By definition at resonance the forces due to stiffness are cancelled by inertia (the "force" due to mass). It doesn't matter whether the stiffness forces are large or small they have no influence on the magnitude at resonance. It is only the force due to damping that will determine that.
The definition of a resonance is not going to change because you posted in haste and the quote from speaker dave was referring to resonance.
At low frequencies below the resonances the forces due to stiffness are dominant and a stiff cabinet is going to be a quiet cabinet. At high frequencies well above the lowest frequency resonances the forces due to stiffness are small compared to inertia and it is mass that will reduce the motion of the cabinet (actually stiffness can have an indirect effect via the speed of bending waves but we will keep things simple). At the lowest resonant frequencies it is damping that will reduce the magnitude of the resonant peaks and not thick walls of MDF as many speaker DIYers seem to believe. I am pretty confident this was speaker dave's point.
For a linear system energy cannot move between frequencies and so the motion at resonance is completely unaffected by the motion at other frequencies. For example consider a stiff glass being shattered by singing the right note. It works because the tiny amount of energy that is not reflected by the large impedance mismatch at the glass/air interface of the small amount of energy in sound stays at the resonant frequency and builds up. It cannot shift to other frequencies until the motion becomes large enough to be non-linear.
Agreed -- for steady state conditions (non-musical).
Also agreed -- I posted in haste, especially the words 'by definition'. The statement was wrong on several practical levels, but not by definition. I was more interested in the amplitude of cabinet vibration at all frequencies under musical signals, which does reduce when stiffness increases.
To be fair to both sides they are both correct, but in very different ways.
High Stiffness does reduce the sound transmission through panels and helps to reduce the total vibration level, but not specifically at the resonances. It does this but decreasing the density of resonances in a specific band. But clearly at a resonance it is only the damping that matters.
This points to something that I have said for a log time, that one wants a very stiff, light and well damped enclosure for the lowest sound radiation. The idea of using very dense/heavy materials is wrong.
There will not be a difference in how something responds to steady state versus "musical" signals in the sense that they would ever go in different directions. Reducing the steady state response will also reduce the response to musical signals. They both have to obey the same physical equations for the system - just with different inputs.
High Stiffness does reduce the sound transmission through panels and helps to reduce the total vibration level, but not specifically at the resonances. It does this but decreasing the density of resonances in a specific band. But clearly at a resonance it is only the damping that matters.
This points to something that I have said for a log time, that one wants a very stiff, light and well damped enclosure for the lowest sound radiation. The idea of using very dense/heavy materials is wrong.
There will not be a difference in how something responds to steady state versus "musical" signals in the sense that they would ever go in different directions. Reducing the steady state response will also reduce the response to musical signals. They both have to obey the same physical equations for the system - just with different inputs.
Last edited:
True, except in the special case of at resonance (as you said at the top of your post). In that special case, stiffness has no bearing under a steady state response, but it does have a bearing under musical signals. That's where I came in: I think we should be discussing the best cabinet under musical signals, at resonances.
I don't get how musical signals are any different than any other signal. Once you know the impulse response you know how the system will react to steady state signals, musical signals, noise, whatever. And since one can derive the impulse response from the proper steady state signal, then all signals become the same.
Unless your point is that a single tone signal at a resonance is different than a musical signal, then of course, one is barrow band and the other broad band. But a resonance will react to a broadband steady state signal in exactly the same way that it reacts to a musical signal. Its not the steady state that matters but the bandwidth.
Dave - I don't think that this is exactly correct. Its like saying that resonances can't be heard with music because they don't get excited, which is clearly not true. The energy, whatever it is, is amplified at a resonance no matter how short a duration it is. The longer the duration, of course, the greater the amplification (up until it reaches the steady state), but it is always there. And lets be honest here, a musical tone is not that short, especially for a MF-HF resonance. It is certainly long enough to fully excite a reasonable Q resonance at say 1 kHz probably even a lot lower. The lower the Q the faster the tone reaches steady state. So only an extremely high Q resonance would not reach a steady state for a musical tone.
I think the basic misunderstanding is this....
Any resonance will store energy, and will then decay. That released energy is not part of the music, and will color the music.
Music is pretty broadband, so it will indeed exite any resonances in the cabinet, yielding that familiar "boxy" sound we all know so well. Thicker cabinet walls doesn't fix that, nor can it.
There's nothing new in this. Linkwitz noted this is the early '70s, when mounting a phono cart in contact with a speaker cabinet, and noting that at some frequencies the cabinet was louder than the drivers.
You can argue this from any angle you like, but a decay waterfall will show you the cabinet resonances.
Any resonance will store energy, and will then decay. That released energy is not part of the music, and will color the music.
Music is pretty broadband, so it will indeed exite any resonances in the cabinet, yielding that familiar "boxy" sound we all know so well. Thicker cabinet walls doesn't fix that, nor can it.
There's nothing new in this. Linkwitz noted this is the early '70s, when mounting a phono cart in contact with a speaker cabinet, and noting that at some frequencies the cabinet was louder than the drivers.
You can argue this from any angle you like, but a decay waterfall will show you the cabinet resonances.
No, i am saying the resonance is not being heard because it is not happening (not getting excited even thou the potential is there)
dave
Well Dave, then I have to completely disagree with you.
Consider testing with a swept sine wave (chirp) versus testing with noise or stepped sine tones. If what you say were true then these two tests would yield completely different results, but they don't. Basically the results are exactly the same unless the sweep is exceedingly fast - way faster than one would actually ever use. But typical sweeps are certainly faster than a musical tone.
Consider testing with a swept sine wave (chirp) versus testing with noise or stepped sine tones. If what you say were true then these two tests would yield completely different results, but they don't. Basically the results are exactly the same unless the sweep is exceedingly fast - way faster than one would actually ever use. But typical sweeps are certainly faster than a musical tone.
How much of that decay has something to do with the speakers and how much to do with the room? How much of what is due to the speakers is due to a resonating driver cone and/or frame and how much to a resonating cabinet?
It is an interesting topic because it is far from straightforward to perform an experiment to determine the level of perceived cabinet radiation.
Unless you guys buy something like this:
http://www.diyaudio.com/forums/equipment-tools/267734-accelerometer-testing-loudspeaker-drivers.html
you are all rowing in the dark blindfolded on a moonless night.
I bought and built one and one thing is obvious: reaction forces are the major enemy and are what induces most vibrations in enclosure panels. Even the reaction forces of a tweeter are visible on the opposing panel. This, by the way, is a strong argument for the best enclosure material to be heavy (and lossy, in other words, not stiff).
http://www.diyaudio.com/forums/equipment-tools/267734-accelerometer-testing-loudspeaker-drivers.html
you are all rowing in the dark blindfolded on a moonless night.
I bought and built one and one thing is obvious: reaction forces are the major enemy and are what induces most vibrations in enclosure panels. Even the reaction forces of a tweeter are visible on the opposing panel. This, by the way, is a strong argument for the best enclosure material to be heavy (and lossy, in other words, not stiff).