Even then we would fall short. There is a considerable amount of research that shows the impact of low frequency audio and infrasonic information conveyed through other body tissues too. The percussionist Evelyn Glennie is amazing evidence of that.
The information is already there to be used. The analysis tools are here already. People choose not to use it all.
Hi tmuikku,
Well, this is where you can't have everything you want. You can optimize for one aspect, but the rest goes to pot. So far, voltage drive is the best way to do things. If it wasn't, the bulk of audio systems would drive a different way. The market would have naturally gone that way. So we can say that we have an imperfect way of doing things, but looking at the overall picture it seems to work the best. What I am saying is that while controlling current through the driver has advantages, they don't solve all the issues. Big picture.
From what you have said, distortion is reduced with high impedance drive, yet we also lose control over the motion of the cone doing this. Hmmm, choose your poison I guess.
The amplifier is actually important, and part of the equation. It must be at least specified in terms of power, voltage compliance, maximum current capability and output impedance in order to test a driver under those conditions. The test results will vary widely.
I can also say that different amplifiers with low distortion do sound different even with budget speakers. However, in these cases the differences are easily measured and quantified. Case in point, a Luxman L-580 and Marantz 170DC sound completely different in a pair of in-wall 6 1/2" two way speakers. The difference surprised me as I didn't expect to hear it in these very modest speakers.
Well, this is where you can't have everything you want. You can optimize for one aspect, but the rest goes to pot. So far, voltage drive is the best way to do things. If it wasn't, the bulk of audio systems would drive a different way. The market would have naturally gone that way. So we can say that we have an imperfect way of doing things, but looking at the overall picture it seems to work the best. What I am saying is that while controlling current through the driver has advantages, they don't solve all the issues. Big picture.
From what you have said, distortion is reduced with high impedance drive, yet we also lose control over the motion of the cone doing this. Hmmm, choose your poison I guess.
The amplifier is actually important, and part of the equation. It must be at least specified in terms of power, voltage compliance, maximum current capability and output impedance in order to test a driver under those conditions. The test results will vary widely.
I can also say that different amplifiers with low distortion do sound different even with budget speakers. However, in these cases the differences are easily measured and quantified. Case in point, a Luxman L-580 and Marantz 170DC sound completely different in a pair of in-wall 6 1/2" two way speakers. The difference surprised me as I didn't expect to hear it in these very modest speakers.
Yes, they would change frequency response by diluting transducer ability to influence current, which also reduces distortion in acoustic domain. Frequency response can be compensated to some degree before amplifier so what is left is reduced distortion, and perhaps some power lost in the parts as heat.
Hi soundbloke,
Sure, then if you also reproduce the sound field the same, that same impact affects your other body tissues the exact same way. I experienced this with a set of Klipsch Jubilee speakers playing concert material at and above the original levels. Simply amazing!
If the information is there, how can you say people simply refuse to use it? If it correlated with subjective listening, any engineer I know would be all over it. Now if you are talking about the ad people, yep. Anything that looks "ugly" will be hidden. Especially if it reveals faults with their product.
Engineers I know are like you. They chase the truth and knowledge. Marketing folks, well there are a different kettle of fish unless the data is in their favour.
Sure, then if you also reproduce the sound field the same, that same impact affects your other body tissues the exact same way. I experienced this with a set of Klipsch Jubilee speakers playing concert material at and above the original levels. Simply amazing!
If the information is there, how can you say people simply refuse to use it? If it correlated with subjective listening, any engineer I know would be all over it. Now if you are talking about the ad people, yep. Anything that looks "ugly" will be hidden. Especially if it reveals faults with their product.
Engineers I know are like you. They chase the truth and knowledge. Marketing folks, well there are a different kettle of fish unless the data is in their favour.
Hi tmuikku,
Well, let's not make conclusions like that. "diluting transducer ability to influence current" is the conclusion. There are other things at play here.
This is a total package. Distortion is but one component of performance. Optimize for that and you allow other problems to become more dominant. Pick your poison, but the market will vote with it's dollars, picking what sounds the best. What is true is that we have a compromise, so make the best trade-offs you can.
Well, let's not make conclusions like that. "diluting transducer ability to influence current" is the conclusion. There are other things at play here.
This is a total package. Distortion is but one component of performance. Optimize for that and you allow other problems to become more dominant. Pick your poison, but the market will vote with it's dollars, picking what sounds the best. What is true is that we have a compromise, so make the best trade-offs you can.
I don't know about the OP, but I'd sure love to hear what you think needs to be done to make a "good sounding" speaker.If the OP is interested, I'll pontificate on what to do for a 'good sounding' speaker.
The first two sentences might appear to contradict each other: You have identified that the output impedance of an amplifier has a significant effect on the driver distortion emitted; Then you say the amplifier has nothing to do with it. I fear more confusion...
As I commented previously, my contributing to this thread has been to show how measurements can better relate to subjective assessments commensurate with the questions posed in the thread opener. In that endeavour, and in particular to add subjective significance to different distortion measurements, amplifier output impedance will determine what you can measure and what you can hear. Testing a driver without a source is a tricky proposition.
Off-topic they might be, but there are some worthwhile nuggets of information in this thread - and many more in others on this forum.
Motion of cone is damped by the driver itself, by allowing electrical damping through back EMF generating current that opposes the movement. This only happens around at driver resonant frequency though, where mass is cancelled by spring and force results velocity that makes back EMF 180deg out of phase from the initial force. But above this, on the mids, the force accelerates mass and velocity lags 90deg behind and is now perpendicular to the initial force, having no electrical damping effect. Increasing inductance of voice coil would add more delay, making the supposed control actually amplify disturbance higher up in frequency. This is very well laid out in this one article here: https://www.edn.com/loudspeaker-operation-the-superiority-of-current-drive-over-voltage-drive/
Please ignore the controversial title of the article and all the fight between amplifier types, it's almost irrelevant, important thing is the mechanisms at play to be able to utilize the info to our advantage.
Almost, except one can optimize for suitable compromise. On an active system, with voltage amplifier, insert an inductor in series with the driver for one low pass filter pole. This would leave low impedance for electrical damping to happen at resonance, the control, and reduce distortion from non-linear Le(x) for example, or hysteresis in the iron. Best of both! or anything in between, what ever suits a particular application.
For sure, amplifier matters for it's own sake, but no matter which voltage amp you have you still could add impedance to reduce transducer distortion, in addition!
It all comes into system design, if there is no DSP for example, one could still make passive speaker with two different crossover variations, which would yield quite similar frequency response, but difference in distortion performance.Whether any of this is audible is another question as it also depends on system design for particular application, but still it's a way to manipulate acoustic output of a system.
Last edited:
Oh, I almost forgot, on the validity of tonal testing:
Instead of swiping a constant amplitude tone across frequencies, lets swipe a constant frequency tone across a range of amplitude, up and down. The spectrogram would look like that:
and the flattening to time:
which is ... somewhat more complicated than some believers in tonal testing may have been used to expect.
Are the THD / harmonics / etc numbers the "scientific" properties of the driver or just the artifacts of a certain archaic methodic, a wild mix of whatever second-order effects?
Instead of swiping a constant amplitude tone across frequencies, lets swipe a constant frequency tone across a range of amplitude, up and down. The spectrogram would look like that:
and the flattening to time:
which is ... somewhat more complicated than some believers in tonal testing may have been used to expect.
Are the THD / harmonics / etc numbers the "scientific" properties of the driver or just the artifacts of a certain archaic methodic, a wild mix of whatever second-order effects?
Hi tmuikku,
Just one thing. The damping is effective well away from the resonance of the driver as well. This is audible for most people. The cone isn't always good at damping itself and any time you have ringing or other resonance, you need electrical damping for sure.
Yes, a compromise is needed. What that looks like is beyond me. What we do need to do is optimize those things most easily heard while keeping the other aspects under control. Your classic engineering problem.
Just one thing. The damping is effective well away from the resonance of the driver as well. This is audible for most people. The cone isn't always good at damping itself and any time you have ringing or other resonance, you need electrical damping for sure.
Yes, a compromise is needed. What that looks like is beyond me. What we do need to do is optimize those things most easily heard while keeping the other aspects under control. Your classic engineering problem.
I don't think anyone should argue with the title without investigating the matter for themselves. Most advantages of "current drive" are overblown and of slight subjective advantage only, if that. Oddly, however, the removal of mid-band motor linearities can be significant audibly but is seldom referred to. As can be the constant frequency balance in the presence of thermal modulation in multi-way designs too.
Cone resonances are different from the fundamental resonance of the cone/coil assembly where electrical damping is applicable. As I mentioned in a previous post, cone (or surround) resonances can also be reflected in the coil impedance, and therefore would do better with a high source impedance not a low one. Ameliorating this distortion mechanism is why I believe Purifi developed their odd looking surrounds?
Hi soundbloke,
Yes, different for sure. Cone fundamental resonance has a larger displacement, and that makes electrical damping more effective.
There is the rub. Sometimes very low source impedance is needed, other times a higher one I guess. This may be done through processing of some kind, or we have to determine which aspect is more important and live with the other until technology catches up. Maybe this problem will remain and a new technology for exciting the air will emerge.
The surround might damp the cone vibrations at the edge, mostly it is there to keep the cone and voice coil centered. I would think that disturbances in the cone itself is better handled with cone profile and materials. I haven't examined surrounds for many years since their effect is understood pretty well.
Yes, different for sure. Cone fundamental resonance has a larger displacement, and that makes electrical damping more effective.
There is the rub. Sometimes very low source impedance is needed, other times a higher one I guess. This may be done through processing of some kind, or we have to determine which aspect is more important and live with the other until technology catches up. Maybe this problem will remain and a new technology for exciting the air will emerge.
The surround might damp the cone vibrations at the edge, mostly it is there to keep the cone and voice coil centered. I would think that disturbances in the cone itself is better handled with cone profile and materials. I haven't examined surrounds for many years since their effect is understood pretty well.
If you look at any transducer datasheet frequency response and impedance plot, they state it's made with constant voltage, which means such driver is made for voltage amplifier, just like you say.
This means the transducer is made for voltage drive, made to regulate current in the circuit, tailored to EQ acoustic output for nice sound and graphs, right.
Add series impedance outside the driver, apply constant voltage sweep, and both plots would look different. One could correct for the frequency response though, by varying the sweep voltage, and have it flat again, use EQ before amplifier, nice!
It's just a system that can be exploited to benefit the end goal, what ever that is. There is no point make black and white division, lot's of gray area in between, all passive speakers with their crossovers are gray! Just a powerful thing to be aware of.
Yeah, exactly. If someone didn't already notice, the common denominator for all you listed here is just impedance in series with the transducer. Current drive is nothing more than added impedance series with the transducer, which reduces or even eliminates transducers ability to affect current in the circuit, which makes the difference in acoustic domain between the two.
Impedance varies with frequency, so do the electronic and mechanical prooerties, so one could almost pick and choose what to address. Electrical damping to keep control and manipulate system Q, check, reduction of midband hysteresis check, what do I need? low impedance in series with voice coil at resonant frequency, high impedance on the mid band, now just figure out if it's doable, and how
Well, simplest is just a regular voltage amp with low impedamce, and an inductor. Want something else, no problem.
Last edited:
Why almost nobody uses current drive? It is not audible i guess..
Yes, i understand how current drive reduces some of the distortion products (not all). But most of sound quality is coming from other terms that current drive can improve. Even no perfect current source exists.
Most of the drivers have designed to use voltage drive. It is easier to use with voltage source and like purify have show as that even distortion can be low using voltage drive.
Yes, i understand how current drive reduces some of the distortion products (not all). But most of sound quality is coming from other terms that current drive can improve. Even no perfect current source exists.
Most of the drivers have designed to use voltage drive. It is easier to use with voltage source and like purify have show as that even distortion can be low using voltage drive.
Say... would you be able to point an interested party to some of those tools? Looking for Wigner Distribution brings up academic papers and an occasional Python script that may or may not be useful for acoustic analysis (plus I don't yet know Python). Any other recommendations of analysis tools would be appreciated.