RMAF 2019

My understanding is…

MBL uses something similar to a conventional electrodynamic woofer motor attached to the end of the blade set, and on excursion of the motor it thrusts axially on the blade set which flexes the centers of the blades outward providing excursing displacement of the bladed driver. For each blade the ends are narrower than the middle, and the blades do not appear to have significantly varied thickness, so each blade will bend more nearer the ends and less nearer the middle (not a constant radius of curvature), which would cause broad area excursion on the blade radiating as a prolate spheroid rather than plain spheroid, which is more toward cylindrical than plain spheroid. The effect would be that measure of vertical polar response of that driver should show much higher SPL nearer the horizontal plane than the vertical plane. The varied blade width would also affect bending modes and any associated resonances.

While that varied axial thrust will induce modal flexure and resonance at higher frequencies, I think the design tries to avoid higher order bending modes in the blades, and that bladed driver would be low passed well below chaotic "break-up" modal resonances. I do not know if the blades include some constrained layer damping, but if they do that would reduce the power conversion efficiency. I expect that blade flexure at higher frequencies of the bladed driver pass band should be more like a textile dome tweeter at high frequencies, where the textile domes are designed to flex with low order bending modes, with coherent phase and not chaotic, without much chaotic breakup in the passband. Likewise the blades would be designed to flex with low order bending modes dominating the flexure.

That would be very different from modal resonance in a DML, which at higher frequencies is a diaphragm intentionally operating in breakup modal resonance … in the passband. Unlike the undesirable modal breakup resonances in a metal cone woofer, that breakup in the DML is utilized as a feature rather than a fault, is better managed for smoother usable response in that wideband breakup. The modal resonances become very complex, chaotic, with distributed phase, radiating non-directional diffuse sound, and therefore do not much interfere constructively or destructively with other sources, and so sum with other sources with largely uncorrelated phase at higher frequencies. You can see some of that in a conventional metal cone woofer by looking at off axis response in the breakup resonance frequencies, where the woofer beams with narrowed directivity below the breakup(s) and then radiation transitions into broad directivity in narrow band in the breakup modal resonances where local patches of diaphragm are radiating separately and chaotically. A good crossover for that woofer would trap the resonances and place those deep enough in the stop band in the low pass to make breakup resonances inaudible in the application. In the DML, that chaotic breakup is desired, and associated peaks can be smoothed by tweaking the panel geometry and the position of the panel exciters on the panel.