I feel like this lines up with what I meant by the idea that it is wise to study TL, the one enclosure that can actually mute the back wave. Once you see what it takes, you can use these tactics in a practical sense for a non TL. Coincidentally, folding the line creates the same affect that our desired bracing would have, that is, raising resonant note of those unintentional resonances. For the enclosure chosen an actual line might not work but, breaking up what would have been a line into segments, using creativity and some insight, and you’ll find the bracing resembles what people have found successful. Damping material, density of it and the distance travelled through it, can be monitored as a system that is most effective in a line, or some type of practical compromise, but non the less dampening material is the real killer of resonance. Attenuating the source stimulation of the resonance is what damping material can do but depending on the design it can only do this so much or not at all if it augments fr undesireably. A higher resonance note is easier to mute with damping material...higher frequency carry less energy than lower which is why bass is heard coming through walls not treble. The q of the resonant note is connected to the surface area of the resonant panel. Breaking up the plane of large panel with bracing raises the resonant note and the q. Just like with Eqing, a high q filter is less perceivable than one with a low q. When it comes to resonant control it seems there are Transmission lines and then subsequently there are compromises.
Without damping the resonances will persist forever so that is easy to answer. If you want the resonances to decay you will have to define the damping present for each of your panels.
As mentioned above, the radiated SPL is not only a function of the velocity of the panel it is also a function of the area (and some other things).
You also haven't mentioned how energy is put into the panel. An impulse? This matters because the spatial location will determine how much energy is received by each mode.
Finally, individual panel resonances are of limited practical relevance since only a small number of modes, sometimes only one, will be of low enough order, driven and radiate efficiently to the listening location. These modes usually involve the speaker deforming as a whole with multiple panels moving together and the drivers bouncing or rocking on the baffle. Studying individual baffles or worse the bracing of individual panels may well be more misleading than informative when it comes to designing speaker cabinets.
lousymusician said:
Modal analysis will answer that for you since it decomposes the whole cabinet motion into a weighted sum of the individual modes/resonance. At low frequencies a small number of the lowest frequency modes will contain pretty much all the energy. Damping will reduce the level of a mode just as much away from the resonant frequency as on it. This is why when you look at measurements/simulations of the cabinet radiation in the KEF or BBC papers linked to earlier damping pulls down the SPL everywhere and not just at the resonant peaks.
The notion that you cannot hear loud cabinet resonances is directly refuted in the BBC paper where they identified which undamped resonances could or could not be heard and how much damping reduced them to inaudible levels. Note the damping used in the BBC study is cheap and easy to apply global damping which is weak compared to what can be done if you first understand which resonances are a problem and target damping at their mode shapes (and stiffness in some cases in order to move the frequency, and geometry in order to add or break symmetry, and...)
So what then is a back wave? Describing the sound pattern from a dynamic loudspeaker driver as a planar wave is oversimplifying things.
Damping TL's is exactly the worst example as it's always too much (line with no acoustic output) or too little (output wit undesired resonances).
Sound power is frequency independent and because power x time = energy this statement is false. Transmission through any material though is frequency dependent, whether it's air or solid concrete.
Try to discover that a string really resonates longer when you shorten it. Take your guitar and play 😉
That high Q filters are less perceivable is true to some extent. It's no universal truth. And while transmission lines may do fine in electromagnetism, in acoustics they're only one another -very complicated- way of achieving a certain goal.
The sound wave coming from the back of the diaphragm
A certain and known approach of TL is to completely mute the backwave, ie no acoustical output. My point was to understand how it is done, in order to control the aspect of manipulating the level of the backwave, and use it in a practical manner for other designs
Practically speaking, treble is going to bounce off a of a solid border, and bass is going to more easily transmit through it. It takes more energy to create bass at equally perceived level vs treble, thus dampening it is likewise. We are working with systems that are presenting power x time=energy on a equal loudness basis, thus, there is going to be lower energy in the backwave at high frequencies.
untrue...the guitar string will simply resonate longest at its materials resonant frequency. Shortening the string may bring the note closer to the resonant note of the wire material.
Agreed. The body of air within the cab as a whole represents a mass/density/volume by itself and has its own resonant frequency, as well as the reflective resonance causes by walls within the cab, as well as the resonant frequency of the material of the enclosure.
its not universal but for what we are dealing with, it is true. Do you have any examples of where this doesn't work out?
I've wondered if there where any other ways to completely kill the backwave, what other ways do you know of?
Last edited:
Here is cool paper, maybe something to try with speaker enclosures as well?🙂 The acoustic black hole: A review of theory and applications - ScienceDirect and another Acoustic Black Holes for Flexural Waves: A Smart Approach to Vibration Damping - ScienceDirect
"One of the most important advantages of the above-mentioned one-dimensional and two-dimensional acoustic black holes as dampers of structural vibrations is that they are efficient even for relatively thin and narrow strips of attached absorbing layers. The reason for this is that wave energy dissipation takes place mainly in a very narrow area near sharp edges. This is in contrast with the traditional techniques employing covering the whole surfaces of structures by relatively thick layers of absorbing materials [1,2]. " - copy paste from the latter.
Some practical stuff in this paper https://pdfs.semanticscholar.org/729e/89900553ce90feb90e3adec1611a0a7f5985.pdf
"One of the most important advantages of the above-mentioned one-dimensional and two-dimensional acoustic black holes as dampers of structural vibrations is that they are efficient even for relatively thin and narrow strips of attached absorbing layers. The reason for this is that wave energy dissipation takes place mainly in a very narrow area near sharp edges. This is in contrast with the traditional techniques employing covering the whole surfaces of structures by relatively thick layers of absorbing materials [1,2]. " - copy paste from the latter.
Some practical stuff in this paper https://pdfs.semanticscholar.org/729e/89900553ce90feb90e3adec1611a0a7f5985.pdf
Last edited:
When is a wave a wave and not just a pressure variation?
Build a TL, stuff it until the port is completely silent. This system behaves about the same as a closed box of the same size and the same stuffing.
I completely lost you here.
I don't get this either. And yes, I know what a Phon is.
Only if you push the loudness button. Though spectral energy distribution in music isn't linear indeed.
Lost you again here. So if I play an E on my guitar, it might not be a resonance, but if I change to g, it might? But that wasn't my point either, I referred to the Q of the note played.
Resonance in it's most simple form requires a spring, a mass and an initial movement from an equilibrium. Or the equivalents, a capacitance, a reactance and a current pulse will do fine too. Or, in air, a certain enclosed volume and a certain mass that can be coupled to that enclosed air. A port, a passive radiator or even a loudspeaker driver. The enclosed air in itself doesn't have a resonant frequency.
Yes, a note played on a guitar. High Q and perfectly audible.
There is no such thing as complete kill. Sufficient killing is perfect though. Several ways are available, a lot of them use some form of conversion from air movement to friction induced heat.
A transmission line de facto isn't meant to be 100% lossy. Loudspeaker designers of the past converted the electromagnetic principle to an idea to reinforce bass output of speakers in their resonant domain. Side effects have proven to be hard to predict, the building of a TL (which is perfectly viable) is always a trial and error process.
Ok, thanks for that andy, but unfortunately that didn't answer my question. I didn't ask about decay. I didn't even mention that the panels are part of a cabinet. I also purposely didn't mention what the material the panels are made from. I'm more interested in how loud the resonances would be relative to the different panel configurations given that they are all made of the same material and are all given the same level of energy input, whatever that might be, although pertinent to the discussion it might be an input signal of sound energy containing 1400Hz. But I would expect to get the same results if the panels are struck too, but perhaps I am wrong there.
Or maybe the answer is so complicated that I'm not even asking the question properly.
jReave - -
If you had selected 3 examples which all had the same dimensions (except for thickness) then the answer is "Yes", without a doubt. The thicker panel has higher stiffness, more mass, and more damping. All three factors will drive the displacement down, and SPL is of course related to displacement. The thicker panel would have a higher mode 1 frequency than the thinner panels.
In the case you have presented, the 3 examples have different shapes and sizes, which you carefully selected so that all 3 panels have the same mode 1 resonant frequency of 1400 Hz.
I don't really want to work through the math here, but I believe firmly that if the input forcing function is an impulse, the thicker panel will have less SPL at 1400 Hz than the thinner panel given the same impulse. For a given energy impulse, like a hammer strike, the higher mass will result in lower acceleration, lower velocity, and lower displacement.
If the input forcing function is a continuous 1400 Hz sinusoidal signal, then I am not sure what happens. Each panel will vibrate at its 1400 Hz resonance, and the magnitude will grow with each cycle until it is arrested by the internal damping of the material. There is no doubt that the thinner lighter panel will reach its maximum displacement in fewer cycles than the more massive thicker panel. In other words, it will take less energy for it to achieve its steady state vibration level.
The larger thicker more massive panel will absorb more energy, more cycles of 1400 Hz forcing function to achieve steady state. It will take a longer time.
But given the greater surface area of the thicker panel, I do not know which would produce the greater SPL once they have reached steady state vibration. It may depend on the level of natural damping in the material.
Does this make sense?
If you had selected 3 examples which all had the same dimensions (except for thickness) then the answer is "Yes", without a doubt. The thicker panel has higher stiffness, more mass, and more damping. All three factors will drive the displacement down, and SPL is of course related to displacement. The thicker panel would have a higher mode 1 frequency than the thinner panels.
In the case you have presented, the 3 examples have different shapes and sizes, which you carefully selected so that all 3 panels have the same mode 1 resonant frequency of 1400 Hz.
I don't really want to work through the math here, but I believe firmly that if the input forcing function is an impulse, the thicker panel will have less SPL at 1400 Hz than the thinner panel given the same impulse. For a given energy impulse, like a hammer strike, the higher mass will result in lower acceleration, lower velocity, and lower displacement.
If the input forcing function is a continuous 1400 Hz sinusoidal signal, then I am not sure what happens. Each panel will vibrate at its 1400 Hz resonance, and the magnitude will grow with each cycle until it is arrested by the internal damping of the material. There is no doubt that the thinner lighter panel will reach its maximum displacement in fewer cycles than the more massive thicker panel. In other words, it will take less energy for it to achieve its steady state vibration level.
The larger thicker more massive panel will absorb more energy, more cycles of 1400 Hz forcing function to achieve steady state. It will take a longer time.
But given the greater surface area of the thicker panel, I do not know which would produce the greater SPL once they have reached steady state vibration. It may depend on the level of natural damping in the material.
Does this make sense?
Thank you very much hifijim. That does indeed make sense to me and is explained in a way that I can understand too. I hadn't originally considered that the panel size might make a difference in resonant SPL, I was more trying to explore the results with the exact same input, so that therefore required different size panels to get the same resonant frequency.
So now I am wondering if it is the increase in mass or stiffness or both that reduce resonance amplitude?
So now I am wondering if it is the increase in mass or stiffness or both that reduce resonance amplitude?
While most TLs, especially if used for bass, attempt to preserve the fundamental quarter-wavr resonance, and kill all the harmonics above that for bass reinforcement, half-wave lines or aperiodic midTLs attempt to completely absorb the rear radiation.
dave
It is important to remember that at most, and only at bass frequencies in effectively undamped enclosures, this only represents at a maximum 1/3 of the energy loaded into the box (unless there is active reactive force cancelation).
dave
I think perhaps FAR too much attention has been paid to stiffness and damping here. Lets not argue about something which can be solved mathematically. Everything is a trade off and with trade-offs there's usually an optimum.
Lets take the millennium bridge. Obviously an incredible stiff structure but without adequate dampening an input at the resonate frequency caused it to oscillate out of control. Its clear that a stiff structure without dampening is inadequate.
Its also quite clear that a less stiff structure requires less energy input to reach a certain amplitude. So it is a trade off between the two and I'm sure it can be solved mathematically but there will be different solutions to the same problem.
I tried solving it mathematically myself but I just dont remember dynamics well enough so I'm planning on contacting my old lecturer.
A piece of software I would recommend people try out is CES. We used it at university alot; its used for determining an ideal material based on requirements. Things to consider would be increased mass lowers resonant frequency but also the energy input required to move it. Higher stiffness raises the resonant frequency but also the force required to move it.
Also to consider is that higher frequencies are more audible, but also more easily damped. The damping force is proportional to velocity and higher frequencies move faster.
The are more ways to remove resonances as well, there's a skyscraper in UAE which was designed with a HUGE pendulum in it designed to absorb the energy in the movement of the building caused by wind. I believe its called a tuned mass damper.
So would someone explain to me the other desired features of a cabinet as its clearly not just reducing vibrations or speaker design would be as simple as buying some nice drivers and putting them in a heavily damped unit.
So what else do I need to look at, I suppose the back wave is also very important?
So obviously this can be absorbed with materials, what other routes do people go down? Has anyone ever used a turbine like system to absorb energy? After all a turbine is designed to absorb energy from a moving fluid and you get some very quiet bearings which I'm sure wouldnt be audible through a cabinet.
Lets take the millennium bridge. Obviously an incredible stiff structure but without adequate dampening an input at the resonate frequency caused it to oscillate out of control. Its clear that a stiff structure without dampening is inadequate.
Its also quite clear that a less stiff structure requires less energy input to reach a certain amplitude. So it is a trade off between the two and I'm sure it can be solved mathematically but there will be different solutions to the same problem.
I tried solving it mathematically myself but I just dont remember dynamics well enough so I'm planning on contacting my old lecturer.
A piece of software I would recommend people try out is CES. We used it at university alot; its used for determining an ideal material based on requirements. Things to consider would be increased mass lowers resonant frequency but also the energy input required to move it. Higher stiffness raises the resonant frequency but also the force required to move it.
Also to consider is that higher frequencies are more audible, but also more easily damped. The damping force is proportional to velocity and higher frequencies move faster.
The are more ways to remove resonances as well, there's a skyscraper in UAE which was designed with a HUGE pendulum in it designed to absorb the energy in the movement of the building caused by wind. I believe its called a tuned mass damper.
So would someone explain to me the other desired features of a cabinet as its clearly not just reducing vibrations or speaker design would be as simple as buying some nice drivers and putting them in a heavily damped unit.
So what else do I need to look at, I suppose the back wave is also very important?
So obviously this can be absorbed with materials, what other routes do people go down? Has anyone ever used a turbine like system to absorb energy? After all a turbine is designed to absorb energy from a moving fluid and you get some very quiet bearings which I'm sure wouldnt be audible through a cabinet.
A turbine transfers energy from moving gases or fluids. Sound propagation can be seen as moving air, but the movement itself likely is too small, variable and undirected to be effectively damped by such a slow reacting device as a turbine. If so, it probably works best on steady state signals. Music of course isn’t a steady state signal.
Take exhaust systems that sometimes use (Helmholtz) resonators to dampen combustion engine noise. This works best on stationary engines that run at one speed. In automotive applications such resonators are less effective because of the ever changing output.
Take exhaust systems that sometimes use (Helmholtz) resonators to dampen combustion engine noise. This works best on stationary engines that run at one speed. In automotive applications such resonators are less effective because of the ever changing output.
Michael5039 - I sympathise that this thread may have gone off in a different direction from what you were expecting when you started it... I have had some threads which I started go completely off the rails.
So assuming you have an enclosure (i.e. not an open baffle design), what are the requirements?
1) The box must have the right volume of air for your woofer, and if you are designing a vented system, the port must have the right dimensions to achieve the desired tuning frequency. Numerous free tools help with this.
2) The box must isolate the back radiation from the front radiation to avoid cancellation and absorb it so that it does not color the sound by being re-radiated through the cone diaphragm. The discussions about midrange transmission lines also are aimed at preventing this, although most of us believe that can be handled by simply filling the box with fiberglass insulation or polyester stuffing, or accoustic foam, or other absorptive material. Sometimes acoustical ceiling tiles or heavy felt is used to line the cabinet.
3) The box must not add any acoustic output which is audible on program material. Box structural resonances are the primary source of box signature. All of our discussions about stiffness and damping relate to the prevention of a box signature.
4) the box must have a baffle shape and driver locations which minimizes diffraction. Another approach is design a shape which is tailored so that the baffle diffraction and the crossover work together to achieve an acceptable frequency response. Several software programs are available for free which help model this behavior.
5) The box must meet the end users physical and aesthetic requirements. A large heavy box may not meet the needs of someone with a small space, or someone who needs a portable speaker. A simply plywood cabinet painted with gray latex paint may not meet the aesthetic needs of someone who is going to place the speaker between fine furniture, and under a work of art.
