Acoustic Horn Design – The Easy Way (Ath4)

There's one detail I omitted in the previous simulations of the isolated suspension - the edge of the compression cavity also should have an excursion as the filler would be changing its shape a bit. It doesn't make much difference though, and if then for the better.

As it seems, for this uniformly driven cavity even a one-slot phase plug could perhaps suffice? (Still, completely ignoring the modes of the diaphragm itself.)

1slot-driving.PNG

1slot.png

1slot-png.png
 
Last edited:
I should think that the ideal would be an open cell foam with a density of say 70-90 PPI which would provide for high damping as well as fill the cavity. This could dampen the entire cavity.

This kind of foam cannot be molded or placed and cured (because of how it's made.) You would have to buy blocks of it and cut it, which would not be a trivial effort.
 
This wouldn't be complicated shape, just a thin flat layer (~ 1 mm) would be fine I guess. This would then be cut and squeezed in the cavity. At least this is how I would do it. I wouldn't cut a spherical shape - that would really not be a trivial effort!

Although I have this cutter so with some well thought out jig even a spherical surface might not be impossible. For straight cuts I have this one. Works like a charm.
 
Last edited:
I should think that the ideal would be an open cell foam with a density of say 70-90 PPI which would provide for high damping as well as fill the cavity. This could dampen the entire cavity.

This kind of foam cannot be molded or placed and cured (because of how it's made.) You would have to buy blocks of it and cut it, which would not be a trivial effort.
Would there be enough absorption at the lower end of the CD passband with this foam? I know you're thinking of a higher PPI, but at ~30 you used it just to attenuate HOMs (for the most part).

I guess my other question would be, is there a frequency below which the "filler" no longer needs to act as a cavity filler, i.e. is this a higher-frequency phenomenon to address?
 
These are open cell foams readily available to me:
Code:
Bulpren S 31048	  90	440 - 520
Bulpren S 31062	  75	520 - 720
Bulpren S 28089	  60	740 - 1040
Bulpren S 28133	  45	1080 - 1580
Bulpren S 28190	  30	1650 - 2150
Bulpren S 28280	  20	2300 - 3300
Bulpren S 32450	  10	3800 - 5200

Type, nominal PPI and sizes of the cells in micrometers.

Then a hot wire works well, just know that the fumes are highly toxic. Should be no problem to setup a jig to cut such a simple and regular piece. Slice and then cut on a turntable, just like I made my foam. Yours are much smaller and need to be much more precise, but it's certainly doable. Inner circle cut is tricky though.
 
Would there be enough absorption at the lower end of the CD passband with this foam? I know you're thinking of a higher PPI, but at ~30 you used it just to attenuate HOMs (for the most part).

I guess my other question would be, is there a frequency below which the "filler" no longer needs to act as a cavity filler, i.e. is this a higher-frequency phenomenon to address?

Tough to model perhaps as the required values are not readily available. Best would be to just try it and modify from there.
 
Then a hot wire works well, just know that the fumes are highly toxic. Should be no problem to setup a jig to cut such a simple and regular piece. Slice and then cut on a turntable, just like I made my foam. Yours are much smaller and need to be much more precise, but it's certainly doable. Inner circle cut is tricky though.
Steel tubing with a sharpened "hot end" heated with blowtorch? Only problem would be ensuring coincident centers for the inner and outer circles.
 
Maybe I misunderstood but why an inner circle? I would simply fill the whole cavity. Any sound passing the nearest exit of the phase plug should be attenuated, correct? All the radial resonances could be damped quite effectively I guess.
To make room for your diaphragm? In this context, "inner circle" refers to the opening in the foam filler ring / gasket that your diaphragm will stick out of.
 
Why would they be so much different. The foam has density and resists air flow. That's where the damping comes from. A solid is just a little more dense, but totally reflective.

Your data shows that this small area accounts for a lot of problems. With the right density of foam, I'll bet you could make a substantial difference.

Maybe your thinking of too open foam cells. That would not do much. But as the cells get smaller they begin to act more like a dense material with a slow leakage resistance.

I think that filling the whole chamber is interesting, but probably very bad on HF output.
 
Hmm, then I'm not sure I can get such dense foam because even the most dense from what I have shown seems too open. I'm tempted to try filling the entire chamber with some less dense foam. That can be tried immediately on any existing driver after all.

It appears to me that most of the "direct" sound leaving the driver travels only a short route through the structure, whereas for any resonance or delayed sound it takes a longer path, i.e. higher damping through the foam, if the foam filled the whole space. The same as for your anti-HOM foam plug. It's a matter of the right density I suppose to not to damp the "direct" response too much.
 
Last edited:
The cutting is really no problem, once it is sliced off from the block which will be the tricky part as the slice would be very thin. With the right wire temperature and cutting speed that should be no problem either. I can imagine cutting the circles even with a compass with a blade attached - probably the easiest way to get the exact radius required.