Acoustic Horn Design – The Easy Way (Ath4)

[TNT]

How far back do a rollback need to go before starting to return diminishing gains?

//

 

[Kevmoso]

What about an American football shape?

Thanks so much for that! I guess I'll have to put that down as another bad idea - like most of them!!

I always wanted to find some way to trip-up the axial hole with maybe some small perturbations of the rim. But your designs don't have that problem so I guess it's a moot point.

 

[mabat]

OK, this is fun.

 

[mabat]

This also might interest some of you - the influence of an enclosure on the polar response:

Axisymmetric free standing waveguide, ⌀436 x 170 mm, 1" throat:

The same waveguide in a cylindrical enclosure:

 

[mabat]

Sorry, the graphs above are not in the same scale. Corrected -

(Pretty good waveguide, BTW)

 

[Brinkman]

Fantastic sims, thank you!

 

[mabat]

What is interesting is that the last tiny bit of perfection is possible to add by rotating the whole profile a little (this was done without changing the source, i.e. this is still for a flat wavefront). I wonder if this would be the better way of setting the throat opening angle after all.

Original waveguide:

Profile rotated +3 deg:

Profile rorated +6 deg:

 

[mabat]

So, rotated +4 deg, the final picture. To sum up, this profile is a modified OS throat blended to an Euler spiral part smoothly followed by another Euler spiral of a different scale. Not so hard to guess

 

[Kevmoso]

OK, this is fun.

Really surprised the soup can wasn't more desirable. Maybe an aardvark?
I jest. Truly amazing work on this.

 

[gedlee]

What is interesting is that the last tiny bit of perfection is possible to add by rotating the whole profile a little (this was done without changing the source, i.e. this is still for a flat wavefront). I wonder if this would be the better way of setting the throat opening angle after all.

I don't quite follow what you did above.

Another problem that has always interested me is how much a radius on the rear "can" makes.

 

[TNT]

"in enclosure" thanks!


//

 

[TNT]

but maybe rather like:

//

 

[mabat]

I don't quite follow what you did above.

You don't see the differences or don't follow the procedure? I simply rotated the whole profile anti-clockwise around [0, R_throat]:

 

[TNT]

I was surprised with the wiggles in the FR and impedance as well with the extended roundover - I would have thought the opposite i.e. even smoother transition. But there was some gains in the top (also a bit surprising). Seems to earn some more attention.

//

 

[mabat]

A short study of different coverage angles. A took a small horn and changed the coverage angle parameter (no other).
The profiles tested (25 - 50 half coverage angle / 5 deg):

Throat impedances:

SPL polars attached.

 

[mabat]

Changing the nominal length: 75 - 150 mm / 25 mm:

 

[jcarothers]

I love how straightforward it is to run studies like this now. Nice work as usual mabat.

 

[Bill Brown]

Yes, I am loving the outpouring of data you are producing of late!

Bill

 

[TNT]

Next project? 1 horn + 2 pcs of 15" / side. 700Hz?

//

 

[mabat]

Varying mouth rollback radius.