Interesting. But 200hz is quite high. Higher than most subs will go, and my guess is that a bit thicker plate would solve that.
And what is a "heavy panel". A heavy panel with have the same issue with not being able to resonate.
I have made 1mm panels, but similar panel with around 3.5mm is about 3x as loud with much fuller sound.
I will do some experiments with different thickness and all things otherwise being the same, to verify I'm not wrong, but I'm guessing that hardly any materials will behave optimally with 1mm plates.
Yes, that would be a problem! In the piston speaker world, when people talk about the cone "breaking up", they mean the cone starts bending (like a dml plate) rather than simply moving as a piston. I though maybe you meant "breaking up" in that sense, rather than simply falling to pieces.
Eric
Haven't seen results from using ready made nomex panels, and while you seems to have a much better grip about the theory than me, I do suspect that it might fare better than you expect.
What does seem to happen as you increase size is that you get more modes, spreading out the peaks in the FR. This is especially obvious in the low frequencies. So it becomes harder to even them out by adjusting material properties and size if the plate is small. In the FR from my printed plates, FR is overall quite uneven, but not really worse between 80-200Hz than 200-1500Hz. I would still say it is certainly very useable down to 150Hz. If I was capable of calculating the fundamental I'm guessing it is way higher than 60Hz, in which case the theory is not really true or we have a very different definition of what is useable.
Yes, low density is key. But thin plates do resonate and create modes. Arguably more of them, in fact.
But mainly, what I wanted to clarify is that the function of cavities is not to act as some kind of resonating chamber. (I'm not saying that's what you were thinking, only that if anyone has that idea, it's not correct. ) Rather, cavities work because they are a way of creating low density, and it's the low density that matters, not the cavities themselves.
For example, say you make two panels, both with the same carbon fiber skins. For one panel, the core is solid polyethylene (PE) plastic of a particular thickness. For the second panel, the core is a heavier, stiffer material of the same thickness as the PE, but has enough tiny holes drilled into it so that it ends up being exactly the same weight and stiffness as the PE core. So the second panel has cavities, while the first does not, but otherwise the panels are virtually identical. In this case, both panels will have virtually the same efficiency, even though one has cavities and the other does not.
Eric
If you made the thinner panel with proportionally thinner skins, the panels should theoretically have about the same efficiency. That way they both have the same effective density. But if the thick and thin panels have similar skin thickness, then the thicker one will have lower density, and hence should have greater efficiency as you observed.
Of course, I know there is a limit to how thin you can print the skins. And as well there is a limit to how thin the skins can be before they would buckle in compression, so at that point the thinner one would not function as a proper skin/core/skin composite.
Eric
Last edited:
Loeb,
I should clarify that my comments and calculations regarding panel size are all based on a panel that is "simply supported", which basically means attached to a rigid frame with an elastic material around much of the perimeter. I suspect you may be using the "hanging from strings" method. My (limited) experience with the strings method is that at the low frequency end, such panels have less overall output and even "lumpier" output than a simply supported panel of the same dimensions. But once they get above the lumpy region, the string type panels can provide just as flat of a response as one with a frame.
Others prefer the strings method, I should note.
Eric
I should clarify that my comments and calculations regarding panel size are all based on a panel that is "simply supported", which basically means attached to a rigid frame with an elastic material around much of the perimeter. I suspect you may be using the "hanging from strings" method. My (limited) experience with the strings method is that at the low frequency end, such panels have less overall output and even "lumpier" output than a simply supported panel of the same dimensions. But once they get above the lumpy region, the string type panels can provide just as flat of a response as one with a frame.
Others prefer the strings method, I should note.
Eric
an exciter is no different from an ordinary magnet and coil on a drive unit.Has anyone ever tried attaching a DML exciter to a large diameter woofer cone (cut from an existing woofer, recone kit, passive radiator, etc.)?
Older big woofers with small dust caps that would otherwise be junk in the normal woofer sense might work.
so you would be replacing one type of coil for another type of coil but less powerful , I suspect ?
the cone is designed to prevent dml modes from happening.
but no doubt it will produce sound though.
Steve.
Voids in a material change it's sound quality. Thats a simple fact which you can verify simply by tapping various materials with your finger and noting that.you can hear the difference where there are voids and where there are not. The various formulas might not change based on voids or size of voids independent of overall density but sound does change. Math is helpful in designing speakers but it's not the whole story.
I suspect the very large voids in the honeycomb plates serve more than one function. The voids are large enough to have their own internal dynamics at audible frequencies...they are chambers to themselves with volumes, borders, reflections, etc...and they are large enough that the they to not effectively more as a solid mass...when the panel accelerates the density of air at one side of the chamber is higher than the other and the air has some fluidity of movement. It's very small, but not zero, and definitely affects the duration of standing waves. Assuming two disparate materials of the same overall density and stiffness will give the same result is silly.
I suspect the very large voids in the honeycomb plates serve more than one function. The voids are large enough to have their own internal dynamics at audible frequencies...they are chambers to themselves with volumes, borders, reflections, etc...and they are large enough that the they to not effectively more as a solid mass...when the panel accelerates the density of air at one side of the chamber is higher than the other and the air has some fluidity of movement. It's very small, but not zero, and definitely affects the duration of standing waves. Assuming two disparate materials of the same overall density and stiffness will give the same result is silly.
Narenaud
I have not tried other exciters but I believe the small 25mm coils like mine have a higher frequency response ?
This has been discussed further back on this site.
It is difficult to tell the exciter response from their plots as they tend to use standard xps as the panel material for testing !!!
Probably because they have been promoting their xps panels with their exciters in the online build in parts express ?
This was done by someone who joined our NXT RUBBISH forum for a while , and thought he new how to build a good dml panel 😧
it works I suppose.
Steve.
I have not tried other exciters but I believe the small 25mm coils like mine have a higher frequency response ?
This has been discussed further back on this site.
It is difficult to tell the exciter response from their plots as they tend to use standard xps as the panel material for testing !!!
Probably because they have been promoting their xps panels with their exciters in the online build in parts express ?
This was done by someone who joined our NXT RUBBISH forum for a while , and thought he new how to build a good dml panel 😧
it works I suppose.
Steve.
So if we agree on that, the question then is why does lower density rather than low mass make the plate easier to drive?
I think it is pretty intuitive that less mass is easier to drive, and lower density helps achieve low mass. It takes energy to move mass, so less mass to move, more efficiency...simple.
But how does lower density increase efficiency if we also increase the total mass?
The exciter should have to give more power to move more mass, so any gains from lowering overall density should be negated.
The answer must be that the signal self-amplifies through resonance. At least I cannot think of any other explanation that doesn't break the laws of physics.
Leob
Many exchanges since this post...
It is perhaps a matter of how we read the FR or the ideas we can have to improve the system
I have 2 types of panel working at the moment. The 41x33cm canvas you mentioned and a 45x125cm 3mm plywood. They have enough bass for a daily use in full range but in my view, the 150Hz region has to be fill in by some mean. If it is by EQ, some additional power is needed (not a problem) to reach higher displacement which might be a problem.
To speak with data, it would be helpful to have the relation have the relation SPL = f(displacement, frequency, area...)
It is a characteristic of DML as explained by Veleric and in papers about the acoustic of thin plate that the modes just above the 1st one have a lower efficiency. Then comes a region of higher efficiency and then again a region of decreasing efficiency.
Christian
I asked the supplier of the 50mm exciters I use if they could provide FR graphs of different models, and if I might get better HF response with the smaller models. They didn't provide graphs, but according to the rep there is no such correlation, and "the most important factor to affect frequency response is magnet and copper coil."
I have nothing to compare my 50mm to yet, so I personally have no idea if that is correct.
Hi Steve
By dispersion, I wanted to speak about different materials hidden behind the same general name to explain why some of us get good results, some not. This afternoon, I just saw that what I bought as XPS 9mm in one DIY store is sold in the one I was under Depron 9mm.
Is there a trap here? I remember (where?) DIYer happy with XPS, other not with Depron.
Christian
So I guess we agree that it is harder to get an even response with a small plate. And since gaps between modes will be most noticeable in the lower frequencies, they can have especially problematic bass response. It doesn't have to be unusable though if you manage to distribute the modes, and seems to be far from a hard limit on bass reproduction based on size as at least I initially got the impression should be the case.
Unfortunately, it's hard, and not worth the effort, to try to make two different panels with exactly the same stiffness and density, with and without voids.
In any event, my claim was that they would have about the same efficiency, and not that they would sound identical. Though I suspect they would sound more similar than you might think.
But my main point was that, while the cavities themselves may or may not affect the sound quality, the primary function of the cavities in a carbon/nomex type panel is to make the panel lightweight and thereby improve its efficiency. I suspect that any other effect that the cavities might have would be considered an artifact, rather than a designed-in feature.
I think by now I have seen all the Tectonic videos, and read perhaps hundreds of technical articles about DML panels, and not once do I recall any reliable source talking about how the voids size, shape, etc was considered in the design. If you have seen something different I would be interested to see or read it.
Eric