If the cochlea did not have a "pressure relief" window at the other end, the low frequency response would be negligable. This release allows the travelling waves at LFs that are required for excitation of the hair cells.
Member
Joined 2009
Paid Member
This makes it even more like a Transmission Line speaker.
p.s. I thought the window at the end was to allow sound from your spouse to flow through unimpeded 😀
p.s. I thought the window at the end was to allow sound from your spouse to flow through unimpeded 😀
Again you couldn't be wronger! 😉 The traveling wave in the cochlea is moving at a speed of about 10-30 m/s (depending on frequency and depth into the cochlea). Try that with a TML. 🙂
Rudolf
Member
Joined 2009
Paid Member
I like being wronger !
Transmission lines aren't restricted in terms of their propagation speed, it all depends on the details.
[many years ago I published a paper on a slow-wave transmission line I made for use at GHz frequencies]
Transmission lines aren't restricted in terms of their propagation speed, it all depends on the details.
[many years ago I published a paper on a slow-wave transmission line I made for use at GHz frequencies]
I'd like to hear those speakers you posted Bigun. Looks like there was a lot of thought put into their design. Thanks for posting those.
Might be they have developed a cutting edge acoustic filter combination at the rear?
Love to get a peek inside , too.
Might be they have developed a cutting edge acoustic filter combination at the rear?
Love to get a peek inside , too.
Member
Joined 2009
Paid Member
if you think those were interesting, you might like to check this out - this manufacturer went so far as to show measurement data proving their boxes have no colourations. The approach won't be a surprise - make it as stiff as you can, and then damp any resonances. Doesn't address the back wave concern though.
Magico Technology | Enclosures
Magico Technology | Enclosures
What a waste of aluminum and machining time. I'm not sure how it would have any damping qualities unless you left the screws a bit loose.
My dad had a JBL D130 in a Hartley Boffle type cabinet. It was perhaps 6 layers of horse hair carpet backing in a 6-8 cubic foot box. I don't remember the sound as it was 40 years ago! But D 130s never have bass (Q of about 0.15) and the boffle wouldn't help.
The damping layers would approach the "limp suspended mass" ideal that gives isolation without resonance, but lead loaded foam sheets (used in room issolation) would do better. Having multiple layers with an air space in between might improve the issolation, the same as with multipane windows. The holes through the various layers are a bad idea as all the bass would make it through.
This brings up a point, the best source of information on sound absorption and noise issolation is in architectural acoustcs. Try the NRC archives on wall constructuion to see the TL curves on various wall types. All the measurements on material absorption and transmission loss are done in the architecture field.
David S
Hi speaker dave. I have improved bracing afterwards with my speakers to get rid of some cabinet resonances. My experience is that everytime you put in a new bracing (to dampen a certain resonance) it introduce a new different resonance somewhere else, altough less in amplitude than the one before. Is this what happens? It seem almost impossible to get a totally dead cabinet, is that possible and how?
That would be pretty typical. Bracing provides stiffness but not damping. Most cabinet panels have multiple resonances and harmonics of those resonances like standing waves in a room. If you put a brace from the center of the front face to the center of the back brace you do preclude a few of the lower resonances (the ones with veleocity maxs in the center of a face, but probably accentuate upper resonances. This is just as a woofer dead center in the room can only drive certain modes.
People use the term "dampen" all the time but generally only mess with mass and stiffness when they build cabinets. Dampening is a property of viscous materials, such as hard rubbers, tar, packed sand, etc. Dampening will reduce the Q of resonances and that is very desirable. Messing with stiffness and mass will not reduce resonant Q, but will typically just raise resonant frequencies. Studies have shown that resonances are more audible at higher frequencies and typical cabinet construction can't possibly push the resonances out of the audible band.
The answer is to forget about stiffness and go for mass provided by flexible damping. Architectural acousticians know about this and talk about "limp mass" sound absorbers, that have no coincidence frequencies (resonances) where transmission loss is degraded.
Mass Loaded Vinyl
David S.
People use the term "dampen" all the time but generally only mess with mass and stiffness when they build cabinets. Dampening is a property of viscous materials, such as hard rubbers, tar, packed sand, etc. Dampening will reduce the Q of resonances and that is very desirable. Messing with stiffness and mass will not reduce resonant Q, but will typically just raise resonant frequencies. Studies have shown that resonances are more audible at higher frequencies and typical cabinet construction can't possibly push the resonances out of the audible band.
The answer is to forget about stiffness and go for mass provided by flexible damping. Architectural acousticians know about this and talk about "limp mass" sound absorbers, that have no coincidence frequencies (resonances) where transmission loss is degraded.
Mass Loaded Vinyl
David S.
I was thinking about this kind of bracing the other day. Wouldn't the brace itself vibrate, and "drive" the panels the way an acoustic guitar string resonates the soundboard?
What if I sprayed the inside of the cabinet with rino lining type stuff? They have some that is like a rubber/tar type I think might work. It would be the last step in the process so the braces get coated as well.
I'm sure any brace would have its own modes as well (all physical structures do) but they should be pretty high up in frequency and won't couple to the outside air very easily. It will be very stiff in compression between the front and back panel. I wouldn't worry about it.
I also wonder if a damping brace between panels could work, perhaps with two overlapping boards with a viscous link between the two.
David S.
I also wonder if a damping brace between panels could work, perhaps with two overlapping boards with a viscous link between the two.
David S.
Like the building foundations in earthquake zones ... in particular Japan. Bracing in these structures are isolation pads .. ie. shock absorbers.
I'd bet that a guy could use a sort of (3) panel articulated bracing where any two other panel(s) motion would be counteracted by the third. They want to move out the others pull them back an equal amount.
... a "Y" with a pivot @the joint of the letter ... and two additional pivots, one @each end of the top of the "Y". Bottom of the "Y" is back panel ... each cabinet side is attached to a top branch of the "Y".
... guessing this wouldn't help with resonance though.
I'd bet that a guy could use a sort of (3) panel articulated bracing where any two other panel(s) motion would be counteracted by the third. They want to move out the others pull them back an equal amount.
... a "Y" with a pivot @the joint of the letter ... and two additional pivots, one @each end of the top of the "Y". Bottom of the "Y" is back panel ... each cabinet side is attached to a top branch of the "Y".
... guessing this wouldn't help with resonance though.
Last edited:
But a high frequency box resonance is less likely to ever get excited. Energy available to excite the resonance is inversly proportional to the square of the frequency. An argument can be made that it is actually more like inverse to the 4th power of f. And if the resonance that high is high Q it is even less likely to get excited, because a lot of continuous energy needs to be pumped into to get it going, and this is very atypical of music.
This is a chart i made of a curve starting at 300Hz = 1 (the hinge point for the 4th power argument), showing the decrease in energy for the 2nd & 4th power reductions (for comparitive purposes only), below that one from Linkwitz that will likely have a greater relation to real numbers. They show that one only needs to get the panel resonace up, not out of band (given that most of my stuff is FR not possibly with practical tech)
If a resonance is never excited then it is if it does not exist.
dave
Attachments
I like it. Damping the central pivots would be great. It could be difficult getting the sides and the back panels into the same plane though.puppet said:
Member
Joined 2009
Paid Member
I believe that viscoelastic energy dissipation is a frequency dependent hysterysis - so moving resonances to higher frequencies makes any damping we apply even more effective. Not sure how this combines with your 4th order chart Dave, but it could make it even better.
But a high frequency box resonance is less likely to ever get excited. Energy available to excite the resonance is inversly proportional to the square of the frequency. An argument can be made that it is actually more like inverse to the 4th power of f. And if the resonance that high is high Q it is even less likely to get excited, because a lot of continuous energy needs to be pumped into to get it going, and this is very atypical of music.
This is a chart i made of a curve starting at 300Hz = 1 (the hinge point for the 4th power argument), showing the decrease in energy for the 2nd & 4th power reductions (for comparitive purposes only), below that one from Linkwitz that will likely have a greater relation to real numbers. They show that one only needs to get the panel resonace up, not out of band (given that most of my stuff is FR not possibly with practical tech)
If a resonance is never excited then it is if it does not exist.
dave
Not buying any of that. I've seen plenty of box resonance measurements and the Q's tend to be the same for resonances in any frequency range. Sowter and Harwood also took measurements of cabinets of essentially the same design but varied panel thickness. As you would expect doubling panel thickness doubles resonances (mass doubles and stiffnes is times 8 so primary resonances double). Most of the discernable resonances just shifted up in frequency, non disappeared.
The cabinet vibration curves I've seen are all done with constant voltage and tend to show pretty constant energy (peak level of resonances) well up through the midrange, so a sine sweep has no problem exciting higher resonances. The spectrum of music is fairly flat, especially if maximum energy is considered so an upper range resonance is as likely to be excited by music as a lower range resonance.
If bandwidth or Q don't change then the resonance width is not a function of frequency. (percentage width or width expressed as a fraction of an Octave, higher resonances are actually wider in total Hz). An upper range resonance is as likely to be excited as a lower range resonance.
I hope you are not talking about missing resonances by placing them off of the notes of a 12 tone scale. For those that are intrigued by that consider:
All transients will be broad band or noise like in nature. Most instruments, and especially percussion instruments, contain a lot of energy "off the notes". Plus people and instruments frequently use vibrato so they will wobble about a key (or, before autotune, just be off key). Plus instruments drift in pitch with environmental effects.
Plus you might pitch a single fundamental cabinet resonance between 12 tone scale frequenciess but harmonics of the resonance are not integral multiples of the fundamental, so they can't all fall between the notes (the same even applies to piano strings).
Finally, Harwood conducted his own listening tests and found that resonances at higher frequencies were more audible rather than less.
David S.
Mechanical impedances are higher at higher frequencies. Making box walls stiffer or heavier tends to make typical damping less effective. Read Harwood's excellent paper on the subject.
David S.
What about the frequencies the box is actually subjected to?
My current project bass box as about 8 cu ft and produces f up to about 300 Hz. The mid is a stacked ply design with a thick plastic cap on the back. Under the theory that the stiffer the box the higher the resonances, isn't it possible to push the resonances so high that the frequencies the box actually produces will never excite the box resonances?
What should be the rule of thumb - build to the highest quarter-wave harmonic?
My current project bass box as about 8 cu ft and produces f up to about 300 Hz. The mid is a stacked ply design with a thick plastic cap on the back. Under the theory that the stiffer the box the higher the resonances, isn't it possible to push the resonances so high that the frequencies the box actually produces will never excite the box resonances?
What should be the rule of thumb - build to the highest quarter-wave harmonic?