Speaker enclosure vibrations - a few measurements with accelerometer

[fineMen]

Me, together with others, have different results. But first of all: resonances are NOT distortion, NOT, mind You? In general, by principle, basically a resonance may occur in the elastic regime with tiny, yet humanly perceptible micro-meter movement. But there is no element that may actually distort here.

I utterly acknowledge the delight that the o/p gained from doing this and that to his enclosure. But that doesn't mean, that his measurements are correctly performed. There is of course tons of confirmation bias to be criticized. His measurements cannot be reproduced to begin with. Conditions are not clarified so far.

Not the least, while just imitating the constrained layer" technology with some magic hand waving (yes, cargo cult) the idea of a "resonating enclosure" isn't described--nothing of a model, an abstract assembly of elements with properties that are understood.

Measurements without a model just do not exist. Here is no model, so there is no measurement.

I'm strict, I know. But actually, I'm a scientist by education. What the o/p did is exactly anti-science. Thank You, and sorry, really.

 

[Thomas_A]

Me, together with others, have different results. But first of all: resonances are NOT distortion, NOT, mind You? In general, by principle, basically a resonance may occur in the elastic regime with tiny, yet humanly perceptible micro-meter movement. But there is no element that may actually distort here.

I utterly acknowledge the delight that the o/p gained from doing this and that to his enclosure. But that doesn't mean, that his measurements are correctly performed. There is of course tons of confirmation bias to be criticized. His measurements cannot be reproduced to begin with. Conditions are not clarified so far.

Not the least, while just imitating the constrained layer" technology with some magic hand waving (yes, cargo cult) the idea of a "resonating enclosure" isn't described--nothing of a model, an abstract assembly of elements with properties that are understood.

Measurements without a model just do not exist. Here is no model, so there is no measurement.

I'm strict, I know. But actually, I'm a scientist by education. What the o/p did is exactly anti-science. Thank You, and sorry, really.

Well, you and others are completely free to replicate the results and present them including a model. Things started once due to audible effects and 17 years later I made the measurements confirming what I heard. There is nothing more to it really.

(And distortion can be linear and non-linear. So what are resonances?)

 

[fineMen]

Well, you and others are completely free to replicate the results and present them including a model. ...
(And distortion can be linear and non-linear. So what are resonances?)

Thanks, You presented "harmonics" as distortion and correlated them to the so called "constr/ lay/". Missing clarity in wording plus model free data is an essential part in destroying science.

But be assured, in all his desperation in seeking for the "ideal" (sic!) the purity obsessed audiophile will replicate the cargo cult ceremonies.

Add.: You may not get what I'm talking about, but I know for sure. It was the very first piece of information we got in the very first class in the very first minutes of our education to being a scientist ... . Enough said. Reiterated and again, sorry for that, really.

 

[Thomas_A]

Thanks, You presented "harmonics" as distortion and correlated them to the so called "constr/ lay/". Missing clarity in wording plus model free data is an essential part in destroying science.

But be assured, in all his desperation in seeking for the "ideal" (sic!) the purity obsessed audiophile will replicate the cargo cult ceremonies.

I have been quite clear that a uniform glued MDF cabinet should not introduce non-linearities. But what you have is also a driver mounted with scews.

 

[Thomas_A]

Add.: You may not get what I'm talking about, but I know for sure. It was the very first piece of information we got in the very first class in the very first minutes of our education to being a scientist ... . Enough said. Reiterated and again, sorry for that,

Again, a model needs more than just linear modal analysis of uniform MDF when observations tell you a different story. And you are free to replicate the experiment. If the results does not fit your modal analysis model, the model needs to be modified.

 

[thewas]

Me, together with others, have different results.

Could you present those?

 

[witwald]

So, how come that the panel gets into a secondary mode, while the force is uniform?!

The answer to the tricky question is that it doesn't. That's because the excitation force will not excite any even-order modes of the panel. A uniform pressure force can be decomposed into a set of odd-order harmonics (n=1,3,5, etc); the even-order ones are missing. Those odd-order harmonics then serve to excite the panel at its odd-order resonance frequencies.

