how to brace a speaker cabinet?

Status
Not open for further replies.
Jump to Latest
#41
lemans said:
The extensive bracing approach that some people here are eschewing is two fold. First, it adds stiffness. This reduces large panel resonances and turns them into small panel resonances. Indeed, these panel resonances would occur at higher frequencies. Secondly, it adds a large amount of mass. No one yet has mentioned mass in lieu of damping. When you create a system which consists simultaneously of high mass and high stiffness, you would in fact get a similar resonant frequency as one with low mass and low stiffness.
Click to expand...


... i guess we already treated that.

E.g. "doubling the walls" and especially bracing makes the walls stiffer a lot and contributes only little mass per area.

Doubling thickness will give stiffness x8 and mass per area x2. Resonances are expected to move up one octave.

With bracing you can push the resonances even higher, because less mass is needed while introducing comparable or even higher stiffness, depending on the kind of braces used.

This is why the resonances go up with bracing. The more braced a cabinet is - given there's no damping added - the more it gets unusable for frequencies from upper bass on upwards.

A heavily braced cabinet has also a very low coincidence frequency, causing very effective radiation of the bending modes of the cabinet as soon as the exciting frequency comes close to coincidence frequency.

- stiffness is you friend in cabinets for (very) low frequencies only ("stiff and compact subwoofer" strategy works).

- for mid and high frequencies damping and mass are your friends, given your cabinet is not too stiff (or heavily braced ...)

- if your cabinet is very stiff ( or heavily braced ...), looking for friends at mid and high frequencies is obsolete, because there is no practical cure for such a cabinet


But you are right, that bracing - if made clumsy - may add little stiffness (not working well in all relevant cabinet dimensions at least) but also considerably mass.

Very often bracing does not yield that reduction in unwanted sound radiation from the cabinet's walls which was intended. When applied to "non subwoofers", bracing is questionable as a whole IMO.
 
Last edited:
#42
My point is that when significantly braced and the excitation band goes higher in frequency, the amplitude is minimal compared to without. The idea that keeps getting thrown around here about high stiffness yielding high frequency audible resonances doesn't hold water when it's the woofers doing the major excitation, not the mid. It takes a lot of energy to excite something massive like a speaker cabinet, and woofers are the only drivers capable of realistically doing so. Put it like this: when you add bracing and move the excitation band up, the amplitude of the resonances falls. If you want to drive the cabinet at its now higher natural frequency, you need much more energy to make it a serious issue.

When discussing coloration, the source does not occur in the upper midrange to treble region. It occurs lower in frequency and then modulates said upper midrange and treble. The amount of energy the cabinet is producing in these upper regions due to pushing resonances up higher is so low compared to the radiation of the driver that the bracing becomes effective. When the cabinet is able to radiate sound easily, you induce coloration.

It is no coincidence that the best box speakers in the world are both a. unbelievably stiff and b. quite massive.
 
#43
lemans23 said:
The extensive bracing approach that some people here are eschewing is two fold. First, it adds stiffness. This reduces large panel resonances and turns them into small panel resonances. Indeed, these panel resonances would occur at higher frequencies. Secondly, it adds a large amount of mass. No one yet has mentioned mass in lieu of damping. When you create a system which consists simultaneously of high mass and high stiffness, you would in fact get a similar resonant frequency as one with low mass and low stiffness.

The difference which everyone is neglecting is the amount of force or energy required to excite the highly massive system/speaker. The drivers in a massive, extensively braced speaker are less likely to be able to excite the high panel resonances due to its stiffness because of the cabinet's inherent mass. The addition of damping material is then to remove the low level excitations that are created internally. Damping material does not get used to stop outward cabinet vibration. The difference between damping and mass loading needs to be stated clearly. They are not the same thing. If said mass loading materials have good internal damping, then so be it.

For you car guys, it's like this. Dynamat is a mass loading material, not a damping material. It stops panel vibrations in this way. It does have damping, yes, but that's not the point. Mass is king. Those who have used Dynamat will also note that road noise, particularly rain, is largely unaffected if your car is built properly. That is when you employ damping material. For normal driving use, damping wins because there is nothing in the car to really excite the large panels so you need to absorb the gunk transmitted via the outside. For car audio enthusiasts, mass wins because the noise is then from the panel itself. For the best, you use both, much like high end loudspeaker companies do.
Click to expand...
Interesting (and helpful). Can you give me some examples of effective yet reasonably accessible damping materials?
 
#44
lemans23 said:
My point is that when significantly braced and the excitation band goes higher in frequency, the amplitude is minimal compared to without.... It is no coincidence that the best box speakers in the world are both a. unbelievably stiff and b. quite massive.
Click to expand...

Hi just to say that i thank you very much indeed for your extremely clear and valuable explanation.
I with a friend made some uneducated experiments on cabinets
The only clear outcome is that we heard big variations on sound compared to the original speaker (one was "treated" and the other not).
Then the conclusion was that the cabinet can make or break the speaker sound.
Now i know in which direction to move. Increase stiffness.
Thanks a lot again.
Kind regards, gino

P.S. may i ask you which are the speakers you prefer ? i am curious.
 
Last edited:
#45
lemans23 said:
Put it like this: when you add bracing and move the excitation band up, the amplitude of the resonances falls.
Click to expand...

... while the efficiency of radiation rises, which will more than compensate the effect you try to rely on.

I am sorry, but i can see another thread wasting time here, spreading statements which are not in line with engineering knowledge on structure borne sound.

Just carry on, make stiff and thick cabinets walls to be used widerange in bass, i am fine with it ( ... as long as i do not have to measure your results or listen to it).


lemans23 said:
It is no coincidence that the best box speakers in the world are both a. unbelievably stiff and b. quite massive.
Click to expand...

If you e.g. find an idiot paying the price for a luxury limousine to get a nice bookshelf loudspeaker it should in fact be "incredibly stiff and heavy", thats for sure.

But that has nothing to do with applied acoustics, not at all ... but it has to do a lot with making use of the mental precepts of the idiot willing to buy that product.

I am now canceling my subscription to this thread, because "been there, seen that" you know.

Kind Regards
 
Last edited:
#46
@admins:

My last 2 posts may be removed, as i confused this thread with a different one, where we already had a discussion about cabinet structure (stiffness, mass, damping of the walls).

Of course "mass & damping is king" but excessive stiffness of the walls is counterproductive for "non subwoofer" cabinets.

I did not want to disturb this thread or "prodding" anyone.
 
Last edited:
#47
I think that we are not correctly addressing one another as all that I am stating is fundamental physics. The entire point of the bracing is to move resonances outside of a speaker's passband and decrease the amplitude of the resonances that occur. I am sure we are agreed on that.

The only reason I bothered to comment here is that there is conjecture that bracing a loudspeaker cabinet creates distortion. We are not working with "neglect X" materials here where when you simply increase the stiffness of the system, the natural frequency rises and excitation amplitude is unchanged. The distortion/coloration is fundamentally the acceleration of the panels outward which creates cabinet borne sound. For woofers to excite a speaker cabinet in the "problem" region to the point that it contributes a level of sound comparable to the driver is going to take a seriously unique setup.

Please show me a speaker with high levels of coloration induced by bracing a cabinet so much that the stiffness pushes the resonances into the clearly audible region AND they are still clearly audible. In my many years of experience, I have seen not a single one.

Also let me make it plain: I am not remotely saying to simply brace the **** out of every speaker cabinet to raise its stiffness sky high. I'm talking about just a hollow box vs. a rigid and well damped speaker enclosure.
 
Last edited:
#49
LineArray, I believe I understand what you're saying and it is not what I am arguing. You're envisioning say a midwoofer in a cabinet which has been braced extensively to purposely raise its stiffness. The thought therefore being by intuition that it's not going to radiate sound as effectively because it's "stronger". I agree that this is a stupid idea, much like a cabinet made from say sheet aluminum.
 
#51
I've just found this and the FEA analysis along with (some of) the conclusions made thereafter are what I've been trying to express with words the entire time.

Loudspeaker Cabinet Bracing: Finite Element Analysis - Part I | Audioholics

You will note that as I've said, the fundamental panel resonance amplitude lessens, and while F0 does rise in frequency, the majority of the energy goes towards higher order panel modes which are easily dissipated. If you can create a node by having a properly placed brace, then it is not possible that you've allowed the energy of the system to shift up in frequency AND remain high in amplitude. The FEA results and years of loudspeaker measurements show just that.
 
#52
lemans23 said:
You will note that as I've said, the fundamental panel resonance amplitude lessens, and while F0 does rise in frequency, the majority of the energy goes towards higher order panel modes which are easily dissipated. If you can create a node by having a properly placed brace, then it is not possible that you've allowed the energy of the system to shift up in frequency AND remain high in amplitude. The FEA results and years of loudspeaker measurements show just that.
Click to expand...
If you double the frequency and halve the displacement is the radiated sound from the panel going to be louder, the same or quieter?

If you have a driver vibrating at, say, 500Hz at what frequencies will the cabinet vibrate?
 
#53
Both questions are too simple. Your question assumes that the panel moves like a piston, but it does not. Its output is the summation of all of the eigenmodes that can be excited.

For question 2, this is far too simple as it neglects the fact that a speaker cabinet doesn't behave like a pulsating sphere. It's many small areas radiating their own sound, each of which have their own characteristic properties.
 
#54
lemans23 said:
Both questions are too simple. Your question assumes that the panel moves like a piston, but it does not. Its output is the summation of all of the eigenmodes that can be excited.

For question 2, this is far too simple as it neglects the fact that a speaker cabinet doesn't behave like a pulsating sphere. It's many small areas radiating their own sound, each of which have their own characteristic properties.
Click to expand...
There are no assumptions in the questions but assumptions and physical principles are needed in order to answer them.
 
#55
You can't discuss how to properly brace a speaker until you have a couple of facts

1 what are the box dimensions?
2 what are the box materials?

A generic discussion is simply too vague to suit all boxes of different designs. Eg A 30x20x18cm box made of 18mm birch ply will have different needs to a similar sized box made of 3mm hardboard. In the 18mm birch box you'd Have to play a test tone at levels beyond usual listening and mabe beyond the drivers capability before there was an issue. Playing music with this system I'd bet you'd never come across a problem.

A big 3-way on the other hand......but this still comes down to individual size/construction before bracing is thought about

It's all relative to the design you're adding bracing to
 
#56
andy19191 said:
There are no assumptions in the questions but assumptions and physical principles are needed in order to answer them.
Click to expand...
...What purpose then would that discussion serve other than to steer further off course? Plate, bar, membrane and pistonic acoustic theory are all readily accessible in any good acoustics text. I myself am partial to Beranek's works and Raichel.

OP, I apologize for derailing the thread. Please proceed. 🙂
 
#57
My attempt to answer the questions

andy19191 said:
If you double the frequency and halve the displacement is the radiated sound from the panel going to be louder, the same or quieter?
Click to expand...

If you use the analogy of a speaker driver, halving the excursion will reduce the loudness by 6dB. However this assumes frequency and radiating area of has not changed. If the panel resonance frequency was increased by bracing it, the vibrating area of might be smaller, further reducing the loudness.

However because the frequency has doubled you would also need to consider the Fletcher and Munson curve which shows perceived loudness at different frequencies for equal power. This might mean that even though the radiated power is less you might hear it as louder

andy19191 said:
If you have a driver vibrating at, say, 500Hz at what frequencies will the cabinet vibrate?
Click to expand...

My guess is that unless there is a panel in the speaker cabinet with a resonance of 500Hz (or close to it) the amount of vibration would be very small.
 
#59
lemans23 said:
Fiberglass, mineral wool, denim insulation, and long hair wool are good. Polyfill is decent when used correctly.

Others may disagree, but no foam has ever impressed me (that has been able to fit inside a speaker cabinet anyway).
Click to expand...
Foam probably doesn't collapse or get displaced as speakers are moved about but it does still seem strange then that the BBC chose to use foam in their design of the LS3/5a when they were so fussy about ply, beech and bitumastic panel damping sheets.

I suspect that fibreglass roof and wall insulation was in use at the time and I am absolutely certain that Shepherds had been watching their flocks at night for a VERY long time 😉
 
#60
lemans23 said:
You will note that as I've said, the fundamental panel resonance amplitude lessens, and while F0 does rise in frequency, the majority of the energy goes towards higher order panel modes ...
Click to expand...

I can agree to this so far.


lemans23 said:
which are easily dissipated.
Click to expand...

But i do not agree to this, because the Qs do not change considerably - in a given panel material or cabinet structure - between lower and higher order modes.

Furthermore there often seems to be a misinterpretation on cabinet "quality" with all investigations looking just at "cabinet wall displacement" e.g. measured by interferometry.

- displacement of the walls (even on average) is not "level of unwanted sound radiation"

and furthermore

- "level of unwanted sound radiation" is not "audibility"



When a modally vibrating structure like a loudspeaker cabinet is excited by constant driving force, there is also a fairly constant - frequency independent - average (over area) velocity of the vibrating walls, given the modal density is high enough (e.g. MOF > 3).

The fact that the measured velocity is tilted down by 6dB / octave is due to the driving force being low pass filtered:
The driving force F = B x L x I which acts upon the cone and the driver's basket in equal strength but opposite sign is mass hampered by the basket's mass in acting upon the cabinet's baffle, as long as the basket does not start vibrating in itself at higher frequencies.

This lowpass effect in cabinet excitation has nothing to do with "dissipation" in a structure rising with frequency: This is simply not true, because Qs remain fairly constant with mode number (and frequency) in a modally vibrating structure.

The reason of (observed) low pass filtered wall velocity is just the driving force of the main mechanism of cabinet excitation being low pass filtered.

But when looking at sound radiation - not only looking at "wall displacement" - this low pass effect is "eaten up" and by far overhauled by two different effects introduced by bracing (without damping):

- bending waves have increasing phase speed, which causes more efficient sound radiaion from panels (cabinets) when phase speed of bending waves comes close to the speed of sound in air (coincidence).

- audibility of panel radiation rises with frequency due to the equal loudness curves especially at the lower frequencies.



Bracing and negative effects on unwanted sound radiation:

> rises the phase speed of bending waves, by increasing stiffness overproportionally over mass per area of the cabinet's walls

> thereby the coincidence frequency of the structure is lowered and the efficiency of - unwanted - sound radiation is enhanced dramatically (please refer to standard textbooks on structure borne sound)


> rises the frequencies of the modes of given mode number and thus increases audibility by increasing perceived loudness.

> rises the Q of modes, as "reactance per resistance" of the vibrating cabinet rises : Stiffness is increased without increasing damping.


All these effects by far outweigh the 6db/octave low pass effect, resulting from mass hampering the exciting force by the basket's mass.

BTW: At higher frequencies - where the driver's basket may not be seen as a lumped mass anymore - advocators of bracing (increased stiffness) loose also this effect due to their "pushing modes higher in frequency".


Increased stiffness is only a valid strategy in subwoofer's cabinets, where one can push the modes of the cabinet completely out of the used bandwidth of the enclosure.


Kind Regards
 
Last edited:
Status
Not open for further replies.
Top Bottom