A Study of DMLs as a Full Range Speaker

The patent you posted is the same as I posted on post 1402.
I've lost a few panels over the years to other dogs(RIP)bull in a China shop is a good description,luckily the panels are not expensive and sometimes free,from skips ,I've been very tempted over the years to buy vh grade eps on line but it means buying a lot of 8ft panels,but as mostly nowadays I make small light thin panels it's not a priority.
Sometimes I spend a day going through all my panels listening to each one just to remind myself where I am,it always surprises me how good even the old ones sound ,am I going to be able to choose which to live with?
I'm thinking of having 3 sets of panels large-ish one medium and a small and switching between them or running them all at the same time ,depending on how I feel,the dream is to stop playing with them and just listen ,in a tidy un-cluttered room ,just music!!
Steve
 
  • Like
Reactions: 1 user
1. Long and thin ply panels, in this case 0.3 M by 1.2M by 3mm thick, seem to reach down to 30Hz
2. Long and thin also seems to provide better mode distribution - hat tip to Veleric for that one
3. Non-parallel sides may also provide benefits - hat tip to Spedge and it is claimed by Farad Azima in his 1997 patent
4. Edge treatment via absorbing elastomeric layer to absorb edges reflection and simulate an ideal infinite plate
5. Single exciter per panel appears to be preferable - validated by various sources
6. And finally, multiple panels to allow higher SPL at 30Hz without reaching exciter excursion limit.

Please throw rocks and suggestions as you see fit.

Burnt, well, no rocks, but I will share a few thoughts opinions!

#4: Damping is one area I'm really curious about. Surely some damping is needed but how much? And how do you know if you've got too much or too little? Wish I knew.

#5: I have also read fears that multiple exciters will interfere with each other. But I personally have looked at combinations multiple exciters (up to 3 at least) many, many times and have never seen any bad effect. Correction:the only time I've seen any bad effect is when I've wired them out of phase.
Now, that's not to say that using multiple exciters does anything magical either. But I have found that certain pairs tend to complement each other well. I've mentioned before that my favorite combination so far is the DA Ultra and the Tectonic TEAX32C30. The Ultra extends the low frequency and the TEAX extends the high end (for me at least).

#6 Multiple panels should certainly help. I do wonder if it would help or hurt imaging, but I don't know what to expect in that regard. My own hope is to find a better panel/size/mount combo that will give plenty of clean flat(ish) SPL with a single stereo pair.

Eric
 
Burnt, well, no rocks, but I will share a few thoughts opinions!

#4: Damping is one area I'm really curious about. Surely some damping is needed but how much? And how do you know if you've got too much or too little? Wish I knew.

#5: I have also read fears that multiple exciters will interfere with each other. But I personally have looked at combinations multiple exciters (up to 3 at least) many, many times and have never seen any bad effect. Correction:the only time I've seen any bad effect is when I've wired them out of phase.
Now, that's not to say that using multiple exciters does anything magical either. But I have found that certain pairs tend to complement each other well. I've mentioned before that my favorite combination so far is the DA Ultra and the Tectonic TEAX32C30. The Ultra extends the low frequency and the TEAX extends the high end (for me at least).

#6 Multiple panels should certainly help. I do wonder if it would help or hurt imaging, but I don't know what to expect in that regard. My own hope is to find a better panel/size/mount combo that will give plenty of clean flat(ish) SPL with a single stereo pair.

Eric

Thanks for the feedback Eric.

I am going to try a variation of the ABH method we discussed a few posts back. I am going to rout the edges to thin the section and fill that section with silicone filler. Thats the the thick black edge strip around each panel in the second image. It won't be a true ABH, I don't have a custom cutter, but I will get the cut as close as I can get. I will try one panel first and see how that works.

I will also see if the first panel needs additional damping and then replicate that if it works.

On exciters my comment was based on the long and ultimately circuitous journey taken on a couple of the other long threads by early experimenters. However, they were looking at well spaced exciters where you have your two very tightly packed together and thats another new arrangement I have not seen anywhere else. I can imagine that they are so close they act like one hybrid exciter so my comment doesn't really apply to what you do.

I am hoping/ assuming that the wide dispersion combined with the panels being in the same plane and very close together ( I am aiming for a gap no bigger than 5mm) that they act as one big panel. It will be interesting to see what happens.

Thanks again for the input.

Burnt
 
Hello group,
Has anyone try to measure the surface distortion of their speakers? I happen to catch a ghost hunting show where they used a laser grid to "capture" a sprit presents.

Hmmm skeptical that that might work for them but if one used the laser grid on a panel and using small circles cut from a space blanket A.K. a 1st surface aluminized mirror one could capture the speaker sprit.

One would place the circles where the laser grid appears on the panel. the laser shout be reflected off the mirror and projected on to a screen to see the Lissajous figure.
I believe one could even improve the panel measurement by modulating the laser so to have a stroboscopic effect on the panel similar to stopping a rotating fan blade. This method should allow one to monitor damping in real time as well has making a interesting laser light show.

Cheers
Steve
 
Veleric.
Is it the tin can effect of the primary radiation area that you do not understand?

Burntcoil.
I have a rather large bag of rocks,but don't intend to throw them at you,basically I don't do heavy panels ,large or small or arrays.
I think you already know what I think about damping ,why ,is usually my answer,what is the problem you are trying to fix? What are you willing to sacrifice to fix it?
Think I'll use the rocks to make a rock garden !
Steve
 
For Veleric

@Veleric.

Sorry I forgot to post these yesterday as promised.

White noise tests in the same room with the same amp and source using the Audio Tools App.

1) B&W CM1 reference speakers. Bookshelf bass reflex design
2) Tall panels with no EQ
3) Tall panels with EQ

Burnt
 

Attachments

  • B&W CM1 White Noise Test .jpeg
    B&W CM1 White Noise Test .jpeg
    247.5 KB · Views: 977
  • Blondie White Noise Blondie No EQ.jpeg
    Blondie White Noise Blondie No EQ.jpeg
    247.8 KB · Views: 979
  • Blondie Panel White Noise Test with EQ.jpeg
    Blondie Panel White Noise Test with EQ.jpeg
    247.1 KB · Views: 995
@Spedge,

Much appreciated Steve- notable lack of rock so far.

On damping I think it's important to differentiate between damping the panel and providing edge damping. As I understand it damping at the edges creates a condition where reflections are reduced in amplitude or, preferably stopped completely. This emulates an 'infinite plate' condition which is believed to be ideal. It's in a lot of the research papers on DMLs. I am aiming for this effect, not panel damping. I hope that makes sense.

Burnt
 
Last edited:
Burntcoil.
It makes no sense to me at all,there is no difference between the edge of the panel and the panel,they are the same.
Hold the panel firmly by the edges as you are playing music,and move them around to different positions,the sound will change considerably.
Reflections are very important for dml, in paper's on dml I have never seen any mention of how damping or any other manipulation of the panel improves the sound ,only it gets rid of this spike or dip or gives a smooth looking response.
I have been down this road a long time ago ,yes you can extend the frequency and make the response look good,but at the end of the day they just sounded like any other normal speaker,all I'd manage to do was suck the life out of it!!!
On the other hand some like that smooth sounding cosy sound ,the shops are full of speakers like that,and have been for many years, especially in the UK.

I'd better also say that touching the edge of a light panel such as eps with just your two fingers will have a noticeable difference in sound.
If a panel sounds good to start with ,you are on to a winner.

Is boosting the LF by what looks like about 10db below 50hz a good idea?
Steve
 
  • Thank You
Reactions: 1 user
As I understand it damping at the edges creates a condition where reflections are reduced in amplitude or, preferably stopped completely. This emulates an 'infinite plate' condition which is believed to be ideal.


Burnt,
I think you are correct that sufficient edge damping will reduce reflections and approach an infinite plate condition. But what I'm not so sure about is that emulating an infinite plate condition is preferable, or ideal.

I'm also not saying I think it isn't ideal, just not sure about either.

It's true that the infinite plate is mentioned in research papers. But my impression is that it's often discussed because it's a case for which the math is less complicated. So it is surely an idealized case, but not necessarily ideal.

Also, the foundation of all the NXT patents is modal behavior. The the ideal panel aspect ratios are based on optimizing the distribution of the modes and the recommended exciter locations are based on maximizing the number of modes excited. But there are only modes if there are reflections.

It makes sense that some damping is good, to flatten the resonances. And if it's at the edge, rather than "in the panel" that at least should allow the whole panel to radiate. But I'm not at all sure that the ideal is to stop reflections completely.

It'd sure be interesting to know what happens if you can, however.

Eric
 
Hi Spedge,

Have a look at the Gobel panel design on edge treatment. Their method is different , using slots with silicon filling the slots, but the objective is the same, to prevent the return of refections from the edge to emulate an infinite plate.

Try section 3.1 in this paper.

https://odr.chalmers.se/bitstream/20.500.12380/154618/1/154618.pdf

Its a brief comment on the method but helpful I hope.

Burnt
 
Burnt,
I think you are correct that sufficient edge damping will reduce reflections and approach an infinite plate condition. But what I'm not so sure about is that emulating an infinite plate condition is preferable, or ideal.

I'm also not saying I think it isn't ideal, just not sure about either.

It's true that the infinite plate is mentioned in research papers. But my impression is that it's often discussed because it's a case for which the math is less complicated. So it is surely an idealized case, but not necessarily ideal.

Also, the foundation of all the NXT patents is modal behavior. The the ideal panel aspect ratios are based on optimizing the distribution of the modes and the recommended exciter locations are based on maximizing the number of modes excited. But there are only modes if there are reflections.

It makes sense that some damping is good, to flatten the resonances. And if it's at the edge, rather than "in the panel" that at least should allow the whole panel to radiate. But I'm not at all sure that the ideal is to stop reflections completely.

It'd sure be interesting to know what happens if you can, however.

Eric

Thanks for the comments from you both.

I will merely stress that what I am attempting is an experiment to see what happens. I listen to all feedback which I value so please don't think I am ignoring anyone. I am combining a number of untried things with the array like the ABH method and the multi panel approach. Yes its probably stupid but I always like to push a design envelope until it breaks.

Sometimes you get to a new place.

Sometimes you just end up with matchsticks.

Whatever happens you will all get to hear.


Burnt
 
DML Patents

I said

We need a list of patents we can reference in one post

So I made a start.


I hope you find it useful


DML Patents

I have made this list as comprehensive as I can but inevitably there will be omissions.
I have tried to collate the patents that contain useful information relevant to understanding DML’s and can be implemented by us all.

If you find any you think should be included let me know.

ELECTRONIC RESEARCH ASSOCIATES Inc 1961 - Early DML like loudspeaker using flat strips

Abstract

This invention relates to a loudspeaker system and more particularly to an improved loudspeaker System of a type employing a flat extended, rigid vibratile pis ton driven by an electric voice coil situated in a mag netic flux gap, which flat vibratile structure is particu larly adapted for faithfully reproducing the lower fre quencies of the acoustical spectrum from a structure which is exceptionally flat in depth, employs no front to back interceptive baffling system, and which provides effective double sided radiation.


US3236958A - Loudspeaker system
- Google Patents




Bertagni 1968 - Flat diaphragm for sound transducers and method for manufacturing it
Abstract
A flat diaphragm for sound transducers and a method for producing it, including providing in a cellular platelike member a pretensioned front face and a rear face defining an irregularly shaped figure portion including a central stiffened portion at a zone out of the geometric center of said figure portion, and a marginal vibration damper portion substantially circumscribing said figure portion, said central stiffened portion defining a center for capturing vibrations. The flat diaphragm is preferably made of a granular expanded-bonded cellular, strong, stiff, imperforated plastics platelike member.

US3596733A - Flat diaphragm for sound transducers and method for manufacturing it
- Google Patents




Bertagni 1971 - Flat loudspeaker with enhanced low frequency
Abstract
Flat loudspeaker diaphragm with means for enhancing the low frequencies, wherein the diaphragm has a marginal vibration damping portion surrounding a sound producing figure portion and at least a portion of said marginal vibration damping portion is connected to a plate-like member adapted to vibrate only in the low frequency range.

US3767005A - Flat loudspeaker with enhanced low frequency
- Google Patents



Bertagni 1972 - Flat diaphragm loudspeaker with channels for suspension

Abstract
Flat diaphragm for sound transducers having a flat front face and rear face defining a central figure portion connected to an electromagnetic assembly and surrounded by a marginal vibration damping portion of larger thickness than the adjacent peripheral zone of said figure portion, an endless channel member in said diaphragm, surrounding said figure portion to better acoustically separate said figure portion from said marginal vibration damping portion and if certain tones are to be emphasized a substantially straight metal band resting on said figure portion arranged between said electromagnetic assembly and said endless channel member.

US3722617A - Flat diaphragm for sound transducers
- Google Patents



Bertagni 1972 - Electoacoustic transducers and electromagnetic assembly therefor


Abstract
An electroacoustic transducer comprising a diaphragm and an electromagnetic assembly in driving relationship with the diaphragm, the assembly comprising a magnet, a core defining a gap in the magnet, a coil supported by a coil holder movably arranged in the gap with an end portion facing the diaphragm, a cap on said end portion having a top surface directed toward, and in driving relationship with, the diaphragm, the surface being preferably secured to the diaphragm by means of an adhesive. The adhesive may be soft or hard and the diaphragm is preferably flat. The cap seals the coil holder and a perforation is provided connecting the space formed between the cap, coil holder, core and gap with another surrounding whereby a rectilinear to and fro movement pump is defined by the cap, coil holder and core, the pumping action moving air to cool the coil. The lower end of the coil holder has a flange to prevent the coil from being shockremoved from the coil holder, and the upper end of the holder has a flange which is adhered to the cap.

US3801943A - Electoacoustic transducers and electromagnetic assembly therefor
- Google Patents




Bertagni 1974 - Flat diaphragm loudspeake with asymmetric faces


Abstract
A planar diaphragm, preferably for loudspeakers where, on the rear face side there is an asymmetrically shaped sound producing portion, surrounded by marginal and channel shaped portions to assure that sound waves generated by said sound producing portion will not trespass said portion, and said marginal portion is provided plugs of different density than said diaphragm, to modify the vibratory characteristic of the sound producing portion and thereby improve the response curve

https://patents.google.com/patent/US4003449A/en


Bertagni 1978 - Planar diaphragm and supporting frame assembly

Abstract
A substantially planar diaphragm and supporting frame assembly, preferably for loudspeakers, where on the rear face side of said diaphragm there is an asymmetrically shaped sound producing portion defining the central portion of said diaphragm, which central portion is surrounded by a marginal portion including spaced-apart teeth-like projections which define the periphery of said marginal portion and an endless channel in said marginal portion between said teeth-like projections and said central portion, a resilient cushion-like damping washer having two opposite faces one of which is linked to said teeth-like projections and the other face being linked to said supporting frame, and said washer having blind cuts on said other face, at least in those portions which are located in-between said spaced-apart teeth-like projections.

https://patents.google.com/patent/US4184563A/en

NITO Corp 1986 - Peizoelectric activation of flat panels

Abstract
In this piezoelectric oscillation assembly there are included several piezoelectric oscillation devices, each including a first oscillation plate, a piezoelectric oscillation element attached to the first oscillation plate, and a second oscillation plate layered against the first oscillation plate so as to define an acoustically enclosed space between them. At least one of the oscillation plates is common to all of the several piezoelectric oscillation devices. It may be the first oscillation plate which is common to these several piezoelectric oscillation devices, or the second oscillation plate, or both. And the acoustically enclosed spaces of these several piezoelectric oscillation devices may be mutually isolated, or alternatively a one common acoustically enclosed space may be defined between the first oscillation plate and the second oscillation plate for all of these several piezoelectric oscillation devices.


https://patents.google.com/patent/US4751419A/en



W. Lee Torgeson 1989 - Spatial arrangements of separate flat transducers responsive to two or more frequency range


Abstract
A loudspeaker characterized by a symmetrical arrangement of planar mid-range and/or low-range drivers about an essentially point source tweeter. A flexible mounting of a speaker diaphragm to a rigid support serves to reinforce the advantages achieved thereby or is utilized independently, for instance to provide an improved sub-woofer.

https://patents.google.com/patent/US5081683A/en


HMG Patent 1990 - Initial Filing for HMG

Abstract
A panel-form loudspeaker has a resonant multi-mode radiator panel which is excited at frequencies above the fundamental frequency and the coincidence frequency of the panel to provide high radiation efficiency through multi-modal motions within the panel, in contrast to the pistonic motions required of conventional loudspeakers. The radiator panel is a skinned composite with a honeycomb or similar core and must be such that it has a ratio of bending stiffness to the third power of panel mass per unit area (in mks units) of at least 10 and preferably at least 100. An aluminium skinned, aluminium honeycomb cored composite can meet this more severe criterion easily.


https://patents.google.com/patent/WO1992003024A1/en


Bertagni 1993 - Planar diaphragm loudspeaker


Abstract
A planar diaphragm loudspeaker incorporating a planar diaphragm having multiple sound producing regions, a support frame, and one or more electromagnetic drivers coupled to the sound producing regions on the rear surface of the diaphragm. A first sound producing region is formed generally centrally on the rear surface of the diaphragm, and second and third sound producing regions are formed on opposite sides thereof. The sound producing regions comprise raised portions on the rear surface of the diaphragm which are symmetrical about both first and second orthogonal axes for balance. Cross members extend between opposite sides of the support frame so as to overlay all of the sound producing regions to selectively mount separate electromagnetic drivers for coupling to each sound producing region, as well as to mount one or more anchors extending from the cross members to the perimeters of the second and third sound producing regions in order to isolate them from the first sound producing region. The support frame is constructed from hollow rectangular channel frame members which are joined together by L-shaped corner connectors.


https://patents.google.com/patent/US5693917A/en




NXT 1995 - Early NXT Application to establish invention date.

https://patents.google.com/patent/WO1997009842A2/en

Abstract
Acoustic device (81) including a member (2) extending transversely of its thickness and capable of sustaining bending waves at least over an intendedly consequentially acoustically active area of the transverse extent of said member, the member (2) having, by reason of orderly design methodology disclosed and claimed, a distribution of resonant modes of its natural bending wave vibration at least over said area that is dependent on values of particular parameters of said members, including geometrical configuration and directional bending stiffness(es), which values have been selected to predetermine said distribution of natural resonant modes being consonant with required achievable acoustic action of said member for operation of said device over a desired operative acoustic frequency range.


NXT 1995 - Structural combinations of separate transducers or of two parts of the same transducer and responsive respectively to two or more frequency ranges

Abstract
An acoustic device, e.g. a loudspeaker, comprising a resonant multi-mode acoustic radiator panel having opposed faces, a vibration exciter arranged to apply bending wave vibration to the resonant panel to produce an acoustic output, a cavity enclosing at least a portion of one panel face and arranged to contain acoustic radiation from the portion of the panel face, wherein the cavity is such as to modify the modal behaviour of the panel.
Also disclosed is a method of modifying the modal behaviour of a resonant panel acoustic device, comprising bringing the resonant panel into close proximity with a boundary surface to define a resonant cavity therebetween.

https://patents.google.com/patent/US6553124B2/en



NXT 1996 - Full NXT Patent-lots of detail and diagrams

Abstract
Acoustic device including a member extending transversely of its thickness and capable of sustaining bending waves at least over an intendedly consequentially acoustically active area of the transverse extent of said member, the member having, by reason of orderly design methodology disclosed and claimed, a distribution of resonant modes of its natural bending wave vibration at least over said area that is dependent on values of particular parameters of said members, including geometrical configuration and directional bending stiffness(es), which values have been selected to predetermine said distribution of natural resonant modes being consonant with required achievable acoustic action of said member for operation of said device over a desired operative acoustic frequency range.

https://patents.google.com/patent/US6332029B1/en


NXT 1997 - Exciter Design.

Abstract
An electro-dynamic inertial vibration exciter (1) for exciting a member (2) having capability to sustain and propagate input vibrational energy comprising a motor coil assembly (3, 4) having a coil (3) rigidly fixed to a tubular coil former member (4), a magnet assembly (5, 6, 7) disposed concentrically with respect to the motor coil assembly, and means (11) suspending the magnet assembly for axial movement relative to the motor coil assembly, characterised by a carrier (9) supporting the motor coil assembly (3, 4) and adapted for connection to the member (2) to be excited.


https://patents.google.com/patent/WO1998034320A2/en


NXT 1997 - Further details on exciter mounting method


Abstract
A loudspeaker comprises a panel member (11, 21, 31, 41, 61, 71, 81, 91) as resonant acoustic radiator relying on bending wave action and a driver (12, 22, 32, 42, 62, 72, 82, 92) coupled to the panel member to cause bending waves therein. The panel member has its mass and/or bending capability of the panel member locally altered or otherwise different, particularly locally of the driver as coupled to the panel member. Local holes (16, 26, 36, 46), affixed mass (66) or affixed damping material (76) are effective to reduce high frequency due to drumming effects at the driver coupling (17, 37, 47, 67, 77, 87).

https://patents.google.com/patent/WO1999002012A1/en



NXT 1997 - multiple exciters mounted to small sub-panels

Abstract
A panel-form loudspeaker comprises at least two resonant distributed mode acoustic panels with a single driver provided to excite all the panels such that bending wave energy is efficiently transferred to each of them to excite distributed mode resonance in them to provide acoustic output. There may be two panels one larger than the other - the larger having a lower bending wave stiffness than the smaller. The smaller panel is positioned eccentrically of the larger and the driver eccentrically of the smaller panel. A plurality of smaller panels may be associated with a single larger panel. One embodiment provides a driver mounted to the or each smaller panel which is adhesively attached face to face to the larger panel. Another embodiment provides the or each smaller panel formed within and attached to the larger panel, a driver being attached to the or each smaller panel. The attachment of the panels uses a material forming a controlling compliant coupling. A third embodiment provides the or each smaller panel mounted on one end of an associated coupler passing through and attached to the larger panel, a driver being attached to the other end of the or each coupler. The attachment of the or each coupler to the larger panel uses a material forming a controlling compliance coupling. The or each coupler is tubular in section.

https://patents.google.com/patent/WO1998052381A2/en


NXT 1997 - Standard DML loudspeaker operation with detail of modes, suspension and exciter locations.

Abstract
A loudspeaker comprising a resonant panel-form member adapted to produce an acoustic output and a vibration exciting system on the panel-form member and adapted to apply bending wave energy thereto, characterised in that the vibration exciting system is adapted to apply a bending couple to the panel-form member.


https://patents.google.com/patent/EP1108345B1/en


NXT 1997 - Use of masses to control resonance peaks

Abstract
A loudspeaker comprises a panel member (11, 21, 31, 41, 61, 71, 81, 91) as resonant acoustic radiator relying on bending wave action and a driver (12, 22, 32, 42, 62, 72, 82, 92) coupled to the panel member to cause bending waves therein. The panel member has its mass and/or bending capability of the panel member locally altered or otherwise different, particularly locally of the driver as coupled to the panel member. Local holes (16, 26, 36, 46), affixed mass (66) or affixed damping material (76) are effective to reduce high frequency due to drumming effects at the driver coupling (17, 37, 47, 67, 77, 87).


https://patents.google.com/patent/WO1999002012A1/en



Slab Technology Ltd 1998 - DML panel with differential stiffness on perpendicular axis

Abstract
A flat panel rectangular or elliptical loudspeaker has one or more anisotropic diaphragms (14, 15), made of double skinned fluted polypropylene copolymer (core flute) or corrugated cardboard, having a longitudinal bending strength greater than the transverse bending strength. The diaphragm is vibrated by a driver unit (16) which comprises a magnet (17) and voice coil (18). In this version the magnet (17) is mounted on the rear diaphragm (15) and the voice coil (18) is mounted on the front diaphragm (14). The two diaphragms (14, 15) are mounted at their edges to a frame (11) with the driver unit (16) mounted in the space enclosed by the frame and the diaphragms.

https://patents.google.com/patent/WO1999067974A1/en


NXT 2001 - Use of Panel Stiffening methods

Abstract
A bending wave panel-form acoustic radiator formed from sheet material to define an acoustically active area and comprising at least one integral stiffening member in the form of a corrugation extending out of the plane of the sheet and at least partially across the acoustically active area of the radiator, which stiffening member is of substantially U-shaped cross section. A method of making a bending wave panel-form acoustic radiator, comprising forming a sheet into a panel having at least one integral corrugation member extending out of the plane of the sheet and at least partly across the sheet and of substantially U-shape cross-section, to stiffen the sheet to have a desired ability to support and propagate bending waves.

https://patents.google.com/patent/WO2002078391A2/en


Harman Audio Electronic Systems Gmbh 2001 - DML using curved panel

Abstract
The invention relates to a flat-panel loudspeaker comprising a lightweight panel (1), a retaining element for holding the panel and at least one driver that is connected to the panel in an oscillatory manner to generate mechanical oscillations in accordance with electric control signals. According to the invention, the panel is configured with at least a one-dimensionally curved shape.

https://patents.google.com/patent/WO2002060218A1/en


Philips Sound Solutions Belgium N V 2001- DML with resonant cavity ( boxed DML)


Abstract
A panel-shaped loudspeaker comprising a panel (1) having a first main surface (1A) and a second main surface (1B) parallel to the first main surface and further comprising an exciter (3) positioned at the first main surface. When the exciter is energized the panel is capable of sustaining bending waves and has a distribution of resonant modes of its natural bending wave vibration, which results in an acoustic radiation. The loudspeaker has a tuning element (17) extending near the second main surface of the panel and positioned opposite the exciter so as to form a resonant cavity (18) between the panel and the exciter.

https://patents.google.com/patent/US7155021B2/en



NXT 2005 - Comprehensive DML patent including use of shaped panels


ABSTRACT

Acoustic device including a member extending transversely of its thickness and capable of Sustaining bending waves at least over an intendedly consequentially acoustically active area of the transverse extent of Said member, the member having, by reason of orderly design methodology disclosed and claimed, a distribution of resonant modes of its natural bending wave vibration at least over Said area that is dependent on values of particular parameters of Said members, including geometrical configuration and direc tional bending stiffness(es), which values have been selected to predetermine Said distribution of natural resonant modes being consonant with required achievable acoustic action of Said member for operation of Said device over a desired operative acoustic frequency range.

https://patentimages.storage.googleapis.com/bb/ea/e0/90906b2a128e4e/US6904154.pdf


NXT 2006 - Use of multiple exciters

Abstract
An acoustic device and method of making said acoustic device. The acoustic device comprises a diaphragm (10) having resonant bending wave modes in the operating frequency range, and a plurality of electro-mechanical transducers (12) coupled to the diaphragm. The positioning and mechanical impedance of the transducers are such that at least a selected number of the resonant bending wave modes are balanced so that the net transverse modal velocity over the area of the diaphragm tends to zero with the balancing of the resonant bending wave modes being achieved substantially by the positioning and mechanical impedance of the transducers. The parameters of the diaphragm may be such that there are a plurality of nodal grouped locations at or around which the nodal lines of a selected number of resonant modes are clustered. Each transducer may be mounted at one of the plurality of nodal grouped locations.


https://patentimages.storage.googleapis.com/20/12/af/e97a13224985b9/WO2007083127A2.pdf



Apple 2014 - Use of multiple panels and frequency bands

Abstract
An electronic device includes an enclosure or housing panel that is used as part of an acoustic system. The panel is divided into several sub-panels. For each sub-panel, the device includes one or more actuators attached to vibrate the sub-panel. The actuator and its attached sub-panel convert an audio signal to acoustic output. Each actuator and sub-panel combination may receive a separate audio signal. The device includes a digital signal processor for controlling each of the sub-panel driving audio signals. The device may further include one or more backing frames that are attached to the panel to provide boundary conditions to the sub-panels. The boundary conditions define a resonance frequency for each sub-panel.


https://patents.google.com/patent/US9525943B2/en
 
From BurntCoil's link: A critical review of bending wave loudspeaker technology and implementation

https://odr.chalmers.se/bitstream/20.500.12380/154618/1/154618.pdf

"By examining the result of IACC and the SPL distribution in a well-damped surrounding, the property that DML is with diffuse sound field is confirmed"

This is what I wanted from DML, a wide soundfield that is resilient to room reflections. It's the primary comparative advantage DML has over cone speakers.


"In addition, the listening test unveils that after equalization the audiences may prefer DML to electro-dynamic loudspeaker, since the problem of lacking low and high frequencies has been fixed by equalization, which also give a possibility that one may improve the frequency response by the use of an equalizer."

Equalization was always inherent in my build strategy because I wanted a virtual CBT array which requires digital signal processing. As a result, I didn't face the same struggles others face when building their DMLs. I tuned with signal processing rather than physical adjustments.


"The simulations also pointed out the effect of the ratio of length/width to the modal density, saying that the length and width should never be the same for the sake of modal optimization and one may have the best ratio by simulation."

True, but that's because most research papers use a single homogeneous panel material, e.g. a sheet of aluminum. I use a plastic honeycomb panel with resin and threads of carbon fiber on one side while the opposite side is the cloth fiber without any modification. A panel's aspect ratio may not matter as much when you have a complex panel structure.

On the other hand, rigid foam is an inexpensive yet homogeneous material DIYers can use to make their DMLs. In which case they'd benefit from shaping their panel edges to create chaotic edge reflections (reduce standing waves.) Even if they want to create an overall square shape as in the attached image.

IMO, the primary reason to build a DML is to achieve a wide diffuse soundfield. These things are the opposite of a high directivity wave guide approach that tries to overcome room reflections by keeping the sound away from walls as long as possible.

Do you get really good intelligibility out of your DMLs in most of your room? Great, you accomplished the primary goal and enjoy a comparative advantage over traditional speakers. Now you can tune with physical materials and shaping then move on to DSP and equalization.
 

Attachments

  • 81792FC4-3973-424F-8AA5-D6D384ABEF7A.jpeg
    81792FC4-3973-424F-8AA5-D6D384ABEF7A.jpeg
    200.6 KB · Views: 992