You might better have placed the butyl between the enclosure panels and the braces, as KEF with their FEM analysis have shown. Anyway, resonances of enclosure panels are mainly excited by the chassis counterforces and (very much) less so by air pressure variations.
Through the years I have become a bit sceptical to a ‘broadband’ approach of panel resonances of loudspeaker enclosures. Of course extensive measures won’t hurt. But the elegance of separating LF enclosures from MF enclosures and thus minimizing the net result of panel resonances in a much more effective way seems forgotten too often to me.
Last word on this for now: non-resonant transmission through enclosure panels hardly seem problematic to me. They cause only smal linear distortion effects (with -theoretic- effect on directivity) which might hardly be measurable, let alone perceivable.
Through the years I have become a bit sceptical to a ‘broadband’ approach of panel resonances of loudspeaker enclosures. Of course extensive measures won’t hurt. But the elegance of separating LF enclosures from MF enclosures and thus minimizing the net result of panel resonances in a much more effective way seems forgotten too often to me.
Last word on this for now: non-resonant transmission through enclosure panels hardly seem problematic to me. They cause only smal linear distortion effects (with -theoretic- effect on directivity) which might hardly be measurable, let alone perceivable.
Last edited:
Do you have some source? Would be interesting to see some analysis results that show where most of structure vibration comes from, I really do not know for sure. I had guessed most critical is modes of the sound in the enclosure, where the pressure knots sit on the cabinet walls and excite them. Of course when you size the inner dimensions sufficient small that the drivers passband stays below the modal range, this situation does not occur.
I made good experience to support woofers and mids with a bracket at the magnet to the back panel (with some butyl smeared into the tight gap to couple, helps also to damp). Significant less vibration of the front panel, at least "measured" with my fingers at higher volume. Back panel vibrates more, but overall sound gets cleaner.
I also care that the brackets divide the panel area so that the remaing "Plates" between the panels have a plate resonance that is significan higher and out of the passband of the driver. As you said, increasing the mass by e.g. butyl sheets or similar that the resonance is lowered so that it reaches the driver's passband again would be contraproductive.
Some rough example of a typical free area part of panel shows that these resonances occur mostly in the midrange. So the mid woud be my preferred canidate to decouple from the enclosure, if such measurements come in consideration. Or give the mid a separate enclosure to decouple the structural resonances from the large bass cabinet that has much more wall area, and has so small panels that their resonances are much higher than the passband of the mid.
Yes this is also a good idea. I had put his thing here in a speaker tensed between the side walls and glued on the Aluminum-Butyl sheets after assembly. Cabinet got very quiet.
Not by hand but I will do a quick search. From a force equilibrium pov it makes sense anyway, the forces are a magnitude bigger than those of the enclosed air on the panels.
Two things: if you have such high amplitude resonances, you’re doing something wrong in the internal damping that probably will emerge through the cone or port, rather than the enclosure wall. And it’s not that likely that such resonances excite panel resonances at exact that spot and frequency range. So in essence mass law transmission loss applies for perpendicular incidence (very simply put) and any reasonably heavy panel will dampen enough.
But like I said earlier, too much attention won’t hurt here. It’s not that expensive and often only costs labour. So please continue…
The butyl-alu sheets you use appear to be rhe same i used. But no longer, as over time the butyl softens to a thick paste.
There is a lot of what if that goes into panel vibration and noise.
Much of frequency dependent vibration shows up in the Harmonic Distortion Plots as second or third HD. Personally I do not believe that it is always audible. It can often be measured.
Back when the 6-1/2 inch Purifi drivers were new I fabricated a closed test enclosure and made some test measurements and found some peaks in the FR and HD curves at aprox 450Hz, 900Hz and 1350Hz. Scratched my head and sinched down on the screws into the T-nuts pulling the driver tighter to the enclosure front. The distortion peaks got worse. I loosened the screws and the distortion peaks improved. I completely removed the driver and added a 4 mm closed cell neoprene gasket and pulled down the screws "just right" and the "distortion peaks" all but disappeared into the measurement floor.
I also ran some Rub and Buzz plots of the above test setup, there were spikes in the plot that were also removed with the thicker closed cell neoprene gasket.
Thanks DT
Much of frequency dependent vibration shows up in the Harmonic Distortion Plots as second or third HD. Personally I do not believe that it is always audible. It can often be measured.
Back when the 6-1/2 inch Purifi drivers were new I fabricated a closed test enclosure and made some test measurements and found some peaks in the FR and HD curves at aprox 450Hz, 900Hz and 1350Hz. Scratched my head and sinched down on the screws into the T-nuts pulling the driver tighter to the enclosure front. The distortion peaks got worse. I loosened the screws and the distortion peaks improved. I completely removed the driver and added a 4 mm closed cell neoprene gasket and pulled down the screws "just right" and the "distortion peaks" all but disappeared into the measurement floor.
I also ran some Rub and Buzz plots of the above test setup, there were spikes in the plot that were also removed with the thicker closed cell neoprene gasket.
Thanks DT
Nothing.
It was not full on rubber isolation of the driver.
Another driver and another box I did use a rubber grommet with a washer to keep the screw head from passing through the grommet.
The rubber gasket did wonders all by its' self if the screws were torqued down "Just Right"
DT
KEF sold their constructors’ B110’s with gasket and grommets. Back in the eighties… So history repeats itself.
Hello,
Not to loose, not too tight but somewhere in between.
Too loose flops around.
Too tight flattens the springs (closed cell neoprene).
Measurement is the method.
The vibration will show up as "harmonic distortion" in a distortion plot. It will show up as spikes in a Rub and Buzz plot. It may also show up in the impedance plot as resonance peaks. Turn off all the processing, curve fitting and smoothing in the impedance measurement software.
Noise and vibration isolation was a large part of my day job. lots of mass, isolation springs and damping.
Speaking of mass, I think that I may try concrete wall board glued to the inside of the enclosure, you know the concrete wall board behind the tile walls in your shower.
Thanks DT
https://www.ap.com/news/raising-the-bar-for-rub-buzz-defect-detection
https://pearl-hifi.com/06_Lit_Archi...ication_Notes/AN_23_Rub_and_Buzz_Learning.pdf
https://pearl-hifi.com/06_Lit_Archi...ication_Notes/AN_23_Rub_and_Buzz_Learning.pdf
And from Audioexpress
Last edited:
@DualTriode Thanks for your elaboration. Need to digestief the info in the links you provided.
We share working experience, mine is from way back starting as chief engineer structures for the walrus class submarines. Lateron i focused loudspeaker design and 1/8 gas powered model racing cars. I am now retired but still avtive in loudspeaker design and optimisation.
Mass , spring, vibrational energy paths, absoption, isolation are interesting aspects to combine.
Reason asking for what in between screw head and chassis was that that in particular can cause spikes when other side of chassis is resilient mounted.
So stiff but not too stiff is a way out of that trap. Still not sure if that is a lasting solution.
We share working experience, mine is from way back starting as chief engineer structures for the walrus class submarines. Lateron i focused loudspeaker design and 1/8 gas powered model racing cars. I am now retired but still avtive in loudspeaker design and optimisation.
Mass , spring, vibrational energy paths, absoption, isolation are interesting aspects to combine.
Reason asking for what in between screw head and chassis was that that in particular can cause spikes when other side of chassis is resilient mounted.
So stiff but not too stiff is a way out of that trap. Still not sure if that is a lasting solution.
This is also a good reference, a AES publication.
https://www.researchgate.net/public..._Using_a_Simplified_Auditory_Perceptual_Model
An added thought:
The typical Rub and Buzz conversation focuses only on the driver. I do not feel that there is fault in extending the conversation to include the attachment to the enclosure even including air leaks and enclosure wall vibrations.
Thanks DT
https://www.researchgate.net/public..._Using_a_Simplified_Auditory_Perceptual_Model
An added thought:
The typical Rub and Buzz conversation focuses only on the driver. I do not feel that there is fault in extending the conversation to include the attachment to the enclosure even including air leaks and enclosure wall vibrations.
Thanks DT
More construction photos:
Below is the inner carcass fully glued up. The first layer of the woofer baffle is attached.
Below is a trial fit of the baffles with all three drivers - nothing glued or attached at this point. There are three sections to the baffle, the mid-tweeter section, the transition block, and the woofer baffle. The mid tweeter baffle and the transition block were completed some time ago. The parts for the woofer baffle were just machined. In the photo it appears that the woofer is positioned ahead of the midrange, but this is an illusion. All three drivers are in the same plane.
Next steps are to glue up the woofer baffle and sand it flush with the main carcass. Then I need to do finish sanding of the mid-tweeter baffle and the transition block. Next I veneer the main carcass and woofer baffle, then I attach the mid tweeter baffle. Finally, I apply some trim around the perimeter of the mid tweeter baffle.
j.
Below is the inner carcass fully glued up. The first layer of the woofer baffle is attached.
Below is a trial fit of the baffles with all three drivers - nothing glued or attached at this point. There are three sections to the baffle, the mid-tweeter section, the transition block, and the woofer baffle. The mid tweeter baffle and the transition block were completed some time ago. The parts for the woofer baffle were just machined. In the photo it appears that the woofer is positioned ahead of the midrange, but this is an illusion. All three drivers are in the same plane.
Next steps are to glue up the woofer baffle and sand it flush with the main carcass. Then I need to do finish sanding of the mid-tweeter baffle and the transition block. Next I veneer the main carcass and woofer baffle, then I attach the mid tweeter baffle. Finally, I apply some trim around the perimeter of the mid tweeter baffle.
j.
Here is a close up side view of the transition block, with the woofer hole on the left side. The M-T baffle is recessed by about 1 mm, but that will be taken care of during trim and finishing. Other than that, all three drivers are in the same plane.
It is not just photos that create the illusion. Seeing it in real life there is a definite impression that the woofer baffle juts forward of the mid and tweeter.
It is not just photos that create the illusion. Seeing it in real life there is a definite impression that the woofer baffle juts forward of the mid and tweeter.
It is the darker colour of the transition piece and the MT part that create this illusion.
Here is the same kind of illusion: I just superimposed semitransparent white or grey triangles on the pic in a photo editor on a completely flat baffle and the illusion of a bevel is very strong.
Here is the same kind of illusion: I just superimposed semitransparent white or grey triangles on the pic in a photo editor on a completely flat baffle and the illusion of a bevel is very strong.
Last edited:
My progress has been slowed a bit by another project. It is not an audio project, but it may have some crossover interest to some people on the forum, considering the fabrication and woodworking aspects.
I am making an anchor pulpit for my boat. It is made from 13 mahogany boards (or strips), 2" x 1" x 72". Each strip was fully encapsulated in epoxy, and then the strips were epoxied together. The resulting beam is 70" long, 2" thick, and 13.5" wide. I cut the circular radius on the end by hand with several different hand saws, then sanded it to the finished arc.
The electric windlass (winch) can pull the anchor chain with up to 1300 lb of force, so this cantilevered beam needs to be rather stout.
Back in the world of active 3-way speakers, here is some progress... I am almost finished with veneering the cabinets. The photos below show the mid-tweeter baffle and the transition block set in place, but they will not be permanently attached for some time yet. The profiling of the MT baffle is going well, but I need to do a bit more shaping and smoothing. Overall, I am pleased with how it is turning out.
j.
I am making an anchor pulpit for my boat. It is made from 13 mahogany boards (or strips), 2" x 1" x 72". Each strip was fully encapsulated in epoxy, and then the strips were epoxied together. The resulting beam is 70" long, 2" thick, and 13.5" wide. I cut the circular radius on the end by hand with several different hand saws, then sanded it to the finished arc.
The electric windlass (winch) can pull the anchor chain with up to 1300 lb of force, so this cantilevered beam needs to be rather stout.
Back in the world of active 3-way speakers, here is some progress... I am almost finished with veneering the cabinets. The photos below show the mid-tweeter baffle and the transition block set in place, but they will not be permanently attached for some time yet. The profiling of the MT baffle is going well, but I need to do a bit more shaping and smoothing. Overall, I am pleased with how it is turning out.
j.