Bracing material?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
What is sufficient energy?

You should be telling us. You are telling us in no uncertain terms it can & will happen. How can we hope to cure the problem you say always exists unless you know how to start the problem off?

Heres how: You need sufficient volume, a badly designed & built cabinet and maybe a disc of test tones.

All of which are easily remidied by 1) not playing at 100db (85db to keep your hearing ;)), 2) don't be a wally when designing & building your cabinet and 3) listen to music not test tones

In other words: Its not an issue in a large percentage of cases in the real world. (There are always exceptions to the rule)
 
Wrong. No matter how much energy you put into it, the resonance is always there. And, at resonance, the panel is more likely to receive the energy, because both impedances, mass and stiffness, cancel out and there's only the resistance (the damping) working.
The impedances cancel out?

Consider a lead ball hanging on a string. It will have inertia due to its mass, and to move the ball we need enough force to overcome the inertia. Now, replace the string with a steel wire, fixed rather than hinged at both ends. The wire will act as a spring, and the spring will add to the inertia of the ball, thus even more force will be required. Increase the stiffness of the wire, and the force requirement will increase further. At no point does the stiffness 'cancel out' the inertia of the mass.
 
What is the complex impedance of ANY resonant circuit (regardless if electrical, mechanical or acoustical) at resonance? The phase is zero, so the complex impedance is purely resistive. The impedances of mass and stiffness (or inductance and capacitance) cancel out, and only the damping resistive part works.

Come on, that are basics!
 
What is the complex impedance of ANY resonant circuit (regardless if electrical, mechanical or acoustical) at resonance? The phase is zero, so the complex impedance is purely resistive. The impedances of mass and stiffness (or inductance and capacitance) cancel out, and only the damping resistive part works.

Come on, that are basics!

A resonant system as i understand it, is at its simplest a 2nd order transfer function M-C-K system of mass M, damper C ,and compliance K. Much akin to a car suspension system. I must say that i thought the phase was 180deg not 0, but ive been wrong many times.

You are correct, just like in a parallel LC at resonance Xc and Xl cancel leaving R, resistive loss.

The real question is what level of loss is optimal, how to achieve it in a given panel in order to coincide with the resonance, and damp it optimally.

At all other frequencies, our attentions can then be turned to reflective and diffractive absorption.
 
Last edited:

I see math not music. Can you hear it? Or can you only measure it?

I've said it before, I'll say it again. Music has been around since someone could repeatedly grunt at the same pitch. It exists now away from hifi and will continue to survive without hifi. The same cannot be said of hifi.

What are the differences between test tones and music?

um let me see, Mozart, Sandy Denny, Joe Bonamassa etc etc? I could go on but I don't know if you'll get my drift (where can I buy test tones 5th or greatest hits?)

Come on, that are basics!

But then so is listening to actual music on your stereo and being content, instead of using an a accelerometer to measure something you in all probability cant hear
 
I see math not music.

Then you should probably do not post on this topic...

um let me see, Mozart, Sandy Denny, Joe Bonamassa etc etc?

That's no difference. You know, I could use Mozart instead of test tones. It doesn't matter, I can still measure the resonances. That's signal theory.

But then so is listening to actual music on your stereo and being content, instead of using an a accelerometer to measure something you in all probability cant hear

You were talking about excitation of panel resonances. Audibility is another topic, and should be discussed separately.
 
What is the complex impedance of ANY resonant circuit (regardless if electrical, mechanical or acoustical) at resonance? The phase is zero, so the complex impedance is purely resistive. The impedances of mass and stiffness (or inductance and capacitance) cancel out, and only the damping resistive part works.
The "impedances" can't cancel out, because neither is a simple number. This isn't electronics or theoretical physics. There will be a resonance at some frequency, as Dave specified, but you seem to be arguing the "purely resistive" part is negligible.


Dave said:
Fighting this is the natural energy dissipation of the material which is related to the relative dimensional differences between the material thickness and the wavelength in question.
To which you replied, "What?" The question hasn't been answered, but it would behoove you to figure out to what he was referring.


BTW. The measurements you posted ( http://www.diyaudio.com/forums/multi-way/218981-enclousure-strengthening-needed-3.html#post3151455 ) merely show that what you refer to as "heavily braced"... isn't.
 
The "impedances" can't cancel out, because neither is a simple number. This isn't electronics or theoretical physics.

Yes, it is. And you should really take a lesson in it.

The "natural energy dissipation" is the resisitve part in the complex impedance, so at resonance only this works. But there is (nearly) nothing non-linear in the damping over a wide range of stress and strain, so every resonance is excited during playback of music even at low levels.
 
Yes, it is. And you should really take a lesson in it.

The "natural energy dissipation" is the resisitve part in the complex impedance, so at resonance only this works. But there is (nearly) nothing non-linear in the damping over a wide range of stress and strain, so every resonance is excited during playback of music even at low levels.

This is what I was saying. The 'reactive' parts at resonance cancel, leaving resistive loss.
I.e. At resonance only the resistive loss has to be overcome for a resonance to establish.
Material such as MDF which is supposedly lossy, exhibits a greater loss 'bottom-line' than that of a harder material that is less internally damped.
 
Yes, it is. And you should really take a lesson in it.
No, it's a very practical application of materials science.

The "natural energy dissipation" is the resisitve part in the complex impedance, so at resonance only this works. But there is (nearly) nothing non-linear in the damping over a wide range of stress and strain, so every resonance is excited during playback of music even at low levels.
I guess that's why musical instruments sound the same whether they're played loudly or softly. ;)

Look, you're really confused. You seem to be conflating resistance and damping, and while the terms are interchangeable in electronic filter theory, they have very different meanings in materials. Electricity is essentially a one dimensional phenomena, yet we need to invoke an imaginary term (i) to keep track of the phase of a transmitted waveform. How many imaginary terms do you think it takes to track the phase relationships of waves propagating in a three dimensional material? And do you think they're going to conveniently cancel out?

You are suggesting that the tensile strength is not a factor, but as I explained before, it never goes to zero. It will have a minimum at a certain frequency, but it's still the dominant force resisting motion (assuming we have overcome the inertia of the mass). And given we're talking about non-homogenous materials, usually of rectangular shape, with the edges constrained, you think the resistance is going to be linear?

Not, you know, that we want it to be linear. That's the whole point of bracing. Creating areas of increased tensile strength, preferably in a random pattern, that prevent a uniform motion over the entire surface. (see how I sneaked back on topic? :D)
 
Then you should probably do not post on this topic...That's no difference. You know, I could use Mozart instead of test tones. It doesn't matter, I can still measure the resonances. That's signal theory.....You were talking about excitation of panel resonances. Audibility is another topic, and should be discussed separately.

So far you have repeated over and over that it can happen and can be measured but haven't answered at what levels it becomes an issue, or even if its audible under normal listening conditions with music which is all any speaker builder would be interested in. Audibility is the only subject in a loudspeaker forum.

BTW its very rude to tell someone not to post in a thread especially as this is not a theoretical physics forum and no practical application has been suggested by you in relation to the opening post.
 
keriwena, I do not think myself an expert materials science and doubtless there is much for me to learn, but the analogies with electrical and acoustic and mechanical eng are very well documented. That is accepted theory. Where there certainly MAY be more than 2 poles acting, there can only be an odd or even number. At resonance 2 of these cancel, leaving the real term. If 2 poles still remain they too can be conjugates. If 1 pole remains then we have reached a point where damping is the only recourse. I believe. I qualify my statement by stating that I do not have the knowledge to say for certain. But then neither does anyone else here, and i cant give a numerical answer, also similarly to the others here. But baseballbat has his control sys knowledge right on. Dismissing theory because it doesnt fit a personal belief is reckless, unless there is a 'better' theory. In which case, please share,im sure we could all benefit.

My greater point being that there is no material which satisfies those requirements. Since we want the rigidity of a high MOE solid, with the damping of a far softer one. The answer? Dissimilar bracing and cabinet materials, or CLD. A debate is about discussion, and comparing ideas, but at some point a ref to theory or the reasoning that formed your conclusion is necessary. Its not good enough to proclaim anothers error, without bringing the vinyl to the disco.
 
Last edited:
keriwena, I do not think myself an expert materials science and doubtless there is much for me to learn, but the analogies with electrical and acoustic and mechanical eng are very well documented.
This is true. Consider the case of one Neville Thiele, who ignorantly applied electrical filter theory to bass reflex boxes. His calculations generated errors of nearly 30%. It's all well documented in the AES journals, as is the work of Small, Keele, and others who accommodated the losses that Thiele overlooked.

I feel I've stated my case as clearly as I can. As I said, I think you guys are conflating the terms "damping" and "resistance", and it occurs to me that further, you are ascribing the behavior of a resistor, which is by definition linear, to the resistance to motion of a speaker cabinet panel. My contention is that "resistance" is primarily the tensile behaviour of the material, i.e., its behaviour as a spring. And springs are not linear, unless a fair amount of engineering is applied to make them that way.

What your "theory" doesn't explain is why a cabinet, whose panels exhibit losses better than -16dB below the driver's output at +86dB, can howl like a banshee at +106dB. If what you appear to be saying was true, then we'd expect linear behaviour from the panels, and they'd still be -16dB down with a five watt input (I've mentioned I'm playing with loudspeakers, haven't I? ;)).

For a theory to be accepted, it has to explain and predict observed phenomena, and in this case, I don't believe it does.
 
An interesting debate to be sure.
However for some of us the point regarding the use or non use of MDF is moot at best.

Where I live the types of ply that some are recommending are simply not available. They are northern hemisphere products and as such are simply not imported. What little high quality speaker building grade ply there is is generally used in the marine industry and can exceed mdf by triple the cost.

Therefore mdf really is the only material priced well enough for speaker hobbyists and builders, In fact commercial speaker builders either make there own ply or have special runs done to their specs.

With the speakers I build, I borrowed a design idea from my own industry ( The Glass Trade ), and use the same principles used in the manufacture of sound reducing glass. Thats simply two substrate of differing thickness and densities (where possible) laminated into a single piece (laminated).

Using this idea I use mdf for all the inner casing , with hardwood cleats, corner braces and screw points. The exterior surface is 8mm native hardwood strips, recycled 4 x 2 and 6 x 2 that I machine in my workshop. Each piece is liberally coated with glue and clamped into place. That done and glue dried and you have an very solid wall that resists resonance very nicely, running at 89dB or close to it, yields very little cabinet vibration (resonances) for both sealed and vented enclosures. You also end up with a nicely mannered enclosures that have a much warmer sound than standard mdf. Mounted on spikes for further isolation again improves the overall enclosure performance.

Stringed instruments sound very natural and accurate when using this method of construction.

Of course enclosure design also plays a large part as well and I prefer an cabinet that's either curved or 6 sided with enclosures above 12litres. The result is a heavy dense enclosure with good ( not perfect) resonance characteristics. A recent build.
DSCF2682.JPG


DSCF2683.JPG


DSCF2692.JPG
]
 
I guess that's why musical instruments sound the same whether they're played loudly or softly. ;)

And that has nothing to do with Fletcher-Munson, higher harmonics due to higher excitation of, for example, the string.

Electricity is essentially a one dimensional phenomena

Yeah, right. The theory of an electromagnetic waveguide is one-dimensional, as is the propagation of electromagnetic high frequency waves on a copper plane.

How many imaginary terms do you think it takes to track the phase relationships of waves propagating in a three dimensional material? And do you think they're going to conveniently cancel out?

Please, get the waveform theory right. It is the same regardless if acoustics, electromagnetics or mechanics.

You are suggesting that the tensile strength is not a factor, but as I explained before, it never goes to zero. It will have a minimum at a certain frequency, but it's still the dominant force resisting motion (assuming we have overcome the inertia of the mass).

No problem. I'll just say that this non-linearity is so small, that it is not relevant in practice. Look at my measurement. The vibration of the panel was measured at approx. 90 dB@1m. This level is easily reached in practice.

And given we're talking about non-homogenous materials, usually of rectangular shape, with the edges constrained, you think the resistance is going to be linear?

Here's a picture: http://upload.wikimedia.org/wikiped....png/450px-Stress_Strain_Ductile_Material.png

Please show me in which area of the curve we're working.

Not, you know, that we want it to be linear. That's the whole point of bracing. Creating areas of increased tensile strength, preferably in a random pattern, that prevent a uniform motion over the entire surface.

Bracing can be easily seen as a step in the local impedances. Propagating waves are reflected at this step. In result, the eigenfrequencies of the panel change, and you get another resonance pattern. Might be useful or harmful, depending of where the new resonances are (in a more audible frequency region?) or the level (higher Q?).

In bass region, below the fundamental resonance of the panel, the bracing increases the strength, reducing motion of the panel, because it is stiffness controlled at that frequencies. In that frequency region you are right, and I never denied it. But I was tallking about modal behaviour.

My contention is that "resistance" is primarily the tensile behaviour of the material, i.e., its behaviour as a spring.

A spring is never a resistance, because a resistance is dissipating energy, while a spring is storing energy.

What your "theory" doesn't explain is why a cabinet, whose panels exhibit losses better than -16dB below the driver's output at +86dB, can howl like a banshee at +106dB. If what you appear to be saying was true, then we'd expect linear behaviour from the panels, and they'd still be -16dB down with a five watt input (I've mentioned I'm playing with loudspeakers, haven't I? ).

Pics or it didn't happen. You've made an observation that is not explained by a widely accepted and proved theory. So you have to present valid measurements about observation, which either shows that the theory is completely wrong, or are capable of extending the theory.
So, was it really the cabinet that was resonating at a much higher level than expected? Or was it just a driver at overload?
The latter could be interesting: assume you have panel resonances above 500 Hz, and the woofer is crossed at 400 Hz. In normal operation the resonances will certainly not excited by the woofer. But if he produces high distortion, those resonances will be excited. Have you considered that?

You see, you have a lot to do to show that the theory is wrong. As long as your data is not valid, the rest of the world assumes that the well known and proven theory is right.

BTW, by admitting that the panel has a damping better than 16 dB you prove I'm right. The panel IS vibrating, regardless of excitation level.

@JRKO:
I've already said something about audibility. Above 30 dB below the drivers sound panel resonances may become audible under perfect conditions and certain signals (which also includes some music), and above 20 dB below the drivers sound they surely become audible. I've also said that it is a mainly linear process, and therefore the excitation level is irrelevant. And this is not just my view, but the view of the rest of the world except a few people.
 
I consider a panel as a very lossy spring. Loss in the form of friction. The loss can NEVER overtake the reactive part, without the spring ceasing to be a spring... i love the way my own text is quoted by others and used as a veiled insult. You clearly believe you know something that I dont, please share. I have been quit modest in my interpretation of the theory, while you argue against without a shred of math to back it up. Damping Zeta is part of the 2nd order (or higher) transfer function. I am not confusing this with resistance, i am also not confusing the spring k. As BBB said, springs store energy in the main. This forum is becoming troll heaven. Leaving in search of a forum where your words arent picked apart, regurgitated as an insult to your intelligence, WITHOUT the responder adding to the discussion in a meaningful way (facts, correction n citation for when im wrong). No, rather than a little humility, these folk just perpetuate the 'black art' image of Audio.
 
Last edited:
Pics or it didn't happen. You've made an observation that is not explained by a widely accepted and proved theory.
Which is what I've been saying. That theory isn't comprehensive enough to explain the real world. There's more to it. I describe what is essentially a thick drumhead, and you reply with a two-dimensional graph of Young's Modulus? You're over-simplifying everything.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.