Speaker Cabinet Design Considerations

[jsy]

Well I am an engineer so I am referring to useful structures, ie a box with a big enough volume to correctly load the bass unit and panels stiff enough for their first mode of resonance to exceed 20kHz.
Not sure your mile high box of steel would comply but I certainly can't be arsed to do the sum.

It definitely would not comply with your, or any other reasonable definition of useful. Agree with your conclusions on arseing

 

[Biblob]

Based on the Gedlee patent posted in my previous post, bracing like this would be more effective than what I modelled in a previous post.


Edit: such design does make me worry about reflection waves coming back at the woofer that bounce back on the woofer in the box. Stuffing it might solve that though

 

[jsy]

Based on the Gedlee patent posted in my previous post, bracing like this would be more effective than what I modelled in a previous post.


View attachment 49029

Trying to visualize...these are sandwiches of rigid braces with a layer of damping in between them? With one brace connected to each panel, in 3 axes?

 

[Biblob]

Trying to visualize...these are sandwiches of rigid braces with a layer of damping in between them? With one brace connected to each panel, in 3 axes?

Correct!

 

[Rick Sykora]

Based on the Gedlee patent posted in my previous post, bracing like this would be more effective than what I modelled in a previous post.

View attachment 49029

Edit: such design does make me worry about reflection waves coming back at the woofer that bounce back on the woofer in the box. Stuffing it might solve that though

is the damping material still sorbothane or something else?

 

[Biblob]

is the damping material still sorbothane or something else?

Well, I don't know what to try out. In post #74 of the diyaudio thread a lot of possible materials are mentioned: some types of glue that don't hard and other damping products. The problem with Sorbothane is that it doesn't post some important technical properties.
The problem for me is that some stuff is either expensive or not available in Europe.

 

[Rick Sykora]

Well, I don't know what to try out. In post #74 of the diyaudio thread a lot of possible materials are mentioned: some types of glue that don't hard and other damping products. The problem with Sorbothane is that it doesn't post some important technical properties.
The problem for me is that some stuff is either expensive or not available in Europe.

Thanks for sharing the thread! here is Gedlee’s answer...

The best damping material that I have found is 2K polyurethane - the soft stuff, shore hardness A I believe. It dries even in large panel sandwiches because it cures without air or moisture. It is quite expensive though, but works great. I mix in a lot of glass beads to increase the damping.

 

[jsy]

The text on Dr. Gedde's website is really informative. Just started going through it and the explanations for the models are fantastic

 

[jsy]

Googling for 2k poly damping brings me to decidamp brings me to geddes brings me to diyaudio brings me to Arnandsway's CLD thread here. There seem to be many small loops of the same people online trying to make soft sandwiches.

 

[Biblob]

Thanks for sharing the thread! here is Gedlee’s answer...

The best damping material that I have found is 2K polyurethane - the soft stuff, shore hardness A I believe. It dries even in large panel sandwiches because it cures without air or moisture. It is quite expensive though, but works great. I mix in a lot of glass beads to increase the damping.

Yes, I do wonder which brand. Some seem to be very expensive. In Europe Bison has a 2k one that is affordable. The Sikaflex is very expensive, for the amount you get.
But honestly, I don't have the technical knowledge to know what glue would be best. I mailed Swedac for pricing of the DG A2 in Europe.

 

[jsy]

Correct!

If your sandwich braces are only connected to one panel on each side, how do you keep the motion between the panels constrained to a shearing effect along the length of the damping? Wouldn't the assembly want to deform out of axis? The sandwich "bread" would be pushed apart.

What if you split each brace in half along the length, and connected the halves to opposite panels, but mated to the same damping layer? Each "face" of your sandwich would be connected to both panels. You could even do multiple fingers

 

[jsy]

If your sandwich braces are only connected to one panel on each side, how do you keep the motion between the panels constrained to a shearing effect along the length of the damping? Wouldn't the assembly want to deform out of axis? The sandwich "bread" would be pushed apart.

What if you split each brace in half along the length, and connected the halves to opposite panels, but mated to the same damping layer? Each "face" of your sandwich would be connected to both panels. You could even do multiple fingers

Also, in this setup you could get your damping layer between the fingers. The cross section of the damping would be cross-shaped. More damping, more shearing, same volume

 

[Biblob]

If your sandwich braces are only connected to one panel on each side, how do you keep the motion between the panels constrained to a shearing effect along the length of the damping? Wouldn't the assembly want to deform out of axis? The sandwich "bread" would be pushed apart.

What if you split each brace in half along the length, and connected the halves to opposite panels, but mated to the same damping layer? Each "face" of your sandwich would be connected to both panels. You could even do multiple fingers

Because you want the dampening effect to be handled by the constraining layer, because of it's visco-elastic properties. If one where to connect the brace to both panels, the vibration of the two panels wouldn't be able absorbed by the layer itself. It would make the panels stiffer though.
I took the inspiration from this picture of the Gedlee Patent, there are more ways to do this though:

I do have trouble visualising what you're trying to describe though. I'ts bedtime for me
Maybe you can make an example?

 

[jsy]

Because you want the dampening effect to be handled by the constraining layer, because of it's visco-elastic properties. If one where to connect the brace to both panels, the vibration of the two panels wouldn't be able absorbed by the layer itself. It would make the panels stiffer though.
I took the inspiration from this picture of the Gedlee Patent, there are more ways to do this though:
View attachment 49037

I do have trouble visualising what you're trying to describe though. I'ts bedtime for me
Maybe you can make an example?

See where the arrow is pointing marked 30-1? Just the one brace adjoined to the panel closest to you, the viewer? Remove it. Cut in half along the length. Now you have two finger shaped braces instead of one. Affix one to the near panel, and the other to the far panel, where you used to have just one brace. Now you have two finger braces, each connected to only one panel, but they are opposing panels.

Now remove the brace to the left and do the same thing. You'll have 4 braces instead of two, each brace still touching only one panel.

 

[jsy]

Because you want the dampening effect to be handled by the constraining layer, because of it's visco-elastic properties. If one where to connect the brace to both panels, the vibration of the two panels wouldn't be able absorbed by the layer itself. It would make the panels stiffer though.
I took the inspiration from this picture of the Gedlee Patent, there are more ways to do this though:
View attachment 49037

I do have trouble visualising what you're trying to describe though. I'ts bedtime for me
Maybe you can make an example?

Like this. Apologies for the freehand drawing

 

[Biblob]

View attachment 49041

Like this. Apologies for the freehand drawing

Aha! Like this?:
And am I understanding you correctly that this way the dampening factor increases?

 

[jsy]

Aha! Like this?:
View attachment 49043

And am I understanding you correctly that this way the dampening factor increases?

Almost! Now that you've sliced the brace in half, glue each half-brace to opposing panels

 

[Biblob]

Almost! Now that you've sliced the brace in half, glue each half-brace to opposing panels

Sure, will do it tomorrow. I guess this would be doable to do for the other side as well? So instead of 3 'fingers' you have 4? Or is that what you're trying to say all along? Effectively doubling the amount of braces.

 

[jsy]

Yes, 4 and now the damping shear applies in two planes for the same brace. You don't need the fingers to be uniform widths. You don't need to stop at 4 fingers, either.

 

[Biblob]

Yes, 4 and now the damping shear applies in two planes for the same brace. You don't need the fingers to be uniform widths. You don't need to stop at 4 fingers, either.

Something like this?