Acoustic Horn Design – The Easy Way (Ath4)

[turk 182]

in a single dimension it's starting to resemble the folding done to a line array waveguide.

 

[mabat]

Try that in ripple tank

 

[AllenB]

In this latest example the throat is around 4 wavelengths, this seems a lot.

 

[miero]

I don't have the time for that

- This is a comparison between an original throat and the vanes calculated based on the same throat. What still puzzles me is that the leading wavefronts in both cases are pretty far from following the elliptical coordinate lines (the vanes end all at u=1.5) but I'm sure there's an explanation for that as well.

Try to place a shorter sound source, just between focal points of the ellipse.

 

[mabat]

I think this is actually an applet issue, where the leading wavefront is just not correct - see the wavefront of the subsequent zero pressure (the black contour), which seems to follow the coordinates much better and I suppose this should be the case for the leading wave as well... It starts to get somewhat complicated (maybe not)

(As for your suggestion - the source is always shorter than the focal points distance.)

 

[mabat]

It is really closer for the second wavefront -
It's probably best to rely simply on the calculated average path lengths.

 

[mabat]

For a little higher frequency -

 

[mabat]

1-inch driver, 24 kHz, k=5 (not optimized)

 

[tmuikku]

Cool :O

 

[mabat]

If there was a suitable direct-radiating driver, it would be straightforward to make it a compression device with a phase plug made this way, directly in the form of the vanes, separated at the source. Flat disc diaphragm, if strong enough, would be ideal but I guess a regular dome (or any other shape for that matter) would work as well, it only adds the complexity of the compression chamber modes. One could also take a naked compression driver... Or just use a flat disc directly, without compression.

BTW, it seems that the algorithm for the optimized meandering vanes calculation can be automated almost completely, giving the right wavefront at the entrance of the rest of the waveguide. The OS coordinates make a convenient starting point for this.

 

[gedlee]

Flat disc diaphragm, if strong enough, would be ideal

The first OS waveguide that I ever tested used a flat honeycomb diaphragm made by Panasonic. It worked exceptionally well with no phase plug at all.

 

[Patrick Bateman]

They weren't a big hit, but Tangband made an aluminum honeycomb driver that plays all the way out to 20khz and beyond.

Really nicely made (I bought one to mess around with.)

 

[gedlee]

That would make a nice driver for a lower powered (and cheaper) system - without the grill I should think.

 

[mabat]

How did you handle the surround? That seems to be quite a complication.

 

[gedlee]

The biggest complication, of course, and one major reason that CDs are made like they are, is to minimize this complication. A flatter surround can be made to have a smaller cavity at the surround, but it can be done with a concave one, with the right - let's see 3D printer.

 

[Mazza]

CDX1-1745 on ST260/B

Measured with Omnimic at 1m driven by DATS V2.

 

[TNT]

The gap/slit you have in the WG, is that sealed with something cured and transparent or is it void space?

//

 

[Mazza]

I've not glued the two halves together, its just a gap. One half didn't print perfectly, the other did. The CD bolts are all that held it together.

 

[454Casull]

If there was a suitable direct-radiating driver, it would be straightforward to make it a compression device with a phase plug made this way, directly in the form of the vanes, separated at the source. Flat disc diaphragm, if strong enough, would be ideal but I guess a regular dome (or any other shape for that matter) would work as well, it only adds the complexity of the compression chamber modes. One could also take a naked compression driver... Or just use a flat disc directly, without compression.

BTW, it seems that the algorithm for the optimized meandering vanes calculation can be automated almost completely, giving the right wavefront at the entrance of the rest of the waveguide. The OS coordinates make a convenient starting point for this.

Ring radiator tweeters may be the closest off-the-shelf compromise?

 

[mabat]

I've not glued the two halves together, its just a gap. One half didn't print perfectly, the other did. The CD bolts are all that held it together.

All I can add is that these waveguides are so good that even as bad prints and in conditions they were not intended to work in, they are still usable

I wonder how close to the free standing version you will get after you place them in a baffle: Acoustic Horn Design – The Easy Way (Ath4) (#7506, the free standing version measured by Zvu et al.)