This is definitely so. I even look specifically for your posts (and those from other known guru’s) , as I already learned a lot from it. Still trying to completely understand your book on audio transducers. Currently following the MIT_opencourseware calculus course in the hope this will fill in some of the blanks.
I'm sorry for quoting myself, but I'd really like to know if there is some empirical evidence to support this idea. Has anyone ever done experiments with coupling and decoupling a driver to an enclosure. I do have some anecdotal evidence, but imho that's the worst kind of evidence to put forward.
Happened to see this while browsing around. For your reference:
Onken 360
Sort of minimum bracing (don't know if optimal?). Triangle, uneven support points, simple and elegant... Seems "well equipped"😀 Very nice ply, and beautifully built.
Wondering if solid oak could be even better for the bracing? And the rear 'cover' seems to be a weak spot...
Onken 360
Sort of minimum bracing (don't know if optimal?). Triangle, uneven support points, simple and elegant... Seems "well equipped"😀 Very nice ply, and beautifully built.
Wondering if solid oak could be even better for the bracing? And the rear 'cover' seems to be a weak spot...
It is not an easy question. There are many factors that need to be taken into account. Size and weight of the driver, baffle size, thickness and material, number of driver fixing points etc etc.
In a thin, badly braced design, a great deal of what you hear may come from the enclosure walls.
Think about the vibration from a 300mm driver vs a 100mm driver. I could answer your question for a specific build but your general question cannot be answered.
I have researched mechanical coupling, but only in midbass units up to 200mm. As a result I do prefer to use rubber gaskets but as with most things (including bracing) it is easy to go too far.
Hope this helps
Terry
You are right, there is no one answer tot the question. In my own experience the sound in the box can be dampened relatively easily, but the mechanical energy is a lot more difficult to get rid of. When you think of it, there is a lot more energy in movement of the membrane of a driver than there is in the sound it radiates.
The first time I noticed this was some five years ago when I built my first speaker. It was a 2-way Scanspeak Kit, which I built as sturdy as I could with very thick walls and lots of bracing. Lo and behold, at certain frequencies the box rang like a bell! The only way to stop it from resonating was to simply lay the box on its back and remove the crews that kept the midwoofer in place. Then the driver would come up a little as the rubber gasket decompressed. This way the the drivers' basket was better decoupled from the enclosure and even at higher levels the box could hardly be felt to vibrate anymore! Quite an epiphany.
The first time I noticed this was some five years ago when I built my first speaker. It was a 2-way Scanspeak Kit, which I built as sturdy as I could with very thick walls and lots of bracing. Lo and behold, at certain frequencies the box rang like a bell! The only way to stop it from resonating was to simply lay the box on its back and remove the crews that kept the midwoofer in place. Then the driver would come up a little as the rubber gasket decompressed. This way the the drivers' basket was better decoupled from the enclosure and even at higher levels the box could hardly be felt to vibrate anymore! Quite an epiphany.
And there is a lot more mass to an enclosure than there is the membrane of the driver, so this offsets the energy coupling a great deal.
Siegfried Linkwitz has or had a link to this article on his website.
Loudspeaker Driver De-Coupling,
A Preliminary report
A.Jones, Pioneer Electronics Tech.
Loudspeaker Driver De-Coupling,
A Preliminary report
A.Jones, Pioneer Electronics Tech.
But is the enclosure sufficiently rigid to consider it as just a mass? The entire enclosure probably won't rock back and forth due to its' greater mass, that's true. But don't you agree that the mechanical vibration of the driver's basket has a greater chance of causing the box and its panels to resonate than the sound inside the enclosure?
Thanks! www.linkwitzlab.com/Driver Decoupling.doc
EDIT: It's a nice article. Not definitive by any means, but it indicates my reasoning may not be far off.
Last edited:
It would be interesting to see how effective running two bass units of a 3 way, on opposite faces with a bar between the magnets, is at preventing the baskets driving the cabinet
Just about everything has it's trade-offs, and it stands to reason that the price to pay for a decoupled driver is the loss of dynamics (a portion of the driver's cone excursion will be absorbed/counteracted by the movement of the basket on a decoupled driver), and that a better solution would be to fix the enclosure via damping, mass, or bracing if it's resonating. To my way of thinking damping and mass are more effective solutions than bracing.
McCormack, I doubt if that effect would be very serious. Look at the article Agent327 mentioned. There is hardly any influence on the frequency response.
As with all resonant systems the enclosure has both mass and stiffness (and damping) characteristics and each dominates in a different frequency region.
The actual problem is grossly oversimplified here (I've said that before, but maybe it needs to be said again). At LF the air coupling is likely to be the greater effect, and at HF the structural is likely to be the greater effect. But than that all ignores the radiation effect, which again has its own frequency dependence. All in all its far too complex to "guess" at what the answer is and likely very specific to the exact design. Hence, you have to do tests. I've done those, and from what I can tell the air coupling at LF (say < 300 Hz) is about the only place that the enclosure shows signs of radiating. But a reasonable amount of bracing and damping and those are elliminated droping the whole enclosure thing to below the level of detectability.
Maybe this has been covered since I did not read all 13 pages of posts, but I tend to agree with the camp on the "window" style bracing. For me, the idea is not to stiffen the cabinet as much as transfer energy to an opposing panel. Pressure wants to push/pull cabinet walls out and in with driver movement. having a brace that ties one wall to the opposite wall is good, but to be able to tie all 4 cabinet sides together will result in better cancellation of any modes. Just loading a cabinet up with braces does not mean you will kill all modes...it just means you will move them to another range. Many equally space their braces as well...which does not work as well as keeping bracing non uniform.
Potentially..
But even if correct for all individuals in all circumstances, the perspective is "stuck" on the cabinets as a form of additional sound source(s).
Directly related, though of far greater importance, is the effect on VC motion. Generalized claims of "good enough" for bracing, damping, etc.., might not be with regard to minimizing the effect on VC motion. 😉
Very effective as Dave said concerning the opposing forces in effect canceling the lateral movement of the cabinet in a push-push alignment but I see a bit of a hitch there. It would cause twice the resonance in the panels and as I see it that's what the thread starter was asking, not about keeping the speaker in place.
Why?
Energy available to couple into the box structure is dramatically reduced (assumming good direct coupling -- in situations where the box itself is used for coupling (ie the new paradign woofers), i expect not nearly as much effectiveness). Air space energy is the same (as a different arrangement of 2 woofers)
The above represents an ideal case where the woofers have been designed with coupling in mind (PEARL PR-2 woofers)
dave
Just one consideration of many.
Some of the cabinet considerations that I believe to be roughly as important as bracing:
1. Internal standing waves. Are your internal dimensions non-integrally related? Or are the walls non-parallel? A sloped back can really help. That's why pyramid shaped enclosures are used by some...
2. Woofer and port height above the floor. Evaluate at your normal listening height and distance. You can partially minimize floor bounce cancellation effects.
3. Internal reflections; suitable stuffing (I like Acousta-stuff, but wool is better),
and a thick, soft, high wool content felt lining the cabinet. Sloping back helps here too.
4. On the midrange, or a highish crossed over woofer, is the driver cutout beveled, so the interior side of the opening is larger? For mids, is the cutout lined with thin felt? At it's worst, you can get an odd "tunnel" effect with tight vertical cutout walls.
5. Does the port have rounded, reduced-turbulence ends? My tests showed this to be surprisingly effective.
6. If the driver has a flimsy basket, damping the ribs with a non-outgassing, non-drying caulking compound sometimes helps. For a stamped basket, build up the flat rib to a rounded shape; for mids the reflection and diffraction around the rib may be audible.
7. Diffraction reduction! A felted front panel, 4" minimum variable radius rounded edges, help a lot. For a quick test, drape a nice open weave wool blanket over a conventional cabinet, rounding the cabinet edges and covering the sides, but trying not to get very near the driver. With considerable testing, I was surprised that even with rounded front corners, the rear corners also had lesser contribution.
And to reiterate some of the things mentioned in earlier posts; make sure your braces do not divide the cabinet panels into equal sized (worst) or integrally related sub panels.
Shelf braces work awfully well. Thiels used them to great effect for many years. Instead of right angled cutouts, round the cutout of the crosspiece where it attaches to the panel brace; in any event make the wall-contacting pieces with increasing thickness from end corner to the crosspiece, putting more material where it does the most good. (Sorry, didn't move my CAD program to this computer. )
The decoupling mounting I'm of two minds about. Sonus Faber paid a lot of attention to this, with rubber gaskets, and I believe neoprene well nuts. Vance Dickason also recommends. I tried this, on a midrange, and at best got no improvement.
For woofers, it seems a moot point. There is usually very little front panel to the sides of the woofer; the woofer is mounted very close to the enclosure sides, and perhaps the bottom and top. Very stiff arrangement, and any decent woofer won't be flexing enough to separately excite the middle of the panel. Basically, the woofer is trying to move the entire cabinet. If living arrangements allow, spiked feet will help.
I'd suggest not getting too hung up on one detail. It's a time consuming process if you're rolling your own design, and the most time (IMHO) should be spent on driver selection, crossover parts selection, and fine tuning.
Some of the cabinet considerations that I believe to be roughly as important as bracing:
1. Internal standing waves. Are your internal dimensions non-integrally related? Or are the walls non-parallel? A sloped back can really help. That's why pyramid shaped enclosures are used by some...
2. Woofer and port height above the floor. Evaluate at your normal listening height and distance. You can partially minimize floor bounce cancellation effects.
3. Internal reflections; suitable stuffing (I like Acousta-stuff, but wool is better),
and a thick, soft, high wool content felt lining the cabinet. Sloping back helps here too.
4. On the midrange, or a highish crossed over woofer, is the driver cutout beveled, so the interior side of the opening is larger? For mids, is the cutout lined with thin felt? At it's worst, you can get an odd "tunnel" effect with tight vertical cutout walls.
5. Does the port have rounded, reduced-turbulence ends? My tests showed this to be surprisingly effective.
6. If the driver has a flimsy basket, damping the ribs with a non-outgassing, non-drying caulking compound sometimes helps. For a stamped basket, build up the flat rib to a rounded shape; for mids the reflection and diffraction around the rib may be audible.
7. Diffraction reduction! A felted front panel, 4" minimum variable radius rounded edges, help a lot. For a quick test, drape a nice open weave wool blanket over a conventional cabinet, rounding the cabinet edges and covering the sides, but trying not to get very near the driver. With considerable testing, I was surprised that even with rounded front corners, the rear corners also had lesser contribution.
And to reiterate some of the things mentioned in earlier posts; make sure your braces do not divide the cabinet panels into equal sized (worst) or integrally related sub panels.
Shelf braces work awfully well. Thiels used them to great effect for many years. Instead of right angled cutouts, round the cutout of the crosspiece where it attaches to the panel brace; in any event make the wall-contacting pieces with increasing thickness from end corner to the crosspiece, putting more material where it does the most good. (Sorry, didn't move my CAD program to this computer. )
The decoupling mounting I'm of two minds about. Sonus Faber paid a lot of attention to this, with rubber gaskets, and I believe neoprene well nuts. Vance Dickason also recommends. I tried this, on a midrange, and at best got no improvement.
For woofers, it seems a moot point. There is usually very little front panel to the sides of the woofer; the woofer is mounted very close to the enclosure sides, and perhaps the bottom and top. Very stiff arrangement, and any decent woofer won't be flexing enough to separately excite the middle of the panel. Basically, the woofer is trying to move the entire cabinet. If living arrangements allow, spiked feet will help.
I'd suggest not getting too hung up on one detail. It's a time consuming process if you're rolling your own design, and the most time (IMHO) should be spent on driver selection, crossover parts selection, and fine tuning.
- Status
- Not open for further replies.