Acoustic Horn Design – The Easy Way (Ath4)

It would be interesting to compare DI of square "AthABECDemo" with any version of pure axisymmetric OS horn. If you have ABEC3 project of axisymmetric OS, you can upload it I will calculate "full sphere" DI.
I have plenty of those :) That would be great. Attached are some designs (full ABEC projects) I would be especially interested in -
 

Attachments

  • DI_study.zip
    226.6 KB · Views: 35
Last edited:

Mazza

Member
Paid Member
2013-07-18 2:55 pm
I'm compelled to say that, whilst we are way off topic here, I am simply dazzled by the wonderful contributions all are making here and the plentiful facts presented to both substantiate the claims and illuminate the novice, like me!

Thanks to all!

Mabat - more jewel like waveguide images and polars please!


Cheers :cheers:
 
The curved baffle is undoubtedly effective and the final design probably required quite a few simulation hours.

I'm not 100% certain, but I *think* that a baffle that progressively curves backwards produces a polar response that's similar to a horn.

In a horn, we get a beamwidth that gets progressively wider because the angle of the walls gets progressively wider.

A baffle that's curved does something similar I think. As the wavelengths get longer and longer, they're radiating into a beamwidth that's wider and wider.

The downside to a baffle that's curved backwards is that you're throwing away efficiency. The LS50 is kinda infamous for being strained at midbass and low frequencies.
 
In a horn, we get a beamwidth that gets progressively wider because the angle of the walls gets progressively wider.
I'm not sure I would put it that way. I would say that the beamwidth gets wider as the wavelength gets longer so the waveguide itself cannot affect it anymore. It is the same for the baffle (or a box for that matter).

BTW, the mentioned KEF doesn't look like a particularly controlled directivity design to me...
 

bmc0

Member
2019-08-26 2:09 am
The real effect cannot be determined without a sharp-edged comparison cab, true.
However, basic acoustics implies that towards the low end of the spectrum the wavelengths won't even 'see' the roundovers of your cabs (1000 Hz = 13.56 Inches).


True, not much high frequency energy makes it to the edges. There may still be some measurable effect far off axis though since the direct sound would be highly attenuated and the diffracted energy would not be. I doubt the difference would be audible, however. The roundovers just seemed like Good Engineering™ :D.
 

bmc0

Member
2019-08-26 2:09 am
BTW, the mentioned KEF doesn't look like a particularly controlled directivity design to me...

Here's some more data on the LS50 (from audiosciencereview.com, but graphed by me 'cause I think my graphs look nicer ;)). Personally, I think Kef could've done better.
 

Attachments

  • directivity_contour_h.png
    directivity_contour_h.png
    511.6 KB · Views: 223
  • directivity_h_pos.png
    directivity_h_pos.png
    352 KB · Views: 219
  • directivity.png
    directivity.png
    174.6 KB · Views: 250
Last edited:
I'm not sure I would put it that way. I would say that the beamwidth gets wider as the wavelength gets longer so the waveguide itself cannot affect it anymore. It is the same for the baffle (or a box for that matter).

BTW, the mentioned KEF doesn't look like a particularly controlled directivity design to me...

KEF%20LS50%201.jpg


KEF%20LS50%20V%20Contour%20Plot.png


KEF%20LS50%20H%20Contour%20Plot.png


Here's the LS50 and it's vertical and horizontal polars

GedLee%20Nathan%201.jpg


GedLee%20Nathan%20V%20Front%20Contour%20Plot.png


GedLee%20Nathan%20H%20Front%20Contour%20Plot.png


Here's the Gedlee Nathan and it's vertical and horizontal polars

The Kef isn't perfect, but it's good. And it's also small, affordable, and it's radiation is largely symmetrical. Not the greatest speaker in the world, but for the money it's pretty great.

I think a challenge with a two-way waveguide speaker, like the Gedlee Nathan, is that it's hard to get the vertical polars to be well behaved, due to the center-to-center distance.

Data courtesy of Princeton University : Index of /3D3A/Directivity
 

bmc0

Member
2019-08-26 2:09 am
Interesting that the LS50 vertical polars are much better behaved than the horizontal


Eh, I don't know if the vertical is much better behaved. There's less of a diffraction artifact (curved feature 700Hz-2kHz, 60°-140°), probably because of the baffle curvature. I've attached the same data, presented with a perceptually uniform color map.
 

Attachments

  • directivity_contour_h.png
    directivity_contour_h.png
    511.6 KB · Views: 558
  • directivity_contour_v.png
    directivity_contour_v.png
    500 KB · Views: 445
Talking about loudspeakers, these are coming soon :)

It's an all-passive three-way with 1,4" CD, 12" midrange and 15" woofer inside a 4th order band-pass (that's the port under the midrange). Initially there was supposed to be a passive radiator but there's no off-the-shelf suitable - it would be too much effort.

attachment.php
 

Attachments

  • T-skica-02.png
    T-skica-02.png
    12.1 KB · Views: 105
  • T-skica-0.png
    T-skica-0.png
    12 KB · Views: 102
  • Tritonia+.png
    Tritonia+.png
    62.9 KB · Views: 1,144
So, I have my first free-standing simulation ever :) I took the 12" demo OSWG.

Infinite baffle:

attachment.php


Free-standing with 5 mm wall thickness (nothing but the waveguide):

attachment.php


Acoustic impedance of the free-standing one:
(And no, I haven't mixed that up, this is the free-standing waveguide.)

attachment.php
 

Attachments

  • os-0.png
    os-0.png
    22.4 KB · Views: 647
  • OS-demo-free-standing-imp.png
    OS-demo-free-standing-imp.png
    15.1 KB · Views: 638
  • OS-demo-free-standing-polars.png
    OS-demo-free-standing-polars.png
    22.2 KB · Views: 639
Last edited: