OP
- Thread Starter
- #141
Just build the cabinets and do it your way. Ok?
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Speaker enclosure vibrations - a few measurements with accelerometer
- Thread starter Thomas_A
- Start date
I should *think* a waterfall measurement would illustrate late arriving sound. Although as pointed out already, the cone is only one escape route for the rear wave.
Sure, I advertise to reduce the internal air column resonances for good measure (pun intended). The remaining "resonances" seen on cabinet panels will become negligible.
- what is the relative phase of "panel resonances" on opposing walls?
- are there "higher modes" of "panel resonances" indentified by phase / amplitude?
- what is the spectrum of "panel resonances", basically is it harmonic (1,2,3 ..) or different?
- how do "panel resonances" react to stiffening without changing the mass, or vice versa changing the mass without changing the stiffness?
- how do "panel resonances" react to pointwise counter bracing opposing walls
- how do "panel resonances" react to dampening the internal air column resonances effectively?
Wich model of "panel resonances" aligns with the findings from investigating the questions posed above? Easy, right?
TU so much
- Joined
- Jan 15, 2020
- Messages
- 6,901
- Likes
- 16,910
As long as you don't present any comparative measurements this is just a claim, especially about the magnitudes getting negligible, as even if the airborne transmission will be reduced, the structureborn transmission from the drivers won't.
Just to muddy the waters, I bet a hard cone driver doesn't pass nearly as much sound as a soft cone, except at the cone resonances.
OP
- Thread Starter
- #146
These are small bass reflex boxes (9liter internal volume) stuffed with fiberglass to about 40% volume. Not useless polyfill.
You can do it.
I only issued an advertising intermission; please go ahead with You constr/lay/damp/ for the better feel of it. You deserve CLD, nothing about that from my side.
Nope, I don't feel using up the time once I relate the effort to the expectable outcome--here. Or what do You think would be the result discussing the topic with people who really and deeply love discussing audio as an exercise in idealistic purity? I can't do it, so You win.
Anyway, one question added to my post https://www.audiosciencereview.com/...rements-with-accelerometer.36254/post-1365890
- what is the difference between a "knocking test" resonance spectrum and the spectrum when the enclosure is driven by the speaker?
.. You didn't even bother to read the questions, right? Not to mention answers ... there You have it ...
OP
- Thread Starter
- #148
I have read them. Still what do you say about Purifi and the 380 Hz resonance of driver basker and magnet? Are they wrong? It is an internal standing wave of the box or air column?
Reiterated, that is not how it actually works. I acknowledge, that You follow the common habitual standards in amateur audio. I do not criticise You personally. It is only so, that on a board that has the term "science" in its name, there should be room to point to a more systematic approach.
Answers to the questions stated in post https://www.audiosciencereview.com/...rements-with-accelerometer.36254/post-1365890 would reveal very much more than:
- speculating about one reason of many for some singled-out effect observed by somebody else using unknown instruments with an object known only by its name with unknown origin without any comparison to another specimen of the same type ... c'mon!
Reiterated, a measurement needs an idea what the object of measurement is about. That's called a "model", some would say a "theory". I already gave You the example of measuring a distance. That needs an idea of space. Is space a continuum? No! Is space static? No! What about a ruler then?
More related to the topics discussed here: the Klippel is less about the funny machinery, it is about a model of the speaker and the room it sits in during the measurement. Without that model the data is useless rubbish. The Klippel IS the mathematics derived from that model.
You kind of playing around with the accelerometer didn't relate the data to a model. Hence the data is useless. Even Your conclusion is kind of debatable at best. Why should less amplitude of panel activity as it was kind of detected by the accelerometer be of any relevant benefit? Btw, is the accelerometer, to put it easy, "linear"?!
Sorry for being that verbose, but I mean it and don't want to bother You. What drives the panel movement?! See questions above. The answer will for sure impact Your view on the, if You will, problem and lead to more practical solutions, if You still mind it.
Only in German, sorry:
http://www.waveguide-audio.de
Über Lausprecherbau mit Waveguides, also mit Schallführungen für Hochtöner
www.waveguide-audio.de
That's systematic, isn't it? Result: once the box is stiffened by braces and damped with wool, the (in)famous constr/lay/ doesn't do much more ...
Last edited:
OP
- Thread Starter
- #150
Again. I have been quite clear on the points made throughout the whole thread.Reiterated, that is not how it actually works. I acknowledge, that You follow the common habitual standards in amateur audio. I do not criticise You personally. It is only so, that on a board that has the term "science" in its name, there should be room to point to a more systematic approach.
Answers to the questions stated in post https://www.audiosciencereview.com/...rements-with-accelerometer.36254/post-1365890 would reveal very much more than:
- speculating about one reason of many for some singled-out effect observed by somebody else using unknown instruments with an object known only by its name with unknown origin without any comparison to another specimen of the same type ... c'mon!
Reiterated, a measurement needs an idea what the object of measurement is about. That's called a "model", some would say a "theory". I already gave You the example of measuring a distance. That needs an idea of space. Is space a continuum? No! Is space static? No! What about a ruler then?
More related to the topics discussed here: the Klippel is less about the funny machinery, it is about a model of the speaker and the room it sits in during the measurement. Without that model the data is useless rubbish. The Klippel IS the mathematics derived from that model.
You kind of playing around with the accelerometer didn't relate the data to a model. Hence the data is useless. Even Your conclusion is kind of debatable at best. Why should less amplitude of panel activity as it was kind of detected by the accelerometer be of any relevant benefit? Btw, is the accelerometer, to put it easy, "linear"?!
Sorry for being that verbose, but I mean it and don't want to bother You. What drives the panel movement?! See questions above. The answer will for sure impact Your view on the, if You will, problem and lead to more practical solutions, if You still mind it.
Only in German, sorry:
http://www.waveguide-audio.de
Über Lausprecherbau mit Waveguides, also mit Schallführungen für Hochtöner
That's systematic, isn't it? Result: once the box is stiffened by braces and damped with wool, the (in)famous constr/lay/ doesn't do much more ...
- The cabinet is a small vented cabinet with 7 sides, non-braced. It is filled with glass fiber (not polyfill!).
- A glued MDF cabinet should behave linearly at least in the xy dimension (within reasonable forces). Driving force is the driver motion and its inherent non-linearity!
- You also add a metal frame (inserted with wooden screws!), a basket and magnet with its own resonant behaviour.
- What I found 17 years ago (on three built cabinets), was an audible "hardness"/distortion around 500-600 Hz. And a significant vibration of the cabinet at those frequencies.
- Sfuffing more fibre-glass or Rock-Wool helped none.
- A constrained layer cabinet was made, and I did not hear those distortions any more. It vibrated significantly less.
- 17 years later I made some measurements. Are they anyhow inconsistent with the findings 17 years earlier? Yes or no?
As already mentioned in the first post, the "problem" may or may not have been fixed with e.g. bracing. Constrained layer was another solution that fixed it.
There are secondary effects that can be discussed, as e.g. lowering an already "inaudible" distortion to even lower levels, etc. But then you are going into the "audiophool" territory.
OP
- Thread Starter
- #152
Brace! Brace! Brace!
All time happy landings!
Yes, bracing may or may not be a solution, as mentioned.
It depends on many things, e.g. the size and shape of the cabinet. Note that these are very small 9 liter enclosures, 5 inch woofers, with asymmetric shape. Whether bracing would help was not that obvious with such asymmetric shape and small size, but it might. These were the originals, 19 mm MDF:These were generation two (DIY, active), still only 19 mm MDF (I built two of them, with slight change in tweeter position). But as mentioned I do not have those cabinets left.
And generation three (passive, with 9+9 mm MDF).
- Joined
- Jul 23, 2019
- Messages
- 480
- Likes
- 505
Thank you for sharing that link. Google Translate seems to work well on that website, making it quite readable. It contains a lot of measurement data of the DUT (device under test), in a wide variety of configurations.Only in German, sorry:
http://www.waveguide-audio.de
Über Lausprecherbau mit Waveguides, also mit Schallführungen für Hochtöner
That's systematic, isn't it? Result: once the box is stiffened by braces and damped with wool, the (in)famous constr/lay/ doesn't do much more ...
Although filling a sealed enclosure with sound absorption material is a viable proposition, it's not really possible to do the same with a vented enclosure. Of course, that could point to avoiding using vented boxes.
For the vented box enclosure, the constrained layer damping (CLD) approach is still likely to be quite helpful in dampening the panel vibrations. I don't know the loss factor of the bitumen pad material, and this would be worth knowing. The constraining layer in the DUT seems to be quite stiff and might be expected to be doing a reasonable job. Keep in mind that the thickness of the bitumen damping layer is quite thick, more in line with what would be used with an unconstrained damping layer. A thinner layer of bitumen damping would likely provide higher damping when used in a CLD configuration, due to the increased shear stresses that serve to produce the damping effect. It is apparent that the CLD treatment needs to be properly dimensioned to give its best in any loudspeaker application.