The 3rd mode shape of a simply supported MDF panel (0.3m x 1.0m and 19mm thick) obtained from a modal analysis (an eigenfrequency solution with no force applied) is shown below:

Below is the 3rd mode shape for the simply-supported panel as obtained from a harmonic analysis using a uniform (upward) pressure as the excitation force:

 

[witwald]

A significant body of work on vibrations of loudspeaker enclosure panels can be found in the following report:

Harwood, H.D.; Matthews, R. (1977). Factors in the design of loudspeaker cabinets. Report 1977-03, January 1977.
https://www.bbc.co.uk/rd/publications/rdreport_1977_03

Harwood and Matthews studied the mechanical properties of timber and other materials that could be used in the construction of loudspeaker cabinets. They undertook extensive measurements and presented details of the results. Their work included assessing a variety of commercially-available damping materials. They described a method of testing that could be used for the cabinets of completed loudspeakers, as well as making an attempt at defining a performance specification.

Does anyone know of any other publications on this topic?

 

[witwald]

But I had a bunch of materials left over from my experiments and I threw them together for the heck of it. I made sandwich of MDF (one side thicker than the other), some wall-damping material, a sheet of steel, and more wall damping material. I put it under my butcher block as well. I didn't end up measuring it with the vibration app, but my goodness that constrained layer damped piece was that one solid item! Rapping my knuckles on the MDF yielded no "wood" thonk sound whatsoever - it was like rapping rock!

And that indicates that the use of constrained layer damping is a very worthwhile approach to damping cabinet vibrations. It is an approach to damping of flexural vibrations in plates that was described by Kerwin back in 1959:

Kerwin, E. M. (1959). Damping of Flexural Waves by a Constrained Viscoelastic Layer. The Journal of the Acoustical Society of America, Vol. 31, pages 952–962.
https://doi.org/10.1121/1.1907821

But if you had speakers and could actually open them up, and layer them as I just described above, what type of plausible benefits might one hear? And what would be the liabilities? Might you change some carefully controlled Q-factor resonance in a speaker for the worse? Might you change internal cavity size in a way that affects the design negatively, etc?

I doubt that there are any cabinet resonances whose changed Q factors would degrade the sound reproduction coming from the speaker. Panel resonances are just bad in general. I agree that reducing the size of the internal cavity could affect the design in a negative way. For example, changing the effective enclosure volume would alter the enclosure tuning of a vented box or a sealed box. However, it would likely be a relatively small change, as long as the constrained layer damping treatment was applied in a modest manner.

 

[witwald]

Thanks, You presented "harmonics" as distortion and correlated them to the so called "constr/ lay/". Missing clarity in wording plus model free data is an essential part in destroying science.

Would it be correct to say that all acoustic output from the walls of elastic loudspeaker cabinets can be deemed to be distortion? After all, an undesired output signal, be it of a harmonic nature or otherwise, when produced in combination with the desired signal is distortion.

 

[witwald]

So what are resonances?

In so far as we are talking about the vibrations of elastic loudspeaker enclosures, resonances are localized frequency-dependent vibrations of the enclosure panels that achieve high amplitudes relative to other parts of the vibration response spectrum. As these resonances are usually relatively lightly damped (high Q), they have the propensity to ring after the removal of whatever signal led to them being excited/forced into motion. This feature, coupled with the fact that they at all times are adding undesired acoustic output in response to music program material, means that they can become audible in an annoying manner. In this way, panel resonances add an unwanted distortion signature to the sound reproduction characteristics of a loudspeaker system.

 

[witwald]

I have been quite clear that a uniform glued MDF cabinet should not introduce non-linearities.

And it is quite likely that any non-linearities in the behavior of the glued MDF panels will be very small, and similarly for the raw MDF material itself.

 

[Thomas_A]

In so far as we are talking about the vibrations of elastic loudspeaker enclosures, resonances are localized frequency-dependent vibrations of the enclosure panels that achieve high amplitudes relative to other parts of the vibration response spectrum. As these resonances are usually relatively lightly damped (high Q), they have the propensity to ring after the removal of whatever signal led to them being excited/forced into motion. This feature, coupled with the fact that they at all times are adding undesired acoustic output in response to music program material, means that they can become audible in an annoying manner. In this manner, panel resonances add an unwanted distortion signature to the sound reproduction characteristics of a loudspeaker system.

Agreed. My point was the use of terminology since this was very important for Finemen. Distortion can be linear and non-linear and resonances casuing frequency response errors are linear distortion. With respect to the non-linear effects you must add not only the MDF but also the driver, cage-magnet resonance, and interaction between frame and enclosure. The driver can introduce non-linearites and is as far as I know, not included in the modal analysis that Finemen refers to. As Purify also notes there is a 380 Hz cage-magnet resonance in their driver that may introduce non-linearites if not dealt with (special nuts and torque force). This exists in all drivers as soon as you fasten it to an enclosure.

 

[Thomas_A]

And it is quite likely that any non-linearities in the behavior of the glued MDF panels will be very small, and similarly for the raw MDF material itself.

Yes as long as it is isotropic. Question is what happens when adding a driver with its own resonant behaviour. I can’t see that the combination is isotropic but rather anisotropic.

 

[DanielT]

In this video it is tested with damping glue (constrained layer damping). Measurements 36:45 into the video with accelerometer are performed, first on only MDF and Playwood, then 42:10 into the video measurements with damping glue (constrained layer damping).

It's just a shame that he seems to have different thicknesses on the pieces of wood when he tests. In any case tips are given on a brand of damping glue (constrained layer damping) which could be obtained.In the US when the video was made anyway

And nope it's not the world's second best speaker, quite the opposite. Such pipes, like the one in the video, are notoriously difficult to get a good FR with. A lot of damping material is usually needed in the pipe, so the effectiveness ends up being reduced, but better FR can then be the case ( if you are lucky). After many tests that is. Which ends with the fact that you could just as easily build a traditional sealed speaker and avoid all the worries that a pipe entails.The pipe in itself then adds nothing compared to a sealed box in terms of efficiency, the speaker's sensitivity, and FR. It was a bit OT but still just thought I'd mention it considering his bragging in the video.
(what I have read about those who have tested these type of pipes)

Voigt pipe - Wikipedia

en.wikipedia.org

 

[thewas]

Here also some knock test audio recordings on different board combinations

DUO-DXT

hifi-selbstbau.de

 

[NTK]

MDF, like most composite materials, is anisotropic. E1, E2 (elastic moduli parallel to the sheet, say, in x, y directions) are very different from E3 (perpendicular to the sheet, z direction).

https://www.researchgate.net/publication/285495335_Mechanical_properties_of_MDF_as_function_of_density_and_moisture_content

Also, glued joints behave very differently in compression (when being pressed together, compressing the MDF) than in tension (when being pulled apart, stretching the glue). In fact, screwed joints can be more linear if the screw preloads aren't exceeded, i.e. the joints are under compression at all time (may not be easy for MDF/wood as they tend to relax with time).

"Constrained layer damping" is also inherently nonlinear as it depends on the shear modulus of the viscoelastic layer being a strong nonlinear function of frequency (loss/damping is strongest in the transition region).

To summarize, the structural behavior of loudspeaker cabinets should not be assumed to be linear.

 

[Thomas_A]

MDF, like most composite materials, is anisotropic. E1, E2 (elastic moduli parallel to the sheet, say, in x, y directions) are very different from E3 (perpendicular to the sheet, z direction).

https://www.researchgate.net/publication/285495335_Mechanical_properties_of_MDF_as_function_of_density_and_moisture_content

Also, glued joints behave very differently in compression (when being pressed together, compressing the MDF) than in tension (when being pulled apart, stretching the glue). In fact, screwed joints can be more linear if the screw preloads aren't exceeded, i.e. the joints are under compression at all time (may not be easy for MDF/wood as they tend to relax with time).

"Constrained layer damping" is also inherently nonlinear as it depends on the shear modulus of the viscoelastic layer being a strong nonlinear function of frequency (loss/damping is strongest in the transition region).

View attachment 240618

To summarize, the structural behavior of loudspeaker cabinets should not be assumed to be linear.

But the point is xy dimension of MDF which becomes quite altered when bolting a metal frame onto it.

And I am not at all convinced that a cabinet with screws works better than glue due to the reasons of force at the joints vs other points (and over time).

 

[NTK]

And I am not at all convinced that a cabinet with screws works better than glue due to the reasons of force at the joints vs other points (and over time).

No. I am not saying it is better. I am only saying it may be a more linear structure. Linear may not be what is needed, e.g. CLD works because of its nonlinearity.

 

[Alexium]

There is some discussion of cabinet knock test here. What about the speaker cone tap test? What affects the sound you hear when gently tapping the center of a base or midbase driver? I have a pair of floorstanders and a pair of more or less matching bookshelves. On the bookshelf the tapping sound is what I expect it to be, but on the floorstanders it's colored. Both 6.5", but the drivers are different models. Is it affected by the speaker parameters, by the enclosure type (sealed, BR, TL etc.)? Is this "test" useful for anything?