sphere shaped speakers

[MAB]

For some reason those phase plug drivers are always referred as point source drivers

I’ve never heard that general reference! Who says this? Perhaps I am uninformed.

Both coaxial and single drivers can be considered point source.

 

[gordoste]

I would try a coaxial driver instead of a full-ranger - we're on ASR, after all
The budget is rather restricting (quite impossible if crossovers are accounted for) but the SB Acoustics SB12PFCR25-4-COAX gets pretty close

This exact driver was used for a Silicon Chip article recently. (You'll need to pay the price of a sandwich to get the full thing, but it has complete measurements and plans).

 

[JSmith]

Weltron did these back in the 70's;

JSmith

 

[ppataki]

It's the best external shape, and the worst internal shape.

Can you please elaborate that?
A sphere has zero parallel sides - I was always under the impression that the fewer parallel sides you have the fewer room modes will form

 

[voodooless]

I was always under the impression that the fewer parallel sides you have the fewer room modes will form

This is exactly the problem . Because you only have one dimension in al directions, you will have one pronounced resonance. This tends to be much worse than spread-out resonances.

 

[rokr]

Here are some tests of small, full-range drivers you may find helpful:

Timothy Feleppa's Pages: Speaker Measurements - Midrange/Fullrange Speakers 4" and smaller

Zaph|Audio

I saw this page which is great (this guy work a lot), but most of his favorite drivers are hard to find or not manufactured anymore.
I was thinking about dayton rs-100-4 which is from the same line of the rs75-4 he really liked.

 

[rokr]

I’ve never heard that general reference! Who says this? Perhaps I am uninformed.

Both coaxial and single drivers can be considered point source.

Look at this page: https://www.daytonaudio.com/category/123/full-range-drivers?pagenum=3
while all are full range drivers, the ones with the plug are called point-source...

 

[rokr]

This exact driver was used for a Silicon Chip article recently. (You'll need to pay the price of a sandwich to get the full thing, but it has complete measurements and plans).

Thanks. Bought it and will go through it today, although for this project I wanted to try and skip the use of of a crossover and see what can drivers like mark-audio's do (but now I'm not so sure). From a quick glance saw some very good ideas I can implement no matter what driver I choose.

 

[voodooless]

while all are full range drivers, the ones with the plug are called point-source...

That must be a Dayton-only thing... Silly really.

The RS100 you're about to opt for should do well. It's clean and has low distortion. I would probably go for the 4 Ohm version. Preferably though, I would use a reflex alignment for (any of) these drivers. It will just lower excursion and make them get deeper. Note that for the RS100-4, you will not need 9 liters. More like 3 liter with a 70 Hz tuning should look quite good.

 

[Killingbeans]

Can you please elaborate that?
A sphere has zero parallel sides - I was always under the impression that the fewer parallel sides you have the fewer room modes will form

Like @voodooless say, it will have a single very pronounced resonance/standing wave.

Although, I'm thinking that if you could fit a very effective Helmholtz absorber inside, tuned to the same frequency, you'd be close to "perfection".

Not sure how big a problem it is in actuality though.

 

[MAB]

Look at this page: https://www.daytonaudio.com/category/123/full-range-drivers?pagenum=3
while all are full range drivers, the ones with the plug are called point-source...

That’s Dayton Audio marketing.
Phase plug doesn’t make them more or less point-source. Not one bit.

 

[ppataki]

Like @voodooless say, it will have a single very pronounced resonance/standing wave.

Although, I'm thinking that if you could fit a very effective Helmholtz absorber inside, tuned to the same frequency, you'd be close to "perfection".

Not sure how big a problem it is in actuality though.

Thanks @Killingbeans
One more question on this please: would that resonance be visible on the frequency curve as well or just on the waterfall/decay graph (time-domain)?
I am just speculating here: if it was apparent on the FR curve too then taming that with adequate DSP would theoretically tame the time domain too? Maybe not fully but somewhat...
Can you comment on this please?

 

[Killingbeans]

One more question on this please: would that resonance be visible on the frequency curve as well or just on the waterfall/decay graph (time-domain)?

I imagine it would. It's honestly not something I have researched a lot. As far as I've been told, the best internal shape would be an "infinitely" tapered tube like a seashell. That's why the top models from B&W look the way they do. Probably complete overkill, and it's a shame that they seem to choose to f¤¤k other things up deliberately, making me not want to touch their speakers with a ten-foot pole.

I think the reason why we don't see real life spherical speakers having a nasty spike in the frequency response, is because none of them probably have a truly spherical cavity. Once you add the geometry of the drivers, binding posts, crossovers, damping materials and other bits needed to get actual sound from the thing, it probably gives enough scattering and absorption to make the resonance a non-issue.

I am just speculating here: if it was apparent on the FR curve too then taming that with adequate DSP would theoretically tame the time domain too? Maybe not fully but somewhat...

I suppose it would. But again, I'm not sure it's an issue in real life implementations.

 

[rokr]

Thanks @Killingbeans
One more question on this please: would that resonance be visible on the frequency curve as well or just on the waterfall/decay graph (time-domain)?
I am just speculating here: if it was apparent on the FR curve too then taming that with adequate DSP would theoretically tame the time domain too? Maybe not fully but somewhat...
Can you comment on this please?

This is what the guy who built this kind of speaker (silicon chip) had to say about this:
"Those who make speakers will be quick to comment that a sphere should be highly resonant; however, our tests show this is not the case. The fact that the driver forms a significant portion of the surface area of the sphere results in the Q of the internal resonance being relatively low. As a result, our measurements don't show resonant peaks in the response."

 

[Thomas_A]

Internal standing wave: you make two halves with a board with an smaller opening. Like in the Mirsch Rondo.

 

[mhardy6647]

Well...

Omnidirectional Sphere Loudspeaker Systems | Soundsphere

SoundSphere creates advanced omnidirectional speakers in the audio industry. Shop hemispherical and full-range speaker and loudspeaker systems.

soundsphere.mseaudio.com

...or, perhaps, an engineer's approximation of a sphere.

Design Acoustics D-12 speaker (Apr. 1973)

 

[ppataki]

I went to IKEA today (had to buy some furniture anyway) and I have checked the 28cm (the biggest version) of the bamboo bowl
I have put them together and put my hand there for scale:

Actually I really got intrigued by this idea and I will definitely try this in my next project in 2024
I will use CHN110 drivers and plan to create a 4.0 system using these bamboo bowls + I am going to repurpose/rebuild my MCLA speakers (not being used now) to become subwoofers below the couch

 

[flowolf]

Would it be possible to fit a (sub)woofer on the back, like the Cabasse Rialto? Maybe a Dayton Epique?

 

[ppataki]

Would it be possible to fit a (sub)woofer on the back, like the Cabasse Rialto? Maybe a Dayton Epique?

Well.....you would need X liters of volume for the full range driver + Y liters of (ideally) separated volume for the sub
With a total volume of approx. 9 liters (in case of the 28cm bamboo bowl) this can be a challenge but I guess it is not impossible

 

[BenB]

That must be a Dayton-only thing... Silly really.

The RS100 you're about to opt for should do well. It's clean and has low distortion. I would probably go for the 4 Ohm version. Preferably though, I would use a reflex alignment for (any of) these drivers. It will just lower excursion and make them get deeper. Note that for the RS100-4, you will not need 9 liters. More like 3 liter with a 70 Hz tuning should look quite good.

I don't like the idea of porting speakers at frequencies much above 50 Hz. Although you lower excursion near the porting frequency, excursion increases very rapidly below tuning. 40 Hz content is often the most demanding from an excursion perspective, since it's often recorded with high volume (in contrast to content in the 20s and 30s that's typically lower level). If you tune your port to 70 Hz, then 40 Hz content could give you excursion issues if you don't apply a high pass filter. If you're using a subwoofer and apply a high pass filter due to that, then you probably don't need to port the satellites.