I finished my first basic test on the Canvas DML Design. It was an exciting endeavor with a lot of things learned along the way.
I took 27 measurements on birch plywood panels with different shapes (circular, square), thicknesses (2, 4, 6.5mm) and surface areas (450, 250, 50cm²), added holes and rounded corners, too.
Subsequent measurements were completely robust = repeatable with the same frequency response outcome. The only thing I found introducing minor errors in the measurement data, was my somewhat unsophisticated way to make sure, when exchanging panels to orient them in exactly the same way relative to the microphone (perfectly perpendicular).
So just by looking at the graphs of my limited sample set of a specific type of material with no claim of validity or universal applicability (so, don't quote me!
), I come to the following conclusions with regard to my assumptions.
First, due to the limitations using gated measurements, I can't say anything about differences in bass extension.
What I found to be true:
I took 27 measurements on birch plywood panels with different shapes (circular, square), thicknesses (2, 4, 6.5mm) and surface areas (450, 250, 50cm²), added holes and rounded corners, too.
Subsequent measurements were completely robust = repeatable with the same frequency response outcome. The only thing I found introducing minor errors in the measurement data, was my somewhat unsophisticated way to make sure, when exchanging panels to orient them in exactly the same way relative to the microphone (perfectly perpendicular).
So just by looking at the graphs of my limited sample set of a specific type of material with no claim of validity or universal applicability (so, don't quote me!
), I come to the following conclusions with regard to my assumptions.First, due to the limitations using gated measurements, I can't say anything about differences in bass extension.
What I found to be true:
- the bigger the panel's surface area, the more bass SPL (<2kHz); no difference in SPL above 2kHz
- the thinner the panel, the more overall SPL over the complete frequency response
- assuming that the thinner the plywood panel = the higher the compressive strength, the more treble extension (weak indications for frequencies starting from 16/18kHz up)
- the more panel holes (idea introduced by raph1978) relative to the panel's surface area, the more compressed (vertically & horizontally) the impulse response
- shapes (circular/square) and rounded corners
- the smaller the panel's surface area, the more treble extension
Last edited:
This exciter is supposed to handle 50 watts?
What are you attaching them to that is making you overdrive them enough to heat them up ?
The problem attaching the exciter to different panels is the thin film which covers the tape on the coil foot.
If you used double sided tape the thin film could very easily pull off?
You could use a thin bead of bluetak around the foot and gently hold or support the exciter or as mentioned rest the exciter facing down on the panel.
This should prevent the buzzing and maybe give you an idea of how the panel sounds?
But weather the thin film will still come off is sill going to be a problem.
Steve.
Pixel1.
Instead of hanging the wires down ,you could use the wires to support the exciter weight,which will take the weight off of the thin foil.
Steve.
How To Build "The World's Best Speakers" - Are Flat Panel Speakers Really Any Good? - YouTube
Instead of hanging the wires down ,you could use the wires to support the exciter weight,which will take the weight off of the thin foil.
Steve.
How To Build "The World's Best Speakers" - Are Flat Panel Speakers Really Any Good? - YouTube
do you think that in my case to do a serious test I have to get a carbon panel with 4mm total tectonic honeycomb type or maybe with an aluminum panel?
all those vibrations as well as the poor adhesion of the exciter the panel is also ddocuto to the fact that I did not have a frame?
Thanks again
all those vibrations as well as the poor adhesion of the exciter the panel is also ddocuto to the fact that I did not have a frame?
Thanks again
pixel1,
I've been wondering what tectonic panels are made of. Are you saying you believe that they use 4 mm carbon/honeycomb panels? Do you recall where you got that information? I'd be very interested to know.
What panel material are you using now? If you want to play really loud you really need a very light and stiff panel. Few materials really meet this requirement. The only inexpensive material I know of that is sufficiently stiff and light to play really loud is polystyrene foam (EPS or XPS). Trying it will give you an idea of whether or not it's possible to make speakers as loud as you desire using your chosen exciters.
You may or may not like the sound quality with polystyrene foam (I don't, but others do). But at least you will know if it's possible to get the sound volume you want.
If the foam plays loud enough but you don't like the sound, then it may be worth investing in a more exotic material like a carbon fiber/honeycomb.
I don't think aluminum/honeycomb is likely to be loud enough for you. I'm not saying it's a bad material for a DML, just probably not super loud.
Aluminum is similar in stiffness to carbon fiber, but about 70% heavier. That extra weight will really limit the sound volume possible, I suspect.
Eric
Hi Eric,
Just my experience, my aluminum panels (with a single exciter on each) can play far louder than I can bare.
Granted, I have them in a domestic (if pretty large) rather than commercial setting.
Yes, except the materials (aluminum vs carbon fiber) have differences that may significantly affect their use as DMLs.
Unlike carbon fiber, aluminum, as isotropic, (e.g., uniform) has the same properties in every direction.
Carbon fibers strength and rigidity come from a careful positioning of the fabrics in a specific ways for specific applications. I don't know of any (all though I'd certainly like to) that have isotropic properties.
As with all engineering design, the physics & materials present tradeoffs.
I have no doubt that (at some point) one will have the ability to optimize carbon fibers for use in DMLs.
I just don't think anyone can know (at this point) if commercially available carbon fiber composit panels by pure coincidence happen to work better than aluminum ones.
All said, I REALLY want to find out ;-)
All of stand at the beginning of this grand experiment.
What fun!
Just my experience, my aluminum panels (with a single exciter on each) can play far louder than I can bare.
Granted, I have them in a domestic (if pretty large) rather than commercial setting.
Yes, except the materials (aluminum vs carbon fiber) have differences that may significantly affect their use as DMLs.
Unlike carbon fiber, aluminum, as isotropic, (e.g., uniform) has the same properties in every direction.
Carbon fibers strength and rigidity come from a careful positioning of the fabrics in a specific ways for specific applications. I don't know of any (all though I'd certainly like to) that have isotropic properties.
As with all engineering design, the physics & materials present tradeoffs.
I have no doubt that (at some point) one will have the ability to optimize carbon fibers for use in DMLs.
I just don't think anyone can know (at this point) if commercially available carbon fiber composit panels by pure coincidence happen to work better than aluminum ones.
All said, I REALLY want to find out ;-)
All of stand at the beginning of this grand experiment.
What fun!
Regular woven carbon fiber is 25% stiffer than aluminum, unidirectional is more but less pretty than the woven.
Carbon fiber has better damping than aluminum
. High modulus carbon fiber exists thats up 2.5 stiffer than the normal high strength carbon fiber. Its more brittle expensive and weaker but its way stiffer.
Carbon fiber has better damping than aluminum
. High modulus carbon fiber exists thats up 2.5 stiffer than the normal high strength carbon fiber. Its more brittle expensive and weaker but its way stiffer.
Mpte,
Interesting.
Does unidirectional carbon fiber have isotropic properties?
Not certain how it would follow from having the fibers running in a single direction.
Just trying to understand your thoughts on this.
Also, how does carbon fiber's
I haven't considered this, so I'd like to understand as I start thinking about my next set of panels.
Found this Rockwest Composites link Carbon Fiber Grades: It's All About Tensile Modulus that might help explain some of this.
I wonder how much strength one needs in a panel?
I guess enough that an exciter doesn't shatter the panel.
Still looking for a simple model one could hold in their mind relative to optimizing panel material selection.
Interesting.
Does unidirectional carbon fiber have isotropic properties?
Not certain how it would follow from having the fibers running in a single direction.
Just trying to understand your thoughts on this.
Also, how does carbon fiber's
come into play in all of this?
I haven't considered this, so I'd like to understand as I start thinking about my next set of panels.
Found this Rockwest Composites link Carbon Fiber Grades: It's All About Tensile Modulus that might help explain some of this.
I wonder how much strength one needs in a panel?
I guess enough that an exciter doesn't shatter the panel.
Still looking for a simple model one could hold in their mind relative to optimizing panel material selection.
How DMLs Work Part 2 - YouTube
Product of the Week: Tectonic Audio Labs DML500 - Tectonic | Tectonic
you can get information directly from their videos and site, i think they use 4 exciters on a 4cm total honeycomb panel. It would be useful if someone here heard these speakers playing or maybe bought some similar panels.
As for exciters, which ones do you think are the best on the market?
Product of the Week: Tectonic Audio Labs DML500 - Tectonic | Tectonic
you can get information directly from their videos and site, i think they use 4 exciters on a 4cm total honeycomb panel. It would be useful if someone here heard these speakers playing or maybe bought some similar panels.
As for exciters, which ones do you think are the best on the market?
pixel1,
Thanks for those links! I've seen both of those before but did not recall that those details. Indeed, they say they are using a panel with carbon fiber skins and a honeycomb core that is about 40cm x 50 cm and 3.5 to 4 mm thick.
Some details are missing, like exactly what honeycomb or what thickness and layup of the carbon layers are being used, and these details can make a very big difference. But my best guess would be they use 3.2 mm Nomex honeycomb with a 0/90 6k carbon fiber fabric/epoxy skins, each about 0.25 mm thick. That would result in a total thickness of 3.75 mm, which is smack in the middle of the 3.5 to 4 mm range they mentioned. And would also have a very good "stiffness/mass^3" ratio, which should be favorable for high output.
If you are looking for panels to play really loud, I suspect this is what you really should be looking for.
Interestingly, even thinner carbon/epoxy skins (theoretically) would provide an even higher stiffness/mass^3 ratio. But this requires a 3k carbon fabric, or special weave, that runs about 3 times the cost of of the 6k fabrics.
Eric
It does not, for exactly the reason you mention. Unidirectional carbon is actually the least isotropic type of fabric.
You can come close to isotropic behavior with 3 layers of unidirectional layed at 60 deg angles, or 4 layers are 45 deg angles, or 5 layers at 36 deg angles, etc. These are actually referred to as "quasi-isotropic" layups.
Most CF fabrics are biaxial (layers at 0 and 90 degrees). So you can make a quasi-isotropic skin with 2 layers of 0/90 fabric layed at 45 degrees.
That said, I'm not sure if isotropic is necessarily better for a DML than anisotropic.
I'm not sure if this is simple or not, and I'm sure it's not the whole story, but I believe the starting point is maximizing "stiffness/mass^3".
Eric
I don't doubt that a bit. I can say the same of my plywood panels (two exciters on each). But I had the impression that pixel1 had a commercial setting in mind.
Well, given the same density honeycomb core, and the same skin thickness, a carbon fiber panel will be stiffer and lighter than the corresponding aluminum one, and hence have a significantly higher stiffness to mass^3 ratio. So, arguably a CF panel is better than aluminum.
But it all comes down to the details of construction. It's possible to make a great panel with either material, I believe. And I don't think anyone is designing commercial panels intentionally optimized for use in a DML. So it really all depends on what construction you can find to buy, or build yourself.
Eric
Just to mudded the water a bit on the current topic of honeycomb and thin stiff plates. It appears that most of the waves being generated are transverse and not longitudinal operating in the Rayleigh region. This is more like a Surface Acoustic Wave or SAW phenomena. In this case the wave velocity is quite high compared to air and should be generating sonic booms since the sound velocity is on the order of 343 m/sec in air and several thousand of m/sec in aluminum and CF. Maybe one should focus on how to generate Lamb waves since these are longitudinal in nature and couple more efficiently in to the air and have a slower velocity which gives a better impedance match or better coupling of acoustic energy.
Cheers Steve
Cheers Steve
yes I work in the professional sector, I think we should start from what tectonic did to be able to apply quite high powers and contain resonances, for the sound quality if they have chosen carbon and not aluminum obviously it is not for the weight but maybe for the sound, probably the high end has been reduced so they thought of adding a ribbon tweeter (sharable choice for quality over compression)
Aluminum will surely sound too but at least for me finding aluminum with honeycomb is impossible
Aluminum will surely sound too but at least for me finding aluminum with honeycomb is impossible
My point is that the sound is largely because of the weight (or rather low weight). So, in effect, they really did choose it for the weight. Or at least for the combination of high stiffness and low weight. As they said in the Youtube vid.
Eric