@Tattoo
Yes,
in fact - when used as a wideband enclosure - your proposed concept (thin wood, thick dampening sheets) would work well, as those thick and massive dampening sheets will also prevent the cabinet from bending too much at very low frequencies - just by damping and their own stiffness wich usually is not neglegible.
But as a pure subwoofer cabinet this version would be too large, massive and costly:
You can have that "cheaper and better" by making a subwoofer just "compact" and with "good stiffness" using more or less conventional materials like MDF.
A good subwoofer's "enemy" is size in my view:
When it gets too large, it is more difficult to make a cabinet, that is virtually free of sound radiation.
And also the inner Dimensions may come into the range of (half-) wavelengths (maybe at crossover frequency or even close above ...) causing resonances of the enclosed air volume.
Yes,
in fact - when used as a wideband enclosure - your proposed concept (thin wood, thick dampening sheets) would work well, as those thick and massive dampening sheets will also prevent the cabinet from bending too much at very low frequencies - just by damping and their own stiffness wich usually is not neglegible.
But as a pure subwoofer cabinet this version would be too large, massive and costly:
You can have that "cheaper and better" by making a subwoofer just "compact" and with "good stiffness" using more or less conventional materials like MDF.
A good subwoofer's "enemy" is size in my view:
When it gets too large, it is more difficult to make a cabinet, that is virtually free of sound radiation.
And also the inner Dimensions may come into the range of (half-) wavelengths (maybe at crossover frequency or even close above ...) causing resonances of the enclosed air volume.
Last edited:
I agree to that ... this is definetely the case.
But to make a "low radiation cabinet" it is not so important to know how it is excited. The "preferred wall mix" stays more or less the same.
Of course - once you know the exciting mechanism - you can introduce a lowpass filter there consequently by e.g.
- adding mass to the driver's motor
- decouple the driver from baffle by viscoelastic elements
Vice versa in a subwoofer especially the cabinet's stiffness at and around the driver's baffle is important to keep bell modes "high" in frequency ... and thus out of the used bandwidth.
Eh, and let's not forget about the quality of the driver's basket itself.
Magnet mounting is also an option to think about - if the cabinet is prepared for that - to avoid a basket ringing somewhere in the bass or midrange. But again:
IMO this is usually not a subwoofer's problem, not even with properly made stamped steel driver baskets used.
But to make a "low radiation cabinet" it is not so important to know how it is excited. The "preferred wall mix" stays more or less the same.
Of course - once you know the exciting mechanism - you can introduce a lowpass filter there consequently by e.g.
- adding mass to the driver's motor
- decouple the driver from baffle by viscoelastic elements
Vice versa in a subwoofer especially the cabinet's stiffness at and around the driver's baffle is important to keep bell modes "high" in frequency ... and thus out of the used bandwidth.
Eh, and let's not forget about the quality of the driver's basket itself.
Magnet mounting is also an option to think about - if the cabinet is prepared for that - to avoid a basket ringing somewhere in the bass or midrange. But again:
IMO this is usually not a subwoofer's problem, not even with properly made stamped steel driver baskets used.
Last edited:
Where "professionals" become Members, then I would always suggest they declare their proficiency and in which field/s, in their profile and not remain anonymous
Andrew,
just to avoid misunderstandings, i am not the professional in question.
But would you think of this "Biography" entry in a profile as being sufficient:
"A long term loudspeaker designer who has worked for JBL, KEF, Meridian, McIntosh, Snell and PSB. "
My opinion is that with rude, arrogant and ignorant people we never talk about their "anonymity", do we ?
What if everyone would conform to a lowest level of politeness, so that we all could be regarded potential "professionals" ?
I mean, that's just a thought ...
just to avoid misunderstandings, i am not the professional in question.
But would you think of this "Biography" entry in a profile as being sufficient:
"A long term loudspeaker designer who has worked for JBL, KEF, Meridian, McIntosh, Snell and PSB. "
My opinion is that with rude, arrogant and ignorant people we never talk about their "anonymity", do we ?
What if everyone would conform to a lowest level of politeness, so that we all could be regarded potential "professionals" ?
I mean, that's just a thought ...
Last edited:
....they also remain vulnerable to error, if at a much lower rate. Jim Rogers was a legendary and accomplished designer but the thin walled, highly damped JR149 in my experience radiated as much from its aluminum cabinet as it's 110. Bracing helped, ultimately not as much as giving it away.
What about a sandwich of two panels with a bitumen layer between the panels?
This sandwich could be for every major panel of the speaker.
As the innermost panel flexs then the outermost panel is forced to follow and the bitumen is sqeejeed between the panels absorbing energy.
The problem is that the 100litre cabinet (clone of the Tannoy Berkley) I made from doubled 18mm chipboard ended up about 45kg. The whole speaker weighed in at over a hundredweight.
ps.
I didn't know that MDF existed way back in 1979ish.
This sandwich could be for every major panel of the speaker.
As the innermost panel flexs then the outermost panel is forced to follow and the bitumen is sqeejeed between the panels absorbing energy.
The problem is that the 100litre cabinet (clone of the Tannoy Berkley) I made from doubled 18mm chipboard ended up about 45kg. The whole speaker weighed in at over a hundredweight.
ps.
I didn't know that MDF existed way back in 1979ish.
Last edited:
AndrewT said:
...which is known as "constraint layer damping" sure, this is a way to use a dampening layer effectively.
e.g. MDF-Bitumen-Aluminum has been sucessfully tested, Alu used as a thinner inner layer.
But also e.g. MDF 16mm - Bitumen 4mm - Chipboard 4mm
I have to look whether i can find some pics or links ...
yes i can
Gehäusevibrationen Nachlese - waveguide-audio
http://www.picosound.de/GehMatAll.png
Last edited:
Thanks for that search criteria.
There is an old Thread discussing it.
http://www.diyaudio.com/forums/multi-way/153419-constrained-layer-damping-mdf-ply.html#post1951209
and google throws out lots of links.
There is an old Thread discussing it.
http://www.diyaudio.com/forums/multi-way/153419-constrained-layer-damping-mdf-ply.html#post1951209
and google throws out lots of links.
@midrange
If you choose dense damping mats like e.g. polymer-bitumen, "thinner wood and thick lining" is the better choice for a multiway cabinet, where the woofer also operates in upper bass and midrange.
Especially in the upper bass and midrange this is the better choice.
Just in case of a dedicated subwoofer, "thick wood" and thin (or even void ..) lining is an option, given the lowest resonances are far above the used bandwidth. That should be - depending on the slope of the crossover - typically more than just one octave ...
Have a look at the post and links above here, so you can choose a material that meets your requirements ...
To be clear: Thickness of the wood is relative to size of cabinet. E.g. 15mm MDF in a Bookshelf Speaker may be rather "thick", while in a huge PA Cabinet it is "thin".
In fact you choose a range of resonance by appropriate thickness (of the wood), and then apply the damping to bring the Q down, simply speaking.
I would say usually a cabinet should be dominated by stiffness up to about 1-2 octaves above the lower cutoff frequency ...
In a subwoofer i would align for dominating stiffness up to 1-2 octaves above the upper cutoff frequency ...
When stiffness does not dominate the cabinet's walls anymore, then mass and compliance will have equal reactance and you get the lowest resonance there ...
In an ideal world even that first resonances had to be suppressed by damping to have inaudible level of sound radiation through the cabinet's walls.
In case of the "very stiff" subwoofer cabinet proposed above, you can "muddle through" by saying:
"Cabinet resonance will not be excited, because input voltage is already down to 24dB at NN Hz." Nevertheless excessive Q at resonance should be avoided IMO to ensure even here, that it is inaudible.
@David: Maybe you could drop a few lines, whether those "rules of thumb" meet your understanding of matters ...
It is hard in fact, to give "common rules" here IMO.
If you choose dense damping mats like e.g. polymer-bitumen, "thinner wood and thick lining" is the better choice for a multiway cabinet, where the woofer also operates in upper bass and midrange.
Especially in the upper bass and midrange this is the better choice.
Just in case of a dedicated subwoofer, "thick wood" and thin (or even void ..) lining is an option, given the lowest resonances are far above the used bandwidth. That should be - depending on the slope of the crossover - typically more than just one octave ...
Have a look at the post and links above here, so you can choose a material that meets your requirements ...
To be clear: Thickness of the wood is relative to size of cabinet. E.g. 15mm MDF in a Bookshelf Speaker may be rather "thick", while in a huge PA Cabinet it is "thin".
In fact you choose a range of resonance by appropriate thickness (of the wood), and then apply the damping to bring the Q down, simply speaking.
I would say usually a cabinet should be dominated by stiffness up to about 1-2 octaves above the lower cutoff frequency ...
In a subwoofer i would align for dominating stiffness up to 1-2 octaves above the upper cutoff frequency ...
When stiffness does not dominate the cabinet's walls anymore, then mass and compliance will have equal reactance and you get the lowest resonance there ...
In an ideal world even that first resonances had to be suppressed by damping to have inaudible level of sound radiation through the cabinet's walls.
In case of the "very stiff" subwoofer cabinet proposed above, you can "muddle through" by saying:
"Cabinet resonance will not be excited, because input voltage is already down to 24dB at NN Hz." Nevertheless excessive Q at resonance should be avoided IMO to ensure even here, that it is inaudible.
@David: Maybe you could drop a few lines, whether those "rules of thumb" meet your understanding of matters ...
It is hard in fact, to give "common rules" here IMO.
Last edited:
@David: Maybe you could drop a few lines, whether those "rules of thumb" meet your understanding of matters ...
It is hard in fact, to give "common rules" here IMO.
It is really hard to give rules in terms of stiffness or mass. The whole notion of stiffness and mass and resultant resonance is a simplified look at a single resonance system. Cabinets clearly have lots of resonances (see the previous curves by Harwood) and so we are really into modal behavior where dimensions and speed of sound give rise to multiple modes of rising complexity.
No matter, damping can deal with all the modes so the real question is what ratio of damping to other mechanical impedances are needed. If we ask "how thick should the damping be?" we run into "depends on how thick your cabinet walls are". Harwood found, as you would expect, that the thicker walls had generally higher mechanical impedances and required even greater damping to get a certain amount of acoustic breakthrough.
I would rather think of it as starting with a target for cabinet weight or cabinet wall thickness. Lets say we want our walls to be 1" thick in some combination of MDF and bitumen padding. We can make the wood thinner and the damping thicker, or vice versa. Which is best? Under that scenario I would shoot for the thinnest practical walls and the greatest damping thickness. We have to maintain some cabinet rigidity for integrity and to support our drivers but certainly 1/2" walls with 1/2" damping would be reasonable and practical.
At this point someone will say, "Then wouldn't 1" thick with that 1/2" damping be even better? Thicker, you know." The answer is decidedly no. Growing the wall thickness may increase stiffness but it just leads to a higher mechanical impedance with less resultant benefit from the damping.
Practical experience supports all this. We have all seen the wooden cabinet with incredibly thick walls. I used to give them the knuckle rap and admire the high pitch ring. "Solid!" This is just a sign that their thickness made the modes undampable. The one cabinet I have owned that really sounded different was a BBC LS5/1ac. It was 1/2" or 5/8" plywood with heavy felt and tar pads lining the interior. No solid Rock Maple sound to it, just a dull thump.
Is all this the most important matter in diy Audio? No, not really. It is still more important that we design our crossovers correctly and achieve smooth and flat response. It isn't that frequent that I listen to a system and say: "Pity about the cabinet resonances spoiling the sound". (Infrequent but "not never".)
Harwood has it right in showing that there is a threshold of audibility to resonances. If we can push the level below the threshold or lower the resonance frequencies to a region where we are less perceptive, then we will reach a point where the problem is effectively solved and we should focus our attention elsewhere.
Regards,
David S.
David, i am a real fan of the way you put things together !
You keep up that idea, that we can in fact solve problems to a manageable - and most often even affordable - technical limit and that we can find solutions that are in line with our auditory perception of "quality".
It is not necessary to bring every measurable disturbance "down to zero" at insane cost, but we do the best job possible if we make those oddities inaudible, by considering the limits of perception to any special kind of distortion.
Often the "extra costly" ("extra thick walled"...) solutions are the poorer ones in measurable and perceived quality.
Cheers
You keep up that idea, that we can in fact solve problems to a manageable - and most often even affordable - technical limit and that we can find solutions that are in line with our auditory perception of "quality".
It is not necessary to bring every measurable disturbance "down to zero" at insane cost, but we do the best job possible if we make those oddities inaudible, by considering the limits of perception to any special kind of distortion.
Often the "extra costly" ("extra thick walled"...) solutions are the poorer ones in measurable and perceived quality.
Cheers
Last edited:
Linearray, thank you for your very fulsome answer. Whilst I sort of understand about resonances and panels and damping, I was really looking at, apart from output from the panels due to vibration, is there any difference between the levels of attenuation to the sound passing through the panel if it is thick, or thin and damped. Or does sound not actually pass through a panel?
The rationale I hear for deploying rigid cabinets is to purposely raise the resonant frequencies up on the belief that the damping materials and mechanisms are more efficacious in those higher frequencies than at the lower frequencies exhibited by less stiff cabinets. I.e. that the damping loss is frequency dependent, rising with frequency
DDF said:
maybe, but this belief is not supported by reality.
The Q of bending modes (resonances) in a panel material tends to be constant: The higher order modes do not have lower Q than the lower ones.
The belief behind thick and underdamped wall is simply a misconception, and David's last post describes also the economic aspect - "how do i invest a given cabinet weight best" - on point IMO.
Compare the level of the low frequency resonances with the level of the high frequency resonances in the two plots David posted a couple of pages earlier. Why are the high frequencies resonances a lot lower in level than the low frequency resonances?