them, e.g. order of filter, number of amplifiers, high/low/band-pass etc, as well as an alphabetical list of names (yet textbooks are far from consistent in their use of names, and drilling down to see the circuits in such a list is a slow way of identifying a circuit you have in front of you). Also, there are details like mathematical treatment - zero/pole approaches - should that be the way to classify them (e.g. some books define second order in terms of the highest power of "s") - yet isn't that level of theory too far above the reading level this is pitched at? To go that way (which would be helpful to some readers) there would have to be lots of stuff added to wiki, like an easy explanation of all-pole for example.

Maitchy (talk) 07:11, 4 August 2009 (UTC)

How about having a simple image for a low-pass and a band-pass active filters.--Diraniyoussef (talk) 06:08, 20 January 2021 (UTC)

Leapfrog filters
A type of active filter that I've come across, is the leapfrog filter. It doesn't seem to be mentioned on Wikipedia yet. It is similar to the state variable filter, I think. It's probably worth both a mention on this page, and a separate article. --HelgeStenstrom (talk) 08:59, 26 October 2010 (UTC)
 * @Helge, Do you have a link to something? That would be more helpful, as you have noticed nobody have picked up this thread in the last 9 years. Jahibadkaret (talk) 17:35, 26 April 2019 (UTC)
 * Leapfrog filter Constant314 (talk) 21:43, 26 April 2019 (UTC)

Twin T filter
If the Twin T filter is a fully passive filter, why is it listed in this article which is about active filters?

ICE77 (talk) 17:29, 11 May 2011 (UTC)


 * I took it out. Dicklyon (talk) 20:38, 11 May 2011 (UTC)

Active over passive filters
I guess the article should list the reasons why active filters are used and what are their advantages or disadvantages over passive filters.

ICE77 (talk) 20:51, 13 December 2011 (UTC)
 * Apparently someone you never heard of at Bell Labs 40 years ago decided active filters were a bad idea and that is why they are totally extinct today. --Wtshymanski (talk) 16:51, 18 December 2011 (UTC)
 * Actually, the edit you're flaming about was questioning the "stability and predictability" with the comment goes against Jimmy Tow's at Bell Labs investigation of active vs passive filter sensitivity in the 1970s. Tow's finding was that passive filters are less sensitive to component variations that active filters." Certainly this is true of the sorts of filters he investigated.  Active filters are not known for their exception stability or insensitivity to component value variation.  They're used mostly because they don't need inductors.  Probably we can find sources that actually examine these issues.  This book covers some of the issues. This one mentions inductor size and cost.  Many of J. Tow's papers can be found via Google scholar.  This one talks about how to get some of the advantages of classic passive filter topologies into active filters, doing much better than the usual cascade of second-order sections.  So it's like he's against active filters, if that's the impression you got from the report of some of his results.     Dicklyon (talk) 16:57, 18 December 2011 (UTC)


 * Dicklyon, thanks for understand the tag and digging out some references. (I gave Tow credit for Fleischer's proof, and I had forgotten about Friend.) The lead (and article) should explain the advantages of active filters without making dubious claims about their superiority.
 * Wtshymanski's absurd belief that Tow was opposed to active filters is regrettable. Part of Tow's career was making better active filters -- which involved understanding how they compare with other designs and how to push their limits.
 * Sadly, the three points in Electronic Instrumentation (1. Gain and Frequency Adjustment Flexibility, 2. No Loading, 3. Cost), appear confused. I'd put cost first. Adding gain blocks to a passive filter is not difficult; the gain aspect is a side effect of the amplifier already being present. The adjustment flexibility argument is not clear: pot core inductors are easily tuned with a slug. Active filters can be tuned with inexpensive resistors, but that is a cost issue. Active filters can be designed to have independent adjustments for CF and BW or Q, but that has a loading/unilateral spin to it; independence has a negative sensitivity impact; Williams, Electronic Filter Design Handbook, describes passive tuning procedures (section shorting) to make independent filter adjustments; Agilent offered tools to make independent tuning adjustments to avoid iterative tweaking). The no loading comment is tied to a unilateral gain argument. The book's earlier comments about application of active filters are good (e.g., active filters used at < 100kHz -- but the statement is dated). Also the GBP limitation. The list doesn't mention size or weight.
 * I objected to the implication that active filters are better in every way than passives. Passives are certainly better on a power consumption metric (no batteries required). Active filters are not used at the output of high power RF transmitters (yes, I know that amplifiers are used for linearizers). As Electronic Instruments points out, active filters are not replacing crystal filters. Nor transmission line filters. Lots of people use crummy inductors in CMOS RF processes. Horowitz and Hill, page 281, compare a $2 switched capacitor filter to a $10 AF100 active filter; cap ratios are cheaper than precision caps. There are trade offs, and active filters do not always win.
 * Similarly, a huge advantage of digital filters is the "filter components" have extraordinary precision that shame 1 percent resistors and can avoid tuning. And digital coefficients have a 0 ppm temp co.
 * I do not consider Lancaster a WP:RS on the issue (if he even made the purported statement).
 * I believe the dubious tag or a further edit is in order. Just dropping stability is not enough; sensitivity is also about predictability.
 * The challenged statement is also backward. Design requirements set the performance, stability, and predictability goals. Amplifiers make complex pole/zero filters without inductors feasible. The notion that one starts with a crummy real pole/zero passive RC filter and "improves" it with an amplifier is odd.
 * Glrx (talk) 20:56, 29 December 2011 (UTC)

The reply to my initial comment by Wtshymanski is clearly not directly related to what I was trying to find out. In cases like this one, I suggest starting a new thread. Either way, active filters are not extinct. In fact, they are used in audio applications and a 3-way Linkwitz-Riley crossover is a very common example of active filters.

Dicklyon seems to point out the disadvantages of active filters by saying that they are more sensitive to component variation (I have read that in a book by Douglas Self on active filters where he was also saying that higher quality factors can be obtained). Dicklyon also says that the advantages of active filters are that inductors do not need to be used (so space and money can be saved). Before I start to take a look at the links Dicklyon provided, is there any other advantage/disadvantage for active filters?

ICE77 (talk) 23:13, 28 December 2011 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Active filter. 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/20080511220632/http://www.national.com/an/AN/AN-779.pdf to http://www.national.com/an/AN/AN-779.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 07:27, 26 June 2017 (UTC)