Some simulation help understand directivity behavior and why lower is 'better' in this case
https://www.diyaudio.com/community/threads/tachi-coaxial-point-source-bem-modelling.386711/
That said following Dave's design and Fluid's simul i experimented with principle and i second Dave's comment about force cancellation: quiet box/cabinet is very nice to have especially when it work even with very simple design ( no need for thick or heavy cabinet wall).
If you can limit to below 150hz it's even better ime.
Last edited:
Normally for a three way I choose based on baffle step unless mid distortion is an issue, but in this case, as you say, dispersion of the woofers is the deciding factor. Unfortunately, I think that mid distortion may indeed be an issue. I’ll need to take new measurements with the rears closed and see. The ND105’s only model to about 100 dB or so, so maybe I can push them down closer to resonance.
Some of this is only going to become obvious if you take measurements at different points around the speaker.
Otherwise if both ways have a range where they disperse the same, it shouldn’t matter where the cross is with regard to the baffle step.
Otherwise if both ways have a range where they disperse the same, it shouldn’t matter where the cross is with regard to the baffle step.
Thanks for that. A lot of helpful info in that thread that I hadn’t seen.
I understand why lower is better from a directivity standpoint, but I wonder how much it maters in practice considering room effects and our aural processing in those regions.
Nothing compares to first hand experience: prototype a box, make your own observation and report here. 😉
All i can say relative to 'holes' in power transfer ( non constant directivity pattern) is i'm sensitive to it. If schroeder freq of your room is high enough it might not matter to you but only you could tell.
Experiment/prototype with dirt cheap drivers. Wood for box might cost more than a pair of Visaton or Grs drivers. And no need for ply or mdf as you'll have force cancelation at play.
Last edited:
Indeed. I don’t have the woofers and PR’s yet. The point of this thread for me was to see if it made sense to buy them or figure something else out. It sounds like I may end up with a final sensitivity of 88-89 dB with near wall placement which is more than I hoped for honestly.
Besides that, I get force cancelation, anechoic f3 around 50Hz, Sd of just 102cm plus 4 PR per side and max output of 100-102dB. All in about 12 total liters. Would you spend the $180 for those things?
In a 3-way speaker the bass radiator position(s) and baffle/box dimensions are practically irrelevant.
Some mild changes (in anechoic environment and simulations) start happening above 200Hz, more relevant above 400Hz (with normal size baffles).
In-room boundary reinfocement and room modes are much more difficult to handle below 400Hz. Typical home listening rooms and non-nearfield listening have Schröder limit around 400Hz, below which we cannot localize sound source. Even the so called floor bounce around 200Hz is quite difficult to hear, but a measurement mic can show drastic dip.
So, the answer to OP question is - no effect. Many very well designed 3-way speakers use single or more down- , side- or backside woofers only ( D&D 8c, Kii-3, Gradient, Genelec)
With dsp the backside woofer of bipole can be delayed to create cardioid or dipole pattern in two octaves (eg. Gradient W371A and many diy speakers)
Some mild changes (in anechoic environment and simulations) start happening above 200Hz, more relevant above 400Hz (with normal size baffles).
In-room boundary reinfocement and room modes are much more difficult to handle below 400Hz. Typical home listening rooms and non-nearfield listening have Schröder limit around 400Hz, below which we cannot localize sound source. Even the so called floor bounce around 200Hz is quite difficult to hear, but a measurement mic can show drastic dip.
So, the answer to OP question is - no effect. Many very well designed 3-way speakers use single or more down- , side- or backside woofers only ( D&D 8c, Kii-3, Gradient, Genelec)
With dsp the backside woofer of bipole can be delayed to create cardioid or dipole pattern in two octaves (eg. Gradient W371A and many diy speakers)
Last edited: