Acoustic Horn Design – The Easy Way (Ath4)

Oh, I see. BTW, this is the throat impedance of the WG used. So is there a chance that in a waveguide that has a higher throat impedance to a lower frequency, the dip between 1 - 2 kHz won't be as pronounced?



  • SWG33-throat-impedance.PNG
    15.4 KB · Views: 2,452
AS you might expect, the coupling of the elements around resonance can get complex and unintuitive. Models are usually required.

But at these L(ish)Fs the waveguides impedance looks like a simple mass. The dual humps in the electrical impedance curve are due to the coupled oscillator defined by the gap compliance and the two masses - waveguide and diaphragm. At the lower impedance you have the two masses in phase and the above peak they are out of phase, i.e. 180 degree phase shift. This coupling makes the whole problem much more complex.
This evening my friend Jokerbre and i measured Peerless DFM2535R00-08 on Mabat's ST260 waveguide.

Measurements are done from 1m distance with calibrated Behringer ECM8000, Steinberg UR12 and gate frequency of 4.4ms so enough to capture trends of this waveguide. What is shown here are measurents 0-90 degrees with ≈7.5 deg increments in between. Since i did not calibrate REW and my mic for loudness, SPL levels shown aren't representative of actual SPL level. Off axis measurements 60 degrees and up are smoothed 1/48 to avoid "hair" at higher frequencies.

0-90 degrees
0-90 степени.jpg

0-45 degress
0-45 степени.jpg

45-90 degrees
45-90 степени.jpg

75dB vs 85dB loudness
75 85.jpg

measurement on axis (red) and averaged 0-90 degrees (purple)
мерење на оси и п&.jpg

printed waveguide
звуковод напре&#10.jpg звуковод назад.jpg

I must say that i've never seen better and more extended frequency response from a smaller waveguide and a cheaper compression driver.
Last edited:
Maybe this could be a way of printing it with minimal warping.
For even larger horns it could the the throat section + "petal" pieces of the rest, printed one by one diagonally across the print bed. Could be tedious to put it all together, that's the drawback.


  • ST260-pieces.PNG
    130 KB · Views: 363
  • ST260-throat-bed.PNG
    69.3 KB · Views: 623
  • ST260-mouth-bed.PNG
    59.2 KB · Views: 302
Last edited:
What is the maximum dimensions this way? My printer is tiny . . .
As even the smaller printers can do 250 - 270 mm diagonally across the bed, I guess it could be pretty big, actually, like 15" at least.
A good thing about this approach would be that a one failed print doesn't mean doing it all again.

An actual object optimized for this approach could be made a lot more sturdy/massive, of course, especially around the main joint. But it wouldn't be particularly cheap either if made in a print shop, I'm afraid. Someone would have to try.
Last edited: