Acoustic Horn Design – The Easy Way (Ath4)

Hi audiokinesis

excellent approach. So then this makes the project quite easy. I will accept this (more theoretical) resonance issue then and just make a front insert. Sorry for the ms paint illustration ;-)
Will of course keep an extending cross section area. The integrator is then left untouched. Aiming for 1" makes the standard adapters useable 🙂
471f9ebd-3288-4b41-a33c-53d3204dc81f-1_all_32034.jpg
 
Is the black line the whole modification you're going to make? In that case I'd be sceptical.
I don't quite like the shape of the original air path, as the wavefront area seems to be shrinking along the way (at least in the middle section), but you would make it even worse, IMO, by definitely shrinking it even further.

As long as it's not an irreversible mod, it's probably always worth a try, but I'm sceptical. I think you would need to come up with a completely different "combiner", somewhere very deep in the driver - something close to a regular BMS phase plug perhaps. And start a narrow-throat horn at that point. But of course that would be a lot more demanding, as you would need to replace the whole central thorn.

- With the BMS 4554 I start the horn at ~17mm but the wavefront area is only increasing from there (I'm just about to publish the model).
 
Last edited:
  • Like
Reactions: Arez
Replace the central thorn was the orignal idea. Indeed I missed that the diameter and therefore cross section surface decreases. So a insert only will not work. I have already scanned the thorn so will check what is possible at the beginning in combiner section. But I am scared by the necessary precision. Distance thorn-membrane is a few 1/10mm...
 
I guess you are considering 3D printing. The dimensions can be iterated until it's good enough as long as it's repeatable, which it typically is - the accuracy errors of 3D prints tend to be more systematic than random, I think. I'd be more scared of the design itself but I can imagine it could be a lot of fun.
 
Perhaps there's some reason, but it's not quite clear to me why they don't make both voice coils the same diameter. The diaphragms could be different but the outlets from the compression chambers could then simply face each other, being closely packed together, making the "dead" volume for the HF waves as small as possible. I have yet to see a coaxial driver with the HF section as good as I would actually like. The mids are typically very good.
 
Last edited:
Yep output from the tweeter diaphragm / slit would propagate also toward the mid ring slot and reflect there and interfere. Sound propagates to all directions.

I think size of the hf diaphragm needs to be smaller due to all kinds of resonances and weight and modes happening and what not. Conversely mids could be smaller, but it's volume displacement would suffer and other compromises would happen. This is and always will be a compromise, because high frequencies wavelength is smaller than mids so the transducers must be physically different (size), and coaxial or not must necessarily be in physically proximity to each other affecting each other acoustically, and it will never be optimal as long as the transducers have physical dimensions to them, as long as they are physical objects.
 
Ok. I see I do not earn encouraging response to the basic idea. Well ....I have another approach to discuss.

I have dismantled a BMS4591. Very easy to use this as a midrange in a MEH. Just a new "channel distribution" is necessary. Left is a 4551 then this original channel unit and then right the magnet and diaphragm assembly. The midrange channels could be inserted right after the 4551 outlet (so in mabats design this is maybe within first 3cm of the horn with 17mm throat).

What do you think?
IMG_20241209_134712.jpg
IMG_20241209_134702.jpg
IMG_20241209_134356.jpg
IMG_20241209_134353.jpg
IMG_20241209_134344.jpg
IMG_20241209_133819.jpg
IMG_20241209_133836.jpg
 
  • Like
Reactions: isitone
Conversely mids could be smaller, but it's volume displacement would suffer and other compromises would happen.
Or there's almost enough space for a second mid diaphragm that would be above the first one and mirrored, the two facing each other push-push, like in the JBL D2 / D2430K.

Speaking of which, I've been thinking that construction would lend itself nicely for electrostatically driven ring diaphragms. I believe that flex PCBs would work as stators, and that would be just like 0.1mm extra thickness on the walls of the chamber, and very flat chambers possible, and that overall DIY at home manufacturing would be quite possible, like getting the flex PCBs manufactured, cutting the film and assembly 3D printed jigs and parts, with fairly low material costs as well. (I think easier and cheaper than making a magnet based DIY driver.)

 
  • Like
Reactions: TNT