A Study of DMLs as a Full Range Speaker

André,
I don't think the model used for an open baffle works for a DML. In a DML, there are sound emissions on all the surface, the emitting point changing with the frequency according to the mode shape (in a frequency representation meaning in a steady state with a sinus input). In some papers, the DML are said "self baffled" to say the membrane and the baffle are the same object. I point that because more than one year ago I made a comparison between a canvas panel (ok this is probably not exactly a DML) and a small full range on a same dimension open baffle). the FR in the low freq (below let say 500Hz) are completely different. No bass from the OB, bass from the canvas. Some 2D or 3D simulations of the front and rear wave propagation and recombination should help.
Christian
Yes I did indeed put in a disclaimer that those were baffle simulations, not DML. I was simply pointing out the issues with comb-filtering in round baffles which may or may not apply to DML's.
 
  • Like
Reactions: homeswinghome
@Andre Bellwood

And, here
IMG20230119114750.jpg
IMG20230119130153.jpg

the midrange are sunk in basins on the front 'baffle', tweeter is jutting out in an inverted horn like tube, long inside the box and the side woofer is also sunk in, but the tube in which it is, juts out. The so-called baffle, thin one in front and the side curvy one are made of such materials, so the so-called edge effects. All the 4 speakers are simply isolated from the 'baffles.' No surface is parralel to each other. That's why I said they have massive R&D departments. I collect somewhat unusual speakers. The shown speakers are placed few cms away from the wall as they have bass-reflex hole on the back. They have a 'sound stage' above the top of them and a bit away behind them and spread to sides. In other words, I don't hear them as coming from the speakers. Other listeners too.

This is quite an interesting pair. You may have noticed that the midrange doesn't appear to have any connection with the voice coil behind them. It is the same with the woofer. If you want, you can watch the dismantling of a newer version.

Looking at the membrane, doesn't it look like a DM membrane, but concave? Or, distributed Balanced Mode membrane? NXT patents also talk about curved membranes as DMLs. 🙂
 
Last edited:
Further to the post #8,683 and #8,655

The woofer in the video appears to be under suspended. Could be the reason of that interesting 'sound stage.'
View attachment 1132070 View attachment 1132071
I won't be dismantling mine to check, by the way. 🙂
Under suspended? You mean the inverted surround material? If so there are speakers that utilize it. The interesting sound stage is from the drivers being placed on the side of the cabinet, its nothing new.

Kef's Blade speakers has side firing drivers.


 
Last edited:
When the exciter excites the physical structure it is attached to, it will induce a bending wave that travels along / in the physical media. This wave travels with the "sound speed" of that material - it is much harder than air and therefore travels much faster. As this wave propagates in the physical media, sound pressure is created thanks to the bending of the surface - the generated sound wave travels with the speed of air.

So two different speeds in play ;-)

//
Hi TNT, (@TNT)
I must admit I'm not sure if you are agreeing or disagreeing! I can't be sure if you read my whole (long, sorry) post or not.
For sure I agree that with respect to a DML panel, it is the speed of the panel's bending waves that matter, as it is the bending waves in the plate (and not longitudinal waves) that generates the sound waves in the air.

But my point was that when most people talk about the "speed of sound" in a solid material, they are talking about the speed of longitudinal waves, not bending waves. But the two are quite different things.

Longitudinal wave speed is given by this equation:

1674167251653.png


While bending wave speed is given by this:

1674167694887.png


The one similarity is the dependance on E and rho. But the speed of longitudinal waves is simply a constant depending on the material properties, while the bending wave speed is a function of frequency and of the thickness of the plate.

But many people don't realize the two are (very) different things.

Eric
 
Hi TNT, (@TNT)
I must admit I'm not sure if you are agreeing or disagreeing! I can't be sure if you read my whole (long, sorry) post or not.
For sure I agree that with respect to a DML panel, it is the speed of the panel's bending waves that matter, as it is the bending waves in the plate (and not longitudinal waves) that generates the sound waves in the air.

But my point was that when most people talk about the "speed of sound" in a solid material, they are talking about the speed of longitudinal waves, not bending waves. But the two are quite different things.

Longitudinal wave speed is given by this equation:

View attachment 1132261

While bending wave speed is given by this:

View attachment 1132267

The one similarity is the dependance on E and rho. But the speed of longitudinal waves is simply a constant depending on the material properties, while the bending wave speed is a function of frequency and of the thickness of the plate.

But many people don't realize the two are (very) different things.

Eric
So would that be analogous to wave speed underwater vs wave speeds on the surface?
 
Further to post #8,655

Even though the exact words were not mentioned, BMR was invented by NXT, patent WO 2005/101899.
bmr1.jpeg

bmr_abstract.jpeg


Interestingly, in that patent, NXT moved somewhat away from the asymmetric placing of one transducer on a flat rectangular panel, to symmetric placing of few transducers in regard of the centre of mass of the panel, vertically or diagonally, such as,
vertically.jpeg
diagonally.jpeg


And, this one,
this one.jpeg and this, and this one.jpeg
confirms my thinking of using a conventional cone speaker with a thin quite stiff cone on the back of a thin under suspended membrane, whether circular or rectangular that membrane is.

Now, thinking about the LG's concave shallow speaker drivers (#8,683, #8,684), I might try to find another pair to dismantle them to get those drivers out to use them behind panels, maybe filled with some foam as Hitachi did once. 🙂
 
Last edited:
Further to post #8,655

Even though the exact words were not mentioned, BMR was invented by NXT, patent WO 2005/101899.
View attachment 1132395
View attachment 1132403

Interestingly, in that patent, NXT moved somewhat away from the asymmetric placing of one transducer on a flat rectangular panel, to symmetric placing of few transducers in regard of the centre of mass of the panel, vertically or diagonally, such as,
View attachment 1132404 View attachment 1132406

And, this one,
View attachment 1132411 and this, View attachment 1132412
confirms my thinking of using a conventional cone speaker with a thin quite stiff cone on the back of a thin under suspended membrane, whether circular or rectangular that membrane is.

Now, thinking about the LG's concave shallow speaker drivers (#8,683, #8,684), I might try to find another pair to dismantle them to get those drivers out to use them behind panels, maybe filled with some foam as Hitachi did once. 🙂
You cant just use any type of driver, they have to be full range drivers.
 
Be gentle

"True enthusiasm is a fine feeling whose flash I admire where-ever I see it."

Charlotte Bronte

😁👍
Eucy
I am all for enthusiasm, but when one does nothing and does not accomplish anything time after time the enthusiasm turns into annoyance. If he is as smart as he thinks he is he should be able to build a better sounding DML panel then everyone here.
 
@aagas ...just a thought in case the prices or spec are better https://signtradesupplies.co.uk/col...luminium-composite-sign-blanks-standard-sizes

This company produces signage blanks as an Aluminium/EPS/Aluminium composite material. There are many stock sizes and the price looks reasonable. The Aluminium Skin is very thin at 0.21 but the blanks are described as very rigid. I am tempted to test a small sample
This is a very interesting find! The cheaper ones with the thinner aluminium facings look like very good value. I do wonder how they would compare with my 2mm ceiba plywood panels though. £9.95 delivery charge for <£150 orders.

I will be rehousing my 15" Monacor PA drivers in H frames when the weather gets better (i.e. I can get the panels cut at B&Q and into the car without being rained/snowed on). I will then be in the market to replace my current 30x40cm ceiba panels with something a bit larger. My previous 55x40 poplar panels seemed to have better midrange detail, so 55x40 ceiba seems like a possibility.

With active crossovers & EQ, my current hybrid setup is +/-1.5db from 35hz to 18khz with a drop off above that. Whilst the overall tonality is fine, things seem to get a bit confused with complex orchestral passages. it is, however, very revealing of recording quality (and lack thereof).