I agree with that but if you are crossing the coax over to midwoofers at 400 Hz or so, you aren't down to Shroeder.
Room modes is (are?) one thing. Floor reflection null is another. a line of woofers below the coax will definitely help with the floor null. Placing the woofers per Double Bass Array guidelines will help with room modes. i/e locating the line at the 1/4 and 3/4 room width marks gives symmetry against side to side standing waves. Placing the vertically at 1/4 and 3/4 height gives symmetry against floor to ceiling modes. that is, if I recall correctly, I might be off on those fractions.
With a 100 Hz up point source for example on a baffle wall with damping at first side reflection points, you hope to get a single narrow floor null that is largely inaudible because it is somewhat filled in by ambient sound field and because of how the ear works.
You can model and see the floor and sidewall nulls in Vituix..
You can get some idea of the room modes in REW room sim
Room modes is (are?) one thing. Floor reflection null is another. a line of woofers below the coax will definitely help with the floor null. Placing the woofers per Double Bass Array guidelines will help with room modes. i/e locating the line at the 1/4 and 3/4 room width marks gives symmetry against side to side standing waves. Placing the vertically at 1/4 and 3/4 height gives symmetry against floor to ceiling modes. that is, if I recall correctly, I might be off on those fractions.
With a 100 Hz up point source for example on a baffle wall with damping at first side reflection points, you hope to get a single narrow floor null that is largely inaudible because it is somewhat filled in by ambient sound field and because of how the ear works.
You can model and see the floor and sidewall nulls in Vituix..
You can get some idea of the room modes in REW room sim
The original blade (which i got these drivers from) crossed over at 320hz 18 db octave, so it’s not that far off schroeder.
I view my midbass setup as a kind of d’appoloti setup which in practice makes it a point source.
Kef reference 5 is also a 3 way setup in the same way, with 4 drivers in total, not 8 which i have.
though, i have been concerned lately that the two top and two bottom midbass drivers may have a too far distance to make this effective a point source in the way i want it.
I have been contemplating modifying the setup to a 3.5 way with the two closest hpf ~ 350-400hz and the rest be crossed over at hpf ~ 200hz.
Also considering adding two midbass drivers on top to make the impedance load easier for the amp and improving vertical dispersion by having bass drivers close to ceiling also.
Here is a picture now before finishing the baffle wall. Dont mind the rough design, it will end up where everything is hidden in the end.
I view my midbass setup as a kind of d’appoloti setup which in practice makes it a point source.
Kef reference 5 is also a 3 way setup in the same way, with 4 drivers in total, not 8 which i have.
though, i have been concerned lately that the two top and two bottom midbass drivers may have a too far distance to make this effective a point source in the way i want it.
I have been contemplating modifying the setup to a 3.5 way with the two closest hpf ~ 350-400hz and the rest be crossed over at hpf ~ 200hz.
Also considering adding two midbass drivers on top to make the impedance load easier for the amp and improving vertical dispersion by having bass drivers close to ceiling also.
Here is a picture now before finishing the baffle wall. Dont mind the rough design, it will end up where everything is hidden in the end.
Attachments
You can listen here (from 16.00 min) from Jack Oclee-Brown (Head of Acoustics) how they did the 3 way in kef reference to mimic a point source with the d’appolito configuration.
In text from the same Stereophile review :
"
Oclee-Brown explained that, with a conventional enclosure, the ways to create a single sound source are limited, but one thing they can do is use a symmetrical driver array. "It is exactly the same thing [as a D'Appolito configuration], except that now you don't have a tweeter, you have a Uni-Q instead, and that means that [the] crossover frequency for the D'Appolito array can be much lower. . . . f you go low enough, you get lobe-free vertical dispersion."
"
In text from the same Stereophile review :
"
Oclee-Brown explained that, with a conventional enclosure, the ways to create a single sound source are limited, but one thing they can do is use a symmetrical driver array. "It is exactly the same thing [as a D'Appolito configuration], except that now you don't have a tweeter, you have a Uni-Q instead, and that means that [the] crossover frequency for the D'Appolito array can be much lower. . . . f you go low enough, you get lobe-free vertical dispersion."
"
Update on the project :
I will have to admit i was too "nonchalant" about midwoofer placement in this design which annoys me, which i can't accept. I was too busy with other thing in the house build i neglected it.
But after doing some "deep diving" the last few days, i have found a solution which is beautifully simplistic and effective!
Back-story , see previous post and video from Kef.
After KEF having designed the Blade as a 100% point source in the whole frequency area, with the Reference, they solved it by placing the midwoofers in a symmetrical line above and under the coax using the D'appolito theory keeping within 2/3 of the wavelength from center woofer to coax.
See also cut from White paper : (especially, notice the polar response)
In short, a symmetrical D'appolito midwoofer setup setup together with a coax is brilliantly simple and effective!
It shows with the success of the reference 5, which is a more consumer friendly design of the Blade two. The Blade has less diffraction issues due to curve and a better woofer design for reducing bass distortion since the woofers are mounted together back to back which eliminates resonance.
But all in all they perform very similar as a point source, though the Reference would perform poorly when not in a seated position due to the driver positoning.
This reasoning validates the D'appolito configuration for emulating an effective point source together with a Uni-Q coax (when seated)
As for why this matters in my design ;
On the Ref 5 they spaced the midwoofers quite far since they have a lot of playroom without reaching the 2/3 of wavelength at the crossover. They could have added another set of woofers. under/above, but for the customers the speaker would become too tall and the addressable market would decrease.
After calculating my distances i am too far out on my midwoofer spacing to keep within 2/3 wavelength, but LUCKILY i am able to get withing 2/3 wavelength from W to C with placing the RS180 mids as close as physically possible.
This puts me at 70cm from Woofer to Coax and translates to 327-328hz as a max crossover point between coax and woofer.
My Blade UniQ drivers was designed for a crossover of 320hz in the original Blade, so i should be fine.
In addition, my baffle wall setup will theoratically add up to 6db to the lower band of the uni-q, this should give me enough headroom.
Yesterday i cut out front baffle of the speakers and made new ones, i will update with pictures when i get the time to finish the modifications.
I also braced the enclosure more implementing ideas from the Reference white paper.
I made the bracings 2mm short and adding a good chunk of elastic glue to absorb resonance from front to back wall. Also considering supporting the magnets of the woofers in the same way absorbing resonance.
I am quite happy about the way i solved this mishappening
On a side note of design benefits, my hidden baffle wall setup enables me to put dampening material in a strategical area around the Uni-Q to eliminate tweeter diffractions.
PS, i also considered going 3.5way with the outer two midbasses spaced out further and crossed lower than the two closer.
But i feel this setup is beautifully simplistic and effective which would i think would give warmth to a KEF engineers heart
I will have to admit i was too "nonchalant" about midwoofer placement in this design which annoys me, which i can't accept. I was too busy with other thing in the house build i neglected it.
But after doing some "deep diving" the last few days, i have found a solution which is beautifully simplistic and effective!
Back-story , see previous post and video from Kef.
After KEF having designed the Blade as a 100% point source in the whole frequency area, with the Reference, they solved it by placing the midwoofers in a symmetrical line above and under the coax using the D'appolito theory keeping within 2/3 of the wavelength from center woofer to coax.
See also cut from White paper : (especially, notice the polar response)
It shows with the success of the reference 5, which is a more consumer friendly design of the Blade two. The Blade has less diffraction issues due to curve and a better woofer design for reducing bass distortion since the woofers are mounted together back to back which eliminates resonance.
But all in all they perform very similar as a point source, though the Reference would perform poorly when not in a seated position due to the driver positoning.
This reasoning validates the D'appolito configuration for emulating an effective point source together with a Uni-Q coax (when seated)
As for why this matters in my design ;
On the Ref 5 they spaced the midwoofers quite far since they have a lot of playroom without reaching the 2/3 of wavelength at the crossover. They could have added another set of woofers. under/above, but for the customers the speaker would become too tall and the addressable market would decrease.
After calculating my distances i am too far out on my midwoofer spacing to keep within 2/3 wavelength, but LUCKILY i am able to get withing 2/3 wavelength from W to C with placing the RS180 mids as close as physically possible.
This puts me at 70cm from Woofer to Coax and translates to 327-328hz as a max crossover point between coax and woofer.
My Blade UniQ drivers was designed for a crossover of 320hz in the original Blade, so i should be fine.
In addition, my baffle wall setup will theoratically add up to 6db to the lower band of the uni-q, this should give me enough headroom.
Yesterday i cut out front baffle of the speakers and made new ones, i will update with pictures when i get the time to finish the modifications.
I also braced the enclosure more implementing ideas from the Reference white paper.
I made the bracings 2mm short and adding a good chunk of elastic glue to absorb resonance from front to back wall. Also considering supporting the magnets of the woofers in the same way absorbing resonance.
I am quite happy about the way i solved this mishappening
On a side note of design benefits, my hidden baffle wall setup enables me to put dampening material in a strategical area around the Uni-Q to eliminate tweeter diffractions.
PS, i also considered going 3.5way with the outer two midbasses spaced out further and crossed lower than the two closer.
But i feel this setup is beautifully simplistic and effective which would i think would give warmth to a KEF engineers heart
Last edited:
Yeah .. well, the cabinets came from eBay! They came from a Lamborghini showroom (they are painted in some lambo metallic black? I think .. ) or offices and the building manager asked one of his staff to get rid of them after being on display.. he put them on eBay (£350 about 6 years ago . ) .
I did try asking KEF for spares and explained the situation but they refused .. sadly!
I instead am exploring using other drivers, non-coaxial like Bliesma domes with Dayton 8" drivers in the sides perhaps.. although the flat Kef drivers suit it much better.
I have mine horizontal during movies and I don’t experience any lobing during movie playback. Probably a dolby thing? Or, maybe I just get into the movie and don’t listen as I would with music? However, when I listen to 3 channel or 5 channel music I set my center channel vertically in front of my TV. Kind of Blue in three channel is great this way.