Talk:Damping factor/Archive 1

This is an archive of past discussions about Damping factor. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

Archive 2

Change to source and load

Note to self: change these to source and load - Omegatron

Done. - Omegatron 13:39, Aug 4, 2004 (UTC)

Definition/lede

This is not now accurate becuse to have damping factor, there must be something to damp [1]- like an oscillatory system. No such system is described in the lede para! Therefore Im proposing a merge with the damping article --Light current 22:05, 22 April 2006 (UTC)

Large Damping Factor is desirable for audio amplifiers

I have a real problem with this statement... "A large damping factor is no advantage beyond a certain point, probably around 10." This is totally not true... and totally obvious to anyone who works in the professional audio field, as I do. A large damping factor (100 or greater, preferably 400 - 1000) is highly desirable, and mandatory for quality bass reproduction, given the way that loudspeakers work. Please read the referenced article from Crown Audio. Anyone who believes a large damping factor is no advantage is confusing the issue of feedback (and possible negative effects of feedback) with damping factor. Tvaughan1 22:33, 21 April 2006 (UTC)

Moved from User talk:Light current

Light Current... please read the "Description of damping factor" article referenced at the bottom of damping factor. If you understand how a loudspeaker works, and I trust that you do, you will realize that the article is correct (due to mechanical factors... the spring effect of the diaphragm, the mass and inertia of the diaphragm...). Tvaughan1

Please sign your posts anon, then I know who Im talking to!--Light current 00:16, 22 April 2006 (UTC)

THe article you reference is very badly written and inaccurate. Let me ask you a question: How well damped would a speaker be if the amplifier output could be represented as a pefect S/C. Would it be perfect - or less than perfect. Why?--Light current 03:11, 22 April 2006 (UTC)

Moved from my talk page--Light current 03:24, 22 April 2006 (UTC)

I'm sorry... I thought I had signed my post on your talk page. Thanks for replying. OK... if the amplifier output impedance was zero (it's in series, so it isn't a short circuit or you would have no voltage), the speaker would be well damped. It would be perfectly damped, to the extent that is possible without mounting an accelerometer on the speaker and using feedback to attempt to eliminate all unwanted speaker motion (which, as you may know, has been attempted in some woofer and subwoofer designs). Tvaughan1 13:32, 22 April 2006 (UTC)

Opening disscussion

It's perfectly possible for the output impedance of the amp to be zero (in theory) but still drive a current and thus develop an emf (voltage) across the load. In fact, the lower the impedance, the better it is able to do this. To the speaker, the output impedance appears across the speaker (i.e. in parallel). For the case of the speaker's back-emf, the amp's output impedance is the load. The question is whether an impedance here of 0.001 ohms or 0.0001 ohms makes any real difference, when the coil (source) impedance is 4 or 8 ohms. My view is that it doesn't, because the amount of current that the speaker can push back as a result of its back-emf is not in the order of hundreds of amps, even momentarily. No matter how good the speaker, it's not an equally efficient transducer in both directions - a mechanical movement of the cone of x doesn't induce a reverse current equal to the input current that moves the cone by x. Thus the back-emf is vastly smaller than the driving signal, and so even a poorish damping factor is probably "good enough" in practice. Good mechanical design of the speaker is probably vastly more important here. The problem is that this is one of the main areas that the audiophile magic fairy dust merchants go to town on, so unfortunately simply because one article claims factors of 10,000 "are necessary" for good listening doesn't make it so. The only way to determine the truth is double-blind listening tests, and in those, no clear cut correlation is shown between damping factor and listening quality, once a basic d/f level of about 10 is present. Graham 14:23, 22 April 2006 (UTC)

Graham, I tend to agree with what you have said here. THe only thing I would comment on is your statement about the non reciprocity of the loudpeaker as a transducer. What I think a lot of audiophiles would do well to remember is that current in the LS coil produces a force that acts against the surround stiffness to cause a defined displacement of the cone. This is the normal operating mode of the speaker. In the reverse direction (mechanical to electrical) it is the velocity of the cone/coil that produces an electromotive force. This emf will naturally produce a current of V/R where R is the total series resistance in the circuit (including the coil resistance) but it is aeay to see that the voltage generated at the amplifier o/p terminals is attenuated in the ratio of the damping factor (approx). But, used as a microphone, properly loaded with a highish Z, you might find that the LS works better than expected. By using the correct calcualtion for efficacy, you may indeed find that the LS is just as efficatious whichever way round its used.--Light current 15:22, 22 April 2006 (UTC)

You're quite right, a speaker can make a decent enough microphone when interfaced appropriately. However, when connected to a power amp output, it is definitely not interfaced correctly for its microphone mode, which is what I'm getting at. Having read the arguments here, it seems that we are all pretty much in agreement except for the matter of degree. Obviously, all else being equal, a high damping factor is a Good Thing, but what you and I seem to be arguing is that once it's "good enough" other factors begin to dominate. Tvaughan seems to be saying that for tight, controlled bass, much higher damping factors are needed. However, his own equations don't bear this out, because even with an infinite damping factor (short circuit) the mechanical resonance/inertia will dominate, so I can't see how increasing the damping factor from, say 20 to 200 will make any appreciable difference to the controlled-ness or otherwise of the cone. As long as there is a low driving impedance (df >10), and the speaker is built lightly, as they are, the amp will keep the cone under tight control. I challenge Tvaughan to identify the lower of two damping factors in a double-blind listening test where this is the ONLY difference (easily set up by inserting a very small (0.01 ohm) resistor in series with the speaker.) Graham 06:04, 23 April 2006 (UTC)

I totally agree with you here Graham. If there was an amp of 0.1 ohm output impedance and a speaker of 8 ohms this would give a DF of 80. If the speaker was now changed for a 16 ohm unit (doubling the 'so called' DF), could TVaughan1 tell the difference between the two? My bet is that he nor anyone else could.

Unfortunately, TVaughan1 has not yet realised that most of the damping resistance lies withing the speaker and that therfore even a zero output Z amp will give imperfect damping. Of course this is becuase he has ignored my previous posts guiding him toward this obvious conclusion. Ah well!--Light current 19:14, 23 April 2006 (UTC)

Light Current - please don't speak for me, and what I have or have not realized. I have conducted more than a few double-blind listening tests, debunking a number of audiophile myths (green magic marker on CDs, mastering at 1X vs. 2X, clock jitter vs. pit jitter, etc.). I know how to tell the difference between real and imagined audio phenomena. I'm not going to be "guided" towards your conclusions when my ears, equipment, and experience are guiding me to different conclusions.

Graham - I will take that challenge (although I don't happen to have a high power 0.01 ohm resistor lying around, and I certainly didn't say I could hear the difference between a damping factor of 799 and 800... but I'd bet I can hear the difference between a DF of 8 and a DF of 80, so let's use a 1 ohm resistor).

Let me make a couple of points. First, damping factor is most important for bass drivers (woofers, subwoofers). The highest quality low frequency drivers are anything but "lightly built". The moving mass of my JBL 2245H is 0.185 Kg, about 6 ounces. Of course you may be using tiny little speakers incapable of quality sound reproduction in the lowest octave of the human hearing range... if so why are you discussing damping factor? Second, speaker impedance will vary... it isn't always 8 ohms. Speakers are often driven in parallel (4 ohm nominal) configuration, which increases the available power from a given amplifier, but halves the damping factor. As others have pointed out, the series resistance of the speaker cable degrades the damping factor. In physically large environments, you can't always locate the amplifier next to the speakers (the speakers may be hung from the ceiling, for instance), and it can be impractical to use very large gauge speaker wire. If you start with an amplifier that has a damping factor of 1000, and you run 100 feet of 12 gauge speaker cable, you have added 0.32 ohms of series resistance, and you are down to a damping factor of 24. If you are using an amplifier with a damping factor of 10, and you are using 16 gauge cable, you add 0.80 ohms of series resistance, and your damping factor has dropped to 5. If you are using 4 ohm speakers instead of 8, these factors are cut in half (to 2.5 in the latter example). Third, you need to consider the effect of electrical damping both the time domain and the frequency domain. Consider that mechanical resonance causes unwanted motion and sound, and that this distortion causes destructive interference (most prominantly at the resonant frequency of the loudspeaker). We humans can hear over a very wide dynamic range. Audible distortion isn't always masked by the program (as the program may be silent while the distortion continues!). A speaker may be asked to reproduce a very loud signal at very high SPL, and then immediately be silent (a gated kick drum at 600 watts into a single speaker). Yes, there are many factors to consider with this problem, but all things being equal, you want a power amplifier with a high damping factor (much greater than 10). Lastly, the mechanical resonance will not be so dominant if it is electrically damped. There is a big difference between the speaker resonating mechanically when it has no electrical connection (the speaker is connected to an infinitely high impedance), and a speaker that is trying to resonate when it is connected to an amplifier that is capable of forcing the signal voltage to the desired level (say, for instance, zero volts immediately after the gated kick at the start of Queen's "We Will Rock You"). It is the same phenomenon you could observe if you open the circuit on an electric motor, which will continue to spin, versus short circuiting the motor's terminals (it will stop quickly). Tvaughan1 20:44, 23 April 2006 (UTC)

Perfect Loudspeaker Damping

Well User:Tvaughan1 I have to say that even with a zero o/p impedance amplifier, the speaker is not IMO perfectly damped and never can be unless, of course, in the electrical equivalent circuit of the mechanical/electrical system:

$\zeta =\omega _{0}\Rightarrow RC=4{L \over R}\,$ or

$R=2*sqr{L \over C}$

ie L represents the cone mass, C the cone compliance and R the total equivalent electrical damping. How and if this could be arranged, Im not sure yet. But it could not be acheived by the amplifier and would have to be engineered into the speaker by the manufacturers. BTW, the output impedance of the source can be zero and you can still have a voltage from it! In fact an amp with lots of NFB approximates a perfect voltage source. Let me ask you another question: What actually do you have to reduce to zero in order to obtain perfect damping? Is it:

a) the LS terminal voltage or
b) the emf generated by the coil?

What is the reason for your answer?--Light current 14:45, 22 April 2006 (UTC)

I have to agree with you, FWIW. The problem ultimately is that the speaker cone has mass and therefore inertia. Even if you could short circuit the coil, there is still the mechanics to consider. Graham 15:03, 22 April 2006 (UTC)

Where is the energy of a moving cone dissipated when its leads are s/c? I think some of course will be radiated into the air as sound (or absorbed in the cabinet wadding), but the remainder must be dissipated in the internal resistance of the coil. Another important factor in terms of damping is of course the nature of the acoustical loading: a speaker in a sealed box will be more highly damped than one in free air (and its resonance freq will be increased). Also, is low damping factor necessarily a bad thing? It can increase the bass output at the expense of lack of tightness/control; but what is the min damping factor required for a good bass drum sound anyway? I suppose it depends on how tight the bass drum is! --Light current 16:15, 22 April 2006 (UTC)

This is all very nice, but I think you are missing my point. As an engineer, I don't disagree with the equations. As an audio professional, I have heard the difference between high quality amplifiers with high damping factors, and low quality amplifiers with low damping factors. Sure, there are many other factors to consider when evaluating an amplifier, such as distortion ratings. In any case, any audio professional who has critically evaluated amplifiers will tell you that a high damping factor is desirable... unless you don't care about accurate bass reproduction. So I ask you... have you listened and compared amplifiers with low damping factor to amplifiers with a high damping factor? The referenced Crown Audio article, in my opinion, is far more informative than this Wikipedia article in its present form.

Light current asked "Also, is low damping factor necessarily a bad thing? It can increase the bass output at the expense of lack of tightness/control; but what is the min damping factor required for a good bass drum sound anyway?". Yes... low damping factor is a bad thing. Any increase in bass response can only be described as noise or distortion... unwanted sound that does not represent the audio signal that we desired to reproduce. If you want accurate sound reproduction, you need an amplifier with a high damping factor. Tvaughan1 03:55, 23 April 2006 (UTC)

How many perfect speaker systems ITO low freq response do you know? They all have a lower 3 dB point!--Light current 13:38, 23 April 2006 (UTC)

Because you have declined to answer my original questions about what is necessary for good damping, I fear you may have missed an important part of my reasoning. If you now go back and look at them, youll see that I did not ask them in vain but to try to illustrate a point ablut the ineffectiveness of super hig DFs.--Light current 18:54, 23 April 2006 (UTC)

DF more critical for larger, more massive loudspeakers

My JBL 2245H 18" subwoofers sound pretty darn good... flat response down to 30 Hz, and good response down to 20Hz. But they sound horrible if they are driven with an amplifier with a damping factor of 10. That is why I use a Crown DC-300A Series II. The manual for this amplifier describes the following... "Your amplifier can provide an excellent damping factor of 750 or more from DC to 400 Hz in Dual mode with an 8 ohm load. In contrast, typical damping factors are 50 or lower. Higher damping factors yield lower distortion and greater motion control over the loudspeakers. A common damping factor for commercial applications is between 50 and 100. Higher damping factors may be desirable for live sound, but long cable lengths often limit the highest damping factor that can be achieved practically. (Under these circumstances, Crown’s IQ System® is often used with Crown’s premium Macro-Tech,® Com-Tech® and Reference series amplifiers for easy monitoring and control of units located very near the loudspeakers.) In recording studios and home hi-fi, a damping factor of 500 or more is desirable."

My ears tell me that the Crown Engineers are right, and that the statement "A large damping factor(greater than about 10) is no advantage" is totally incorrect. I believe the Crown Audio engineers have done their homework, and are an authoritative reference on the subject. I have worked in professional audio for 19 years. I have designed, constructed and operated quality assurance studios for record companies and CD manufacturers. Professional engineers in the recording studio industry and in the sound reinforcement business understand the difference that a high damping factor makes. It's easy to understand if you use your ears. It's also easy to understand from a mathmatical and a practical standpoint. The lower the damping factor, the more you will have unwanted motion on your low frequency (high mass) drivers. Unwanted motion of the cone means unwanted sound. Unwanted sound is distortion... it does not represent the originally recorded signal, it is noise. Arguing that "a large damping factor (greater than 10) is no advantage" is a disservice to those who would reference the article, looking for good advice when choosing an amplifier or building a sound system. Tvaughan1 15:45, 23 April 2006 (UTC)

You seem to be missing my point!. Whilst your system may be able to reproduce anything down to sub sonic, most system cant. In these cases, low damping can enhance an inadequate bass response. THats all I'm saying. I cant see your argument against! BTW Youre ears, as are most people's, are unreliable! THats why we use test gear.--Light current 18:38, 23 April 2006 (UTC)

I get the point, you want "one note" bass. Any bass frequencies all turn into the resonant frequency of the speaker. It's intermodulation distortion. Look, my ears have spent years being the reference standard for quality at a major record label. I'm a musician, and I can hear more than most people can due to years of critical listening. It's a blessing and a curse (as it is with video, by the way, as I see MPEG compression artifacts like the kid in the 6th sense sees dead people). The point is, if ANYONE can hear a difference between system A and system B (in a valid, double-blind listening test), then there is a difference, and there must be an explanation. Audio is an art and a science. If the artist tells you there is a difference, the scientist better figure out how to explain it, measure it, and deal with it. As I've pointed out, I've dealt with many situations where recording engineers and audiophiles would tell me there was a problem, and I would conduct tests to debunk the myth... always successfully. It just seems as if you have never heard a bad sound system with low damping factor. Why else would you argue that zero damping factor would be not only OK, but preferred? Find a really cheap DJ with a really cheap sound system, and ask him to crank up the dance music. Or listen to any teenager's car stereo system with 1000 watt bass drivers. You'll hear what low damping factor is. Muddy bass. Tvaughan1 22:44, 23 April 2006 (UTC)

No I dont want "one note" bass. Im suggesting that in a system with restricted bass response, like a small bass cab, or cheapo woofer, lowering the damping factor can give more o/p at the expense of flabbyness. But lots of people like a flabby bass. THis is well known in the bass amp business! Im surprised youve not heard of it! And wher on earth do you get the notion that it causes IM dist?--Light current 04:07, 24 April 2006 (UTC)

Mounting method and possible effect on DF

BTW you dont tell us how your 18" JBLs are mounted. Are they in a sealed cab, reflex or TL structure.And why do you need TWO? I think this may have some bearing on DF--Light current 04:25, 24 April 2006 (UTC)

They are in a ported enclosure. Tvaughan1 03:20, 25 April 2006 (UTC)

Aha! Now if they are in a reflex enclosure, then below the system resonant frequency, there is no acoustical damping and the cones will flap about with any small lf signal that the amp provides. I think there are 2 poss explanations for your findings:

You really can hear the diff in DF between the 2 amps. OR
The LF response of the cheaper (low DF) amp is actually more exteneded than the expensive (high DF) one and is giving you more rumble or equivalent lf 'noise'.

Im sure I dont need to tell you which I think it is!--Light current 03:31, 25 April 2006 (UTC)

"Flabbyness"? You concede that a lower damping factor is audible. Tvaughan1 21:00, 24 April 2006 (UTC)

When its very low of course. Were not talking about low DFs tho@!--Light current 21:29, 24 April 2006 (UTC)

The frequency response of both amplifiers is excellent... flat from 20 to 20,000 Hz. It's not a factor. Tvaughan1 16:44, 25 April 2006 (UTC)

What about below 20 Hz though?--Light current 21:15, 25 April 2006 (UTC)

Unless you are working with elephants (and specially designed speakers), infrasonic or subsonic signals are undesirable, as they will just waste power, distort your audio and possibly burn out your speakers. The frequency response of even the best subwoofers falls very steeply as you go below 20 Hz (for most speakers it starts to fall off much higher... at 30, 40, or 50 Hz... the vast majority of consumer speaker systems can't reproduce the lowest octave in the human hearing range with any accuracy). Most audio systems that provide the signal (tuners, pre-amplifiers, CD players, etc.) have high-pass filters to prevent subsonic signals from being passed to the amplifier. If you suspect that your system doesn't have this (for instance, for a live sound reinforcement system with microphones and a mixing console), amplifier manufacturers such as Crown recommend that you add a high-pass filter to remove subsonic or DC signal components. Tvaughan1 13:14, 26 April 2006 (UTC)

You will of course be aware that any o/p below 20 Hz, whilst not being directly audible, may affect the sound of the audible bass frequencies. Whilst the o/p response of the subwoofer systems will fall rapidly below the cutoff frequency (is it at 12 dB/oct for reflex cabs- I cant remember) the cones are effectively in free air and so any o/p from the amps will cause the cones to move in a largely uncontrolled manner at subsonic frequencies thus adding doppler distortion to the rest of the bass o/p. With direct coupled amps, the potential for dc and subsonics is there!

I'm therefore trying to determine if each amplifier has the same gain charactersitic below 20Hz. So do you use a high pass filter to take out the subsonics with both amps? (you cant rely on the signal sources necessarily --disc turntables for instance ) and are both amps identical below 20 Hz?--Light current 15:32, 26 April 2006 (UTC)

Personal Experience / Expertise

Ahh, now you tell us that you have Golden Ears! We used to have someone over here who said he had golden ears. He was blind and his name was Angus something (I think it was McKenzie). He could hear a pin dropping on the moons surface-- but no one else could. It's true that you can train your brain to hear things that others can't. Maybe you can hear differences in DF - that doesnt mean that everyone can. And it means therefore that its generally not important to most people if they can't hear it. I think that's what the article says. If it doesnt - it soon will.--Light current 01:33, 24 April 2006 (UTC)

Don't misquote me. I told you I have good hearing, and I told you that I have some training in music and critical listening. I told you a little bit about my professional background and job responsibilities. I think you've hit the nail on the head, however, when you say that maybe I can hear differences in DF that not everyone can... that is almost certainly true. That doesn't mean that a high damping factor isn't important. Look, the majority of consumers don't know how audio systems work, and they don't know how to measure their performance (objectively or subjectively). They don't know how to interpret the manufacturer's specifications. That doesn't mean that they wouldn't benefit from one system versus another. All things being equal, they would benefit from a system with more accurate sound reproduction. Some of these consumers may be coming to Wikipedia to look up damping factor. Let's give them the facts. Tvaughan1 03:24, 24 April 2006 (UTC)

(deleted LC material)

It should be fairly obvious why this setup won't work... you propose changing the volume level along with the damping factor. Tvaughan1 20:47, 23 April 2006 (UTC)

(deleted LC material)

Going from 0 to 10 ohms changes the impedance from 8 to 18 ohms. That's more than a fraction of a dB. The point is, if the volume changes, it kills the validity of a double-blind listening test, as the subject could easily hear the volume difference and reliably identify setup A from setup B when presented with 10 random samples. You wouldn't know if the subject was hearing the difference in damping factor or the difference in volume. Tvaughan1 22:33, 23 April 2006 (UTC)

Zero electrical damping factor

There is also a good case to be made for a system with zero electrical damping. In the case of a speaker in a small sealed box, there is no way to reduce the system resonance below a few hundred Hz (even if the speaker itself has a very low free air resonance) because of the low compliance of the air in the box. However, because of the high system resonance, other losses being equal, the cone damping will be high. All that is needed is then to electrically drive the speaker from an amplifier with a high o/p Z (ie a current source) for a flat response all the way down to dc.

One advantage of the system is that whether it is being operated above or below the system resonance, the cone excursion is current controlled and is therefore linear (independent of surround nonlinearities). The only slight problem with this scheme is that the efficiency may not be too high. But hey, watts are cheap these days - yes?--Light current 19:53, 22 April 2006 (UTC)

The above post was included as a general comment and not as a specific response or goad to any User currently involved in this disscussion--Light current 22:51, 24 April 2006 (UTC)

If you like noise, sure... go with a low damping factor. Let's say you are setting up a sound system to amplify the performance of a band. You set up a microphone to the bass drum (the kick drum), and this is wired to an amplifier, which is connected to a low frequency driver (a woofer). The band may prefer a noise gate on the kick drum, to create a sharp attack and decay. Since the woofer cone has a relatively large mass, it will continue to resonate, even after the audio system has gated the sound to zero. Another way to look at this is as the impulse response of the system (although obviously the system only has a frequency response of 20 - 20,000 Hz, or 20 - about 250 Hz for the low side of a bi-amplified sound system). The objective is to accurately amplify the sound that was captured (and gated). While we can not eliminate unwanted resonance of the speaker cone entirely (without using feedback on the motion of the cone), we can minimize it by using an amplifier with a high damping factor. Again... use your ears. If you haven't heard the difference, this conversation is a waste of time... you will never be convinced of the value of a high damping factor. Tvaughan1 15:55, 23 April 2006 (UTC)

You say:

While we can not eliminate unwanted resonance of the speaker cone entirely (without using feedback on the motion of the cone), we can minimize it by using an amplifier with a high damping factor.

I partly agree. But it seems like you have not read my post properly. I said it can also be minimised by using a sealed box and a current source drive. I think its easy to be fooled by ones ears! Also you appear to be limiting the discussion to one of amplification of live music. There are other applications of amplifiers/loudspeakers. BTW the cone resonace in free air does not only depend upn its mass but on the compliance of the surround. But perhaps you forgot to mention that. --Light current 17:57, 23 April 2006 (UTC)

I used live sound amplification as an example. Certainly the same is true for the playback of recorded music. Tvaughan1 22:24, 23 April 2006 (UTC)

The burden of proof

When you can report results of double blind listening tests that show that extreme damping factors are preferred by the subjects to modest or low damping factors, I will believe you.--Light current 18:02, 23 April 2006 (UTC)

I see... the burden of proof is on me, even though I've cited experts who explain that a high damping factor is valuable. I think that those who say "a damping factor greater than 10 is of no value" have a burden of proof. Tvaughan1 22:24, 23 April 2006 (UTC)

Yes Im afraid the burden of proof does lie with the minority or those who state things that are contrary to common sense or the accepted mathematical analysis of things! This is the way WP works. It generally weedles out the untruths! But it should not be a problem for you as you are so certain about your belief.--Light current 23:05, 23 April 2006 (UTC)

I believe the engineers at Crown Audio. Have you read the article (before changing this wording of the external link)? Do you to claim to have greater expertise than the engineers at Crown Audio? You haven't offered any reason for anyone to believe that you have more knowledge or experience on the subject than the #1 professional audio amplifier manufacturer in the world (who have been studying the issue for decades). You offer no references or citations for your claims, and you make some patently absurd claims. I'm just trying to help out the folks who wouldn't know any better, who come to Wikipedia for some good information. You simply dismiss the experts (the Marantz article is poorly written... the Crown article is sales literature... in other words, they are knowingly, unethically making false claims). Instead of substantiating your own claims, you throw out one red herring (zero damping factor is good!) after the next (perfect electrical damping isn't perfect damping... free air resonance depends on the spring rate of the surround... yeah, no kidding!). So, damping factor remains hijacked with a minority viewpoint (outside the professional audio majority). It's Light Current and friends against Crown Audio, Marantz, and many others. I'll stick with Crown, thanks. It's clear that you have never heard the effect of low damping factor on a system. This is most likely due to limited experience with sound reinforcement systems. Do you own any professional amplifiers or loudspeakers? I do... many. Have you ever done live sound? What is important in a professional sound system may not be as important in a small home stereo system. Have you done any controlled listening tests? I'm guessing "no". I have... many, to AES standards. Audibility (which, by the way, is what we are discussing) can not be shown mathematically. It is a psychoacoustic phenomenon. Are you now, or have you ever been a member of the Audio Engineering Society? I'd recommend it to you.

The issue is whether an amplifier's damping factor effects the sound quality in an audio system, and if so, at what point does the damping factor become large enough so that further increases have no audible effect on the sound quality. The issue is not whether an amplifier can "perfectly damp" a speaker. You claim the Wikipedia reading world should believe your opinion and analysis on the importance of damping factor, instead of Crown Audio and Marantz. Quite a bold claim, in my opinion. Cite your sources for the audibility of damping factor. Again, keep in mind that the amplifier's DF is the starting point, and if you use low impedance speakers and long and/or thin speaker wire, it goes down from there. Advising readers of this article who might be selecting an amplifier that a DF greater than 10 is insignificant shows a lack of experience on the topic, in my opinion. Mathematical analysis of the phenomenon is instructive, but not helpful to answer the real question. Audibility is subjective. Sound quality is subjective. The ONLY way to answer the question is to do double-blind listening tests to AES standards. Tvaughan1 03:07, 24 April 2006 (UTC)

Statements such as "Note that modern amplifiers, employing relatively high levels of negative feedback, generally exhibit extremely low output impedances — one of the many consequences of using feedback. Thus "damping factor" figures in themselves do not say very much about the quality of a system. Given the controversy that has surrounded the topic of feedback for many years, some may see a high damping factor as a mark of poor quality becuase it implies a high level of NFB in the amplifier." do not belong in the article. These statements confuse the issue, as much of the discussion on this page has done. Feedback may or may not be a good thing, but that is a different topic for a different wikipedia article. No one has claimed that damping factor figures say everything about the quality of a system... that is absurd. Damping factor is but one measurement of an amplifier's capabilities.

The following statements "A large damping factor(greater than about 10) is no advantage. Thus provided that the return path of the cables measures less than about 0.8 Ω , thicker or better cables will make no perceptible difference. The difference in damping with a factor of 10 is the difference between 8 Ω and 8.8 Ω, which is unlikely to give more than a fraction of a dB difference in sound pressure level at the low frequency resonance of the speaker." are incorrect. No proof whatsoever is offered, and no citations are given for the statements, particularly with regard to perceptibility. I pointed out some very good articles that contradict these statements. I asked those who made the statements to prove it, and you replied with questions, math and red herrings. Tvaughan1 00:22, 24 April 2006 (UTC)

Well if I may, Id like to interject here and reply to your initial point.

I believe the engineers at Crown Audio are a fine bunch of fellows who have designed damn good amps not only useful for audio but also for driving other voice coil systems etc. I even remember using one for something!.

However, the text (which I have read) in link you refer to will, (although having originally been written by engineers) have been spiced up considerably by the sales/marketing people. After all, if your company makes an amp that has a DF of 2000, why not trumpet it and pretend it's important. I would, whether true or not (its not):So whilst the Crown art is accurate, it does rather tend to overblow the importance of high DF from an audio perspective.

The Marantz article if I can remember it properly, has some glaring errors in the first few paras that led me to reject the integrity of the article as a whole. I will relook at it if you want me to list them. Obviously written by someone who knows nothing about the subject, or been grossly mangled in the translation by the sales dept. (See separate thread on the Marantz page on DF below).

You say:

they are knowingly, unethically making false claims

Of course, it happens all the time. You probably cant sue them over it but they sail very close to the wind in an effort to persuade the innocent. Looks like they succeeded with you!

My zero electrical damping factor stuff has been referenced- draw your own conclusions. It exists. Whats the problem? Also if you had any appreciation of electronic or systems engineering, you would realise that a speaker cannot be damped perfectly just by having a zero o/p impedance because it has its own internal resistance that cannot be ignored. I tried to tell you this before. Are you denying that the free air resonance of a LS relies not only on the cone mass but on the compliance (the restoring force) of the surround? Why do you say that these staements are 'red herrings'?

If you are a typical AES memeber, I think I may be too radical to be been a member of the Audio Engineering Society.

More later.--Light current 01:05, 24 April 2006 (UTC)

I hesitate to add more fuel to the fire, but the arguments seem to be getting a bit heated on both sides here. Is there something we can agree on to the betterment of the article to come from any of this? I have a few comments/observations. Crown Audio are claiming that their amps give DFs of 2000 or so. I don't dispute that, but how did they achieve this? It must be due to having a large degree of parallelism in the output driver stages, coupled with a fair amount of NFB within the amp. This approach, particularly the parallel output drivers with load sharing, is definitely the way to go for a good sounding amplifier, according to numerous references I've read. So could it be that IF the Crown equipment sounds good, it is due to its overall design rather than this one particular measure of performance? That's why I suggest the double-blind test so that the amp can remain the same one, the speaker can remain the same one, and the only thing that varies is the series resistance inserted. If we are arguing about the difference between a DF of 20 and 2000, then inserting 0.4 ohms (for an 8 ohm speaker) will suffice. This will change the amplitude by 0.21 dB SPL which should be inaudible (and well within any variation caused by very small changes in the head position of the listener). If there is an appreciable difference to the resonance/tightness of the bass as a result, this should be detectable where the amplitude difference is not. However, this is something of an academic point, since it's not reasonable that we can get together to set this up and carry it out in practice, so whoever the "onus of proof" is on, there is no way to independently verify the results and come to any conclusion, or even know that the test was ever done. However, I must disagree with Light Current that objective measurements are the only way to determine the quality of a system, because there are many things that straightforward measurements cannot tell us about whether a system sounds good or not. Measuring DF alone is not going to tell us much; however, before I get accused of shifting my stance, my own view is that it's hard to see why it WOULD make any difference, because the physics of the system seems to dictate otherwise. The mechanics of the speaker are sure to dominate, and even with perfect damping there will be mechanical resonance and inertia. TVaughan correctly points out that a motor left to run open circuit and one short circuited will behave very differently, the short circuited one coming to rest almost instantaneously. However, this is the difference between infinite and no damping at all, which I'm sure is totally audible (in the case of a speaker). However a motor subjected to a braking load of 0.4 ohms will stop as if short circuited (assuming its windings are in the tens of ohms range), with no discernible difference to its mechanical behaviour compared with a s/c. So applying this analogy to a speaker seems to suggest that this wouldn't make any audible difference there either. As far as I can see the argument is at stalemate, because the only way forward is to perform real-world tests that all of us take part in and agree on a result. That isn't going to happen, so where can we go from here? Perhaps instead we should focus on what we can agree on (which appears to be a lot in fact) and see where the article can be improved specifically. Graham 01:51, 24 April 2006 (UTC)

I didnt say that test gear measurements are the only way. I said ears are unreliable so we shouldnt rely on them. They may give indications but Test gear is the easier way because you are not hampered by the vagaries of peoples mood, health, intoxication, previous exposure to loud sounds, physiology, psychoacoustic differences etc.etc.--Light current 02:13, 24 April 2006 (UTC)

Double-blind listening tests are the only reliable way to determine if something such as the effect of damping factor on audio is audible or not. Test equipment can measure signal properties, but not audibility. BTW, I didn't promise to do anything. I don't feel that the burden of proof to establish that a damping factor greater than 10 is insignificant is on me. I didn't make the statement in the article, I questioned it. Still, I'd love to know the facts... where does DF become large enough to be inaudible? Of course, this is highly dependent on the speakers used (again... large, professional woofers are probably more reliant on a high DF to avoid unwanted resonance / distortion). A test wouldn't be too hard to design, but to be valid you would want to null the effect of volume. I'm sure you would see a big difference between trained subjects and untrained. Tvaughan1 02:48, 24 April 2006 (UTC)

Interesting table from one of the links

For Damping factors between 50 and 2000 there is less than 1ms diff in decay time. There is 1 ms diff between DF of 20 and DF of 50. Can TV1 hear 1ms difference? [2]--Light current 22:27, 23 April 2006 (UTC)

Please sign your posts. This is an interesting reference... I added it to damping factor. Did you read the rest of the article? The tables in this article were made using certain assumptions (speaker with a Q of 3, etc.). In any case, the audibility higher damping factors is certainly an open question. I don't claim to have the ability to hear the difference between a DF of 50 and 2000, nor even 20 and 50. I'm reasonably sure I could hear the difference between 10 and 100... but I'd have to conduct a valid experiment, and such and experiment would take some careful set-up (especially if you are using a series resistor, which changes the volume, and if compensated by increasing the amplifier's gain it increases the noise... although in most cases by a trivial amount). Tvaughan1 22:20, 23 April 2006 (UTC)

Merge

No — Omegatron 01:12, 23 April 2006 (UTC)
No — Tvaughan1 03:58, 23 April 2006 (UTC)

Dont be shy - say what you really mean! Theyre the same thing!!! Have a (long) think about it! Ive only just realised it myself--Light current 13:35, 23 April 2006 (UTC)

Archiving

This page is overdue for archiving. As replies and posats have been interleaved somewhat, does anyone want me to put the posts now into logical order unser separate hdgs (refactoring) . Or do you want to leave them in this (rather unsatisfactory) order before archiving? See WP:REFACTOR--Light current 03:38, 24 April 2006 (UTC)

It is not. Leave it until a few weeks after discussion has died down. — Omegatron 13:45, 24 April 2006 (UTC)

What about the page size limit?--Light current 14:43, 24 April 2006 (UTC)

What about it? It's archaic, and only a recommendation. Any decent browser that is less than five years old can handle much more than 32K text. Graham 14:46, 24 April 2006 (UTC)

Yes Ive heard that, and Ive also heard the other argument that pages should not exceed 32k for readability etc. THe policy is uncertain. However, I do beieve that a page longer than 32k can be difficult to navigate and follow esp with multiple respondents.--Light current 15:15, 24 April 2006 (UTC)

THIELE - SMALL PARAMETERS: for JBL 2245H

fs: 20 Hz
Re: 5.8 ohms
Qts: 0.27
Qms: 2.2
Qes: 0.31
Vas: 820 L (29 ft3)
Sd: 0.130 m2 (200 in2)
Xmax: 9.5 mm (3/8 in)
Vd: 1,230 cm3 (75 in3)
Le: 1.4 mH
no (Half space): 2.1%
Pe (Max): 300 W Continuous Sine Wave

--Light current 23:17, 25 April 2006 (UTC)

Marantz page mumbo jumbo

Refer to this [3]

The reason that a speaker might require more - or less - current given a particular applied voltage is that the impedance of the speaker system changes depending on what direction the speaker drivers are moving when the voltage is applied, how fast they are moving that way, and where they actually are in the first place. It helps to visualize that the intent of placing a voltage on a speaker system is to put the speaker cone(s) in (a) particular position(s). How difficult that is to do depends on the factors just mentioned (and others, some subtle.)

my bolding

Well, this is IMO just complete and utter horse droppings (with extra added BS). Where did this guy get these ideas? And who the hell is he anyway to be peddling this stuff? The speaker impedance does not depend on the direction or the amplitude of motion in a linear system. It may (will) depend on frequency.

Also voltage at speaker terminals does NOT determine the position of the cone. It is the voice coil current that does that. THe voltage comprises a portion due to I*Z and a portion due to the emf generated by the moving cone. This is simple High School physics!

This sort of false info published on the net is most dangerous rubbish. Some people actually believe it and quote it as gospel. --Light current 22:21, 25 April 2006 (UTC)

Light current... open your mind! It's not rubbish... it's just different way of describing the problem. Not the ideal description perhaps, but that's what we're trying to accomplish with this article, right?

A speaker is a driver. It has a motor. The piston has mass and inertia, and it acts against acoustic load and a mechanical spring (the suspension). The audio signal we are trying to accurately amplify and reproduce is represented as a varying voltage. What the author was trying to describe was the difficulty in getting the piston to move in response to the audio signal with precision. He is correct in describing that we want the position of the piston to be the mechanical equivalent (proportional) to the voltage of the electrical signal... just as the microphone diaghragm was moved by the acoustical signal (sound pressure variation), generating a voltage proportional to the movement.

Of course, Ohm's law tells us that V = I * R. A speaker doesn't have a fixed resistance, it has a voice coil with a reactive load, and this is measured as an impedance (Z). Nonetheless, if we assume the impedance is constant (it isn't... but this is a standard simplification), the current is proportional to the voltage. The EMF generated by the moving cone is the problem we are describing... the author was leaving it aside when he described what we wanted under ideal circumstances.

On to the part about more or less current being required. As you acknowledged, speakers have widely varying impedance. See Electrical characteristics of a dynamic loudspeaker. Why is this so? It is due to the natural resonant frequency of a speaker. In other words, the varying impedance is due to mechanical factors. The author of the Marantz article was trying to describe the need to electrically damp mechanical overshoot due to the inertia of the piston, or due to the mechanical resistance of the spring (suspension)... not averaged over time, but instantaneously. Of course, the only way to do this is to use feedback circuits to oppose the reverse EMF generated by the moving cone. This feedback will supply increased voltage (and current) as needed to compensate for varying speaker impedance and resonance... as the Crown engineer mentioned (higher damping characteristic gives improved frequency response). Tvaughan1 01:58, 26 April 2006 (UTC)

I stand by my statement that this guy is talking thro his *****le. I reiterate my statement that:

voltage at speaker terminals does NOT determine the position of the cone. It is the voice coil current that does that. The voltage comprises a portion due to I*Z and a portion due to the emf generated by the moving cone.

I invite anyone to try to prove me wrong (or right if you so desire). Comments from other educated editors are welcome. :-|--Light current 02:05, 26 April 2006 (UTC)

I think YOU don't know what you are talking about. Instantaneous current (and position) will give you the instantaneous FORCE. Solving the equation of motion, you can find the position given the instantaneous force history, current history, or VOLTAGE history. It many ways its all the same thing. - TheShagg

For any value of steady current through the coil, a force is produced proportional to that current. THe force acts to move the cone until it is balanced by the stiffness of the suspension.--Light current 17:40, 12 October 2006 (UTC)

Critical damping

I think Im right in saying that what is required from a low frequency speaker system is critical damping. That is, not underdamping (where you get hangover) or over damping (where the cone just wont move fast enough in response to a step voltage input). Now it is my contention that a speaker is a mechanical system with a mechanical to electrical transformer (the voice coil) attached and this transformer is driven by an amplifier of usually very low output impedance.

Transformation analysis

Since the mechanical force on the coil, f_m is given by:

f_M=Bli

and the emf developed by the moving coil is

e= Bl dx/dt

It is therefore evident that:

Z_EM= (Bl)^2/Z_M

where Z_EM is the electrical impedance due to motion, and is termed the 'motional electrical impedance'. The mechanical to electrical transformation ratio for impedances is therefore (Bl)^2 and is determined solely by the voice coil/magnet assembly.

Now in analysing such a system, one may choose to convert the mechanical components (mass, compliance, acoustic resistive or reactive loading etc) into electrical components or vice versa by referring the elements to one side or the other of the 'transformer'. It is usually easier to do the former and analyse the performance in the electrical domain. The mechanical/electrical transformer is of course an electric motor/generator. The transformation constant (Bl)^2 is a scalar and therefore does not change the phase angle of the impedance when transformed, only its magnitude. Therefore inductive components (masses) transform to inductances, and capacitive (compliances) transform to capacitors, whilst resistors (mechanical dampers) transform to resistors.

The total electrical impedance seen at the voice coil by the amplifier, Zen, is given by the sum of the damped electrical impedance and the motional electrical impedance ie:

Z_en =Z_e1 +Z_em

Now Z_en can be represented as a series combination of R,L and C. The L represents the transformed mass of the cone in the mechanical cct, and the C represents the total transformed compliance of the driver. The R represents the sum of the transformed motional resistance plus the electrical (DC) resistance of the voice coil.

This analysis appears to be original work. Wikipedia is not the place to publish original work. Tvaughan1 00:47, 15 October 2006 (UTC)

Damping methods

It should be evident that damping of this (essentially second order) system can be accomplished either by mechanical means or by electrical means or by a combination of the two. There is however an inherent limitation in the 'electrical only' damping method. That is the finite resistance and inductance of the voice coil. This can only be reduced by the use of a 'negative' resistance in the amplifier and this is usually not employed for stability reasons (ie it tends to turn the system into an oscillator)

The complete electrical equivalent circuit of a loudspeaker in a cabinet is shown in the figure.

For the system to be critically damped, we can see that in the electrical equivalent circuit, R must be of a certain (critical value.

To be continued... Please do not interrupt this post >:-{

--Light current 17:02, 26 April 2006 (UTC)

Yes. We do. Across the full frequency range of the speaker. But the article is about damping factor, so it needs to focus on how you design or evaluate an amplifier (which may have any speaker connected... cheap or well designed... in various configurations (alone, series, parallel), through wires with unknown length or resistance. Tvaughan1 22:38, 26 April 2006 (UTC)

OK then watch this space!--Light current 22:14, 27 April 2006 (UTC)

Holy bejesus, cleanup?

I can't believe this page! Its like 5 times as long as the article. Does anyone think this talk page needs some cleaning/archiving or something? Fresheneesz 04:17, 19 May 2006 (UTC)

Ive archived about 40k s worth but we're still in the middle of a heated discussion so I think the rest should stay for a while. Just having a cool off between rounds!--Light current 13:22, 19 May 2006 (UTC)

audio only?

Is this page only applicable to audio amplifiers? Comming to a page on the damping factor, I would expect a more general treatment. For example, I was looking for the definition of the damping factor in terms of any general circuit. Does the definition of source impedance vs load impedance work in general, or just for audio amplifiers? Fresheneesz 00:22, 5 June 2006 (UTC)

No, its not restricted to audio amps ATM! It should be a very general treatment.We could put a note at the top of the page saying this page only applies to audio and redirecting others to damping? The definition 'source impedance vs load impedance' is specifically reserved for audio amps and is not the general defn of damping factor. Please see damping 8-)--Light current 00:30, 5 June 2006 (UTC)

I think that perhaps this subject needs a bit of a reorganization. We could change the name of this article to damping (electronics), and have a little disambig for damping factor that can redirect to either page (damping or damping (electronics)) - since it seems like the damping page doesn't cover anything about circuits. Fresheneesz 02:43, 7 June 2006 (UTC)

Or.. yes it does have stuff about electronics. I didn't even see that. Fresheneesz 02:44, 7 June 2006 (UTC)

Zero electrical damping factor

Whoever wrote this nonsense doesn't understand how a speaker works.

Firstly, once you place a speaker inside a sealed box, there is less damping on the cone at resonance. The speaker has to overcome the air in the box, this is why the system Qtc is always larger than the Qts.

The cone excursion *is* controlled above resonance, thus it is linear until the Levc becomes significant. Below resonance it is reduced due to the order of the system.

The major issue about control is not due to the Levc, but at resonance. So even though a zero damping factor may correct for the inductance, we will be left with a huge cone excursion (and frequency response) peak at resonance due to the large electrical losses (very high Qts and Qtc). With a zero damping factor, the frequency response will not extend down to dc.

I believe you can model the effect of a very small damping factor in loudspeaker modelling programs, (such as Winisd Pro which you can download for free) simply by entering in a large value for the series resistance. Or you could go out and test it for yourself.

If you wish to correct for the second order roll off, you simply need to equalize the response with a 12db/octave filter. In reality, you won't want to do this all the way to dc because you will get negligible output. But you can use a filter that extends the response to a lower frequency (at the expense of efficiency) There are a few different filters that you can use, but you should consider the linkwitz transform filter as a possible solution, rather than this zero damping factor nonsense.

Arch.

What are your references? — Omegatron 13:40, 27 June 2006 (UTC)

This was based on my own understanding, I don't have time to sort this out properly. But if you want serious references for the article, then you should look at all of the usual sources - Richard Small's papers, books like "The Loudspeaker Design Cookbook" and so on. There is information about the Linkwitz transform filter, http://www.linkwitzlab.com/filters.htm#9 and http://sound.whsites.net/linkwitz-transform.htm

The losses through Re are not critical as it is easily possible to design a driver so that the Qts is less than 0.5 (critical damping). (It has to be significantly less than 0.5, so you can account for the enclosure too...)

If you add series resistance, then you are dramatically reducing the electrical damping, thus raising the Q dramatically (obviously making things much worse).

Oh, and have a look at this page too: http://www.linkwitzlab.com/thor-design.htm --Arch 11:25, 28 June 2006 (UTC)

Firstly, once you place a speaker inside a sealed box, there is less damping on the cone at resonance. The speaker has to overcome the air in the box, this is why the system Qtc is always larger than the Qts.

Could the author explain why theres less damping in a sealed box? I thought there was more due to the reduced compliance of the air load at the back.--Light current 13:37, 27 June 2006 (UTC)

Yes, there is reduced compliance of the air load, so the resonant behaviour of the loudspeaker is increased. (Hence why the Qtc also increases) --Arch 11:25, 28 June 2006 (UTC)

Yeah So there is more damping. You said less! 8-(--Light current 11:32, 28 June 2006 (UTC)

there is less damping. Qtc > Qts. oscillations in the system don't loose energy as quickly, thus less damping. less air, more pressures. boxes smaller than Vas will have the suspension provided by the air volume instead of the speaker. as box volume increases to values much larger than Vas, Qtc approaches Qts by definition. —The preceding unsigned comment was added by 68.75.154.192 (talk) 01:37, 9 February 2007 (UTC).

I've removed that section since the explanation was clearly incorrect. If you wish to mention the effects of current drive vs voltage drive, negative impedance etc, I'd suggest using Rod Elliot's findings as a guide - since he provided clear empirical results and a more reasonable explanation. http://sound.whsites.net/z-effects.htm Arch 13:25, 13 September 2007 (UTC) —Preceding unsigned comment added by Architectonic (talk • contribs)

Page name

I propose this page is called Loudspeaker damping factor to differentiate it from the other sort related to the detailed issue of damping.--Light current 15:46, 27 June 2006 (UTC)

Damping has a section about loudspeakers, too. Definitely need to either merge or disambiguate, depending on whether they're related enough — Omegatron 15:51, 27 June 2006 (UTC)

My initial reaction is that they are not related except by name. But Ive been meaning to make sure.8-|--Light current 20:14, 27 June 2006 (UTC)

THey cant be the same thing as audio damping factor is just a ratio, whereas real damping factor has dimensions of frequency.8-|--Light current 21:25, 27 June 2006 (UTC)

Critical damping of loudspeakers

I believe that critical damping of a loudseaker cone by electrical methods is, in almost all cases, impossible due to the non zero resistance of the voice coil.--Light current 10:30, 28 June 2006 (UTC)

it is possible. the poles are in the left half plane and can be canceled. many speakers have a Qts < 0.5, and will be overdamped unless put in a box that is not "large" compared to Vas. —The preceding unsigned comment was added by 68.75.154.192 (talk) 01:39, 9 February 2007 (UTC).