A Study of DMLs as a Full Range Speaker

latest products.jpeg


Here is the latest offer. You can import all of them, or one of them, to dismantle and discover how it ticks. I won't.
There are a lot of dismantling videos on YouTube; yours could be one of them, attracting tens of thousands of views. All you need is a clutter-free room for that.
 
Sandasnickaren said:
I've been testing 3D print panels with intresting results. I prefer to keep my speakers fairley small in order to use them in the living room without to much complaints...
I'm using simple PLA for the panel and frame, The panel has 20 % infill and the top an bottom layers are 0.2 mm. The layout of the panel is inspired by the Goebels high end bending wave.
Click to expand...
Hi Sandasnickaren, These look great! Wondering if the backs of these are fully open, since the discussion of late has been on the potential negative effects from the back wave. Thanks, Bruce
 
Veleric said:
I am working on a little analysis with FEM to see just how using this ratio actually works with respect to the distribution of natural frequencies. But I wanted to be sure that I was using the "rule" properly before I tried assessing it.
Click to expand...
So here is what the finite element modeling says about how the "k" factor from the Herger patent performs with respect to the distribution of natural frequencies.

First, I modeled an isotropic plate (stiffness is the same in both directions) with a aspect ratio of 9/5, as prescribed in the Herger patent. The vertical series of dots on the far left (k=1) represent the frequency distribution for that case. Note that the absolute value of the modal frequencies is arbitrary, but their relationship to each other along the frequency axis will be the same regardless of the plate thickness, total area, etc., so long as the plate is isotropic and the aspect ratio is 9/5.

1734041305498.png


On inspection, it's far from clear to me why this distribution would be particularly good. Herger says it is, so maybe it works just fine for him. To me, it looks like a poor choice, particularly looking at the odd, odd modes, that radiate the most efficiently. Note how far apart the first two of these are (1,1 and 1,3). And then, the 5,1 3,1 and 3,3 are all bunched up at near 500 Hz.

But the question I was really wanting to look at was this: Does applying the "k" correction factor to the aspect ratio result in the same distribution of the natural frequencies along the frequency axis? The answer seems to be "yes", for a few of the modes, but "no" for the majority of modes.

I modeled two additional cases, one where the stiffness ratio (k) was 1.25 and another with k=2.0. The aspect ratios in those case was (9/5)*1.25=2.25, and (9/5)*2=3.6, in the two cases, respectively.

The results show that the relationships if the 1,1 2,2 3,3 and 4,4 modes all stay exactly the same. But all other modes either increase or decrease significantly, depending on the k factor, changing their relationships with each of the other modes.

I should add that the patent never claims that using the k factor to adjust the aspect ratio would keep the relationships of the natural frequencies from changing, but it is natural to assume it's what they were trying to achieve. In any event, it seems to work that way for a small subset of modes, but not the majority of them.

Eric
 
Veleric said:
ekha,
What is the relationship between MescalitoDPKTST and Tefra? Are they both the same organization?
Eric
Click to expand...
What do you think? 😉
Tefra, Mescalito, Sotis AG ... all around TST -- transverse sound technology.
The thing is, practically no one in the local crowd is interested in any kind of DMLs – it’s just a humorous topic in the local forums, wherever that may arise. There may be about 3 to 5 people interested in a flattened coil on a flat panel, but none at all in attaching an exciter to a flat panel or hanging them by strings from the ceiling or anything like that. Patents or not...
 
As bdjohns just mentioned, it would be interesting to take a look back at Göbel's flat composite membrane, even though he abandoned the idea of trying to profit from DMLs. Perhaps one should consider it after reading through the article linked by RMAM, but not focusing on the cuts, which is a misdirection – the suspension around the panel would dampen the edges anyway – but rather on the holes at regular intervals (fig.6 of his patent). By the way, Oliver Göbel is an audio engineer who once worked at Siemens as a loudspeaker and electroacoustic developer.

If you take a careful look at that membrane, you’ll notice that it is made of equal pieces of very thin balsa glued together at the thin side, with the pieces arranged against each other in the manner that carpenters typically glue wood together, with the grains placed against one another. Siemens, by the way, has had considerable experience in using wooden panels as sound diffusers since the beginning of the last century.
 
Last edited:
lekha said:
View attachment 1392798

Here is the latest offer. You can import all of them, or one of them, to dismantle and discover how it ticks. I won't.
There are a lot of dismantling videos on YouTube; yours could be one of them, attracting tens of thousands of views. All you need is a clutter-free room for that.
Click to expand...
Please could you post the YouTube links where you see the opening of these?
 
bdjohns said:
Hi Sandasnickaren, These look great! Wondering if the backs of these are fully open, since the discussion of late has been on the potential negative effects from the back wave. Thanks, Bruce
Click to expand...
Yes, they are fully open and free standing in the room with 1m distance to walls.
However I’m using the same panel in two other cases (home cinema):
First being i an athmos system as ceiling speakers. They are in this case installed in the roof with the backside fully enclosed in a large cavity which means it’s an infinite baffle. This works fine.
Second being in a small closed box as center speaker (unable to make a hole in the wall). (Still working on it not sure it’s usable).

I have earlier been experimenting with various form of damping of the backside. I got the initial inspiration from Goebel high end speakers. They have created an acoustical lowpass filter if I understand it correctly (it is stated in their spec that the frequency response on the backside is 200-4000 Hz, if I remember it correct).
I gave it up as I couldn’t see the value of neither through listening or in measurements.

And there are a couple of issues to consider:
  • The frequency response from whatever you are using. It has to be linear/flat otherwise it will change (like an equalizer) the frequency response of all sound coming from the back in comparison with the front. This will influence the in room response (the sum of direct and reflected sound in listening position, what you hear) in an unpredictable way.
  • There will be sound reflected back (” inside) to membrane/panel from the material on the backside this is due to the difference in impedance between the air and whatever material you are using. This is more or less the same effect as you have i regular box speakers ”box coloration”. Can you hear this? It depends on the amount of reflections…

In conclusion, today I prefer the omni directional characteristic of the open solution combined with DSP equalization as this gives the flattest frequency response from the speaker and also in the room response.
 
pixel1 said:
Please could you post the YouTube links where you see the opening of these?
Click to expand...
Sorry, that you misunderstood what I wrote. It was a just proposal for those who have objections, to buy one of them and dismantling it... and, of course, posting a YouTube video of that. There are none at the moment.
 
lekha said:
I will paste the link where I found that interior image when I locate it again.
Click to expand...
yes of course it would be interesting to understand where they really are given that by now they have all failed, tectonic seems to have sold this system to one of its employees, everything suggests that there is no qualitative solution. In short, something is always missing
 
Sandasnickaren said:
Yes, they are fully open and free standing in the room with 1m distance to walls.
However I’m using the same panel in two other cases (home cinema):
First being i an athmos system as ceiling speakers. They are in this case installed in the roof with the backside fully enclosed in a large cavity which means it’s an infinite baffle. This works fine.
Second being in a small closed box as center speaker (unable to make a hole in the wall). (Still working on it not sure it’s usable).

I have earlier been experimenting with various form of damping of the backside. I got the initial inspiration from Goebel high end speakers. They have created an acoustical lowpass filter if I understand it correctly (it is stated in their spec that the frequency response on the backside is 200-4000 Hz, if I remember it correct).
I gave it up as I couldn’t see the value of neither through listening or in measurements.

And there are a couple of issues to consider:
  • The frequency response from whatever you are using. It has to be linear/flat otherwise it will change (like an equalizer) the frequency response of all sound coming from the back in comparison with the front. This will influence the in room response (the sum of direct and reflected sound in listening position, what you hear) in an unpredictable way.
  • There will be sound reflected back (” inside) to membrane/panel from the material on the backside this is due to the difference in impedance between the air and whatever material you are using. This is more or less the same effect as you have i regular box speakers ”box coloration”. Can you hear this? It depends on the amount of reflections…

In conclusion, today I prefer the omni directional characteristic of the open solution combined with DSP equalization as this gives the flattest frequency response from the speaker and also in the room response.
Click to expand...
Hello Scandasnickaren,
Your post is fully on what I consider at the moment!
In this post, are you refering to the panel 250x190mm in post #12529 ? extract below.
1734099465969.png

I initiated directivity measurements few posts ago mainly to detect possible important lobes at high angles. An update of my last paper should come if I am not going to be lost in new ideas...
From that I think I can conclude that some directivity singularities are detectable quite easily (currently with a gated IR). The need or the way to correct them seems to be an other story (see you post + the general position among the DIYer to use fully open back DML).
The singularities in directivity that are detectable are the lobes due to the coincidence frequency (if the fc is not out of the frequency range) and lobes in the mids that might be the combinaition of the front and back rear in the 40° / 60° range in front.
I observe 2 other singularities which might be with more effects.
One on-axis which is a peak around 500Hz. Going back to your post, I see you have a peak in this range. My test panels are 30x400mm measured at 1m so maybe no relation. Have you found a source of this peak? To be more explicit on why I am scratching my head: I have to explain that I made a set of measurements of 30x40cm DML with different back load, a 8cm large range in an open baffle of same dimension and the same model in a small closed box. The peak is here for the DML and the open baffle with or without a back load (even if it is less with) but not (or very small) with the small closed box. A difference being the front face of the closed box is maybe 10x12cm (and 10cm depth 1.2l).
1734100360736.png

The second singularity is an important emission around 2kHz to the back of the panel. Have you already observed that?

Going further in my last directivity measurements today, I agree it could be according to what you tested that some rear loads with fast and strong impedance change bring other problems, the balance probably depending of the distance to the front wall.
Christian
 
pixel1 said:
yes of course it would be interesting to understand where they really are given that by now they have all failed, tectonic seems to have sold this system to one of its employees, everything suggests that there is no qualitative solution. In short, something is always missing
Click to expand...
I have pasted the link in the post above yours. Here it is again.
"In short, something is always missing" is quite correct.
No company with some reputation would ever resort to this method again, I believe.
 
Last edited:
You must log in or register to reply here.
Back
Top