Acoustic Horn Design – The Easy Way (Ath4)

[mabat]

For the sake of completeness / to compare various profiles?

I have yet to see a high-loading horn that I would consider useful. If you manage to design one, I can put it into the optimization loop and get the last bit of performace out of it

 

[mabat]

Will you do another sand order eventually do you think?

//

No such plans at the moment.

 

[TNT]

OK. Your optimising loop - what are the goals?

//

 

[mark100]

I have yet to see a high-loading horn that I would consider useful.

What do you mean by high-loading?

Thanks!

 

[gedlee]

I think that he means acoustic loading to a lower frequency - the obsolete idea that LF loading improves a horn. Not really true IMO.

 

[mabat]

As Earl says.

 

[mark100]

Thank You

haha. was saying thanks to Earl. now i see i need to say thanks to both of you !

 

[mabat]

OK. Your optimising loop - what are the goals?

Well, to have the best waveguides available

 

[gedlee]

I was curious about the optimization myself. What metrics do you optimize to? What is assumed? Like throat area, etc.

 

[camplo]

I think that he means acoustic loading to a lower frequency - the obsolete idea that LF loading improves a horn. Not really true IMO.

It can add extension and output. Thats beneficial if its what one desires...

 

[mabat]

There are better ways how to add extension and output, without deteriorating the directivity.

Let's keep waveguides doing their best, what can't be gained otherwise - the clean directivity control. For adding extension without controlled directivity, you don't need a beaming horn.

 

[camplo]

I guess theres always the renkus heinz cdt-1 or axi2050 or maybe the coaxials like the bms or b&c.... but I don't know how else to add lower extension to a waveguide?

 

[mabat]

I was curious about the optimization myself. What metrics do you optimize to? What is assumed? Like throat area, etc.

What I realized very quickly is that I actually don't need any such thing. I know what I want, and the whole algorithm is simply me briefly skimming through randomly generated results and picking the most promising ones for another round. For the most part, two rounds are enough when you know already what roughly to start with. It's really that simple and as far as I can tell there's no reason to make it any more complex - i.e. fix what you want fixed in a particular case, let other variables vary randomly within defined limits and see what happens. Typically there quickly emerges what you are looking for.

Let's call it an evolution by expert's eye selection.

 

[camplo]

The output of a high quality compression driver + horns acoustic couple output is cleaner than a dynamic radiator of the same size...

Pretty sure Ill bite me tongue of this one, pretty sure no matter the direction we go we trade pros n cons

 

[mabat]

Could you show an example of what you mean by that?

 

[camplo]

My axi + 350hz horn vs a dynamic driver with the same(ish) sd as the mouth of the horn...in the area of loading and above

 

[aragorus]

Can the simplified way of simulating horns in AKABAK/ABEC (by modelling the CD as a radiating membrane) be trusted over 10kHz?

I've been slowly giving up on trying to make a 2" throat work.

 

[mabat]

You mean simulating the throats of the horns as simple membranes?

We already know that for 1" drivers these simulations typically match the actual measurements pretty well to 12 - 13 kHz. So for 2" one would assume it would be only around 6 kHz (pretty low, isn't it). So far I hadn't a chance to do the same for a 2" driver so I don't know. But I wouldn't hope for the reality to be much better than the simulation.

 

[aragorus]

I guess investing in a pair of 2" drivers and printing a bunch of horns to test is the way to go

 

[aragorus]

I have a feeling that a real compression driver with a 2" throat won't beam as much as a 2" membrane in AKABAK does.
But then again - only a test will verify