Which way is up? (Which way does a loudspeaker driver move?)

[Salt]

Maybe the systems lack the direct recognition of sound generation.
Measurements as done today are usually indirect via measuring sound pressure by a microphone, without imaging the movement of the membrane.
This is like X-Ray/CT/MRT, that gives no obvious sight to what is going on, different to what a surgeon would see (and feel).
So primary a closer look at membrane's movements, let's say by uptake by laser of every 1 mm^2, may give an insight what really is going on with reproduction of electrical signals all over the membrane (not only one point).
Arithmetics and modeling may be adapted to that.

 

[Tim Link]

Maybe the systems lack the direct recognition of sound generation.
Measurements as done today are usually indirect via measuring sound pressure by a microphone, without imaging the movement of the membrane.
This is like X-Ray/CT/MRT, that gives no obvious sight to what is going on, different to what a surgeon would see (and feel).
So primary a closer look at membrane's movements, let's say by uptake by laser of every 1 mm^2, may give an insight what really is going on with reproduction of electrical signals all over the membrane (not only one point).
Arithmetics and modeling may be adapted to that.

I read somewhere a post, maybe on DIYaudio.com, that a guy had checked the motion of a woofer using a laser to see if it's position followed the wave form from the amplifier. He said it tracked almost perfectly above resonant frequency and below breakup, so that's a start. The cone on a woofer is moving very uniformly and like a piston, and in sync with the input signal from the amp. I don't have a laser, but if we know the peak excursion point and can get a tick to occur at that point, it should tell us all we need to know. As a matter of fact, I think we can just include the tick in the recording if the woofer has enough bandwidth to create it. So my prediction is that if I generate a bass waveform with a tick at the top of each wave peak, I'll measure it from the speaker as happening at the zero crossing point regardless of my distance from the speaker, assuming I don't get phase shifts from early reflections.

 

[Salt]

You may ask @Karl-Heinz Fink , as he has the technical backup, to performe some measurements.

 

[René - Acculution.com]

I read somewhere a post, maybe on DIYaudio.com, that a guy had checked the motion of a woofer using a laser to see if it's position followed the wave form from the amplifier. He said it tracked almost perfectly above resonant frequency and below breakup, so that's a start. The cone on a woofer is moving very uniformly and like a piston, and in sync with the input signal from the amp. I don't have a laser, but if we know the peak excursion point and can get a tick to occur at that point, it should tell us all we need to know. As a matter of fact, I think we can just include the tick in the recording if the woofer has enough bandwidth to create it. So my prediction is that if I generate a bass waveform with a tick at the top of each wave peak, I'll measure it from the speaker as happening at the zero crossing point regardless of my distance from the speaker, assuming I don't get phase shifts from early reflections.

It cannot track the same above and below resonance, because of the phase shift around the resonance.

 

[Cbdb2]

The acceleration creates the change in pressure because it creates the change in speed, more specifically the change in velocity.

The (rapid) change in pressure is what we call sound, or you could hear barametric pressure. No acceleration, no sound.
I think Ill go with what these guys say.

about | PURIFI

purifi-audio.com

 

[Tim Link]

The (rapid) change in pressure is what we call sound, or you could hear barametric pressure. No acceleration, no sound.
I think Ill go with what these guys say.

about | PURIFI

purifi-audio.com

Or you could test it! It's one thing to "go with what these guys say." It's another thing to understand exactly what they mean. Yes, I can't argue - the acceleration has to happen to get the cone moving in a way that creates sound. But maybe they can pipe in and tell us if they really believe that peak pressure at the cone's surface always occurs when the cone is at its maximum acceleration. If not always, then under what conditions? I'm convinced that if the cone is moving into a very small space that it can fully pressurize during the wave cycle, then the peak pressure does indeed occur at maximum excursion, which is when the driver is rapidly changing direction, so that should also be peak acceleration. It's important to note that this is just a coincidence. It's not the acceleration that's producing peak pressure in that case. It's the fact that the cone has moved in to the small space and made it smaller, so the air gets squeezed. If the cone were to just stop and sit there, the pressure would remain high. Acceleration itself does not create pressure in this case. Acceleration can cause pressure in some cases, like the pressure of your back against the seat of a car. In that case, if the acceleration is maintained, the pressure is maintained. If the acceleration ceases, the pressure also ceases even though the car is still going very fast.


If the driver is playing into free space, then the sound waves move away from the driver much too fast for the pressure to remain high at the cone's surface at peak excursion. When sound can propagate it moves at the speed of sound. The cone doesn't move anywhere near that fast. By the time it comes to a stop and starts moving in the opposite direction, the peak pressure pulse it made while moving forward has moved outward

 

[Cbdb2]

If the driver is playing into free space, then the sound waves move away from the driver much too fast for the pressure to remain high at the cone's surface at peak excursion.

Your not separating near field and far field. Not all the sound waves move away from the driver, thats why near field is different.
"Very close to the source, the sound energy circulates back and forth with the vibrating surface of the source, never escaping or propagating away."

 

[jackocleebrown]

Or you could test it! It's one thing to "go with what these guys say." It's another thing to understand exactly what they mean. Yes, I can't argue - the acceleration has to happen to get the cone moving in a way that creates sound. But maybe they can pipe in and tell us if they really believe that peak pressure at the cone's surface always occurs when the cone is at its maximum acceleration. If not always, then under what conditions? I'm convinced that if the cone is moving into a very small space that it can fully pressurize during the wave cycle, then the peak pressure does indeed occur at maximum excursion, which is when the driver is rapidly changing direction, so that should also be peak acceleration. It's important to note that this is just a coincidence. It's not the acceleration that's producing peak pressure in that case. It's the fact that the cone has moved in to the small space and made it smaller, so the air gets squeezed. If the cone were to just stop and sit there, the pressure would remain high. Acceleration itself does not create pressure in this case. Acceleration can cause pressure in some cases, like the pressure of your back against the seat of a car. In that case, if the acceleration is maintained, the pressure is maintained. If the acceleration ceases, the pressure also ceases even though the car is still going very fast.


If the driver is playing into free space, then the sound waves move away from the driver much too fast for the pressure to remain high at the cone's surface at peak excursion. When sound can propagate it moves at the speed of sound. The cone doesn't move anywhere near that fast. By the time it comes to a stop and starts moving in the opposite direction, the peak pressure pulse it made while moving forward has moved outward

Shameless plug for this thread which answers some of the questions you pose

How does a loudspeaker cone generate acoustical pressure?

Introduction This post is inspired the thread "Which way is up", started by @René - Acculution.com, and especially by a comment from @BeerBear. In "Which way is up" René shows that when replaying a sine wave, the maximum pressure coincides with the point when the loudspeaker cone is at maximum...

www.audiosciencereview.com

 

[Tim Link]

Shameless plug for this thread which answers some of the questions you pose

How does a loudspeaker cone generate acoustical pressure?

Introduction This post is inspired the thread "Which way is up", started by @René - Acculution.com, and especially by a comment from @BeerBear. In "Which way is up" René shows that when replaying a sine wave, the maximum pressure coincides with the point when the loudspeaker cone is at maximum...

www.audiosciencereview.com

I was just looking at that. Thanks for making those animations!

• When a speaker drives an open space, the acoustic pressure in the space is proportional to the cone acceleration.

This makes sense. Sound level in the space that the driver is playing into is directly proportional to acceleration of the cone. The math works out. But does this answer the OP's question about what's happening with the sound pressure around the driver when it's in various positions and moving at various speeds as it goes through it's cycle? It seems we're not all talking about quite the same thing, or I may have misinterpreted the OP's intent with this discussion. I thought he was talking about how the phase of the generated sound relates to the motion and position of the cone. Does peak pressure generation correspond to the point when the driver is moving outward fastest, or when it's accelerating fastest? Or some other point in its motion? For purposes of our listening enjoyment I don't think it's super important because I don't think we're super sensitive to absolute phase. But for those who do believe in maintaining absolute phase relationships, it becomes an interesting question.

"Very close to the source, the sound energy circulates back and forth with the vibrating surface of the source, never escaping or propagating away."

I think that's a very valid statement for the conditions they were talking about but I don't think it helps us to understand how the speaker cone's position relates to the phase relationships of the generated sound.


I've been thinking about my proposed test, which I'm trying to figure out how to carry out properly. If I make a 100hz sine tone and put a little 2000Hz blip at each peak of the 100Hz wave and play that back through a single driver speaker and record it with a microphone, should I expect those blips to still show up at the peaks of the 100Hz tone of the recorded result, or will they be phase shifted to a different position? Will it matter how close I am to the driver when I measure? My expectation is that the 2000Hz blips will appear closer to the zero crossing point of the 100hz wave, and it won't matter how close or far away I am when I measure so long as early reflections don't screw up the phase. Does this make sense? Does anybody have different predictions, or advice on how to properly do a test like this, or why it might be completely senseless?

 

[René - Acculution.com]

The questions have already been answered in the thread, but okay, one more time:

Mass-like acoustic impedance->Cone acceleration phase is in-phase with pressure phase. (Free-field conditions are not exactly mass-like, but close enough for what I am illustrating here.)

Resistance-like acoustic impedance->Cone velocity in-phase with pressure. (Could be an infinite tube, as already mentioned in the thread)

Capacitance-like acoustic impedance->Cone displacement in-phase with pressure. (Check out my Room Gain post for an example)

The acoustic environment could be any combination of these, just as an electrical impedance can be involved/complicated (and complex mathematically also), but we can calculate the resulting pressure, when knowing the source.

We always do these analyses assuming steady-state! That is basic signal-processing. If you want to look at the relationship between variables under transient conditions, you have to know the input signal, the system, and the initial conditions and solve analytically, if the system is simple enough, or numerically. That is why it makes little sense to think about this in a transient sense, before understanding steady-state phasor assumptions, and how they tie in with Fourier and Laplace. Perhaps my article here helps https://audioxpress.com/article/simulation-techniques-misconceptions-in-the-audio-industry.

Applying a voltage phasor with zero phase will show that the cone displacement phase tracks differently across the frequency range, and shifts around the resonance. I believe that that is what leads to confusion for so many people. Having seen the cone move in-phase with voltage at very low frequencies, or move outwards with a DC battery, where it is stiffness-controlled, means that in its operating range where it is mass-controlled, it will move inwards for the same voltage phase. And I have shown in this thread why we actually want that, based on how the acceleration is in-phase with pressure, not displacement as many think (under mass-like conditions, as already explained). If it were opposite, we would need to change our definition of polarity, and make sure that the cone moves inwards, when applying a battery, so positive to black terminal instead. Remember, all transducers have an underlying transductance principle, and for the electrodynamic driver (Lorentz force), this is just how it is, and it has been known for decades. Other transducers operate differently, and when questions have been raises about how this does not fit with how some microphone or other device works, I have also answered those questions.

There is not much more to say. Read what I write. Read what Lars Risbo writes. Read what Jack Oclee Brown writes. Everything fits together.

 

[Tim Link]

The questions have already been answered in the thread, but okay, one more time:

Mass-like acoustic impedance->Cone acceleration phase is in-phase with pressure phase. (Free-field conditions are not exactly mass-like, but close enough for what I am illustrating here.)

Resistance-like acoustic impedance->Cone velocity in-phase with pressure. (Could be an infinite tube, as already mentioned in the thread)

Capacitance-like acoustic impedance->Cone displacement in-phase with pressure. (Check out my Room Gain post for an example)

The acoustic environment could be any combination of these, just as an electrical impedance can be involved/complicated (and complex mathematically also), but we can calculate the resulting pressure, when knowing the source.

We always do these analyses assuming steady-state! That is basic signal-processing. If you want to look at the relationship between variables under transient conditions, you have to know the input signal, the system, and the initial conditions and solve analytically, if the system is simple enough, or numerically. That is why it makes little sense to think about this in a transient sense, before understanding steady-state phasor assumptions, and how they tie in with Fourier and Laplace. Perhaps my article here helps https://audioxpress.com/article/simulation-techniques-misconceptions-in-the-audio-industry.

Applying a voltage phasor with zero phase will show that the cone displacement phase tracks differently across the frequency range, and shifts around the resonance. I believe that that is what leads to confusion for so many people. Having seen the cone move in-phase with voltage at very low frequencies, or move outwards with a DC battery, where it is stiffness-controlled, means that in its operating range where it is mass-controlled, it will move inwards for the same voltage phase. And I have shown in this thread why we actually want that, based on how the acceleration is in-phase with pressure, not displacement as many think (under mass-like conditions, as already explained). If it were opposite, we would need to change our definition of polarity, and make sure that the cone moves inwards, when applying a battery, so positive to black terminal instead. Remember, all transducers have an underlying transductance principle, and for the electrodynamic driver (Lorentz force), this is just how it is, and it has been known for decades. Other transducers operate differently, and when questions have been raises about how this does not fit with how some microphone or other device works, I have also answered those questions.

There is not much more to say. Read what I write. Read what Lars Risbo writes. Read what Jack Oclee Brown writes. Everything fits together.

Great answer. I'm interested in finding a way to demonstrate some of these principles with experiments and real measured data.

 

[Tim Link]

So I just did some experiments. I created low frequency tones using Audacity and put higher frequency blips on the peaks. I tried these at various upper and lower frequencies, and I measured from various distances to see how well the wave form shape was preserved. Guess what I saw? I'll give you a few hints.

1. I was surprised!
2. There was some variation with frequencies chosen, as one would expect with the phase response of my woofer not being perfeclty flat, but overall the results were basically similar, the blips appearing a little ahead or a little behind the same location on the low frequency waveform.
3. The measuring distance didn't have noticeable effect other than volume and relative noise (waveform looked less clean further away) - distances from a few inches out to a few feet from a 10" midwoofer about 2 feet above the floor. Woofer is on open baffle and crossover disabled. Frequencies were all well within the bandpass of the woofer on the baffle, which is about 220Hz to 5000Hz. This woofer: https://www.parts-express.com/pedocs/specs/294-1201-faitalpro-10fe200-8-specifications.pdf on 2' x 4' baffle. It measures very similar to what the manufacturer's graph shows.

So where do you think those upper frequency blips showed up in relation to the lower frequency wave form?

BTW, these waveforms were somewhat annoying to listen to, and brought housemates in to the room to see what the heck I was doing!

Ok, there's no reason for me to make people guess. The wave forms were very well preserved, with the blips being where they should be, at the peaks. This means that despit all of your efforts so far to explain all of this, it's not coming together yet in my mind, but I'm very happy that it creates the results that it does. So if the driver's motion tracks the electrical signal, and the resulting measured acoustic signal from the microphone again transduces back to something very close to the original electrical signal, then somehow it just works and I'm not going to try to keep visualizing how that can be. I'll just be happy it is. I can't even tell if this result is what you've all been trying to tell me I should expect from the math and simulations. Is it? My understanding is that peak acceleration occurs at max excursion (althought the acceleration is in the opposite direction of the excursion), and if that's true then what you've all been trying to tell me is correct, and that's great, because phase relationships are preserved. This has some implications I suppose for direct radiators vs horns in terms of phase response. The direct radiators playing in to free space should be better at this, although other factors affecting phase I'm sure come into play. And, how important this is in terms of sound quality perception is another issue. And, it has been pointed out that these calculations only apply to steady state conditions, not necessarily transient responses, and the size of the driver relative to the wave length matters. My low frequency waves were steady state, and I tried to make the little blips look like smooth little relatively low level wavelets to minimize transient effets. The 10" driver on the open baffle seems to do what it should do in the range I'm using it in.

edit: The driver's position does NOT track the electrical signal! The acceleration tracks the electrical signal under these conditions, which puts the position completely backwards. That's what I wasn't picking up on. I hope that's right because that makes my experiment make perfect sense.

 

[Tim Link]

Shameless plug for this thread which answers some of the questions you pose

How does a loudspeaker cone generate acoustical pressure?

Introduction This post is inspired the thread "Which way is up", started by @René - Acculution.com, and especially by a comment from @BeerBear. In "Which way is up" René shows that when replaying a sine wave, the maximum pressure coincides with the point when the loudspeaker cone is at maximum...

www.audiosciencereview.com

I think you really helped to get me straightened out on this topic with that thread. I think everything is making sense at this point, including my earlier experiments with the microphone tick. I think that had to do with being close to the port frequency of the speaker and possibly from being in a small space. Results are hard to predict under those conditions so I'm not too concerned about explaining that now. Thank you everybody for your patience and efforts. This has been a great thread and I've definitely learned a lot! Still a lot to learn, but I'm better clued in now, unless it's not true that the driver's excursion position is reversed from the electrical signal in steady state when mass is dominating the motion and not air stiffness or suspension. If that's not true, then I'm still confused.

 

[KSTR]

Ok, there's no reason for me to make people guess. The wave forms were very well preserved, with the blips being where they should be, at the peaks. This means that despite all of your efforts so far to explain all of this, it's not coming together yet in my mind, but I'm very happy that it creates the results that it does.

Consider using the blips to trigger a LED stroboscope and then see that the cone is at the most inward position when those blips are located at the positive peaks of the low frequency tone.

 

[Tim Link]

Consider using the blips to trigger a LED stroboscope and then see that the cone is at the most inward position when those blips are located at the positive peaks of the low frequency tone.

That would be cool. I'm convinced that is correct now, but it'd still be cool to see it. You can also demonstrate with a rubber band and a heavy object hung from it. Start tugging up and down and it ends up moving the opposite direction form the force you're applying. Try a lighter object and it just moves in sync. with your hand. So for a direct radiator intended to play into free space we need to make sure the mass dominates. If the mass gets lower, the suspension must get softer, and the back chamber must be adequate in size. But if we're making a compression driver, the stiffness of the suspension and the air in the compression chamber and the back chamber should dominate, it would seem, or at least be more dominant because we want the pressure peak to match the signal peaks and for a compression driver the position of the cone correlates with the pressure, more or less. It really is kind of a miracle that the mass domination gets the driver moving backwards, but then an entirely different phenomenon involving how soundwaves form from a moving membrane in free space gets the pressure reversed again so the sound waves are correct.

 

[Tim Link]

Some people have dreamed of the ideal radiator as a sphere that is evenly expanding and contracting all over its surface. Would this still act like a piston driver in free field? Let's say it's a half sphere mounted on a flat baffle compared to a conventional dome midrange of the same size, both producing waves much larger than their diameter. My guess is that it would have the same phase response as the regular dome. Its main advantage would be at much higher frequencies, where it would maintain dispersion despite the waves becoming small compared to its diameter. However there would still be a phase shift issue as the waves become small compared to the radiating surface size. So maybe that magic expanding and contracting sphere wouldn't be so perfect after all.

Actually I'm starting to wonder because the piston never directly pushes the air inward or outward to and from the sides. Everything is directly pushed forward and back, while the inward and outward sideways motion is a delayed after effect. With the expanding sphere all directions are pushed and pulled evenly at the same time. Might be very different, more like a plane wave tube, which means the sphere might be more efficient, with none of the sloshing effects.