Horn loaded dome tweeter versus direct radiating

[Joseph Crowe]

I’m sharing my latest blog post on the No. 1159 horn loaded 2-way bookshelf speaker. I show a full set of acoustical measurements and compare it against a regular non-horn loaded speaker baffle.

https://croweaudio.blogspot.com/2020/06/speaker-no1159-horn-loaded-2-way-medium.html?m=1

 

[mhardy6647]

The difference between the two is both interesting and unsubtle.

That said, I am a little distracted by the appearance of a face in the plot on the left.

 

[muad]

For some reason you links aren't working for me.

 

[mhardy6647]

For some reason you links aren't working for me.

It's got something to do with the way the URL is entered (not sure what, though). I think it might be for a 'window' within a blog post and not directly accessible(?).

That said, I copied and pasted and it worked fine.
I also fiddled around a little with the blog's URLs -- here, try this, should work.
https://croweaudio.blogspot.com/2020/06/speaker-no1159-horn-loaded-2-way-medium.html

 

[muad]

It's got something to do with the way the URL is entered (not sure what, though). I think it might be for a 'window' within a blog post and not directly accessible(?).

That said, I copied and pasted and it worked fine.
I also fiddled around a little with the blog's URLs -- here, try this, should work.
https://croweaudio.blogspot.com/2020/06/speaker-no1159-horn-loaded-2-way-medium.html

Awesome thanks, works perfectly

 

[gene_stl]

Very nice Work!

 

[mhardy6647]

Awesome thanks, works perfectly

de nada

 

[KaiserSoze]

Excellent! I'm thoroughly impressed. Now, will you please tell us exactly how you machined that intricate and beautiful piece of wood?

 

[Joseph Crowe]

Excellent! I'm thoroughly impressed. Now, will you please tell us exactly how you machined that intricate and beautiful piece of wood?

Thank you! I machined it on my CNC. It took me about 30 hours of CAD/CAM design time and about 10 hours to make the speakers and stands. I say this because the time lapse video makes it look easy. But there’s a tremendous amount of time and effort involved.
http://instagr.am/p/CBMWNEDJFPG/

 

[dwkdnvr]

Really nice work - definitely like how you integrated the roundovers into a single-piece baffle.

What was your design process for the waveguide? Dome tweeters seem to involve a fair bit of trial and error at this point.

 

[Joseph Crowe]

Really nice work - definitely like how you integrated the roundovers into a single-piece baffle.

What was your design process for the waveguide? Dome tweeters seem to involve a fair bit of trial and error at this point.

The horn flare curvature is what I’ve branded “ES” which is the parametric equation I use within Solidworks to drive the 2D sketch curve. ES stands for Exponential Spiral which is the spiral curve you see all throughout nature, so in a sense it’s biomimicry. The rest are just some basic rules on horn theory such as the depth of the horn determining the low frequency cutoff...in this case 40mm which is a quarter wavelength of 2khz. It’s a proper horn in the sense that it fully loads the tweeter starting at 2kHz. Most waveguides I see do this gentle frequency response bump centered around 2kHz but it loses output by 6kHz. My geometry provides boost up to the 10kHz region so almost 2.5 octaves of bandwidth which a normal horn should do.
As far as integrating with a dome tweeter I chose the SB29RDNC because of the dimple dome...this creates a flatter acoustical wavefront which is critical for good loading in the horn throat. I also sized the horn throat diameter to partially conceal the tweeter’s surround. This helps widen the off-axis coverage in the upper treble which the measurements show was effective. This would be the third dome tweeter model I’ve tried with a horn. So far it responds the best.

 

[KaiserSoze]

Thank you! I machined it on my CNC. It took me about 30 hours of CAD/CAM design time and about 10 hours to make the speakers and stands. I say this because the time lapse video makes it look easy. But there’s a tremendous amount of time and effort involved.
http://instagr.am/p/CBMWNEDJFPG/

That is a beautiful video. It really shows what is possible if you have a CNC machine and the knowledge to use it. I'm envious.

 

[Joseph Crowe]

The difference between the two is both interesting and unsubtle.

That said, I am a little distracted by the appearance of a face in the plot on the left.
View attachment 68581


View attachment 68582

That’s very scary. I will never go direct radiating ever again!

 

[Ilkless]

The waveguide loading and directivity matching are readily obvious, but so is the reduction in diffraction. I'm seeing diffraction peaks at 2.5k, 3.5k and 5kHz that were eliminated. BTW, would you ever consider using a 7" (maybe a Purifi, but SB/SEAS/Scanspeak all have wonderful offerings for not much more cost) crossed to a waveguide dome that can handle a <2kHz XO, for more headroom and extension while being relatively compact and still maintaining directivity matching.

 

[Joseph Crowe]

The waveguide loading and directivity matching are readily obvious, but so is the reduction in diffraction. I'm seeing diffraction peaks at 2.5k, 3.5k and 5kHz that were eliminated. BTW, would you ever consider using a 7" (maybe a Purifi, but SB/SEAS/Scanspeak all have wonderful offerings for not much more cost) crossed to a waveguide dome that can handle a <2kHz XO, for more headroom and extension while being relatively compact and still maintaining directivity matching.

Well, I'm testing out a 1.5kHz crossover right now and it's doing okay. 97dB SPL at 1 meter is showing 0.50% D2, tomorrow I'll push harder to see where the limit is. I should mention as well it's only a passive first order high pass. So this would be an unexpected bonus of the steep acoustical slope that the horn is providing below Fc which is 2kHz. If I went bigger I'd likely machine a 3-way baffle with 800Hz and 5kHz crossover points with the drivers still time aligned. That's the biggest I can go as far as Z height on my CNC. For 800Hz I'd likely have to go with a soft dome midrange.

 

[BenB]

The waveguide loading and directivity matching are readily obvious, but so is the reduction in diffraction. I'm seeing diffraction peaks at 2.5k, 3.5k and 5kHz that were eliminated. BTW, would you ever consider using a 7" (maybe a Purifi, but SB/SEAS/Scanspeak all have wonderful offerings for not much more cost) crossed to a waveguide dome that can handle a <2kHz XO, for more headroom and extension while being relatively compact and still maintaining directivity matching.

It's odd to me that the large bevels didn't do more to mitigate this diffraction. I have measured speakers without waveguides but with large roundovers on the side that had far less evidence of diffraction. I was under the impression that a bevel worked similarly well to the large roundover.

 

[HooStat]

wow. really nice-looking speaker and the specs seem great too. It seems like a great candidate for an all-digital cross-over. I am looking for a DIY project using a digital cross-over, and this looks like a nice candidate. Keep us posted on progress!

 

[Joseph Crowe]

My second blog post on Speaker No.1159 is now published. I continue the comparison with the non-horn baffle and look closely at distortion sweeps as I increase the speaker's output to 100dB SPL. 1159 takes the lead remaining below 1% distortion while the non-horn version hits 5% distortion. I conclude that 1159 has 10dB more headroom even with a simple first order crossover at 1.5kHz.

https://croweaudio.blogspot.com/2020/06/speaker-no1159-distortion-tests-and.html

 

[Ilkless]

Well, I'm testing out a 1.5kHz crossover right now and it's doing okay. 97dB SPL at 1 meter is showing 0.50% D2, tomorrow I'll push harder to see where the limit is. I should mention as well it's only a passive first order high pass. So this would be an unexpected bonus of the steep acoustical slope that the horn is providing below Fc which is 2kHz. If I went bigger I'd likely machine a 3-way baffle with 800Hz and 5kHz crossover points with the drivers still time aligned. That's the biggest I can go as far as Z height on my CNC. For 800Hz I'd likely have to go with a soft dome midrange.

You could try horn-loading a smooth, well-behaved wideband cone driver with extremely good extension and directivity. The Scanspeak 10F/8414 is one such driver that will easily take 800Hz (further measurements). I've not seen a really good midrange dome available to end-users on the market. The overrated and overpriced ATC dome (and various others of its ilk from PMC and Volt) have far too large a mounting flange for good CtC, while SOTA domes like the Neumann dome are proprietary.

 

[tuga]

You could try horn-loading a smooth, well-behaved wideband cone driver with extremely good extension and directivity. The Scanspeak 10F/8414 is one such driver that will easily take 800Hz (further measurements). I've not seen a really good midrange dome available to end-users on the market. The overrated and overpriced ATC dome (and various others of its ilk from PMC and Volt) have far too large a mounting flange for good CtC, while SOTA domes like the Neumann dome are proprietary.

What about the Volt midrange domes?

https://audioxpress.com/article/test-bench-volt-loudspeakers-vm752-3-midrange-dome-driver

https://audioxpress.com/article/test-bench-volt-loudspeakers-vm527-midrange