It's dawning on me that while it might work, it will have to be a one-off permanent installation and the planar baffle needs to have some good mass and rigidity. Also, how to measure and tune/EQ it in-room?
I'm not sure I will actually build it, and I am looking into other options that will be smaller/easier/cheaper to build. For example, I already have a 36" L H-frame that I could easily convert into a U-frame or offset H-frame to obtain less dipole cancellation and more LF output. Or I could supersize the driver. I am currently using an 18" driver but the H-frame can accommodate up to a 24" monster like the Stereo Integrity IB-24.
Looks like you need high SPL and dipole. So some sacrifice has to be done anyway. I would do box with front and rear driver. So that higher it works like dipole, somewhere below ~50Hz monopole, and cardioid in transition region. All control (phase amplitude) of the drivers has to be done in digital domain. Did't do that yet myself, but i don't see any problems. It is in my to do list. Would be interesting to compare sonically with true dipole bass. But low end SPL and distortions would be definitely better for such design.
I did that a couple of years ago using a pair of 12" drivers in 15" cubes plus a DSP AP filter. I even thought about publishing an article on it in AudioXpress, but I would need a way to make very good outdoor LF measurement and that has been holding me back. The concept does work, but you have to be careful choosing the distance between the drivers because this limits how high you can use the pair. The AP controls where you transition to cardioid and then dual monopoles, one delayed.
I have considered using this approach with smaller drivers to span a band from about 100Hz to 500Hz, with a SB subwoofer below and a small mid and tweeter above. But it starts to require many channels: tweeter, mid, front woofer, rear woofer, and subwoofer. That's a 5-way speaker, which is getting complicated when you can do the same thing with fewer moving parts. For example, I can already create a very nice 3-way nude dipole system that can operate down to 80-100Hz, and pair that with a single mono sub below 100Hz. It's simple, sounds good, and has plenty of output so that is the direction I seem to be heading with my upcoming projects. The only detail left to be decided is how to implement the mono sub, and that is what brought me to started this thread.
So i see two options here.
First:
Do dipole with big two drivers in w shape arrangement.Connect them together with some rods so force cancelation will be perfect. Thats rock solid approach. No vibrations at large displacement. You can do U shape baffle with this approach.
Second:
Place two monopole subs with large distance between them and use them as dipole with large separation.Sit somewhere but not at dipole null
You can sit close to one off them so SPL will be more than enough
First:
Do dipole with big two drivers in w shape arrangement.Connect them together with some rods so force cancelation will be perfect. Thats rock solid approach. No vibrations at large displacement. You can do U shape baffle with this approach.
Second:
Place two monopole subs with large distance between them and use them as dipole with large separation.Sit somewhere but not at dipole null
You can sit close to one off them so SPL will be more than enoughRight, sort of.
This thread about dipoles has the usual dead-end fallacy of linear thinking and textbook diagrams and ray-arithmetic instead of statistical thinking. Dipoles work nice because the stuff coming out the back bounces around and meets the stuff coming out of the front in heterogeneous ways. In light of the great wash of reflections all around the back of the speaker, they may be doing so at frequencies a lot lower than basic arithmetic (such as Linkwitz) would predict.
Also, can't say as I've ever understood the appeal of H-frame design. Seems to me the lumber would do more good used somewhere else on the speaker baffle.
BTW, seems folks never remember that rear waves supplement the front waves as often as they diminish, depending on wavelength, of course.
B.
Last edited: