This is similar to some patents I have seen before.
My idea of having the thin film on the front of rigid panels is also similar, but I use mine to augment the sound from the front of the panel, mainly in the mid and higher frequencies.
How this affects the dispersion pattern of the dml panel, I have no idea.
Using a thin film on the back should hopefully prevent the usual boxy sound when enclosing the back of any speaker.
But can the thin material be properly controlled at higher volumes ?
Lower to mid frequencies can cause major problems on such surfaces .
I have had a little experience with this problem.
Even the canvas panel can suffer from these problems.
Steve.
I am also curious to know what an exciter driving the panel on a small area and not a circle of quite large dimension compare to the wave length could produce.
Christian
Hello Lekhe,
I think this exchange will not bring new inputs to improve our DML; nevertheless let me add some precisions.
About the current for sure it changes end even to some tens thousands. It is simply the frequency of the audio signal usualy said from 20 to 20kHz
Is it a question? If yes the answer is in the propagation of the sound in the air. Think the air made of very small adjacent cubes. When one face of one cube is pushed, the 5 others expand, pushing then their neighbors.Of course it is over simplified but it helps to understand why there is sound in area which are in the shade of obstacle, how the rear and front wave of an open baffle interfere.
Only DML or other driver like the Walsh or the Manger speaker are fully based on bending wave. For a standard cone, the main movement is called pistonic, the displacement is the same for all the points. It is the reason also such speaker is said mass loaded because the voice coil sees on almost all the bandwidth the cone as a mass. This defines the bandwidth of the loudspeaker in combination with element like the stiffness (spider or enclosure air) and the voice coil inductance. In a DML, the voice coil see more the panel as a resistance which allows despite the apparent important mass of the panel to go high in frequency.
The voice coil doesn't act mainly as a surface but as a force generator that moves the membrane that is the component with enough surface to induce the sound in the air (see the analogy to the string instrument)
Vibration is a common word to describe some rapid oscillation (wiki : a mechanical phenomenon whereby oscillations occur about an equilibrium point). It is not a physical quantity with a unit to measure it, some laws in relation with other quantities. What is demonstrate here it is the force generated by the coil is able to move the surface that will produce the sound. The mass above creates an inertia so that most of the movement appears on the table
Yes and no... Yes about the difference between rooms, yes the same speaker can sound differently depending of the room. It is exactly the reason of the work of M Toole and others that tried to find criteria for measurements in anechoic condition to make a loudspeaker that will work on most of the rooms. Yes and no about the in room measurements. It is a matter of observation time. If the observation time is long like with a measurement with a pink noise, the room is measured. If a tool like REW is used and only the few first ms of the impulse is considered, the room is not seen. The price to pay is this method is limited in its lower frequency. It is not really difficult to be able to reject the floor bounce by some absorber, the other reflections being around 5 to 6ms.
Christian
The patent is not detailed with measurements or figures but what I understand is one face is the DML, the opposite is a kind of membrane. In between there is an absorber (typically glass wool or similar). The rear wave is no more correlated to the front one so there is no more the effect of the open back. In the bass, the membrane acting like a passive radiator reinforce the bass. The system is adjusted by the volume of air in the enclosure (more or less the depth as it is the DML surface) and the frequency of resonance of the membrane (probably the first mode).
For a DIYer it could be a DML in front with the exciter, a canvas in the rear as a passive radiator, some absorbent in between. Not to difficult to prototype... to have it working may be more tricky! We could even imagine an external access to adjust the canvas tension.
I wonder what is the resulting directivity diagram. As the rear wave is an attenuated and filtered image of the front one, it will be more a kind of omni changing to an hemisphere as the frequency increases.
Below is a picture of the ODA speaker from my notes (was from their site I think)
The dimension is 31.5x31;5x3.5cm so not that much space for the absorber.
Christian
I hadn't seen that patent and will have another read later.
I have found my Dml panels at home sound a lot better and more defined with the rear sound attenuated by an absorbent backing pad.
Wondering if anyone has tried making a baffle around their Dml to limit some of the bass cancelation as in open baffle speakers?
In relation to enclosing the back in a box (I know goes against the idea of Dml openness) a thought I had was if you were to make a panel suspended all around by a membrane as in the patent above, or a canvas type suspension, could it then be measured with an impedance rig and REW to find T/S parameters and create a box to suit?
Sorry if I am missing a discussion already had at some point, I have just made a measurement rig to use an audio interface to measure some other drivers I have so am just learning about this.
I have found my Dml panels at home sound a lot better and more defined with the rear sound attenuated by an absorbent backing pad.
Wondering if anyone has tried making a baffle around their Dml to limit some of the bass cancelation as in open baffle speakers?
In relation to enclosing the back in a box (I know goes against the idea of Dml openness) a thought I had was if you were to make a panel suspended all around by a membrane as in the patent above, or a canvas type suspension, could it then be measured with an impedance rig and REW to find T/S parameters and create a box to suit?
Sorry if I am missing a discussion already had at some point, I have just made a measurement rig to use an audio interface to measure some other drivers I have so am just learning about this.
Any sound readings or measurements taken in a highly cluttered room cannot be regarded as accurate or reliable. The presence of numerous objects can significantly affect the acoustics, leading to distorted results that do not truly represent the sound quality or characteristics of the audio being measured.
Regarding the matter of a voice coil positioned vertically on a table, you can indeed place the aforementioned Aiyima exciter in that manner, albeit without the "shock rod," despite claims that it is not "pressure heavy." It would then function as a voice coil with the magnet system remaining intact. This film would demonstrate how the exciter, with a blocked "voice coil," attempts to jump about.
The magnet section of a conventional speaker will also tend to move about, which is why a basket is employed to hold it in place. This basket is somewhat isolated from the enclosure by a type of absorber gasket. However, it still manages to transfer vibrations to the box. The rear side of the cone, in turn, contributes to producing more sound. Yet, no one is likely to accept this alone, even though every speaker manufacturer strives to harness that energy, resulting in various internal designs such as bass reflex, transmission lines, folded horns, and so on—all hidden at the back. Even open-baffle enthusiasts attempt to keep the rear of their speakers far away from the back wall, thereby denying themselves the potential for better sound.
When it comes to the matter of pistonic motion, explaining the more refined, volumetric rear wave can be quite challenging. There are numerous explanations for that rear wave, subtly moving away from the concept of pistonic motion, making it difficult to clarify how the cone's converse side pushes air and make sound. The rear side of the cone would simply move through the air... like a spear.
In DMLs, you may have noticed that the sound from the back, where the exciter is mounted, produces a better quality of sound!
Hello Joska,
What is an "absorbent baking pad"?
I am making tests around attenuating the rear wave of a DML and the main effect is to avoid it interferes with the front wave. The main effect is not in low frequency (depend on what you add to the rear) but in the 1 to 5kHz where an open baffle has lobes.
Note also that figure of height of a dipole is valid only up to a frequency which quite low. It is why we see open baffle design in multi way with baffle close to the drivers or even no baffle, just the drivers. Above, the behavio of an open baffle is a bit messy, even on axis (dips).
I am no more in the idea that an open back for a DML is a good thing. There are to many constructive (lobes) or destructive (dip, even on axis) due to the rear. The DML seems not completely immune to the problem of to wide open baffle.
All the DML give the same kind of impedance curve independently of its type of suspension. You have examples in Eric's posts. It is a suite of several resonances (peak of impedance, the value in frequency changes with the type of suspension), not only one as for a classical loudspeaker. It will need more investigation in my opinion to see if REW can extract something and if it has a meaning. The T/S parameters are related to a model (equivalent circuit) of the pistonic loudspeaker with a relatively limited number of components (Re, Le for the electrical circuit, an RLC net work for the mechanico-acoustic circuit).
Not true. Close mic measurements are virtually unaffected by the room. Likewise, impedance tests. Frankly, tests performed in any room can be accurate and reliable representations of the conditions under which they are measured. True, in order to compare speakers it would be ideal to make all measurements in an infinite space with no walls, ceilings or floors. Or at least in a well designed anechoic chamber. But to suggest that any measurement made in anything less has no value is simply wrong. You just have to understand the limitations of your test environment and interpret any test results accordingly.
Eric
lekha,
You have used the term "blocked voice coil" several times, but I don't know what it means. Can you explain? I have never heard/read that phrase before.
Eric
Christian,
That is correct, of course, But like JoskaNZ I have wondered if there are not similar parameters that could characterize a DML, that includes the exciter, panel and suspension, particularly in the fundamental mode. Is the fundamental mode of a DML really much different from a pistonic speaker? And couldn't those parameters be used to design a bass reflex DML? Perhaps it would be just as easy to build a box/vent with a Helmoltz resonance at the DML panel's fundamental, and then just modify (lengthen) the vent/port until the bass response is optimized. What do you think?
Eric
In a DML, the rear wave is, in fact, the front wave, as it originates from the surface to which the exciter is attached where it produces a better quality of sound. However, suspending the exciter in front would appear somewhat unsightly.
When the panel is flat and thin, both sides will exhibit bending waves, with the wave at the rear commencing slightly before that at the front. Given the panel's thinness, and by positioning it as far away from the back wall as possible, you may achieve some undistorted sound from the front.
When considering a standard cone speaker, it is difficult to envision how the rear of the cone pushes air. The air would simply slide off the back of the cone. However, most high-end manufacturers believe it is essential to harness the sound emanating from the back, which is often superior to that from the front. They go to great lengths to achieve this and direct that quality sound towards the listener. Here is a cutout of a rather expensive high-end speaker made in Poland.
Therefore, with a DML, the objective is to capture sound from both sides. Since one cannot be on both sides simultaneously, the challenge is to find a way to bring the sound from the back to the front, if possible. This is quite difficult to achieve. No partner/wife would permit panels hanging from the ceiling in the living room, or some contraption sitting in the middle of it. As a result, most DMLs are typically kept in basements, garages, workshops, cluttered junk rooms and similar spaces.
When one discovers what truly generates the "rear wave" of a conventional speaker, one will learn how to harness the "rear wave" of the DML. A cone's backside cannot push air; it can only move through the air, or the air would merely slide over its surface. Perhaps bending waves occur on its surface.
When the panel is flat and thin, both sides will exhibit bending waves, with the wave at the rear commencing slightly before that at the front. Given the panel's thinness, and by positioning it as far away from the back wall as possible, you may achieve some undistorted sound from the front.
When considering a standard cone speaker, it is difficult to envision how the rear of the cone pushes air. The air would simply slide off the back of the cone. However, most high-end manufacturers believe it is essential to harness the sound emanating from the back, which is often superior to that from the front. They go to great lengths to achieve this and direct that quality sound towards the listener. Here is a cutout of a rather expensive high-end speaker made in Poland.
Therefore, with a DML, the objective is to capture sound from both sides. Since one cannot be on both sides simultaneously, the challenge is to find a way to bring the sound from the back to the front, if possible. This is quite difficult to achieve. No partner/wife would permit panels hanging from the ceiling in the living room, or some contraption sitting in the middle of it. As a result, most DMLs are typically kept in basements, garages, workshops, cluttered junk rooms and similar spaces.
When one discovers what truly generates the "rear wave" of a conventional speaker, one will learn how to harness the "rear wave" of the DML. A cone's backside cannot push air; it can only move through the air, or the air would merely slide over its surface. Perhaps bending waves occur on its surface.
Last edited:
HvdZ,
It sure does count! Not exactly what I was imagining, but along the same lines. I only vaguely recall the ODA speaker, and don't recall ever seeing that patent so thanks for sharing. That patent is actually a remarkably easy read, as patents go. Normally, US patents include some specific examples ("embodiments") that this patent lacks, sadly.
Eric
Thanks for your reply, yes makes sense with the many peaks in impedance not being as simple and the formula not working for the T/S parameters.
Here is what I have for my home panel speaker (attached images) it's a frame with 35mm approx of glass wool insulation and fabric covering the whole back area of the panel, from memory it lowers the mid/high frequency output of the back approximately 10db on my rough measurements. Definitely helps in my room as with an open back they lacked imaging and were too washed out with so much reflected rear sound.
Hope this helps someone, interesting idea with the rear passive radiator type closed backing mentioned above too.
No it isn't hard to envision. Frankly, it never even occurred to me that air would "slide" off the back of the cone. For all practical purposes, I suspect that the front and back of a cone drive air equally well, or nearly so. As far as I understand it, the purpose of the cone shape is not because it is better at pushing air in front, but rather because the curvature stiffens the surface, and prevents it from having bending waves, or at least pushes their occurrence to very high frequencies.
Plenty of conventional pistonic speakers have been made with perfectly flat "cones", and they work just as well as cones, as long as they have other means of stiffening.
I think most everyone participating on this forum has read enough to understand that the back wave is important for conventional pistonic speakers, and that many "manufacturers believe it is essential to harness the sound emanating from the back". Nobody needs to be convinced of that.
That's just wrong. A cone's backside can push air, and if it couldn't, there wouldn't be any reason to "harness" it, because there wouldn't be anything to harness.
Hopefully not, at least not over most of the cone speaker's intended frequency range. But yes, bending waves will occur on it's back surface, at sufficiently high frequencies. And at exactly the same frequencies at which they occur on the cone's front surface. Because, for all practical purposes, the front and back of a cone (or a DML panel) move at the same time and in the same direction.
Eric
Christian,
I've tried a few times to see if enclosing the back, with or without damping, had any positive effect. I never saw one. But it's very possible that I didn't know what exactly to look/listen for! I look forward to seeing what you find. I never detected coincidence effects before either, but now I do. So I won't be surprised if you are correct about this too.
Eric
Hi Eric,
Did you also try an unclosed but damping/absorbent backing to lessen the back of panel output?
Did you also try an unclosed but damping/absorbent backing to lessen the back of panel output?
Have another look at this video. The cheerful Brit demonstrates how a blocked (from movement) voice coil transfers sound to a table. It can transmit sound to any object, whether it's another EPS sheet, a kettle, an Ikea bowl, a door, or a window pane—none of which have any suspension to support them or even move to disperse the air around them. It is simply the unseen surface wave.
It is the wire connected to a sound amplifier within a magnetic field that produces sound. The wire itself has a tiny surface area, so the air it disperses won't reach the human ear, which is why it needs to be affixed to a surface. As a single thread, the wire would be very long, and the number of magnets would be quite high (as seen in Magnepan). To make it more compact, the wire is wound around a core, allowing for the use of significantly fewer magnets. The core can be round, square, rectangular, elliptical, etc. All that is required is a wound coil—either three-dimensional or flattened—a magnet system, and a surface for the coil to be attached to. No pistonic motion is necessary. The entire DML concept is based on this principle—no pistonic motion.
Nice!
You have effectively blocked the "rear wave," allowing it to be positioned next to a wall. 👍
Hello Eric,
Good question. The DML at its 1st mode is not different from a cone loudspeaker. The difference being that the mass and the stiffness are not localized as in a cone speaker. So it might be possible to derive its parameters and see where it leads.
Christian
Hello Joska
Thank you for the pictures and the evaluation of the attenuation.
Christian