improving bending wave drivers

Jump to Latest
#81
The airfoil.How it works http://sellhigh.com/impact/products/prodairfoil.html
http://sellhigh.com/impact/products/prodairfoil.html

1735292438321.png


Bernt "båndsei"
 
#82
ondesx said:
@ruba1 :
perhaps I misspoke. What I meant was that I'm still working on this type of magnetic circuit, but no longer using the method shown in this photo for example, the one that most people have been exploring for years, if not for almost a century now. Paddock, for example, was one of those who experimented extensively with this design. In fact, back in 2002, when Airfoils were first marketed, Stereophile published an interesting article: https://www.stereophile.com/floorloudspeakers/648/index.html.

If you look closely at this "driver", you can see that there are no "membranes" as such, but only surrounds...
Of course, this is just my personal view, and that's why I told you that I'm continuing to work on this geometry, but to build a "finalized" loudspeaker!
As for the method of relying solely on your ear to make a transducer, I don't think it's reliable enough. Again it's my personal point of view, but of course I completely respect your way of looking at things.
Click to expand...
Yes, the name membrane is probably inaccurate. I would call it a bending wave membrane. Surround would be a bit confusing. With measurements you will certainly get to the goal faster. Then with hearing, the fine adjustment. Monteverdi, hopefully you didn't fight the manger msw with sheep's wool, as is often done in the DIY area. For my rubanoide I only use a bow, for the "membrane" and not as a dipole. Greetings
 
#83
Membrane is defined as sheet separating 2 mediums acting as a selective barrier like cell membrane or filter membrane. Officially loudspeakers have a diaphragm but most of time it is called membrane, no matter if it is stiff like in pistonic speakers or flexible and independent of exact shape.

I listen to an airfoil speaker decades ago and I thought that rubanoide is a significant improvement especially the Audionec ones I heard at the high end in Munich (not so much my short diy attempts). The main issue with that design that it can not reach the last octave and therefore needs the addition of a high-frequency driver. Rubanoide have damping in the longitudinal direction but not in the top and bottom so they act only in the longitudinal direction similar to bending wave drivers. what would happen if the top and bottom of the cylinder would be damped
 
Last edited:
#84
My rubanoid doesn't reach the high frequencies. But in the photo with the measured values, you can see that up to 20 kHz is possible. What I still don't understand is that you write how difficult it is to prevent reflections at the edge of the membrane. Why is the damping material so narrow? I estimate about 8 mm.
 
#85
Balsa wood is more stable when bent in the longitudinal direction of the grain (parallel to the grain) than in the end grain direction (perpendicular to the grain). The cell structure of the wood, with rectangular and narrow medullary rays, ensures high rigidity in the longitudinal direction, while the greater elasticity of the honeycomb structure in the end grain area leads to lower bending strength[6][7].
 
#86
Wood and more extreme balsa wood is anisotropic in longitudinal cuts. They have a stiffer direction (length of the tree, parallel to the grain) )but are easily to bend perpendicular to the longitudinal direction. An end grain disk is in 2 axis (x&y) close to isotropic with low torsional strength but with quite high compressive strength in the z axis (thickness). The biggest force on a tree is its own weight therefore compressive strength is highest in that direction. So balsa wood by its self would be a poor choice for a membrane but if it is used as a core in a composite structure which has in the x and y direction high tensile strength fibers it generates a very stiff, strong and light membrane.
A rubanoid has an anisotropic structure, it is much stiffer in the longitudinal direction and is designed to bend perpendicular to it so an anisotropic membrane could work. My bending wave design is basically a stiff plate/membrane which is extended by vibration absorbing material. Any experiments to add damping material on the stiff membrane have no effects (like moon gel). What works and is essential is extending the membrane on its rim with vibration absorbing materials.
 
#87
I have to admit that I don't understand the theory of the bending wave membrane well enough. With the rubanoid with paper membrane, printer paper, the direction of the fibers is not important. The paint is the dominant feature. Göbel uses balsa end grain, just like you. The compressive strength is much higher than with longitudinal fibers. But longitudinal fibers increase the stiffness in the longitudinal direction. Isn't that useful? What I'm also interested in: in your opinion, your converter is better than the Manger. What advantages do you hear compared to the Manger? Google translate isn't particularly good
 
#88
The rubanoid design is a cylindrical structure and is basically designed to vibrate perpendicular to the longitude. So the membrane does not need to be isotropic and one can use machine made paper (which has directionality in contrast to hand made paper) or longitudinal cut balsa sheets. My version of a bending wave driver the membrane is a stiff plate/sheet (all the early bending wave divers were based on stiff plates before Mangers version) and having a isotopic structure is a advantage as the plate can vibrate in any direction equally. I contrast to Göbel I avoid corners.
One issue with Mangers design is that the voice coil is concentric and that causes interference patterns leading to the uneven frequency response. Read some of the discussion in https://www.diyaudio.com/community/threads/manger-msw-moongel.208995/. The reason I am using a rectangular motor is that the interference pattern is very minimized as the distances (run length) between voice coil and perimeter are varied as much as possible. One thing I forgot to mention is that I position to voice coil not in the exact center line of the disc but slightly offset by a few mm to vary the distances between left and right side (and also top and bottom).

Sound wise I see much improve intelligibility of voices but also much more differentiation in instruments, better sound stage compared to the Manger. With the MSW I had always issues that its uneven frequency response made certain notes under represented or exaggerated and these hight distortion peaks are quite noticeable.
 
  • Like
Reactions: Freedom666
#90
Monteverdi said:
For the voice coil itself I am using 33 or 35 AWG copper magnet wire with a winding height of 4-6mm generating between 4 to 8 Ω DC resistance.
Click to expand...
Just a note: the video above is not about a rubanoide, but rather about a flattened coil with the magnets on one side of it, which were later positioned behind it. It's more about a rectangular coil, as shown in your post, but flattened and glued flat onto the thin membrane. Two (or more) neodymium magnets placed a few millimetres just behind that coil would effectively function as the motor.

Even with that Manger panel, if we simply flattened that round coil within the central circular area and used a round magnet, it would also function as a sound diffuser DM panel. The only issue would be the high flexibility of the Manger panel.
 
#91
I understand, it is a hybrid between a pistonic driver and a magnetostatic driver. A magnet in by itself has a very weak magnetic field as it radiates in all directions. That is why I made these complicated magnetic structures and look at the contributions showing FEMM.
I tried a flat Halbach array (alternating S/N/S) with the voice coil direct on my bending wave membrane and it was extremely inefficient and had terrible measurements
 
Last edited:
#92
Monteverdi said:
I understand, it is a hybrid between a pistonic driver and a magnetostatic driver.
Click to expand...
One might say that he doesn't care much about "bending waves." This much older experiment is more akin to the Manger one, but you might notice that the panel is quite stiff. The handmade panel is so stiff that it won't bend at all.
The alternating S/N/S method has been experimented with a few times, I'm sure. Hope you'd succeed in yours.

EDIT: By the way, this might be of interest to you.
 
Last edited:
You must log in or register to reply here.
Top Bottom