Acoustic Horn Design – The Easy Way (Ath4)

Hello again, a long time inactive with many life changes.

I have a question for the experts. I printed this Tritonia. After almost a year of not printing in 3D, I used the piece as an experiment to configure the machine. I didn't think it would turn out with what I started with, scrap PETG filament. The thing is, after 19 hours, the piece turned out. It's not perfect, but nothing a little filler can't fix.

IMG_20250603_124436.jpg


I intend to solidify it. I left one side open in some places to pour the filler. I had a bad experience with epoxy resin with some filler. So the question is, what method or type of material have you used successfully?

IMG_20250604_110302.jpg


The next question, since I haven't printed the adapter yet: Can I take the measurements empty, or should I make it solid first? Will there be much of a difference? Perhaps it's better to measure both and compare...

IMG_20250604_1101162.jpg


As a good carpenter, I still don't understand anything about this acoustic world.

Thanks, Marcel, for your inspiration.
 
That's nice. Is it a somehow modified Tritonia-F? I woudn't have thought about print it this way, but it seems it's quite possible without issues.
As for any additional strenghtening or dampening, it's really up to you 🙂 After mounting it to an enclosure, I don't think it will resonate much. You can certainly measure it as it is. I would only give it some coating first.
 
I used a Tritonia configuration you posted earlier as a base, then made some modifications and exported the curves to Fusion (what a great job you've done, it works so well). Then, in Fusion, I created the shape and motor adapter.

Printing it vertically wasn't without its problems, one with warping and another with inertia, which caused the supports to come loose from the bed (a little superglue and off you go).

To try to avoid warping, I designed an elephant's foot to give it better adhesion, with relative success (PETG has a lot of shrinkage). If the measurements are correct for next printing, I'll modify the design to give it better stability and adhesion, or perhaps try another, less critical material.

IMG_20250603_124455.jpg
 
Anyway, it's available on Cults.
(STEP file included - this one can be easily scaled for a 1.4" throat, then it will be 590 x 510 mm, this must be enough.)

And now I need a break from designing horns... 🙂

Does the Cults download include the Ath script? This looks like it could be a good fit for a build I'm planning out for a TD-4002. Would be CNC'd from wood.
 
  • Like
Reactions: N173MM and JoshK
I showed that this was not true in a publication that I did for AES many many years ago. Statistically (STD about the mean) a dipole is not any different than monopoles, they just prefer to be in different places. Optimally placed they both have the same STD of the SPL, but the monopoles is much higher as the frequency falls.
That feels like a bit of an oversimplification. Dipoles are significantly more directional than monopoles — even at low frequencies — with a directivity index around 4.8 dB versus 0 dB.

As a result, they inject approximately 67% less acoustic power into the room’s modal field.

Monopoles can be optimized, sure, but dipoles inherently reduce modal excitation by design, not just by placement or DSP.
 
Spheres aren't cheap. Why not just stretch the epoxy with sand, way cheaper just as strong.

Because there’s been studies on the acoustic effects of microspheres showing positive results for dampening of unwanted resonances and also a decent weight reduction. The cost is not that significant really. I found a 5 quart tub of the stuff for 35$ on amazon. It is recommended to mix up to 40%.
 
  • Like
Reactions: N173MM
Don’t miss cork, a lot cheaper than microspheres unless you buy it in bulk.

I always reference this video to understand what different fillers add to the mix.

And I probably already showed this before.
But I made a series of tests using dry filler in a vinyl hose and added the epoxy from the top to evaluate different epoxies penetration and hardening time and so on.

Infusion epoxy and slow hardener from https://www.easycomposites.eu/in2-epoxy-infusion-resin together with sand with grain size 0,4-0,8mm works really good. And the big thermal mass of sand is beneficial.

1749189751224.jpeg


1749189778514.jpeg


It’s a lot easier than filling a slurry thru a small hole.

Another innovation I tried is adding a separating wall with micro perforation. Sand goes in one compartment and epoxy in the other smaller one.
1749190156149.jpeg

1749190178964.jpeg

1749190199895.jpeg
 
I'm assembling the T520 Rosso adapter for A460G and confused about the 'Part 2' piece & how this comes together — If I insert part 2 where the diagram shows, then it hits the driver and doesn't sit flush. If it's not inserted then there's a small gap between driver & adapter. What's the function of this piece? What am I missing?
 
No work at all with the epoxy I linked to.

Fill the volume with cork tap it so it settles and fill the empty space as much as possible.
Ad epoxy, wait, ad more as it flows down and displaces the air.
The type of epoxy used is the key for it to work. And it's a faster process if you include a channel to the lowest point of the volume you like to fill. So the epoxy is filled from the bottom up.
1749233833999.jpeg

You might have to top off the highest point after a night of it settling, but this isn’t a huge deal.
1749234435008.jpeg


Tapping the final product gives a clear indication that it has high internal damping.

Edit: https://www.carbon-connected.de/Gro...ile/Download/CCF94E8FE2F72A4CAB8EAEB1D7EA2708
Only one example of research in using cork to increase the internal damping. There’s plenty if you’re interested in a deep dive.
 
Last edited: