Omnidirectional Multiple-Entry Horn

[Slothrop]

This is my first post, after lurking for a while.
I'd like to bounce an idea off of your collective brains. I've been sketching an omnidirectional speaker design based on a multiple-entry horn (or coentrant horn, unity horn, synergy horn). Here is a rough drawing showing my thinking, such as it is:

Here are some 3D sketches of one possible embodiment of the horn:

I recently got a 3D printer, so I could finally make an actual horn for this; but I'd like to fix any possible boneheaded mistakes first.
Thanks for taking a look.

 

[andreasmaaan]

Hi and sorry your first post didn’t get more of a response initially

It’s a tough one.

My first thought is that your high frequencies are going to suffer going through that 90 degree bend in the horn. There will be diffraction and possibly loss. Any loss might be correctable with EQ, although the diffraction won’t. Whether it’s a worthwhile trade-off is another question.

My second concern would be with the “phase plug”. I can imagine that the small pocket of air it will create at the throat of the horn will act as a low-pass filter and that it will create a lot of unwanted diffraction at high frequencies. This is why horns for high-frequency compression drivers generally don't normally contain any phase plug external from that of the compression driver itself. I wonder if it wouldn’t be a better idea to omit your phase plug altogether, or to use a high frequency compression driver with a concentric phase plug and to extend your external phase plug right up to meet the centre circle of the compression driver’s phase plug.

The final potential issue I would be concerned about is that the horn will be highly directional in the vertical plane above a certain frequency determined by the wall angles and the height of the horn mouth, and then very omnidirectional below that frequency. So in the vertical plane there will be a very abrupt transition from highly directional to omnidirectional.

It’s a very novel idea though so all my thoughts here are quite speculative. The last concern is the one I’m most certain of, however.

Any further developments since your original post?

 

[HammerSandwich]

@andreasmaaan, the discussion's at diyAudio. Be sure you catch the last post!

 

[andreasmaaan]

@andreasmaaan, the discussion's at diyAudio. Be sure you catch the last post!

Oh course it is! Thx @HammerSandwich

 

[Slothrop]

Thanks for your replies!

I wish I had a better understand of how the shape of a horn (bends, angles, etc.) affects the sound going through it. I will try to find some information about diffraction.

I don't know whether 'phase plug' is the right term, but here are some pictures of the horn I based mine on (Klipsh K-85-K). The opening for the tweeter is a thin ring near the edge of the dome; I just copied this geometry:

The vertical HF dispersion issue was brought up on diyaudio as well, and I'm trying to find a solution. My first model of the horn has 15 degrees. My thinking was, the original horn was 90 degrees horizontally and 60 degrees vertically, so since I'm going to have four times the horizontal (360 degrees) I should have 1/4 the vertical (15 degrees) to keep roughly the same loading on the driver:

Now I see that this is too narrow. I made a drawing where the horn starts at 15 degrees, but opens up to 30 degrees after the bend; but others have told me this will not actually increase the vertical dispersion. I'd love to understand why that is.

Thanks again for taking the time to look at this crazy stuff.

 

[Jaimo]

Tom Danley is the go-to reference for multi driver horns. Check out his Synergy Horn.

I owned a pair of Yorkville U215’s for a while and while they were amazing on some respects, that design was never intended for home audio and had some severe limitations In my use case.

Still, I encourage you to continue your efforts. There are substantial rewards in this approach.

 

[Jaimo]

Your drawings BTW remind me of my HK Aura

 

[andreasmaaan]

I don't know whether 'phase plug' is the right term, but here are some pictures of the horn I based mine on (Klipsh K-85-K). The opening for the tweeter is a thin ring near the edge of the dome; I just copied this geometry:

Ok yes. In fact that is a concentric phase plug much like the type I was suggesting you use. You'll find that it's very difficult to get the phase plug close enough to the dome tweeter's diaphragm and to get the path lengths between each opening in the phase plug and the surface of the diaphram to be similar enough to make it function correctly, but it's not impossible. Beginning with a compression driver that uses a concentric phase plug may make your life easier - or maybe not

Now I see that this is too narrow. I made a drawing where the horn starts at 15 degrees, but opens up to 30 degrees after the bend; but others have told me this will not actually increase the vertical dispersion. I'd love to understand why that is.

It will increase vertical dispersion but not across the full range of the horn. This is a slight oversimplification, but as a rule of thumb, the horn will tend to produce a 15° vertical coverage pattern for wavelengths shorter than the horn height at the point where it transitions from 15° to 30°, then a 30° vertical coverage pattern for wavelengths between that length and the height of the horn at the mouth opening. You'll still, however, have a fairly abrupt transition from 30° to omni at this point unfortunately.

 

[Slothrop]

Tom Danley is the go-to reference for multi driver horns. Check out his Synergy Horn.

Yes, I'd love to get his take on my crazy horn. His Synergy horn was my inspiration for this.

 

[milezone]

Interesting stuff. Obviously a reference/take on the synergy horn. I've been curious about designs like this myself. Edit: I've seen less dramatic versions of this approach applied by some audiophile speaker manufacturers (Decware I believe and some others) e.g.

http://www.asi-resonators.com/super_tweeters.html

https://www.monoprice.com/product?p...MImv7Bt9a84gIVkbfsCh14oAUEEAQYASABEgLAzvD_BwE

I'm curious to try something similar with a single full range driver, compression driver (a horn like yours with a throat opening and horn flare akin to a TAD TH-4001 but in 360 degrees), or ribbons/electrostatic drivers. Despite my skepticism of some applications of the process, 3D printing really seems well suited to a project/product like this where internal waveguide channels/ducts of complex and specific geometries are required to achieve optimum performance.

You seem to have some advanced knowledge and skills. Very inspiring to see. Hope you continue exploring the idea.