A lot of this discussion is theoretically OK but way off base in terms of the real materials in use.
Specifically, damping is like sound absorbers in that you just can't feasibly get enough of it to help much for a woofer (unless you have really too thin walls to start with).
Likewise, you can brace a large side wall of a box to the opposite side and cut the motion substantially. But that leaves smaller patches of wall at higher resonance... but they barely budge at all (let alone respond to woofer frequencies).
Today's hot tip: instead of expensive CLD panels, here's something that does similar wonders for the vibrating walls of dishwashers and elsewhere. It is heavy aluminum foil with thick adhesive rubber layer for mounting. Sold as 6-inch wide pipe wrap, stuff like that is widely available in Canada.
Auto supply stores also have various stuff. And my favourite, at least for treating thin metal panels like in a van, is a product with thick latex paint infused with lots of cork pieces.
Sometimes hard to find fibreglass boards. You can use plastic faced fibreglass ceiling tiles (not the cellulose kind) from Home Depot; you can leave on the plastic face since it has no detrimental effect inside the box.
B.
Specifically, damping is like sound absorbers in that you just can't feasibly get enough of it to help much for a woofer (unless you have really too thin walls to start with).
Likewise, you can brace a large side wall of a box to the opposite side and cut the motion substantially. But that leaves smaller patches of wall at higher resonance... but they barely budge at all (let alone respond to woofer frequencies).
Today's hot tip: instead of expensive CLD panels, here's something that does similar wonders for the vibrating walls of dishwashers and elsewhere. It is heavy aluminum foil with thick adhesive rubber layer for mounting. Sold as 6-inch wide pipe wrap, stuff like that is widely available in Canada.
Auto supply stores also have various stuff. And my favourite, at least for treating thin metal panels like in a van, is a product with thick latex paint infused with lots of cork pieces.
Sometimes hard to find fibreglass boards. You can use plastic faced fibreglass ceiling tiles (not the cellulose kind) from Home Depot; you can leave on the plastic face since it has no detrimental effect inside the box.
B.
Last edited:
Ben - that paste on stuff will do almost nothing to a wood panel. It is made for use on thin steel panels.
If you make the CLD yourself as part of the enclosure then its not too expensive. It is a lot more work but that's why you DIY.
If you make the CLD yourself as part of the enclosure then its not too expensive. It is a lot more work but that's why you DIY.
Right. It's all about that word I used, "feasible". I suppose you could make a box out of cardboard and then fill it with some kind of open-cell rigid foam.
Say, I should patent that.
"Feasible" depends on your purposes, of course, and you may opt for complex but light weight, simple but cheap, attractive but small, and so on. Seems to me the "sweet spot" tends to be MDF panels and properly applied bracing, and good judgment about going isothermal with foam. Unlike the idiotic "Iron Law", there are trade-offs on several parameters.
B.
Last edited:
It's all about shear. If I shear a material then the particles all rub against each other and maximum friction occurs. If I just compress the same material the particles just get closer together, no real friction at all. Hence all of the construction techniques are aimed at putting a damping compound in shear.
Your rubbing analogy is only true for discrete particulate matter, like loose grains of sand or the like. If the constrained layer is isotropic like a viscoelastic it behaves like a continuous substance. Both shear and compression can flex the material, and energy will be converted internally when molecules reconfigure to accommodate the stress. This generates heat and provide the loss mechanism, and damping effect.
If you know what you are doing a woofer cabinet can need no damping just like a subwoofer box.
If you brace two opposite sides together you will introduce a nasty low frequency mode that wasn't present before. This mode consists of the two panels moving together with the brace as a large added mass doing nothing positive but quite a lot negative. On its own the negatives are likely to outweigh the positives but there are one or two additions that can help tip the balance and symmetry should lead the mode to being only being weakly driven anyway.
As Earl mentions, not for cabinets. For CLD to work reasonably the material properties and the constraining layer need to be appropriate. There isn't much point to CLD otherwise. Just use simple extensional damping if that is all that is required.
But it would also be true if the binding medium allowed the particles to move like the loose sand does. If the medium is rigid then no this won't happen, but if it is very loose then it will.
This is correct, but ...
I brace between two opposing sides and as you say this will create a new mode with the brace as a mass. But it also stops several of the plate modes where the two side want to go in opposite directions from each other. From an acoustic radiation standpoint the back and forth modes are dipoles with a very poor radiation efficiency while the modes that where cancelled would be monopoles which have a very high radiation efficiency. So you are basically replacing a set of very efficient radiators with a new one of much lower efficiency.
But then one can also add a brace between another pair of surfaces and connect the two braces together, thus killing the newly created back and forth modes as well. Dampen this cross (+) brace and you have a very effective means of damping the box modes as well as lowering their radiation efficiencies.
Earl, the STC of a cone becomes extremely high due to back emf? I would think the STC is dependent on the cone material properties and significant sound transmission would occur even if the coil was glued in position. Surely this must be true, it's self evident. BTW you have a good point about masking but the cabinet wall radiation area is much much larger than the cones. It would be interesting if any of us could sit in the LS50 listening tests where they optimized cabinet properties with damping. We can only conjecture about what is audible or not.
Charlie, any small speaker in the cabinet will have omni radiation over a significant bandwidth of interest for this experiment so I still think its not a valid test to put a small speaker inside.
hollowboy, this side conversation is in response to charlie, not your original request
I spent several years (20 yrs ago) attempting to find a solution that provided significant dissipation of in-box-energy without having a significant effect on a vented alignment Q. I had no real luck and so continued to avoid venteds. As a side question, have any really good solutions been found that meets both aspects very very well?
That's not exactly what I am saying. There will, of course, be some sound leakage even if you glue the cone and that's probably the way the test should be done. But a free cone will move freely and in a real situation this movement would create a back EMF which gets absorbed and dissipated by the amplifier. So yes, if you glue the cone then it is a fair test, but not if the cone is free. Shorting the speaker may be representative as well.
I don't use ports either, so I haven't looked at the situation you describe.
Hollowboy, getting back to your original request (to ease my guilt from all the sidetracking), some other sources of test (and some sim) data you hopefully find useful.
Mark Sanfilipo published some results on audioholics in 2005 (I saved them as .mht) but they are now gone. I did find part of it reprinted here (eg Fig 6);
Lowering a Loudspeaker’s Mechanical Noise Floor
Measurement and simulation of lightweight cabinet (the references section alone will be a goldmine for you):
http://orbit.dtu.dk/files/132613635...ion_of_Lightweight_Loudspeaker_Enclosures.pdf
From John Atkinson:
"Stanley Lipshitz and his colleagues noted that the accelerometer measurements of a loudspeaker cabinet's walls varied tremendously according to how the speaker was supported while the measurement was being performed. " JA then took measures of cabinet vibration with differing mounts to the stand:
The Sound of Surprise (the loudspeaker/stand interface) Page 3 | Stereophile.com
Earl, regarding audibility, from JA:
"In a paper presented to the 90th AES Convention (Paris, 1991), "An Investigation of Sound Radiation by Loudspeaker Cabinets," Stanley Lipshitz and his coworkers concluded from calculations of the total radiated energy that the sound of the cabinets of the loudspeakers with which they were experimenting would be audible or close to the borderline of audibility. "
From our own DIYaudio (can't vouch for veracity but author claims these are measures):
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-70.html#post1202358
Measurements from Audio Express:
http://forum.vegalab.ru/attachment.php?attachmentid=132494
More measurements, however with accelerometer:
http://www.warkwyn.com/wp-content/uploads/2015/01/enclosure-scanning-ALMA-2015-PPT.pdf
I suspect someone probably already posted these measures:
http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf
Finally, while not exactly what you're looking for, I also found good info on the topic from contractors that specialized in vibration analysis and reduction:
https://www.newport.com/t/about-optical-table-performance-specifications
One golden nugget from this company was (going back 20 yrs) :
"Dynamic Deflection Coeffient" to more
accurately model the actual response of a table top or platform to real
world vibrational inputs and it shows that a high stiffness to weight ratio
increases resonant frequency, as you've said, and results in a lower
dynamic deflection coefficient or less relative motion in the platform. In
Newport's experiments they define " a common misperception that a higher
core density implies increased dynamic rigidity. Eactly the opposite is
true. All other things being equal, a lighter table will have better
dynamic rigidity. In fact, lower table weight is the major reason why
honeycomb core optical tables outperform granite slabs in minimizing
damaging resonances"
So much of this points to damp, damp, damp. Bracing raises Q and Sean Olive's audibility tests showed that high q = lowered audibility, but I never fully trusted that those audibility tests apply to all situations, and that perhaps the "right" test signal will betray audibility of a high q resonance. That's what this high-fi stuff is all about, that last couple percent.
Mark Sanfilipo published some results on audioholics in 2005 (I saved them as .mht) but they are now gone. I did find part of it reprinted here (eg Fig 6);
Lowering a Loudspeaker’s Mechanical Noise Floor
Measurement and simulation of lightweight cabinet (the references section alone will be a goldmine for you):
http://orbit.dtu.dk/files/132613635...ion_of_Lightweight_Loudspeaker_Enclosures.pdf
From John Atkinson:
"Stanley Lipshitz and his colleagues noted that the accelerometer measurements of a loudspeaker cabinet's walls varied tremendously according to how the speaker was supported while the measurement was being performed. " JA then took measures of cabinet vibration with differing mounts to the stand:
The Sound of Surprise (the loudspeaker/stand interface) Page 3 | Stereophile.com
Earl, regarding audibility, from JA:
"In a paper presented to the 90th AES Convention (Paris, 1991), "An Investigation of Sound Radiation by Loudspeaker Cabinets," Stanley Lipshitz and his coworkers concluded from calculations of the total radiated energy that the sound of the cabinets of the loudspeakers with which they were experimenting would be audible or close to the borderline of audibility. "
From our own DIYaudio (can't vouch for veracity but author claims these are measures):
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-70.html#post1202358
Measurements from Audio Express:
http://forum.vegalab.ru/attachment.php?attachmentid=132494
More measurements, however with accelerometer:
http://www.warkwyn.com/wp-content/uploads/2015/01/enclosure-scanning-ALMA-2015-PPT.pdf
I suspect someone probably already posted these measures:
http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf
Finally, while not exactly what you're looking for, I also found good info on the topic from contractors that specialized in vibration analysis and reduction:
https://www.newport.com/t/about-optical-table-performance-specifications
One golden nugget from this company was (going back 20 yrs) :
"Dynamic Deflection Coeffient" to more
accurately model the actual response of a table top or platform to real
world vibrational inputs and it shows that a high stiffness to weight ratio
increases resonant frequency, as you've said, and results in a lower
dynamic deflection coefficient or less relative motion in the platform. In
Newport's experiments they define " a common misperception that a higher
core density implies increased dynamic rigidity. Eactly the opposite is
true. All other things being equal, a lighter table will have better
dynamic rigidity. In fact, lower table weight is the major reason why
honeycomb core optical tables outperform granite slabs in minimizing
damaging resonances"
So much of this points to damp, damp, damp. Bracing raises Q and Sean Olive's audibility tests showed that high q = lowered audibility, but I never fully trusted that those audibility tests apply to all situations, and that perhaps the "right" test signal will betray audibility of a high q resonance. That's what this high-fi stuff is all about, that last couple percent.
Last edited:
The high Q/low Q Harman work was applied to the signal. It is correct. Cabinet resonances are low level distortion. High Q means loud and close to the signal in level and continuing to resonate after the music stops. Low Q means tens of dB below the signal and stopping when the music stops. There is not much doubt about which is preferable.
As I have said before, I tried seriously to measure the amount of sound from my cabinets alone. I was not able to do it. But then my cabinets were always Damped, Damped, Damped - I knew that was the answer from my noise control work at Ford. Maybe I wasn't able to measure anything because there wasn't much to measure.
"Measuring sound output from speaker cabinet walls"
Does this also include the sound that is travelling through the walls?
Does this also include the sound that is travelling through the walls?
- Status
- Not open for further replies.