thanks for the insight guys, really nice little resource here for a much touted, but rarely understood factor (certainly i dont understand it fully).
there has been some conversation over on head-fi about damping factor with headphones, i could not really see how in the most part they could be noticeably effected by DF? loudspeakers its obviously part of the equation, as we have a much greater interaction of electrical, atmospheric and mechanical impedances/resonances, but i'm puzzled how this might have an effect on a 28ohm light, crossover-less, fairly open back headphone system.
would anyone care to cast some light on this? some have put the effect as quite profound, i can see maybe a small effect particularly with closed headphones.
what i mean to clarify is, what effect would overdamping such a headphone/system have? the pundits were promoting the use of current drive for these headphones being superior somehow, i would like to understand why, it would seem to be a fairly isolated case as i have only seen people mentioning it WRT grados/alessandros, which are woodies and they have compared them to such fullrange drivers as Fostex, Lowther etc
there has been some conversation over on head-fi about damping factor with headphones, i could not really see how in the most part they could be noticeably effected by DF? loudspeakers its obviously part of the equation, as we have a much greater interaction of electrical, atmospheric and mechanical impedances/resonances, but i'm puzzled how this might have an effect on a 28ohm light, crossover-less, fairly open back headphone system.
would anyone care to cast some light on this? some have put the effect as quite profound, i can see maybe a small effect particularly with closed headphones.
what i mean to clarify is, what effect would overdamping such a headphone/system have? the pundits were promoting the use of current drive for these headphones being superior somehow, i would like to understand why, it would seem to be a fairly isolated case as i have only seen people mentioning it WRT grados/alessandros, which are woodies and they have compared them to such fullrange drivers as Fostex, Lowther etc
Last edited:
sumaudioguy said:
I assume, Bob, you would agree that an amplifier has only one output terminal impedance, as some people here seem to think the opposite! Whether it is measured by injecting a current or drawing a current the result will be the same.Bob Cordell said:
I would expect so.
Normal operation is injecting a signal to +IN at the input.
Reverse operation is injecting a signal to -IN at the output.
Normal operation is injecting a signal to +IN at the input.
Reverse operation is injecting a signal to -IN at the output.
As usual, I'm left slightly cold by the frequency domain 'proxies' for what amplifier output impedance means. Intuitively, a higher output impedance would reduce a speaker's 'slam' when required to produce a huge transient, causing it to smear a sharp edge, and it would also allow the cones to continue wobbling for longer afterwards. In the frequency domain you could say that the frequency response is still "tolerably flat", or that phase doesn't matter. You could say that electrical damping is "not all that important", but at the same time you'd be telling us that you can hear the difference between different cables or noise at -100dB. I don't get it. In my opinion, if the speaker is designed for zero output impedance then that is what you should give it as close as possible. Anything else just seems like hand waving and making it up as you go along.
I don't agree.
The designer may set out to assemble and test a speaker (combination of drivers and crossover) assuming zero output impedance.
The designer is fully aware that real amplifiers, using real cables and real connectors does insert a resistance into the amp speaker circuit and will also realise what effect that has on his "new" design.
I would expect that designers to then trim/tune the sound of that speaker to suit his preferred range of amplifiers. That choice of amplifiers/cables/connectors will be closely linked to the target market for his speaker.
He/she will make the speaker sound good to their potential buyers.
The designer may set out to assemble and test a speaker (combination of drivers and crossover) assuming zero output impedance.
The designer is fully aware that real amplifiers, using real cables and real connectors does insert a resistance into the amp speaker circuit and will also realise what effect that has on his "new" design.
I would expect that designers to then trim/tune the sound of that speaker to suit his preferred range of amplifiers. That choice of amplifiers/cables/connectors will be closely linked to the target market for his speaker.
He/she will make the speaker sound good to their potential buyers.
Only if the driver/speaker has a high Qms (unfortunately many if not most bass drivers).
dave
I don't understand what you mean by the first sentence. What are "proxies" for the meaning of output impedance?CopperTop said:
Be careful of "intuition". Like conscience, if not properly trained it can lead you astray. If a "huge transient" has a "sharp edge" then presumably we are talking about high frequencies, where amplifier damping factor does not tell us much about speaker damping. On the other hand, if we are talking about "wobbling cones" then maybe you mean low frequencies?
Think that these old tube Amp article could be interesting for this topic to : Missing Link in Speaker Operation
Best Regards for all !
Best Regards for all !
While i totally agree on post #65 as being a perfectly valid scenario for designing
a high quality loudspeaker, in reality enclosure-size and (crossover)cost are also important parameters for some "designers".
You only need to look at the popularity of 2.1/5.1 systems with enclosures the size of earbuds, to realize how important small enclosure-size is for many consumers.
So a "designer" may be tempted to design a system as small as possible (or smaller),
resulting in underdamped lf-response.
When a consumer that does actually use his ears complains about the uncontrolled bass,
a salesman may argue that said consumer needs a "better"(higher df) amplifier,
perpetuating the myth that high df is better by definition.
a high quality loudspeaker, in reality enclosure-size and (crossover)cost are also important parameters for some "designers".
You only need to look at the popularity of 2.1/5.1 systems with enclosures the size of earbuds, to realize how important small enclosure-size is for many consumers.
So a "designer" may be tempted to design a system as small as possible (or smaller),
resulting in underdamped lf-response.
When a consumer that does actually use his ears complains about the uncontrolled bass,
a salesman may argue that said consumer needs a "better"(higher df) amplifier,
perpetuating the myth that high df is better by definition.
Interesting article, thanks. It may have limited applicability nowadays, as most speakers are intended for pure voltage drive. Note that the graphs show that damping factor only has an effect around the bass resonance, as I said.banat said:
The salesman could be right, although in extreme cases a negative output impedance might be needed for a seriously underdamped speaker. The error lies not in the salesman's words, but the inference which some people may draw from them.kvholio said:
Unfortunately, small enclosures may be forced on consumers by their wives.
That "under-damped" response results from a Qbox >1/sqrt(2). (Qbox from 0.9 to 1.1 is very common in small speakers).
This high Q exaggerates the bass response, and some use the increased second harmonic distortion to enhance this "false bass" effect.
The speaker designer can and will deliberately use this "trick" to make small cheap speakers sound as though they are producing adequate bass volume.
This is a design technique specifically aimed at low cost speakers or at ultra small speakers that are not cheap. It's the designer reading his market requirements.
This high Q exaggerates the bass response, and some use the increased second harmonic distortion to enhance this "false bass" effect.
The speaker designer can and will deliberately use this "trick" to make small cheap speakers sound as though they are producing adequate bass volume.
This is a design technique specifically aimed at low cost speakers or at ultra small speakers that are not cheap. It's the designer reading his market requirements.
While this may be the case usually, it's not always true.
i've owned a pair of LS from a reputable British manufacturer
that were neither cheap nor extremely small (20 L internal volume)
Enclosure was "too small", counteracted by a big cap in series with the woofer.
DCR of the series-coil was a whopping 1.4R, DCR of woofer was 4.7R.
Not surprisingly it sounded best with a high DF amp.
Agreed
KEF by any chance?
They assume a good amp to drive their speaker, combined with low resistance connections. Again reading their market.
Yes, diy-version (CS3) of their 103.2 .
Their assumption of using a high DF-amp to drive their speaker
has the benefit of ending up with a smaller enclosure (easier to sell, less costly to manufacture, less costly to ship)
Sorry DF96, it's just part of my on-going quest to understand why the world of audio is expressed solely in the frequency domain when in some cases, the time domain would seem more... intuitive.
Not sure what you mean by the statement that a sharp edged transient only contains high frequencies. A step-like input has a sharp edge, but I'd be expecting my woofer to move too.
It's not the sharp edge that makes the woofer move,
this only makes the tweeter move.
It's the length of the step-like input that makes the woofer move.
Doing impulse-measurements on loudspeakers makes it intuitively easier
for me to understand;
The sharp transient of the stimulus contains no lf-information,
it's the slow wobble that follows the transient.
this only makes the tweeter move.
It's the length of the step-like input that makes the woofer move.
Doing impulse-measurements on loudspeakers makes it intuitively easier
for me to understand;
The sharp transient of the stimulus contains no lf-information,
it's the slow wobble that follows the transient.
Yes, but the original discussion was about a transient with a sharp edge. One poster assumed that a sharp edge meant a sharp pulse, without also considering that a transient with a sharp edge can contain lower frequencies too.
Anyway, now into thoroughly pedantic mode, a tweeter is no different from a woofer or a mid-range (if you're going to mention AC coupling), with a finite high frequency cutoff. With a genuine 'sharp edge' what you just said about a woofer is equally true of the mid-range and the tweeter, just on a different scale.
... and as we are talking, presumably, about AC current, it anyway changes direction many times per second wheter sourcing or sinking. 😉
jan
Hi,
The issue is the standard circuit for attenuating amplifier outputs has
a source impedance (resistance) of around 120 ohms, whilst a headphone
amplifier, an MP3 player, a walkman etc can be much lower effectively.
With old school 600 ohm phones 120R represents a "damping factor" of
about 5 and its voltage drive, with 8 ohm headphones there is very
little "damping factor" and 120R is essentially current drive.
So if your building a headphone amplifier, what do your phones want ?
rgds, sreten.
As BC stated earlier DF is often quoted as a single number, e.g. > 100
at a single, usually low frequency. At low frequencies for typical hifi
speakers the actual applicable DF is limited by the series resistance
of the inductor in series with the bass unit. If this is say 0.3R the
effective DF will be about 27 no matter how big the amplifiers
DF is stated, generally anything bigger than 100 is meaningless.
DF really is pretty much meaningless in the bass with a source
resistance of say less than 0.1R, the actual damping is totally
dominated by the bass drivers coil resistance limiting current.
A poor DF of say 8, i.e. a source impedance of 1 ohm can sound quite nice,
slightly raising the bass Q. Even quite decent valve amplifiers can have
a source impedance of around 2 ohms, they do interact with speakers,
sometimes for the better, sometimes for the worse.
Then of course there are speakers that work better with a DF around 1.
- Status
- Not open for further replies.