Acoustic Horn Design – The Easy Way (Ath4)

Jump to Latest
#16,521
I already have a 4552 adapter optimized - the initial throat diameter has been reduced a tiny bit, as it still seems possible, otherwise it's basically the same as for the 4554.

4552-pmap-360.png
 
#16,523
@matbat I bought and downloaded the Gen 2 Horn as well as the 150mm throat adapter to try to understand what was going on in that transition. It appears to me that in order to make a smooth transition from the 0 degree exit at the driver mouth to the horn you've established a smooth curve that actually necks down to a diameter less than the CD exit - approximately 28.8mm vs. the 36mm. Is this reduction doing something to the loading of the CD, or is this just to make the transition from the 53.7mm horn throat opening to the 36mm CD opening smooth? I.e., if you had infinite control, would something like what I've drawn below be better? To explain - I took the curvature of the horn mouth and moved it, which expands the overall diameter of the horn from 520mm to 571mm, and then established a nurbs curve that matches the exit of the CD (using 8 degrees for the example for a B+C driver) to the tangent curvature of the throat of the horn itself. See the attached image:
 

Attachments

  • ATH Gen 2 Modified to B+C Driver.jpg
    ATH Gen 2 Modified to B+C Driver.jpg
    172.4 KB · Views: 119
#16,524
With the ath tool you can simulate all of this yourself. This is the profile you drew, approximted by eye as a single R-OSSE curve.
As I said, this adapter didn't work very well with a driver that it was tested with. It caused a too big efficiency drop on the lower end of the passband. I don't know what else to say, it was just an idea to try (and it may work better with a different driver, that's why I kept it there).

https://www.desmos.com/calculator/vzgnllxfqm

1734875235821.png


1734875264361.png


And this is the ath sript:
Code: 
R-OSSE = {
  R = 285
  r0 = 18
  a = 28
  a0 = 8
  k = 14
  m = 0.83
  r = 0.47
  b = 0.15
  q = 3.9
}

Mesh.LengthSegments = 120
Mesh.AngularSegments = 8
Mesh.SubdomainSlices =
Mesh.ThroatExtSegments = 24
Mesh.WallThickness = 6

ABEC.SimType = 2
ABEC.SimProfile = 0
ABEC.MeshFrequency = 30000

ABEC.NumFrequencies = 64
ABEC.Abscissa = 1
ABEC.f1 = 200
ABEC.f2 = 20000

ABEC.Polars:SPL = {
  MapAngleRange = 0,180,37
  NormAngle = 5
  Distance = 2
  Offset = 0
}

Report = {
  Title = "horn example"
  Width = 1600
  Height = 900
}
 
Last edited:
#16,526
The final verification of the 4554 adapter on my side. This is what you can expect when everything is fine.
You need to hear that 🙂

I did put some effort into sealing the mounting flange this time. In a design this clean, every detail matters.
I'm pretty curious how the frequency response below 1kHz will look like with the other BMS drivers.

Measured across the listening window:
1734882217862.png


4554-adapter-testc.jpg
 
Last edited:
  • Like
Reactions: DCDK, mjhara, aragorus and 7 others
#16,533
A completely different topic: I appreciate a lot what has been created within this thread. Unfortunately we have now 827 pages and it gets nearly Impossible to search through them in order to find some of the valuable examples by mabat and some other users.

Just an idea: chatgpt with scraper plugin seems to be able to search through websites and collect requested data. Has someone here maybe the paid version and is interested in trying this? A collection of examples would be great.... Even if there will be quite some outdated ones or some not anymore compatible with current ATH version.

What do you think?
 
You must log in or register to reply here.
Top Bottom