6) The box must contain the crossover components, or for active speakers, the electronics.
7) The box must provide access to allow repair. Oftentimes the woofer hole is sufficient for this, but other times the rear baffle is made removable. This complicates the control of box resonances.
I hope others add to my list, as I am certain I have not thought of everything...
So assuming you have an enclosure (i.e. not an open baffle design), what are the requirements?
1) The box must have the right volume of air for your woofer, and if you are designing a vented system, the port must have the right dimensions to achieve the desired tuning frequency. Numerous free tools help with this.
2) The box must isolate the back radiation from the front radiation to avoid cancellation and absorb it so that it does not color the sound by being re-radiated through the cone diaphragm. The discussions about midrange transmission lines also are aimed at preventing this, although most of us believe that can be handled by simply filling the box with fiberglass insulation or polyester stuffing, or accoustic foam, or other absorptive material. Sometimes acoustical ceiling tiles or heavy felt is used to line the cabinet.
3) The box must not add any acoustic output which is audible on program material. Box structural resonances are the primary source of box signature. All of our discussions about stiffness and damping relate to the prevention of a box signature.
4) the box must have a baffle shape and driver locations which minimizes diffraction. Another approach is design a shape which is tailored so that the baffle diffraction and the crossover work together to achieve an acceptable frequency response. Several software programs are available for free which help model this behavior.
5) The box must meet the end users physical and aesthetic requirements. A large heavy box may not meet the needs of someone with a small space, or someone who needs a portable speaker. A simply plywood cabinet painted with gray latex paint may not meet the aesthetic needs of someone who is going to place the speaker between fine furniture, and under a work of art.
6) The box must contain the crossover components, or for active speakers, the electronics.
7) The box must provide access to allow repair. Oftentimes the woofer hole is sufficient for this, but other times the rear baffle is made removable. This complicates the control of box resonances.
I hope others add to my list, as I am certain I have not thought of everything...
Last edited:
Thank you planet10 for clarifying that enclosures are the superior solution, otherwise I might think from this discussion that they introduce many more engineering challenges than they solve. Other than being a convenient and compact container to ship drivers in, what exactly are the benefits that monkey coffins confer to achieving satisfying sound reproduction?
Such a device could be termed an 'infinite line'. (The use of the term TL in this context took the thread on a turn, in my estimation.)
Is someone could tell, please:
what is the level of cabinet panels radiation due to the driver metal basket vibrations that shake them in relation of the panel radiation due to the energy absortion & dissipation of the panels due to the internal back energy wave of a cone? i.e. : importance of damping layer between the driver and the panel and isolation of the external panels from the internal surfaces : i.e. sandwich, i.e. mass spring mass, stif-damp-stif.
I assume the first is less arming but what if the voice coil is a long one for long exursion with low Qm number drivers. Is it disspated into the spider, surround and stiff basket ? Or is it the brutal impedance mismatch of the cone load towards the room that motions the cabinet like a reactor turbine (hence the weight importance with bass drivers) ? Advantage to the horns that creates a more progressive impedance adaptation between the cone/dome and the room ?
what is the level of cabinet panels radiation due to the driver metal basket vibrations that shake them in relation of the panel radiation due to the energy absortion & dissipation of the panels due to the internal back energy wave of a cone? i.e. : importance of damping layer between the driver and the panel and isolation of the external panels from the internal surfaces : i.e. sandwich, i.e. mass spring mass, stif-damp-stif.
I assume the first is less arming but what if the voice coil is a long one for long exursion with low Qm number drivers. Is it disspated into the spider, surround and stiff basket ? Or is it the brutal impedance mismatch of the cone load towards the room that motions the cabinet like a reactor turbine (hence the weight importance with bass drivers) ? Advantage to the horns that creates a more progressive impedance adaptation between the cone/dome and the room ?
Last edited:
The box will also have a diffaction signature which also contributes to the sound of the box.
dave
They are not superior, they are different, a different set of compromises.
A “proper” box has the set of compromises i prefer. Others may not.
I have heard a lot of bad boxes, and one of the boxiest sounding speakers i have ever heard were OBs (with Silver Iris driver).
dave