That is true, but i was not talking about the box moving, but the walls resonating.
dave
Yes, but that movement caused by reaction forces is what in the main makes the walls resonate, much more than the fluctuating pressure differential between inside and outside of the enclosure.
Think about how a sound board in a piano works. The minute reaction forces exerted by a moving string are coupled to the sound board, which gets into sympathetic vibration and radiates sound into the air. You want this sound board to be stiff, light weight and undamped. It is quite intuitive that a heavy sound board will work less effectively. Same thing with enclosure panels, but because you want no sound to escape from them, reverse the specs. They need to be heavy, relatively flexible and well damped. That is why thin sandwich panels with sand in between work so well.
But when it comes to those pressure differentials, also there heavy wins from light. Given a certain force, heavy just moves less.
what if there isn't "one best"?
piano string vibration is strongly coupled by the bridge - it isn't just a air path or string mass reaction force
the bridge is a mechanical transformer for string vibration energy transfer to soundboard high mechanical mobility regions, nodes of its vibration patterns
in speaker drivers the cone motion reaction force is shorted to the relatively massive basket and magnet assembly
for this cone motion inertia coupling possibly a high mass front panel makes some sense - but in a conventional rectangular cabinet the sides would only be actuated by bending moment of the front panel transferred across the joints
mass doesn't seem to be as useful in the side panels in this analysis
stiffness, damping, damping bracing wall-to-wall all do reduce both pressure and front panel bending forced vibration of the side panels
piano string vibration is strongly coupled by the bridge - it isn't just a air path or string mass reaction force
the bridge is a mechanical transformer for string vibration energy transfer to soundboard high mechanical mobility regions, nodes of its vibration patterns
in speaker drivers the cone motion reaction force is shorted to the relatively massive basket and magnet assembly
for this cone motion inertia coupling possibly a high mass front panel makes some sense - but in a conventional rectangular cabinet the sides would only be actuated by bending moment of the front panel transferred across the joints
mass doesn't seem to be as useful in the side panels in this analysis
stiffness, damping, damping bracing wall-to-wall all do reduce both pressure and front panel bending forced vibration of the side panels
It is. But there are 2 phenomenon, 1 is causing the box to rock which mass will help with, the other is energy pumped into the panels that causes resonances. Mass does not help there, as it lowers the potential resonant frequency and makes them more likely to get excited.
dave
a=F/m. Less movement is less sound radiation. JCX, I have a sensitive accelerometer that works across the audio band. It picks up the reaction forces of the tweeter on the back panel, easily verifiable by FFT analysis.
Has any of you guys actually ever done any research with a tool like that, or is it all just conjecture?
Has any of you guys actually ever done any research with a tool like that, or is it all just conjecture?
If speaker enclosures need to be heavy, flexible and damped why aren't the enclosures of high performance speakers heavy, flexible and damped?
Because most speaker designers are ignorant of physics and follow the market beliefs that stiff and solid is the answer.
It is much more complex than that for resonance. Where are the potential resonances located, how well damped is the material, the box walls, bracing to name some complicating mechanisms.
dave
The forces in the cabinet arise from three sources not just one:
- mass*acceleration
- damping*velocity
- stiffness*displacement
So you equation should read: F = m.a + d.v + k.u. Since velocity is proportional to frequency and acceleration is proportional to frequency squared this means m.a is relatively large at high frequencies but relatively small at low frequencies. High mass is not particularly useful at low frequencies.
It is remarkably simple at resonance because the forces due to mass exactly cancel the forces due to stiffness leaving only the forces due to damping to determine how loud the enclosure will resonate.
Well, what I found out is that the reaction forces are the real witch to be drowned at the stake while being set on fire. It is very difficult to accomplish. Reaction forces cannot be contained, they are necessarily coupled into the enclosure. The two ways of mitigating their effects are:
1) more weight, so that the movement as a result of these forces remains limited,
2) dissipation, which requires two things. Something that moves in relation to something else, with a medium in between which dampens the movement and turns it into heat. In other words, flexibility and damping.
Another analogy. A brick wall is a heavy and inflexible thing. If my neighbor uses a hammer drill to make even a tiny hole, this creates a very audible noise on my side of the wall. Now, if this wall were made of wet clay, what do you think would happen (with 1 single brick in it to drill the hole into)? No sound would come through, because of a combination of flexibility and damping.
With real life materials you can best achieve this with combinations of materials, like Wilson does.
1) more weight, so that the movement as a result of these forces remains limited,
2) dissipation, which requires two things. Something that moves in relation to something else, with a medium in between which dampens the movement and turns it into heat. In other words, flexibility and damping.
Another analogy. A brick wall is a heavy and inflexible thing. If my neighbor uses a hammer drill to make even a tiny hole, this creates a very audible noise on my side of the wall. Now, if this wall were made of wet clay, what do you think would happen (with 1 single brick in it to drill the hole into)? No sound would come through, because of a combination of flexibility and damping.
With real life materials you can best achieve this with combinations of materials, like Wilson does.
Great thread. So far no one has mentioned the Speaker builder article about 20 years ago, which ended up espousing a mixture of construction epoxy, granulated rubber (finely ground tire rubber) and very fine silica (sand). The initial assumptions were that the stiffer, the better for low frequencies, and the more damped, the better for high frequencies. The authors found after about 3/4 inches, that the benefits did not decrease.
My own preferences is for a stiff cabinet, i.e. good quality plywood and bracing (not much discussion of bracing, as it is getting off topic), and a dense, damped front panel - MDF or HDF with damping on the inside walls.
My own preferences is for a stiff cabinet, i.e. good quality plywood and bracing (not much discussion of bracing, as it is getting off topic), and a dense, damped front panel - MDF or HDF with damping on the inside walls.
This debate flares up every 6 months on this site without fail.
Each time I point out that this is well studied in the field of architectural acoustics in the context of sound transmission through walls. Every wall construction imaginable has been tried and documented in terms of Transmission Loss (TL). It is well known that mass gives isolation but the best mass is referred to as a flexible (limp) mass sheet. The reason that limp is good is that wall stiffness guarantees frequencies of resonance and at resonance the wall is full transparent unless the damping is high (as Andy says a few posts up). With a limp mass the lack of stiffness leads to an absence of resonances.
So high mass with low stiffness but with high damping would be ideal. The problem is that mass and damping are at odds with each other. If you double the mass per unit area of a wall you will need twice the damping to have similar resonance Q's (changes in stiffness aside). This is what led Harwood and others to realize that thinner wall construction, along with significantly higher damping, gave the best performance. In the limit, if the structure were simply a thin shell and all the mass was from damping material, then you would have the best performance achievable.
The answer to the original question: the best cabinet material would be heavy dense rubber. Thump it with a hammer and it don't ring!
David
Each time I point out that this is well studied in the field of architectural acoustics in the context of sound transmission through walls. Every wall construction imaginable has been tried and documented in terms of Transmission Loss (TL). It is well known that mass gives isolation but the best mass is referred to as a flexible (limp) mass sheet. The reason that limp is good is that wall stiffness guarantees frequencies of resonance and at resonance the wall is full transparent unless the damping is high (as Andy says a few posts up). With a limp mass the lack of stiffness leads to an absence of resonances.
So high mass with low stiffness but with high damping would be ideal. The problem is that mass and damping are at odds with each other. If you double the mass per unit area of a wall you will need twice the damping to have similar resonance Q's (changes in stiffness aside). This is what led Harwood and others to realize that thinner wall construction, along with significantly higher damping, gave the best performance. In the limit, if the structure were simply a thin shell and all the mass was from damping material, then you would have the best performance achievable.
The answer to the original question: the best cabinet material would be heavy dense rubber. Thump it with a hammer and it don't ring!
David
Well Dave, maybe we should be building cabs out of old tires.
Now you're talking.
Pirelli? Michelin? Which sound best......
At low frequencies the lack of stiffness would mean the walls moved in and out like a balloon under the action of the internal air pressure and the drivers would put large bending waves into the front panel. At medium and high frequencies the cabinet might work well.
The best performance was not the objective. Sufficient performance for the cabinet motion to be judged inaudible with a reasonably cheap, reasonably light and reasonably manufacturable structure was the objective. You cannot extrapolate further because the cabinet will lack sufficient stiffness. It is cheap and easy to make a cabinet out of rubber sheets but it is not done because it would not perform well.
It may not ring but it would move around excessively.
The most appropriate material would probably be fairly stiff with high damping for the low and medium frequences. Unfortunately, if we want high damping we are not going to get high stiffness and so will have to settle for reasonable stiffness and more material. At high frequencies it is straightforward to passively isolate the tweeter from the cabinet and so there is no need to optimise the cabinet material for these frequencies.
I tried and tried, but failed to get good readings. I was really interested to measure what was going on with the cabinet walls. None of my DIY pickups or any acoustic guitar pickups I tried gave me enough signal. I needed a better preamp.
An accelerometer might be a better choice for a probe.
These guys here build speakers out of paper faced foam board such as used for signage.
http://www.diyaudio.com/forums/full-range/223313-foam-core-board-speaker-enclosures.html
Face both sides of the board with a denser material like thin plywood, apply internal bracing, and the result should be just about a perfect speaker box an old guy can actually pick up without risking injury.
http://www.diyaudio.com/forums/full-range/223313-foam-core-board-speaker-enclosures.html
Face both sides of the board with a denser material like thin plywood, apply internal bracing, and the result should be just about a perfect speaker box an old guy can actually pick up without risking injury.