Is it because of the direction of the antiphase waves oppose the in-phase and hence cancel out or reduce the velocity of the positive waves depending on the relative magnitude of +ve / -ve vibration??
Eucy
Hello Bruce,
I have also this question in mind. For now, this paper hasn't found for me its place in the full picture... Maybe I should read it again after the directivity tests to search evidence of the coincidence frequency and the tests with a concentrator. What lead me with this part are 2 problems pointed here a long time ago, confirmed by the directivity tests which are the cavity noise and the drum effect. By the way, in a patent, the drum effect is associated to a peak as we see it here usually but also a dip a bit lower in frequency.
The simulations from Dave show a relation between the high frequency we can reach and the wave length of the waves in the coil area (larger the interface, lower the HF limit).
In parallel to that I made tests showing a sensitivity to additional mass at the voice coil. I have no clear picture for that except that the 1.7g concentrator that I tested first on an EPS panel kills its HF... It is from that I decided to try something as light as possible having in mind possible problem of stiffness or of weak contact with the voice coil... Tests to come. No idea of the results.
Christian
+ @lekha
It is the answer I had in mind too but I found it too obvious... The knowledge of how an electromagnetic motor works being a kind of prerequisite, something out of the scope of the thread. It is why I asked to have more precision about the question
Hello Eric,
I understand. I even better understand that it is thanks to exchanges with Dave that the decomposition in modes having each their own frequency response makes sense to me.
Seeing the FR of different panels, I am also o the idea that there are maybe to many modes contributing in some range to the final FR.
This is something that Dave simulation should show as the mode decomposition is the basis of the tool
In addition to that, I came to an other possibility... Sometimes I feel a bit alone seeing our panels as for sure DML but in addition as dipole with the effect of the combination of the front and rear waves. So in parallel to DML, I try to have an eye on what happens on open baffle. Among the sources there are the papers from Martin J King with the plot below. What is interesting here is that it shows the typical response of a dipole (simplified here) with the LF roll of at 6dB and what is called the dipole peak (the dashed blue curve). This is baffle dimensions dependent. All the dips that come after are not shown. The point is the plot show the FR of the loudspeaker needed on an infinite baffle to get a flat over all FR (the dashed red line). Don't consider the low pass, the example of the paer is a woofer. The infinite baffle response is the output of Dave's simulation.
To come to your test, an hypothesis is you have created a "dip" in the dipole peak area to smooth the global FR.
Christian
Hello Eucy,
Dense post!
About the mass, as I said in other posts, not clear for me. What the papers about DML theory tell is the exciter doesn't see the panel as a mass but as a resistance. This is expected when the modes are dense enough. When the modes are isolated, it is a part of the mass that it is seen. In HF, the possible effect of the moving mass of the exciter is then not acting in parallel to an other mass but to a resistance.. This is the theoretical schematic I have in mind trying to find some practical evidence.
My representation of what happens is a bit different. I started from the idea that the central area of the voice coil has its own behavior (drum effect) that is not needed (in pure theory or assumption) for the DML to work. It is like a badly adjusted whizzer cone. The idea of the concentrator is to drive the panel on an area smaller than the shortest wave length or at least small enough. Is the pin needed is maybe an other question. It is at least useful to make the part (to handle it to glue the ribs). On the part glue on a plywood panel, I have removed it.
About a ring, for now I don't now... I have a tendency to think the dimension in number of wave length and with the rule of thumb saying every thing that is smaller that a quarter wave length doesn't exist for the wave...
All of that is a fresh topic. I had hesitations to share the concentrators as for now I have only few tests. Next tests is with a plywood panel already tested with the classical interface. Let's see what it will come.
Thank you for sharing your though.
Christian
Hi Christian
Sorry for being contentious. 😵💫
I think that when we get right down to the small dimensions around the voice coil, the panel thickness itself become an increasingly important factor.
We're trying to approximate the performance of a small extremely light tweeter at these high frequencies with a relatively thick lump of plastic or wood !
The answer lies in a laser interferometer.. I'm sure Dave will have examined this at some stage.
I look forward to the tests
Cheers
Eucy
Sorry for being contentious. 😵💫
I think that when we get right down to the small dimensions around the voice coil, the panel thickness itself become an increasingly important factor.
We're trying to approximate the performance of a small extremely light tweeter at these high frequencies with a relatively thick lump of plastic or wood !
The answer lies in a laser interferometer.. I'm sure Dave will have examined this at some stage.
I look forward to the tests
Cheers
Eucy
Last edited:
In hindsight, I should shelve my cockamamie ideas and take up knitting
Eucy
Eucy
Few chance you see me in a knitting forum! Continue like that on this one. No worry.
Christian
In one of the first patent about DML (EP0541646B1), there is a requirement about the panel thickness to be less that half the wave length in the air for the upper frequency. It was in the case of a sandwich material. I think we can extend to any material has the coincidence frequency will be probably targeted in the same range. So roughly at 20k, the wave length is 17mm in the air. So material of 2 to maybe 5mm might be considered thin, a 20mm EPS thick so with probably some difference in its behavior. In between, in the 10mm is a grey zone.
Loud thinking!
The panel can't be think as thin (or without thickness) in all the cases.
Christian
No. More simply the contact in the movement will stop the iron membrane leading to clipping and distortion.
Why to invoke (again?) a sound energy when it is a magnetic field and the according force on the iron which are at play? In all those example there is a force from a known principle that acts on a surface that moves the air, even if the surface is not a dedicated part (membrane).
Soviet for sure but an additional headphone of a French phone of the 30's on the same principle, yes. It is maybe still somewhere in my stuff.
Hi Eric -
Great question - acoustic loading is a specific "textbook" term for an acoustic impedance proportional to the volume of air being displaced (this would assume a baffled configuration). In the simpler case where we have an enclosure behind the panel, acoustic loading is proportional to the volume of air being compressed. In reality, because it's volume velocity, it's proportional to both how much air is being displaced and how quickly it's being displaced.
I drew a little picture below showing that the volume of air inside the enclosure shrinks when a 1st order mode compresses it, but a 2nd order mode actually doesn't compress the air because the volume in the enclosure doesn't change. This is an ideal approximation that assumes the air can move quickly enough from the +ve peak are to the -ve peak area, and in the real world things are never quite this simple. It's definitely interesting to see how much the (1,1) mode of a panel is affected by an enclosure, but the (2,1), (1,2), (2,2) modes, etc aren't hardly affected at all!
Acoustic impedance is primarily real at low frequencies, or where the wavelength on the panel is much shorter than the wavelength in air. A real impedance like this acts as a "drag", more or less, which acts to decrease the Q of the mode but doesn't affect its resonance frequency. At high frequencies the complex impedance is no longer acting like a drag, but more of a mass, but by that point there's hardly any volume being displaced anyway.
Dave
When I said I was guessing, I meant that I was guessing as to what your actual question was. If you were asking why the coils in the videos you posted were emitting sound, then the answer I gave was correct, I'm pretty certain. Hopefully, if I am wrong, others will correct me.
If you were asking a different question, then I guessed wrong.
Eric
Sure, it would be interesting if the electrical engineer for whom the question was posed would do that. 😉
Lenz's law. We do this experiment with younger kids who come to visit our department - wind a coil of magnet wire around your finger, remove it from your finger, connect it with alligator clips to a music source, place it over a magnet, and listen to the coil make sound. Doesn't matter that the coil is flattened.
More or less, but with a somewhat homegrown twist to it. 🙂
It's more about the gap between the magnets and the gap in the flattened coil.
My prefered answer would have been : Laplace's force.
And now that you have had 3 equivalent answers and the supposed trap is discovered :
"Je donne ma langue au chat" (Is there here a direct translation : I give my tongue to the cat = I give up?)
"S'il vous plaît arrêtez de tourner autout du pot!" (seems to be in English : beating about/around the bush)
By the way, I see what is a gap between magnets, but what is a gap in a coil?
What is surprising to me in those coils it is the number of turns (compare to what I have in mind for an exciter or a midrange). Low magnetic field as we can suppose so need of a long wire?
Hello Lekha,
In the video of your question, I see clearly 2 separated magnets and not what is in between. Is the gap in the coil, the central area of the voice coil? Its distance to the magnets?
In the last video linked, I suppose 3 magnets in vertical maybe linked by some horizontal plate. Correct?
Christian
Lekha, I'm afraid you're a conductor without an orchestra
Eucy
Eucy
Lekha, What are the realistic benefits of this approach over conventional exciters? Flatter response, higher efficiency, lower distortion? Wouldn't cooling of the coil be a significant issue?