I feel that way about almost every experiment I read, even very well regarded peer-reviewed scientific papers (regardless of topic).
But yeah, LOTs of "holes" - especially the notion that the cabinets contribution isn't audible (..and wasn't even checked with a microphone).
Still, I did like the contribution of the Accelermoter - it really shows where and how the resonances were building in that portion of the panel.
Yeah, bracing can reduce the vibration magnitude (and typically "shift" it in freq. somewhat) but it can also increase the time and extend the bandwidth of those vibrations (by generating a more rigid structure).
The accelerometer was also interesting with respect to the upper freq. cabinet interior modes in relation to vibrating the panels, not just a matter of sound leakage but interior pressures vibrating the panel. Plus of course you aren't picking-up driver output (no matter how well isolated) so that's another benefit. Of course problematically you aren't actually determining the spl contribution either, so it's back to the mic and the best driver and other panel isolation you can achieve for some sense of subjective effect.
Basically the accelerometer allows you to better define what some of the problems are with the panel (and construction) itself. In fact, that (notional) portion of the video was enough for me to purchase the same accelerometer (which should be arriving tomorrow).
I just watched it again, did you notice that he's using a gasket to connect the panel to the rest of the box? ..AND also those clamps. That will absolutely effect the results when compared to no-gasket and glue as with most box construction.
While pricey, I believe the "gold standard" (at least for Windows/Mac) is Camtasia (which many have said that it saves so much time that it "pays for itself").
GamefromScratch also has a video on its updated version (2022).
There's some high speed cameras that be used these days. Anyone experimented with them?
1-665 fps:
https://www.parts-express.com/pedocs/manuals/390-814--dayton-audio-as-1-instruction-manual.pdf
Even my phone has Slow-Mo mode that records 1080p @240f ps
1-665 fps:
https://www.parts-express.com/pedocs/manuals/390-814--dayton-audio-as-1-instruction-manual.pdf
Even my phone has Slow-Mo mode that records 1080p @240f ps
Last edited:
If I understood the thread, the CLD is made with thicker 3/4 ply outside on which the driver recess is routed then the driver screwed.
The inner less thick MDF or HDF is glued with a viscoelastic glue (here in France Sica 11 glue seems to be advised as easy sourcable).
Am I correct the drivers screws shouldn't touch the inner layer that will be in contact with the rest of the cabinet ?
What glue for the inner panel to the cabinet edge sides please? Wood glue or also the elastic one?
I am making experiment with removable front (and back) panel BBC style (sides, upper + bottom pannels glued on a 1" wood frame). I hardly know if the wood screws are contributing to the cabinet vibrations as there is an evea rubber between the front panel and the frame or if it is the 1/2 5 layers smooth wood ply that dance ?
I also will experiment on a standalone inner layer but elastic glue for the outside cld pannels on it. Pannels because on per driver to increase isolation of drivers with each others. 1 mm elastic seal between outside pannels with same glue than the cld's.
The inner less thick MDF or HDF is glued with a viscoelastic glue (here in France Sica 11 glue seems to be advised as easy sourcable).
Am I correct the drivers screws shouldn't touch the inner layer that will be in contact with the rest of the cabinet ?
What glue for the inner panel to the cabinet edge sides please? Wood glue or also the elastic one?
I am making experiment with removable front (and back) panel BBC style (sides, upper + bottom pannels glued on a 1" wood frame). I hardly know if the wood screws are contributing to the cabinet vibrations as there is an evea rubber between the front panel and the frame or if it is the 1/2 5 layers smooth wood ply that dance ?
I also will experiment on a standalone inner layer but elastic glue for the outside cld pannels on it. Pannels because on per driver to increase isolation of drivers with each others. 1 mm elastic seal between outside pannels with same glue than the cld's.
Last edited:
My first CLD examples were 1/4" MDF joined to 1/4" MDF with the Weicon, so only 1/2" overall. The last two (that you are probably referring to) were 1/2" joined to 1/4". Weicon was used to join all the panels together, no woodglue.
About the screws not touching an inner layer, I just don't know. Part of me says that makes sense, the other part says that's overkill. I do know the screws attaching the driver carried a surprising amount of energy, and when they were isolated with o-rings (instead of only relying on the sorbothane gasket used on the driver) it dropped panel resonance 3dB.
About the screws not touching an inner layer, I just don't know. Part of me says that makes sense, the other part says that's overkill. I do know the screws attaching the driver carried a surprising amount of energy, and when they were isolated with o-rings (instead of only relying on the sorbothane gasket used on the driver) it dropped panel resonance 3dB.
Rubber well nuts, like for motorcycle fairings, might be a pretty simple way to achieve driver/baffle isolation.
Like these:https://www.amazon.com/Premium-Upgr.../B00RZSIFCU/ref=psdc_7298814011_t1_B01N3U6MLD
I have some rubber well nuts, but never used them for a couple reasons. One, they require a very large diameter hole, and there usually isn't that much wood to do that. Second, while I suppose the expansion of the rubber as you tighten the bolt might provide enough force to keep the driver in place, that is not how they are meant to be used. They are meant to be used with thin sheet metal (1/8" thick at most IIRC), where the expanding rubber is BEHIND the sheet metal, so it's really a rubber clamp. That can't happen with wood.
Well nuts can be used in blind applications no problem. As for diameter concerns, there are a variety of sizes available. A 6-32 has a 1/4" body diameter ... that doesn't seem excessively large.
https://www.boltdepot.com/Threaded_...th_brass_insert_Coarse_(standard)_thread.aspx
If a typical woofer has about 1/2" driver flange, and the same rebate in the wood, that would only leave 1/8" of wood on the inside. With MDF that seems pretty easy to blow out, especially when the fastener expands as these. I suppose I could test it, but I'm pretty short on time. If you do, report back.
I got the time to read a bit more deeply into CLD yesterday.
From what I can tell speakers don't really "fit" the concept of CLD as specifically a panel-shear effect. (..though notably a LOT of the papers could have achieved other effects rather than panel-shearing, even though they proceeded in that context.)
Specifically CLD as a panel-shear effect is *typically studied with the panels to be constrained that are very thin (typically sheet metal), where both the panels have significant flexing (and the panel to be constrained is generally more flexible than the constraining layer panel though more than a few were trying the constraining layer that was MORE compliant - like thick aluminum foil???).
*typically: there are other less rigorous "white papers" out there that attempt to further the concept of CLD with much thicker panels, particularly in the building industry (and not a few of which are presenting themselves as "acoustic experts").
Papers (built on many prior papers) did note that the thickness of the visco layer was a variable to overall success (usually very thin but sometimes thicker depending on the other layers and the compliance mattered a lot) but ironically did NOT differentiate that success with specific panel-shearing vs some other effect. In other words, it could (and perhaps likely is) some other additional effect like transmission loss from *boundary shear moving from the layer to be constrained into the visco layer and then into the constraining layer.
*this is more of a "fluid" shear.
Notably some advocated full constraining layer dimensions to the layer to be constrained, where others had the constraining layer dimensionally smaller than the layer to be constrained.
Perhaps more significantly, there were significant variations is testing. While almost all required a flexible layer (that needs constraining), most did not go into detail with the fact that connected panels were fundamentally VASTLY more stiff near those connections. (..as in: the boxes edges/corners being far more stiff than the center of any panel.)
In the context of speakers then:
We have seen more dramatic effects from simply applying stuffing on the interior to lower pressure modes in the air in the box, that then resulted in lower excitation of the panel to be constrained. Beyond that the effects of attempted CLD have been modest at best.
Suggestions then are:
Reduce the thickness of the panel to be constrained. Perhaps 1/8".
Reduce any rigid coupling between adjoining panels (though Brandon has already done this to some extent with his more recent layering/construction), though what I'm suggesting here is more of a panel and frame construction with the corner/edges of the box done with small square dowels and the panels then adhered to that with the structurally visco adhesive from Weicon.
Consider a more compliant visco layer (like green glue) for the the interface between the main panel and constraining panel. (..Though I don't think it's suitable structurally for panel on frame connection.)
Consider different thickness of the visco layers (perhaps including the structural interface between panel and frame).
Consider different rigidity/stiffness of the constraining layer.
Consider different sizing of the constraining layer relative to the layer to be constrained.
Consider MULTIPLE layers of alternating visco and constraining layer (..and perhaps with increasing rigidity as you move further "outward").
Finally, (as I've previously suggested) consider panel damping in the context of a mass damper - where an opposed weight layer (usually very stiff relative to the layer to be constrained) and with a visco layer that has a sufficient thickness and compliance to generate substantive mechanical loss (damping) of the panel to be constrained.
From what I can tell speakers don't really "fit" the concept of CLD as specifically a panel-shear effect. (..though notably a LOT of the papers could have achieved other effects rather than panel-shearing, even though they proceeded in that context.)
Specifically CLD as a panel-shear effect is *typically studied with the panels to be constrained that are very thin (typically sheet metal), where both the panels have significant flexing (and the panel to be constrained is generally more flexible than the constraining layer panel though more than a few were trying the constraining layer that was MORE compliant - like thick aluminum foil???).
*typically: there are other less rigorous "white papers" out there that attempt to further the concept of CLD with much thicker panels, particularly in the building industry (and not a few of which are presenting themselves as "acoustic experts").
Papers (built on many prior papers) did note that the thickness of the visco layer was a variable to overall success (usually very thin but sometimes thicker depending on the other layers and the compliance mattered a lot) but ironically did NOT differentiate that success with specific panel-shearing vs some other effect. In other words, it could (and perhaps likely is) some other additional effect like transmission loss from *boundary shear moving from the layer to be constrained into the visco layer and then into the constraining layer.
*this is more of a "fluid" shear.
Notably some advocated full constraining layer dimensions to the layer to be constrained, where others had the constraining layer dimensionally smaller than the layer to be constrained.
Perhaps more significantly, there were significant variations is testing. While almost all required a flexible layer (that needs constraining), most did not go into detail with the fact that connected panels were fundamentally VASTLY more stiff near those connections. (..as in: the boxes edges/corners being far more stiff than the center of any panel.)
In the context of speakers then:
We have seen more dramatic effects from simply applying stuffing on the interior to lower pressure modes in the air in the box, that then resulted in lower excitation of the panel to be constrained. Beyond that the effects of attempted CLD have been modest at best.
Suggestions then are:
Reduce the thickness of the panel to be constrained. Perhaps 1/8".
Reduce any rigid coupling between adjoining panels (though Brandon has already done this to some extent with his more recent layering/construction), though what I'm suggesting here is more of a panel and frame construction with the corner/edges of the box done with small square dowels and the panels then adhered to that with the structurally visco adhesive from Weicon.
Consider a more compliant visco layer (like green glue) for the the interface between the main panel and constraining panel. (..Though I don't think it's suitable structurally for panel on frame connection.)
Consider different thickness of the visco layers (perhaps including the structural interface between panel and frame).
Consider different rigidity/stiffness of the constraining layer.
Consider different sizing of the constraining layer relative to the layer to be constrained.
Consider MULTIPLE layers of alternating visco and constraining layer (..and perhaps with increasing rigidity as you move further "outward").
Finally, (as I've previously suggested) consider panel damping in the context of a mass damper - where an opposed weight layer (usually very stiff relative to the layer to be constrained) and with a visco layer that has a sufficient thickness and compliance to generate substantive mechanical loss (damping) of the panel to be constrained.
Last edited:
Regarding using thinner constrained layer, the very first sample I made was using just two sheets of 1/4" MDF joined together, and it performed really well. When you consider the other end of the spectrum are people making crazy thick panels, it struck me that CLD was a good example of "work smarter, not harder".
I also wonder what 3 layers of 1/4" MDF would behave like. I know a lot of people don't like MDF though, but high quality 1/4" plywood doesn't exist in my area.
I also wonder what 3 layers of 1/4" MDF would behave like. I know a lot of people don't like MDF though, but high quality 1/4" plywood doesn't exist in my area.
Maybe 1/8 visco 1/4 visco 1/4?
or perhaps
1/4 visco 1/8 visco 1/4
with the 1/8 panel as the panel to be constrained. ..and in the 2nd example the interior 1/4 should not be the size of the 1/8 panel (..instead it should be smaller and not connected to anything but most of the visco layer to the 1/8 panel.)
I don't think high quality plywood is beneficial - if anything it should be MORE rigid than cheaper ply. (..though I could be wrong.) THOUGH if you are using it in the context of the constraining layer then it might well be beneficial.
Also, it might even be benefical to try other materials - particularly more compliant materials for the layer to be constrained like perhaps a thick semi-rigid vinyl sheet with much higher internal-loss character.
or perhaps
1/4 visco 1/8 visco 1/4
with the 1/8 panel as the panel to be constrained. ..and in the 2nd example the interior 1/4 should not be the size of the 1/8 panel (..instead it should be smaller and not connected to anything but most of the visco layer to the 1/8 panel.)
I don't think high quality plywood is beneficial - if anything it should be MORE rigid than cheaper ply. (..though I could be wrong.) THOUGH if you are using it in the context of the constraining layer then it might well be beneficial.
Also, it might even be benefical to try other materials - particularly more compliant materials for the layer to be constrained like perhaps a thick semi-rigid vinyl sheet with much higher internal-loss character.
Last edited: