The best way to "see" it, is to calculate observation fields. This version is a bit Hyper cardioid at 4K.
AKABAK has the same options as ABEC, a demo mode or a student licence for strictly non commercial use. ABEC projects can be imported, the last one complained about a fatal error but still loaded when I clicked apply. I just had to set the source to be driven and choose the driving weight for constant acceleration which is divide by jw. The meshes can be imported directly and with a basic template saved with the orientation and observations set as you want them that is fairly quick.
The observation scripts don't import very well, but with the point and click options in AKABAK setting it up as you want is a little easier than remembering the text commands in the script. Of course if the pre prepared script runs just fine in ABEC then that is the easier option.
Oooooo! Thanks! I’m all over it like a monkey on a football!Oh, I wouldn't want to do that myself
You can just read the current report: https://at-horns.eu/release/R-OSSE Waveguide rev7.pdf
(It's addressing the axisymmetric horns only. It's possible to desing rectangular horns (made of boards) as well but there's no summary for that yet).
How do you do that? I tried an example a few pages back, but that didn't work.
The script I shared there https://www.diyaudio.com/community/...he-easy-way-ath4.338806/page-565#post-7228015
seems to work OK for me, I have tried to simplify but it did not go well at all...
I have the following set up
- Ath 4.9.0-pre-release-221114
- gmsh-4.6.0-Windows64
- ABEC 3.6.0 b7 - 64 Demo
- Windows 10 Pro (and have the issue with the projects sent by Mabat)
As I mentioned earlier, I am trying to start from a flat terminated WG and add a working termination with the enclosure.
I have number of shapes, but anything is probably possible. I have attached examples.
Just don't care about the WG/Enclosure mismatch, that is just to illustrate I have some much better designs.
The last one was very interesting, and related to what @mabat is implementing, with Besier curves and maybe even more so a ROSSE shaped enclosure. The flat WG might be correctly terminated by the ROSSE enclosure:
But as I said, the current limitations on the custom enclosure plan makes it difficult to perform the full optimization.
Just to give some ideas...
Attachments
I was trying with asymmetrical WG so hopefully a bit better, but you are not wrong...Yeah, but only at four points from infinitely many of them. That's not much
Now that you have gifted us with axisymmetric back, it might be a different story as maybe half of the WG could be correctly terminated a la Autotech Mummy http://hifiknights.com/reviews/speakers/horns-mummy-loudspeakers/
Edit: ROSSE is supposed to have low diffraction which is what we also want/need for the enclosure, isn't it?
At least this the starting point of my reasoning...
Last edited:
I remember that one, that was the first OS I ever "designed". The termination itself was still not so great and I only wonder if the enclosure helped. Probably not that much, as the curvature mismatch was already there.
It is very easy in AKABAK as there is a radiation Field Shell option which does most of the work for you.
Ath can generate the nodes or mesh for the fields but that would require resolving.
You can add another observation script into ABEC and solve the observation without having to resolve the BEM part.
Example script attached. You would need to change the identifier and drive group to reference the source in your script and play around with the node values to get the fields where you want them.
Then just solve the fields with that option.
Isn't this the same that mabat explained how to create from Ath here, an axial field?
https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-7186013
It's to be noted that via Ath this only works in a full 3D model (not in CircSym mode, where it's different and really easy to add manually at an observation stage). The 3D field surfaces generated by Ath/Gmsh are actually regular parts of the mesh (MSH) file. They are also configurable regarding their variable resolution - typically, the further away from the source, the lower resolution you need.
Last edited:
My kids brought home some kind of foam clay, air dry and very very light, it weights almost nothing. It is strong enough to preserve shape, but flexes a bit when pushed and I guess it must be very well damped. Is it a bad idea to make a free standing horn from something like that? Unfortunately I don't know what it was, maybe something like this.
As free standing, attached to a driver, it holds only itself so maybe the lighter the better (?).
As free standing, attached to a driver, it holds only itself so maybe the lighter the better (?).
Last edited:
The totem loudspeaker as something Morel made à la Vivid Audio are very cool to look at in plus of being usefull for difraction control
I had to cancel my one piece ST260 at Hong Kong JLPCB that has a 3D printing service : here their advice for that sizes of WG, if can be usefull for garage printing for anyone, I copy here what they told me as the stl fil can not be updated by them (I chose the 8000 resin type that are strong and has a white mat surface, and not too much expensive) :
"For the part DST260-19-4holes.stl, the models could be shrinking to the center and warping to the diagonal lines and edges.We kindly suggest increasing the wall thickness to at least 3.5m to proceed and adding more ribs and rails along the side surface to strengthen the whole structure, which will be much better, but we can’t fully avoid the risk of the deformation due to the construction."
I had to cancel my one piece ST260 at Hong Kong JLPCB that has a 3D printing service : here their advice for that sizes of WG, if can be usefull for garage printing for anyone, I copy here what they told me as the stl fil can not be updated by them (I chose the 8000 resin type that are strong and has a white mat surface, and not too much expensive) :
"For the part DST260-19-4holes.stl, the models could be shrinking to the center and warping to the diagonal lines and edges.We kindly suggest increasing the wall thickness to at least 3.5m to proceed and adding more ribs and rails along the side surface to strengthen the whole structure, which will be much better, but we can’t fully avoid the risk of the deformation due to the construction."
I thought it has 5 mm wall already, maybe not. I (or anyone else) can prepare a more serious model, the one at my website is still pretty rudimentary for people who want to try themselves (there's a STEP file as well). If all goes fine I'll have the jagged-egde version soon as well.
If I was to offer the waveguides myself (for money), they would look different...
If I was to offer the waveguides myself (for money), they would look different...
Last edited:
...I already know how to do it, now I have to come up with something fancy enough
This is boring -