Wouldn't it be more reasonable to first study the existing theoretical study on compression chamber resonances ?
I don't know, maybe. As I rememeber it, the phase plug openings should be placed as kessito kindly explanied long time ago in this thread (#1898, etc.). This was based solely on the positions of the nodes, also without considering anything else at that point, like bending of the diaphragm or a particular diaphragm/voice coil/gap/surround. So what do I miss when I want to measure these things in real to see how much is the simple simulation off?
Also - Acoustic Horn Design – The Easy Way (Ath4)
Also - Acoustic Horn Design – The Easy Way (Ath4)
Last edited:
Here are some images from the study of Jack Oclee-Brown and Mark Dodd demonstrating an effect of coupling of mechanical and acoustical modes.
Last edited:
It's Anycubic Photon Mono.
Dmitrij_S - so what should I do?
I have an existing compression driver for which I would like to design a new phase plug producing a spherical wavefront, suitable for driving a conical waveguide - that's my goal. Where would you start in that situation?
Dmitrij_S - so what should I do?
I have an existing compression driver for which I would like to design a new phase plug producing a spherical wavefront, suitable for driving a conical waveguide - that's my goal. Where would you start in that situation?
Last edited:
In my opinion, the simplest option is to create a driver 3D model and perform its multiphysics analysis. Nowadays, this is a purely engineering task that does not require any deep knowledge of the methods of theoretical physics and acoustics. Comsol is perfect for this purposes, but it is not free. Although there are also free packages like Elmer FEM.
As far as I know, ABEC3 has option to use some kind of vibration data files (from Klippel laser scanner) to account non-rigid behaviour of the membrane. I suppose one could try to model non-rigid motion of the membrane in any free FEM package and then generate vibration data file accepted by ABEC3 ?
Very true, but "not free" is a gross understatement. These packages can be tens of thousands of dollars. The "free" ones seldom have enough capability to do acoustics problems like ours. Comsol is ideal, but NOT cheap!
I agree with Mbat, he needs a practical way to measure an existing driver.
But what about all the material properties you would have to guess?
If the vibration data is what is needed, I think it's far more straightforward to measure it as the driver is already available. Will try that.As far as I know, ABEC3 has option to use some kind of vibration data files (from Klippel laser scanner) to account non-rigid behaviour of the membrane. I suppose one could try to model non-rigid motion of the membrane in any free FEM package and then generate vibration data file accepted by ABEC3 ?
What if we could do something like this but on te audible spectrum
Visualizing Ultrasound with Schlieren Optics Part II - YouTube
Visualizing Ultrasound with Schlieren Optics Part II - YouTube
What you are looking for are small aberrations of the wave dynamics - they are going to be hard to see by eye.
- Introducing a new series: ATH - Cutting Edge Series
(For all those who contacted me recently about providing a custom profile, this is all I can offer at the moment.)
(For all those who contacted me recently about providing a custom profile, this is all I can offer at the moment.)
Those horns look pretty good!
I was contemplating on the idea of making a horn out of cork and while thinking about it tried simulating 2 versions of the same horn contour.
One is 100% reflective, the other one with 0.6 absorption from 1k-20k and rolling off below 1k.
The results are kind of intuitive, but made me think about making a horn that is half way plywood and half way cork.
I could however not successfully simulate a 50/50 horn.
As soon as I split the nodes into 2 groups, the result gets weird.
Rigid
View attachment 968469
Cork
View attachment 968470
I was contemplating on the idea of making a horn out of cork and while thinking about it tried simulating 2 versions of the same horn contour.
One is 100% reflective, the other one with 0.6 absorption from 1k-20k and rolling off below 1k.
The results are kind of intuitive, but made me think about making a horn that is half way plywood and half way cork.
I could however not successfully simulate a 50/50 horn.
As soon as I split the nodes into 2 groups, the result gets weird.
Rigid
View attachment 968469
Cork
View attachment 968470
Last edited:
A short survey -
If I were to run a crowdfunding project to make an injection moulded free standing waveguides (basically to collect resources for making a steel mould, which is on the order of thousands of USD, and then let produce a batch), what waveguide size, if any, would you prefer? CE360, perhaps, or something bigger?
If I were to run a crowdfunding project to make an injection moulded free standing waveguides (basically to collect resources for making a steel mould, which is on the order of thousands of USD, and then let produce a batch), what waveguide size, if any, would you prefer? CE360, perhaps, or something bigger?
Last edited: