Speaker enclosure vibrations - part II - ABX wanted!

[test1223]

Some are made in the thread I linked to in first post.

Post in thread 'Speaker enclosure vibrations - a few measurements with accelerometer'

Oct 30, 2022

Today I made some acoustic measurements of the two speakers, about 2.9V signal, frequencies 510 Hz, 620 Hz, and 790 Hz, and RTA using REW. Speaker A is the old one with 19 mm MDF and speaker B with constrained layer damping. Note that the two 5 inch drivers are different Peerless drivers, but as mentioned, I do not have the previous conventional cabinets that I made for the newer CSX drivers. Matching levels were difficult, but THD was consistently lower in speaker B, and especially around 600 Hz.

At 510 Hz, speaker A

Speaker B:

620 Hz, speaker A

...

• Thomas_A

• 

Thank you. It makes a lot of sense harmonics between 1kHz and 5kHz are very audible especially if they have a higher order. Even the improved enclosure should have some room for further improvement.

 

[Heinrich]

Thank you. It makes a lot of sense harmonics between 1kHz and 5kHz are very audible especially if they have a higher order. Even the improved enclosure should have some room for further improvement.

This is a typical example of a general observation. DIY topics are included in the Review++ section. If it were a matter of improving ready-made speakers, why not? But the CLD does not go in that direction at all. Rather, the design of ready-made speakers is generally suspected of not getting the most out of them.

Here, too, things go wrong because even DIY enthusiasts sometimes do not fully understand what the measurements mean. And individual cases cannot be generalized anyway. What's more, critical discussion is often lacking. Some traditional—never tested—assumptions spread from the hobby corner into the discussion of professional, systematic work (see below).

If it is meant that cabinet vibrations lead to non-linear distortions, then the question of where these come from is justified. These are accelerations with deflections in the micrometer range. Could it not be that the construction of type A and type B is generally quite different, that the adhesive bonds are different? It would be enough for the driver to be screwed on with varying degrees of force, and so forth. And then there is also the question of whether non-parallel walls change anything substantial in terms of vibration patterns, since this assumption meant that a much more (cost/effort-)effective internal stiffening was not considered here (post #12).

Finally, I wonder what the hearing test is actually for. “The recording was made using a 540 Hz signal, ...” - a single tone?! Comes down to detect the distortion, right? I'e never measured distortion from activated cabinet walls. There's something wrong with the MDF type, which is not inherent to MDF.

 

[Heinrich]

This is a typical example of a general observation. ...

Postscript: Please do not take this personally; it is just that so much misleading information has been written about cabinet vibrations for many decades. This should now come to an end. All myths on this subject are fundamentally incorrect. Of course, you will find a little bit of confirmation if you focus on the desired effects – the behaviour of a cabinet is sufficiently diverse and diffuse.

Here is a more systematic approach: https://www.waveguide-audio.de/vibrationsmessungen-und-nachbetrachtungen.html
Unfortunately, it is only available in German. Once again, all myths have been disproved, literally all.

This is likely to be particularly important when evaluating ready-made loudspeakers. More effort does not help. A single, sensibly placed strut is usually enough. The rest is taken care of by damping of the internal sound orderly. Professional, effective engineering versus tons of "what you can do", but alas, just that.

 

[Thomas_A]

This is a typical example of a general observation. DIY topics are included in the Review++ section. If it were a matter of improving ready-made speakers, why not? But the CLD does not go in that direction at all. Rather, the design of ready-made speakers is generally suspected of not getting the most out of them.

Here, too, things go wrong because even DIY enthusiasts sometimes do not fully understand what the measurements mean. And individual cases cannot be generalized anyway. What's more, critical discussion is often lacking. Some traditional—never tested—assumptions spread from the hobby corner into the discussion of professional, systematic work (see below).

If it is meant that cabinet vibrations lead to non-linear distortions, then the question of where these come from is justified. These are accelerations with deflections in the micrometer range. Could it not be that the construction of type A and type B is generally quite different, that the adhesive bonds are different? It would be enough for the driver to be screwed on with varying degrees of force, and so forth. And then there is also the question of whether non-parallel walls change anything substantial in terms of vibration patterns, since this assumption meant that a much more (cost/effort-)effective internal stiffening was not considered here (post #12).

Finally, I wonder what the hearing test is actually for. “The recording was made using a 540 Hz signal, ...” - a single tone?! Comes down to detect the distortion, right? I'e never measured distortion from activated cabinet walls. There's something wrong with the MDF type, which is not inherent to MDF.

So what is the displacement of a 5-inch woofer playing 550 Hz at 80 dB SPL @ 1 m? In the ±20 µm range? What would a 3rd harmonic at -40 dB be? 25 nm?

 

[Heinrich]

So what is the displacement of a 5-inch woofer playing 550 Hz at 80 dB SPL @ 1 m? In the ±20 µm range? What would a 3rd harmonic at -40 dB be?

I was referring to the movement of the enclosure panel in question. Be that as it may – or precisely because of that – it seems to me that the measurements are not based on a desire to conduct research, but rather on a desire to confirm an optimistic assumption. The design of the measurements is therefore also kept quite simple. Critical cross-checks are missing.

I had already pointed out that I had never measured any distortions in the movement of panels that did not originate from the driver or its mounting. But the distortions are the only difference in the sound files presented in post #1. Unfortunately, the story provides little basis for discussion.

 

[Thomas_A]

I was referring to the movement of the enclosure panel in question. Be that as it may – or precisely because of that – it seems to me that the measurements are not based on a desire to conduct research, but rather on a desire to confirm an optimistic assumption. The design of the measurements is therefore also kept quite simple. Critical cross-checks are missing.

I had already pointed out that I had never measured any distortions in the movement of panels that did not originate from the driver or its mounting. But the distortions are the only difference in the sound files presented in post #1. Unfortunately, the story provides little basis for discussion.

As mentioned bracing and damping material are ways to deal with this. Or as here, CLD. Still there are lots of small speakers that does not follow the logical gradual increase of distortion towards lower frequencies. There is often a "dip" and "rise", as this one measured by Amir. Why do you think it is so?

 

[Heinrich]

As mentioned ...

As said, I don't think your findings are to be debated. Regarding the Elac, it is not my place to speculate about things that are foreign to me.

 

[Thomas_A]

As said, I don't think your findings are to be debated. Regarding the Elac, it is not my place to speculate about things that are foreign to me.

Ok. I can also mention the Revel M106 that Amir measured. It shows the logic increase of distortion against lower frequencies with no peaking in the mid-frequencies. As Amir also mentioned in the review: "I read some place that Kevin Voecks (most visible person behind Revel line) said that there was strong effort to reduce distortion and that they had managed to push it below audibility in mid to higher frequencies."

I read between the lines that this was not only by choosing good drivers, but also cabinet design.

 

[Thomas_A]

I was referring to the movement of the enclosure panel in question. Be that as it may – or precisely because of that – it seems to me that the measurements are not based on a desire to conduct research, but rather on a desire to confirm an optimistic assumption. The design of the measurements is therefore also kept quite simple. Critical cross-checks are missing.

I had already pointed out that I had never measured any distortions in the movement of panels that did not originate from the driver or its mounting. But the distortions are the only difference in the sound files presented in post #1. Unfortunately, the story provides little basis for discussion.

So a bit of thoughts here.

If you have the ideally inert speaker with a driver-cabinet mass ratio of 1:1000 floating in air, you will get a linear displacement of the cabinet which is 1000-fold lower than the driver. That equals -60 dB. The only way to decrease this is to use a higher mass ratio. However a cabinet cannot be 100% inert within 20-20000 Hz, so therefore resonances exists. If stiffnes is high and damping is low, you can expect high Q of the resonances. A 20 dB increase at resonance would take you to -40 dB linear displacement. The problem, as it seems, these may also be non-linear. To this you also have standing waves within the cabinet which add to non-linearities. Damping material within the cabinet helps with standing waves but when a cabinet already has got the damping material, the only thing to do is to increase stiffnes by bracing (moving resonances around) or decrease stiffness and add damping by CLD (reduce Q). All of these cabinet vibrations or distortion should ideally be measured in the direction to driver force. Not at side walls.

 

[Heinrich]

... displacement. The problem, as it seems, these may also be non-linear. To this you also have standing waves within the cabinet which add to non-linearities.

Two forms of "non-linear", namely by amplitude (frequency response) and harmonic distortion (frequencies created, in lack of a better word). These two forms are different in origin and effect.

All of these cabinet vibrations or distortion should ideally be measured in the direction to driver force. Not at side walls.

Depends - alas the above mentioned link is only available in German. I may summarize the findings:

Yes, to use CLD can help to a degree,
but bracing, even of a simple kind helps way more,
while the panels' resonances are driven to the major part by the internal air volume's resonances,
so that bracing does not shift a panels' observed resonance frequency,
and additionally no break-up modes are expected,
which should all be set into critical relation to sound escaping through the driver's cone alone -
not the least harmonic distortion from the panels alone was utterly neglegible.

The author experienced some severe trouble with glueing the CLD to regular panels made from MDF. It separated within weeks in the hide-out of a speaker's enclosure, unobserved.

Once the tech/ parameters are understood, your problem seems to originate in a not-so optimal construction of the type MDF box. Alas, it would be very problematic to investigate the case of the casing ;-) after the fact remotely.

 

[Thomas_A]

Two forms of "non-linear", namely by amplitude (frequency response) and harmonic distortion (frequencies created, in lack of a better word). These two forms are different in origin and effect.


Depends - alas the above mentioned link is only available in German. I may summarize the findings:

Yes, to use CLD can help to a degree,
but bracing, even of a simple kind helps way more,
while the panels' resonances are driven to the major part by the internal air volume's resonances,
so that bracing does not shift a panels' observed resonance frequency,
and additionally no break-up modes are expected,
which should all be set into critical relation to sound escaping through the driver's cone alone -
not the least harmonic distortion from the panels alone was utterly neglegible.

The author experienced some severe trouble with glueing the CLD to regular panels made from MDF. It separated within weeks in the hide-out of a speaker's enclosure, unobserved.

Once the tech/ parameters are understood, your problem seems to originate in a not-so optimal construction of the type MDF box. Alas, it would be very problematic to investigate the case of the casing ;-) after the fact remotely.

My understanding that amplitude is a form of linear distortion while those creating new frequencies are non-linear. Linear distortion in these cases are of no concern since the effect on frequency response is very tiny. It is when non-linear effects occur that the problem may arise.

What is simple or not to construct depends. Bracing the type of speaker box I exemplified is not trivial when using struts bust because speaker walls are non-parallell. Glueing CLD was not difficult and there is no separation after 20 years of use. Using the correct glue is probably a start (here Swedac DGA2).

And you are probably correct that the type of box I exemplified might be a special case. Both the commercially bought speaker and two DIY cabinets showed the same results though, so I do not think this is a construction error from my side. Just a box with quite high stiffness and high Q resonances.

 

[Heinrich]

Bracing the type of speaker box I exemplified is not trivial when using struts bust because speaker walls are non-parallell.

Ah, now I get it. There's a solution for that. What I do: I glue a small ear-sized section of MDF onto the panel (inside). The cuting edge sticks to the panel. It doesn't need to be pressed, the PVA-glue sucks in for itself, just letting it dry. Then I glue a strut onto tha little piece, so that it nearly reaches the opposing wall. I fixate it until cured. In a third step I glue a second ear both to the strut's free end and the panel, which then connects panel-ear-strut-ear-panel, done. No exactness required, easy pleasy, its ugliness (to some) is hidden in the enclosue, but it darn works.

Spoiler: how ChatGPT puts it, the usual praise omitted:

... By using a small MDF "ear" as a mounting block, you avoid the need for complex joins or precision fitting, while still creating a strong, hidden support structure inside the panel. Your step-by-step method makes it very clear:

1. Glue ear to panel – Let the PVC glue wick in and dry on its own.
2. Glue strut to ear – Position it to nearly reach the opposing wall and let it cure in place.
3. Add second ear – Glue it to both the strut and the opposite panel, forming a stable bridge.

... simplicity: it doesn't require perfect measurements, it's repeatable, and the structure ends up robust. Sure, it might not win a design award, but as you said, it works, and it's out of sight anyway.

 

[Thomas_A]

Ah, now I get it. There's a solution for that. What I do: I glue a small ear-sized section of MDF onto the panel (inside). The cuting edge sticks to the panel. It doesn't need to be pressed, the pvc-glue sucks in for itself, just letting it dry. Then I glue a strut onto tha little piece, so that it nearly reaches the opposing wall. I fixate it until cured. In a third step I glue a second ear both to the strut's free end and the panel, which then connects panel-ear-strut-ear-panel, done. No exactness required, easy pleasy, its ugliness (to some) is hidden in the enclosue, but it darn works.

Spoiler: how ChatGPT puts it, the usual praise omitted:

... By using a small MDF "ear" as a mounting block, you avoid the need for complex joins or precision fitting, while still creating a strong, hidden support structure inside the panel. Your step-by-step method makes it very clear:

1. Glue ear to panel – Let the PVC glue wick in and dry on its own.
2. Glue strut to ear – Position it to nearly reach the opposing wall and let it cure in place.
3. Add second ear – Glue it to both the strut and the opposite panel, forming a stable bridge.

... simplicity: it doesn't require perfect measurements, it's repeatable, and the structure ends up robust. Sure, it might not win a design award, but as you said, it works, and it's out of sight anyway.

There are probably several ways to do it. But just order preformed 9 mm boards ready-made, glue them together, and then make a cube, slant it and glue front panel is easy as well.

 

[MattHooper]

Curious: what would the likely effect be of these resonance vibrations if coupled to a large wood floor?

 

[Thomas_A]

Curious: what would the likely effect be of these resonance vibrations if coupled to a large wood floor?

Probably nothing. The putative effects with floor-speaker coupling happens somewhere in the bass region. Movement happens at resonance which is in the bass or upper bass region and also where the floor vibrates due to standing waves in the room. Above those resonances, the speaker is ”by its own”. See it as the fundamental tonearm/cartridge resonance; at resonance the system is moving; at higher frequencies it is either the tonearm itself or cartridge that has intrinsic resonances. These can only be solved with stiffness/damping of the tonearm tube and/or damping the cartridge.

I might pursue an IMD test to look at any effects of speaker to floor.

 

[Heinrich]

Curious: what would the likely effect be of these resonance vibrations if coupled to a large wood floor?

Good question because he does so, see this post:

'Speaker enclosure vibrations - a few measurements with accelerometer'

Aug 4, 2022

There are been quite a number of threads dealing with vibrations in speaker cabinets, both in relation to "speaker feet" and vibrations in enclosures themselves. I recently took out my ACH01 accelerometer to do some measurements, both speaker feet and enclosures. There was some initial measurements posted in another thread, and here I have repeated these with some modifications. For test one I wanted to test speaker feet, i.e. the so-called "spikes" against soft feet. However, I do not want to destroy my wooden floor or my speakers so I choose to put the speakers on the floor and use three...

• Thomas_A

• 

Anyway, the comparison of MDF vs/ CLD is also demystified. Different drivers! And then we wonder why the distortion profile is different, right?

... pursue an IMD test to look at any effects of speaker to floor.

You may want to clarify if type A and type B differ only in the cabinet's construction.

 

[tmuikku]

Dramatic difference between the files on post #1. It's very important to pay attention on constructs.

 

[Thomas_A]

Good question because he does so, see this post:

'Speaker enclosure vibrations - a few measurements with accelerometer'

Aug 4, 2022

There are been quite a number of threads dealing with vibrations in speaker cabinets, both in relation to "speaker feet" and vibrations in enclosures themselves. I recently took out my ACH01 accelerometer to do some measurements, both speaker feet and enclosures. There was some initial measurements posted in another thread, and here I have repeated these with some modifications. For test one I wanted to test speaker feet, i.e. the so-called "spikes" against soft feet. However, I do not want to destroy my wooden floor or my speakers so I choose to put the speakers on the floor and use three...

• Thomas_A

• 

Anyway, the comparison of MDF vs/ CLD is also demystified. Different drivers! And then we wonder why the distortion profile is different, right?


You may want to clarify if type A and type B differ only in the cabinet's construction.

Yes there are different drivers (5 inch Peerless). But as mentioned the same problem with TWO different cabinets that I made using the same new drivers. These cabinets are long gone. The distortion was measured with these as well linked to a Swedish forum post 20 years ago.

 

[Thomas_A]

One can also look at the review of the Aiyima speaker. Does anyone believe that the driver itself has that nasty 400 Hz resonance?

'AIYIMA S600 Review'

Aug 2, 2025

It’s been a while since I reviewed a compact, entry-level bookshelf speaker.
This one comes with a variety of input/output options and even a remote control.
If it gets the fundamentals right, it could be worth a closer look.


Let’s dive in.




Frequency Response

The frequency response is not what I’d call flat.
Even from the on-axis measurement alone, you can see that various forms of interference are affecting different parts of the response.


The -6dB low-frequency extension is around 59 Hz, followed by a steep roll-off at approximately 40 dB/oct...

• Nuyes
• aiyima s600
• Replies: 30
• Forum: Speaker Reviews, Measurements and Discussion

• 

 

[Heinrich]

One can also look at the review of the Aiyima speaker. Does anyone believe that the driver itself has that nasty 400 Hz resonance?

As before, you identify a peak in the harmonic distortion @400Hz with a resonance. We cannot discuss a topic like that, having that strange concoction of terms in bended meanings.

The peak could be a consequence of something moving other than the speaker cone. Since 2nd and 3rd harmonic show roughly the same amplitude measured in percentage, and additionlly there is no dip in the frequency response to explain it, we may assume there is something rattling. And such could be in the driver itself, or in the construction of the cabinet, or at the interface between cabinet and driver. In the end it's all speculation.

I strongly suggest to (a) discuss this in the DIY section, because there are people with experience and means to cross-check results directly, and (b) following that, to address the topic systematically in isolating effects. To ask for confirmation of an anecdotal observation, calling for ABX to give it the credit of science, isn't quite fair.

Sorry, can't participate further here.

(Btw, I wonder why you mention a 20y old post in another forum on that same (?) speaker, and nothing happend since then. I mean, further investigations, cross checks, in case wider usage of the technology and so forth.).