Herb777.
The 6mm has round flutes with thicker walls, the thinner proplex has square flutes and starts to look more like correx.
Proplex has a strong output down to 40hz with my 10 watt excites, But I need more power to push this panel harder.
The thinner proplex starts to look like correx with much lighter and thinner walls.
The correx is heavy compared to EPS.
My exciters can drive the 4mm proplex ok.
Correx is 3mm.
Steve.
The 6mm has round flutes with thicker walls, the thinner proplex has square flutes and starts to look more like correx.
Proplex has a strong output down to 40hz with my 10 watt excites, But I need more power to push this panel harder.
The thinner proplex starts to look like correx with much lighter and thinner walls.
The correx is heavy compared to EPS.
My exciters can drive the 4mm proplex ok.
Correx is 3mm.
Steve.
It's your standard cardboard but I have added 2 more layers of corrugated paper/ cardboard , I will be adding a final piece reversed ( corrugated part facing out) this will give it the appearance of an old washboard 😄. The size of the panel is 130 cm X 60 cm it has 2 exciters per side one 32 mm and a small 25 mm strategically placed after many listening sessions, ribbon tweeter fills in after 2400 k and is on a L pad so it can be adjusted to taste. Sides are held solid only the top and bottom are free to move.
Interesting!
When you have time, please post some images, sketches and everything relevant of your build.
I think this stuff comes in different materials some seem pliable while others seem brittle, have you found this? The pliable one if you do the tap test seems a bit flat/ dull while the other one has a bit of ring to it.
Most people mention honeycomb core and a thin film/paper covering it tightly. Honey comb is used to make the panel stiffer, like cardboard. But, the honeycomb panel is used as the DML panel, the honeycomb, not only makes it stiffer, but there are empty spaces in it. Maybe. Göbel considered that and made that perforation in that balsa sheet. What do you think?
Oh, yeah. You could do the same with the proplex or corflute too if you want to reduce the anisotropy. As long as it doesn't get too stiff. Probably you would start with one of the thinner versions in this case.
Eric
I just rememberd one more input / inspiration that I had when building my first version of my flat speaker. It’s a patent by
BENDINGWAVELOUDSPEAKER+
(75) Inventor: HenryAzima,Cambridge(GB)
you might allready have covered this in the thread, I haven’t checked. My apologies in that case.
It’s about moving from a finite panel to infinite panel. My understanding that its very much in line with the göbel design.
It’s dealing with fact that the incident wave induced in the panel is reflected back from the edges with more or less the same energy. You can see the effect in some of the graphs In the patent.
I’ve been struggling with this problem in another bending wave concept ”the walsh speaker” there it’s very clear that the reflected wave is creating distorsion. The walsh speaker is actually working in both modes bending wave and piston As a panel. But the design is a cone instead of flat. I’ve included a picture and the patent for reference.
I’ve also tested to create a border with a stepwise change of the impedence making it more effective to capture the incident wave at the edge.
It’s the huge difference in impedence between air and the panel that causes the reflection.
BENDINGWAVELOUDSPEAKER+
(75) Inventor: HenryAzima,Cambridge(GB)
you might allready have covered this in the thread, I haven’t checked. My apologies in that case.
It’s about moving from a finite panel to infinite panel. My understanding that its very much in line with the göbel design.
It’s dealing with fact that the incident wave induced in the panel is reflected back from the edges with more or less the same energy. You can see the effect in some of the graphs In the patent.
I’ve been struggling with this problem in another bending wave concept ”the walsh speaker” there it’s very clear that the reflected wave is creating distorsion. The walsh speaker is actually working in both modes bending wave and piston As a panel. But the design is a cone instead of flat. I’ve included a picture and the patent for reference.
I’ve also tested to create a border with a stepwise change of the impedence making it more effective to capture the incident wave at the edge.
It’s the huge difference in impedence between air and the panel that causes the reflection.
No, I'm pretty sure that's not why the perforations are there (assuming they actually are there). Balsa (especially end grain) is really a lot like honeycomb already (light with decent shear stiffness). In the Goebel panel they aren't really even holes in the end, but rather filled with epoxy.
Just to clarify an important point, that may or may not be clear to all. It's not exactly true that the honeycomb makes it stiffer. It is kind of true, but not exactly. What I mean is this, what makes a stiff sandwich structure is (a) stiff skin layers that are (b) separated by a significant distance and (c) both skins attached to a core between them that resists shearing across it's thickness. Shearing deformation is depicted here:
So a honeycomb does make for a stiff composite in that it separates the stiff paper or carbon fiber or fiberglass layers and has just enough shear stiffness.
But what honeycomb really does is make the stiff sandwich panel lightweight. Any rigid core material between the skins would make a sandwich panel that's as stiff (or stiffer) than the same panel with a honeycomb core of the same thckness. But the panel with the honeycomb core would be much lighter.
Eric
Thomas,
I'd be very interested to hear the details of your stepwise impedence change. I've been thinking along those lines too but have not tried anything yet.
Eric
Honeycomb makes the composite panel stiff enough to carry a good load on its surface, while being very thin, which is its main asset. But, when you make certain enclosed, or semi enclosed thin walled vertical structures with those composites, they'd carry lot of weight. In our case, the honeycomb panel is fixed, one way or another on the perimeter, shear deformation will not happen. The exciter is actually acts on the stiffer surface perpendicularly. The so-called transverse bending waves are so tiny that they won't break the sheet.
I don’t understand this. AFAIK, a stepwise impedance change is what causes reflections in the first place. A smooth impedance change allows minimal reflection, no? This is the principle of impedance matching in electric theory or transmission line theory - to prevent reflections. And principle used in acoustic black hole edge treatment.
Or are you talking about something akin to a quarter-wave optical coating?