3-way speaker with double waveguide and cardioid bass

[peniku8]

sealed speakers benefit more from room gain.

A vented speaker doesn't have meaningful output below its tuning frequency anyways, so I think it hardly matters that it isn't able to pressurize a room in the same way a sealed cabinet would. *At* its tuning frequency, the port and the cone are in sync, which means the vented speaker will pressurize the room in the same manner as the sealed speaker would, actually even better since it builds up pressure from the cone and the port. And above its tuning, pressure vessel gain is irrelevant since wavelengths will be acoustically small for the size of a typical residential room (at typical tunings too for that matter) and you'll see modal behavior instead. No matter how you look at it, I think this is a non-argument and such a statement creates more confusion than it solves.

 

[Scgorg]

I made a little writeup about some design aspects of the speaker and the project as a whole. Posting it here for whomever may be interested.

 

[KLang1]

Hi Georg, thanks for your write up!
I'm impressed with your results especially since it's your first own speaker build!
I will probably try something similar soon, as my solution of a cardiod speaker got very heavy with 1x15" and 2x12" for cancellation and it has the usual problem of only cancelling horizontally.
A round or square slotted box with a good coaxial driver (e.g. https://bcspeakers.com/en/products/coaxial/13.5/8/14CXN76) could be cardioid in both axis, but round would be a pain in the ass to build, so the first try will probably be square
How does the dampening material on your speaker work in different humidities?

 

[Dave Bullet]

A couple of small tips. You're likely getting some false ripple above ~ 8Khz in your measurements due to reflections off the mic clip and stand. Put on a boom (e.g. inside some PVC pipe), long enough to be outside your gated measurement response.

Also - remove the polystyrene blocks. They only mask the first reflection (or create additional ones). I doubt they effectively absorb it. Hard surfaces all around if you want a clear reflection marker for gating.

 

[Scgorg]

A couple of small tips. You're likely getting some false ripple above ~ 8Khz in your measurements due to reflections off the mic clip and stand. Put on a boom (e.g. inside some PVC pipe), long enough to be outside your gated measurement response.

Also - remove the polystyrene blocks. They only mask the first reflection (or create additional ones). I doubt they effectively absorb it. Hard surfaces all around if you want a clear reflection marker for gating.

I mentioned in the first post that the picture seen there is not indicative of how the completed speaker was measured. Those were preliminary measurements, and they had issues, as you've deduced.
The finished speaker was measured with the ground plane method. I think the method is specified in the PDF i posted above.

A round or square slotted box with a good coaxial driver (e.g. https://bcspeakers.com/en/products/coaxial/13.5/8/14CXN76) could be cardioid in both axis, but round would be a pain in the ass to build, so the first try will probably be square

I've thought of something similar, but all commercially available coaxes for DIYers leave something to be desired in my opinion. I'd probably look into optimizing multiple-entry horns instead.

How does the dampening material on your speaker work in different humidities?

Quite frankly, I do not know! I've thought about it, but figured that in all likelihood the differences are going to be too small to be meaningful. Believing that is certainly a good way to gain peace of mind.

 

[KLang1]

I've thought of something similar, but all commercially available coaxes for DIYers leave something to be desired in my opinion. I'd probably look into optimizing multiple-entry horns instead.

They're far from perfect, but the approach of a multiple-entry horn has the same drawback as my active cardioid speaker: Weight & size. For that I'm willing to compromise on the directivity of the HF part and that B&C coaxial looks good enough.

Quite frankly, I do not know! I've thought about it, but figured that in all likelihood the differences are going to be too small to be meaningful. Believing that is certainly a good way to gain peace of mind.

I'll investigate this when I get this far.

 

[Jukka]

Very nice construction! Did you get inspiration for the waveguide from Genelec?

This is the 1237A with 1", (almost) 5" and 12" drivers. Looks very much like yours

 

[Scgorg]

Very nice construction! Did you get inspiration for the waveguide from Genelec?

In the beginning the waveguides were elliptical, but there were still some small irregularities I wasn't able to get rid of. I decided, mostly on a whim, to use a similar profile but in a waveguide shaped more like the genelecs. It worked like a charm. The tweeter waveguide can surely be made better still, however (mostly fixing the slight issues in the top octave).