Acoustic Horn Design – The Easy Way (Ath4)

Well done.

In any case, it allows for super fast iteration and testing, which is exciting.
To me, "testing" would mean either a simulation or measuring a prototype. How fast can you do that?
If you managed to incorporate a fast acoustic simulation into the Grasshopper loop, you would be king 🙂

[...] with the segments being able to be pleated or shaped (if that's a good thing or not I literally have no idea - its just geometry to me)
It seldom brings any real (acoustical) advantage, honestly. But it looks cool.
 
Last edited:
The long shape of the 4554 extended throat makes it basicalle impossible to insert a midrange within 1/4 wavelength. Within this region the horn expansion is still very small.

What happens if they are NOT placed within 1/4 wavelength of XO frequency? Active crossed with matched delay of course. For passive the reasons are obvious why this should be avoided.

Some reflection will happen but thr CD also still supports a bit....
 
"negative expansion" - I doubt a bit that the area would actually decrease. I found a drawing here - is it this?

View attachment 1395335 View attachment 1395340

It's hard to tell what the wavefront area expansion is. One could at least calculate the total area of the input and the output quite easily. Maybe it's a linear change in between then (in this case), I'm not sure.
Yes, that drawing looks right. And everything considered it's probably not negative now I see. Although how to calculate!
I suppose it might all come apart - but not from the exit side - would need all dismantling. And I kinda decided to sell them...
 
I checked before for xo 700hz based on your adapter. You see....not much expansion there. Ignore the 90mm..
 

Attachments

  • IMG-20241215-WA0141.jpg
    IMG-20241215-WA0141.jpg
    376 KB · Views: 74
My opinion is that the vertical issues associated with noncoincident sources crossed around 600 Hz are actually no big issues at all. I would not sacrifice anything above 600 Hz just to improve the small region around 600 Hz. It just makes no sense to me.

I don't know of any solution that would only improve everything even further.
 
Last edited:
  • Like
Reactions: isitone
I checked before for xo 700hz based on your adapter. You see....not much expansion there. Ignore the 90mm..
You mean little expansion at 120mm length?
In post #16270 I've sketched a 2:1 combiner for the 5530ND, without expansion because that just easy to draw by hand. The extension length outside of the driver was 67.048mm. Expansion could start after that.

Later I fooled around simming a loft that is circle-triangle-circle instead circle-semicircle-circle and the one with a triangle had a better looking response 🙂
 
The long shape of the 4554 extended throat makes it basicalle impossible to insert a midrange within 1/4 wavelength. Within this region the horn expansion is still very small.

What happens if they are NOT placed within 1/4 wavelength of XO frequency? Active crossed with matched delay of course. For passive the reasons are obvious why this should be avoided.
Assuming the side mounted low (midrange..) cone drivers output output is time aligned with delay to the high frequency compression driver, the on axis distance being more than 1/4 wavelength apart is not an issue.

That said, if the low entrance ports are more than 1/4 wavelength apart from each other horizontally, off axis response will have dips and peaks not conforming to the on axis response.