I presume you saw Q decrease with increasing panel resonance frequency, which would be consistent with many materials' internal damping behavior.
When I found that the bracing wasn't adequate to render panel vibration inaudible I added more bracing until I found the required brace spacing that rendered panel vibration inaudible. Once I achieved that I added damping as well, with no significant benefit realized. That's when I decided to go with bracing alone, rather than a combination of less bracing plus damping, as the former is less expensive in terms of both materials and labor.
Is there an approximate panel resonance frequency at which inaudibility was achieved?
I like your answer, as CLD is a PITA.
Now, instead of using your hand to stop the cone from moving add dense material to the cone, like self-stick EPDM. SPL will drop, and the more material you add the lower the SPL will be, as the cone mass becomes too great a load for the motor to overcome. That's what damping does.
Actually that's mass-loading, which is reactive; damping is resistive.
But I take your point about reducing panel motion.
Earlier I was thinking that moving the freq higher, whose natural effect of reduced displacement with constant input is the same as with a driver, was just moving the problem to higher and likely more audible freq, because SPL would stay the same like for the driver.
However that's only on-axis, and power response should drop for the panels as well and have a reduced contribution to total SPL, at least on axis.
Doesn't constrained layer damping mean that this thin elastic/adhesive barrier is "constrained" meaning that its constrained on both sides not just one. Just adding a damping layer to a panel won't do much because you have to constrain it between two layers to achieve the dampening effect, ie. turning the energy into heat.
Yes, that's what CLD is but it's not the only type of damping; shaking a lossy material will also absorb energy.
Actually I was referring to Bill Fitzmaurice's damping test. It appears that the test enclosure didn't employ CLD technique but just glued the damping to the sides of the finished box. To check the effectiveness of CLD verses braces he would need two separate boxes, one employing just braces and the other sandwiching the damping between two layers of siding material.
I like the idea of using moderate bracing and CLD. With Sikaflex its not that much more work aside from the necessity of using laminated sides.
Sounds like it was supposed to be CLD, but as I said earlier I think the rubber is way too thick to give significant shear strain.
CLD works by having a damping material between the two panel layers that will absorb as much motion of the inner panel as possible, rather than transferring it to the outer panel, thus reducing the motion of the outer panel. That's best realized with a thick damping layer. If the inner panel is sufficiently braced so that it doesn't flex to begin with then there's nothing much to be gained from the damping layer. For that matter, CLD can have worse results. With a single layer panel the restraining force of the brace is mechanically connected to the outside of the panel. With CLD it is not.
What you describe would work by tension/compression deformation of the rubber, which is *not* how CLD works; it works by *shear* deformation.
You're correct that it's ineffective to put damping material near the connection point of a brace because there's essentially no motion to damp.
That's why CLD mfgr's recommend that for maximum economy CLD need only be applied in the central regions of unsupported panels where deflection and shear deformation between the panel and CLD material are greatest.
This observation facinates me.
When I think about the compression/vacuum which the cavity undergoes in a sealed configuration, I think of a spring mass system. I also think about the Q of the box and how bracing raises the resonant frequency, ideally beyond excitation frequency of the driver. It doesn't make sense to me how harmonics can be generated. This appears to me to be an LTI system.
But I've seen THD plots of drivers. I'm not confident on the mechanics that produce the harmonics but I vaguely remember a post some years ago that it is caused by the cone wave rippling, hitting the surround and folding back in. If this is true, then I propose that the cone "waving" is more significant in the design of the enclosure structure than the actual rigidity of the box. The box design requires consideration for the function of absorbing the cone wave, transfering this energy into the box , damping it and prevent it from bouncing back into the cone. Anybody? What causes harmonic generation in any speaker system?
To test a box, I have some ideas.
Drop a ball bearing on the dust cap of the woofer, record the output, and FFT the impulse.
Drop a ball bearing on the side panels of the box, record the output and FFT the impulse.
But I've seen THD plots of drivers. I'm not confident on the mechanics that produce the harmonics but I vaguely remember a post some years ago that it is caused by the cone wave rippling, hitting the surround and folding back in. If this is true, then I propose that the cone "waving" is more significant in the design of the enclosure structure than the actual rigidity of the box.
Does anybody know how to model a passive radiator? It strikes me that one component of the box design resembles a passive radiator
There was also some discussion about decoupling the driver from the box. I don't think this description is quite accurate. I think its better to describe it as absorbing the driver energy, transferring it to the box where it can be damped or mechanically rendered inaudible. So the driver interface is relatively critical to impedence match. Kind of like impedence matching transformers. I don't think I have the competency to study wave mechanics.
as a noob builder its a tremendous insight
For the SUBMAXIMUS, bracing of the chamber I can see and will plan for the benefit of the triangular/semicircle/dihedral smartly spaced internal panel to panel away from the (very firmly) secured edges.
For the long Panel A which is so far secured only by the edges, it would have made more sense to prefab A and B together to get the long braces between them
its like the baffle braces I fussed over when building the THTLP except for access to the end that goes up to the end of panel H at B. there is not much room to maneuver.
and thank you all, especially post 212.
Dude! Where's my pencil?
DIY FAN Denon X5200 , ATI A 2000 for 7.4.6 SCATMOS/DSU/SHARP 80" LED/LCD
L/R: Fusion 15 V2 , C: 88 Special , SL/SR: 88 Special(V2) , RL/RR: F-3, TF/TR: Volt 6's TM: SLX, FH: F4Q4
SUBMAXIMUS, ,Submaximus V3,LOWARHORNCustom Dual Driver VBSS,2 x Inuke6000DSP
There are some interesting measurement here
(Google Translate if needed)
The combination of 16mm MDF, 4mm Bitumen and 4mm Plywood is the best combination.
BBC have done studies of audibility of resonanses both in Q and in frequency range. And their finding suggest that at least for two way systems the best was to have "limp" cabinets to keep the resonanses in the bass range and out of the more critical midrange.
I do understand that the proper solution will vary with the application. The horn throat of a 15" bass driver with compression used up to 80 Hz is very different from a nearfield monitor cabinet.