Another viscoelastic material that could be applicable for use in loudspeaker panel damping applications is this one: DECIDAMP® DC06.
OP
- Thread Starter
- #154
This is the data for the glue DG-A2 that I used (only in Swedish, but you might understand some graphs):
OP
- Thread Starter
- #155
Thank you for sharing that link. Google Translate seems to work well on that website, making it quite readable. It contains a lot of measurement data of the DUT (device under test), in a wide variety of configurations.
Although filling a sealed enclosure with sound absorption material is a viable proposition, it's not really possible to do the same with a vented enclosure. Of course, that could point to avoiding using vented boxes.
For the vented box enclosure, the constrained layer damping (CLD) approach is still likely to be quite helpful in dampening the panel vibrations. I don't know the loss factor of the bitumen pad material, and this would be worth knowing. The constraining layer in the DUT seems to be quite stiff and might be expected to be doing a reasonable job. Keep in mind that the thickness of the bitumen damping layer is quite thick, more in line with what would be used with an unconstrained damping layer. A thinner layer of bitumen damping would likely provide higher damping when used in a CLD configuration, due to the increased shear stresses that serve to produce the damping effect. It is apparent that the CLD treatment needs to be properly dimensioned to give its best in any loudspeaker application.
Another viscoelastic material that could be applicable for use in loudspeaker panel damping applications is this one: DECIDAMP® DC06.
Agreed on several points including;
- vented enclosures are never completely filled with damping material, there is a limit to what you can do
- the efficiency of the "bitumen CLD" in the example appears not to be documented with respect to loss factor; As with the DG-A2 glue the loss factor increases from 0.01-0.02 for a single plywood sheet to 0.2-0.3 for a double plywood sheet. That is significant.
- There is no example with a true CLD approach as comparison to the other examples
- braces work to dampen some peaks and pushing some of them up somewhat/a bit in frequency, nothing strange with that (as already mentioned)
-and one more thing; a shoe-box cabinet may not show the same result as a box with irregular shape (as my example)
Last edited:
Yepp, I discussed the evaluation of the D/U/T with the author in our common native language. My own independent examination of another, way bigger enclosure suggested the same conclusion: it's the air that, as he put it, hammers at the panels making them move, and so peaks in movement correlate quite reasonable with air resonances, but not with original panel resonances. The latter can be excited by that (in)famous knocking against the enclosure--they differ from resonances seen when driven by the active speaker.
I don't want to waste too much of Your time. Bracing plus air damping yields the first 25dB or so of attenuation of resonances, additional CLD adds some mere 5dB to that. It is worth the effort? Me thinks not.
If the air resonances 'hammer' at the panel, and so imprint their vicious behaviour onto the poor, innocent panel, what to do? Punish the panel with tar and feathers? The author reminds people that the gluing for CLD must not be taken for granted, it may get loose without notice ... but there is relief for the panel.
Solution is, to isolate the internal air movement from the outside panel. There was a 'patent' in German that describes what to do and how. Much of Your DIY CLD is most likely something along those lines, not actual CLD. Only that the 'patented' solution doesn't eat up much of the available internal volume of the enclosure. I was the issuer ;-) Tested the solution proved to be very effective, cheap, of small footprint, easy to realise. Starting point is to see the problem from the correct (if You will, scientific) perspective. It is the air, not the panel that resonates primarily.
Are we, considering me, through with this?
Add.: to not stuff ported enclosures properly makes them perform sub-par. To allow for stuffing, without losing too much in bass (which could be equalized together with in-room performance anyway) just make the enclosure bigger. That easy, mostly.
Last edited:
OP
- Thread Starter
- #157
Just to comment; bracing is not a ”lossy” method as such. Damping material and CLD is. The example you linked to did not compare the box to a serious CLD solution. And not either to 50% fiilling material which is better linked to vented boxes.
So better have a real comparison if it should be systematic. Then you can compare numbers.
And again You want to dampen resonances of the air with altering the panel construction. It's up to You, basically.
OP
- Thread Starter
- #159
So how many dB did my CLD construction reduce? Quite substantial effect, I would say.
DanielT
Major Contributor
Building, constructing with very thick material walls may be unnecessary, but when will it become so thin/lightweight that vibrations, resonances become audible? Apparently here. Erin can't stand listening to them.Erin lifts a speaker with one hand and we're still talking about speakers with 12" woofers. They weigh next to nothing.
Also as expected: abysmal measurement results:
Also as expected: abysmal measurement results:
Similar threads
