Discussion arising from Geddes loudspeaker

[gedlee]

I have to respectfully disagree. The rod approach just joins the two surfaces together so they resonate together. It does nothing to minimize the resonance. To do this you need to triangle brace with a rod from the cetre of one panel to the centre of the adjacent.
Easier way is to us a full panel bisecting the space down the middle, from side to side which will brace all sides.

A brace like this would go a long way:

And I respectfully counter disagree. The rod forces the two walls to be one at the center - the oak is virtually infinitly rigid. So if the two walls move together their stiffness has doubled and the resonance WILL go up. But the critical factor is that the two walls CANNOT move apart which is the way that the pressure in the box is trying to move them. In this regard the resonance goes WAY UP - far more than what you have shown. Yours solution is not even remotely effective to an interior pressure mode. While what the 2nd guy posted would work OK from a stiffness standpoint its acoustic effect would be very bad - I would never do that. MDF or plywood is very weak in compression - its a terrible brace material. Almost no wood products are any good in bending - which is why they need to be braced - but natrual woods can sometimes be very good in compression (like oak which is supurb) - that's how trees grow!

And you ignored the part about joining two or three rods at the center - now they can't move in ANY direction. Net result - enormous lowering of cabinet wall motion with very little added weight or loss of interior volume. There simply is no more effective method of stiffening than this - trust me! (or don't I don't care.)

 

[gedlee]

You are only looking at the sound pressure, not the mechanical from the driver itself.
The full panel brace will outperform the rod. No question.

And here again I have to disagree. The mechanical force due to the driver is very small compared to the pressure force from the driver. The mechanical force is the reaction force from the cones motion, but the cone has almost negligable weight compared to the mass of the encloure - there will be very little force transmitted in this manner. But the interior pressures can be enormous over large areas - this force can be very large and simply dwarfs the mechanical reaction force.

But this too I deal with by simply making the front baffle two pieces of MDF 3/4" and 1/2" glued together with a well damped glue. The baffle doesn't move from mechanical forces and the walls don't move from pressure forces. Thats what an enclosure is supposed to do.

 

[soongsc]

Has anyone done FEAs on these various shapes?

 

[MJL21193]

And I respectfully counter disagree. The rod forces the two walls to be one at the center - the oak is virtually infinitly rigid. So if the two walls move together their stiffness has doubled and the resonance WILL go up. But the critical factor is that the two walls CANNOT move apart which is the way that the pressure in the box is trying to move them.

Again, with all due respect, you don't seem to be seeing the picture here. Triangular bracing (on it's own or in the form of a panel) maintains the 90* angle in the corner. The rod approach doesn't. The panels will move in unison, that's all.

MDF or plywood is very weak in compression - its a terrible brace material.

It's not compression forces that are being resisted here, but tension.
Also, To quote myself from post #89 of this thread:

Here's the deal with braces: they are structural, therefore depending on the application, I might not use MDF. I'd use something stronger. Yes, plywood is stronger. So, whatever box it would be, the bracing material would be the same for both.
This is nothing special, just the correct way to build in the first place

And you ignored the part about joining two or three rods at the center - now they can't move in ANY direction. Net result - enormous lowering of cabinet wall motion with very little added weight or loss of interior volume. There simply is no more effective method of stiffening than this - trust me! (or don't I don't care.)

I didn't ignore it, it doesn't improve the situation enough.

And here again I have to disagree. The mechanical force due to the driver is very small compared to the pressure force from the driver.

Why check one and not the other? Better to be sure and cover both, than to ignore one completely as insignificant.
The full panel brace does this easily.

 

[gedlee]

Again, with all due respect, you don't seem to be seeing the picture here. Triangular bracing (on it's own or in the form of a panel) maintains the 90* angle in the corner. The rod approach doesn't. The panels will move in unison, that's all.

It's not compression forces that are being resisted here, but tension.
Also, To quote myself from post #89 of this thread:

I didn't ignore it, it doesn't improve the situation enough.

Why check one and not the other? Better to be sure and cover both, than to ignore one completely as insignificant.
The full panel brace does this easily.

Why is it that when I disagree with you I'M NOT seeing the picture. I think I see the picture, I know I see the picture, and I still disagree with you.

You are obviuosly NOT an engineer. Compression and tension are the same thing!! Just different signs.

Fixing the rods at the center is extremely effective - "your not seeing the piture here".

Why do something just to do it? Thats just a waste of time, materials, cost, everything - if it doesn't make a difference why do it?

At any rate, I'm done with this argument. I made my case, I'll rest on that.

 

[MJL21193]

Why is it that when I disagree with you I'M NOT seeing the picture. I think I see the picture, I know I see the picture, and I still disagree with you.

You are obviuosly NOT an engineer. Compression and tension are the same thing!! Just different signs.

Fixing the rods at the center is extremely effective - "your not seeing the piture here".

Why do something just to do it? Thats just a waste of time, materials, cost, everything - if it doesn't make a difference why do it?

At any rate, I'm done with this argument. I made my case, I'll rest on that.

Ok Earl, I never claimed to be an engineer. Neither did I claim that "Structural engineering is somewhat of a specialty of mine", that was you a few post ago.

Still, why do I see it and you don't? No problem to disagree, but you aren't disagreeing with me.

Compression and tension are the same? Apply that to a rope.

The difference between a discussion and an argument - something else you don't get.

 

[Robh3606]

"So a single Oak rod (maybe 1") across the enclosure will do more than many time that amount placed on the walls, and there is NO interior acoustic effect."

Thanks Earl

I have used the oak dowels used in closets to hang clothing on to good effect in a couple of my boxes. Made a big difference before and after. I crossed them but didn't tie them together. I didn't think the dampening would do all that much. Thanks for the tips.

Rob

 

[gedlee]

"So a single Oak rod (maybe 1") across the enclosure will do more than many time that amount placed on the walls, and there is NO interior acoustic effect."

Thanks Earl

I have used the oak dowels used in closets to hang clothing on to good effect in a couple of my boxes. Made a big difference before and after. I crossed them but didn't tie them together. I didn't think the dampening would do all that much. Thanks for the tips.

Rob

When I say "tie them together" I mean glue or bond somehow so that they can't move - there won't be any damping. Think about this structure - it forms the basis of almost every structural component that I know of - Bridges, space stations, you name it.

 

[soongsc]

I have to respectfully disagree. The rod approach just joins the two surfaces together so they resonate together. It does nothing to minimize the resonance. To do this you need to triangle brace with a rod from the cetre of one panel to the centre of the adjacent.
Easier way is to us a full panel bisecting the space down the middle, from side to side which will brace all sides.

A brace like this would go a long way:

I agree with gedlee on the effectiveness of using the rod. For enclosure surfaces to resonate together as you have mentioned, the mass distribution of the total enclosure must meet quite specific conditions so that the net force will be zero to keep the box stationary.

 

[gedlee]

Ok Earl, I never claimed to be an engineer. Neither did I claim that "Structural engineering is somewhat of a specialty of mine", that was you a few post ago.

Your point is? Should I hide my expertise?

Still, why do I see it and you don't? No problem to disagree, but you aren't disagreeing with me.

Actually I am disagreeing with you. No we don't "see" things the same way. Maybe what you are "seeing" isn't correct, maybe I do "see it" but just don't agree with it.

Compression and tension are the same? Apply that to a rope.

Again, this is incorrect. The rope is weak when you push on it because it bends. If it didn't bend then it would have nearly the same strength pushing or pulling. Don't confuse bending stiffness with compression stiffness. Would a piece of rope glued on the side stiffen the panel? No, but it can carry a lot of force if I can set that force in a way that doesn't allow the rope to bend. the rod idea that I talked about will actually work a little better if you put the rod in tension or compression (either way is about the same) - this is called pre-loading.

The difference between a discussion and an argument - something else you don't get.

Its an arrgument and I don't won't engage in it when people become disrespectful like you have been in this post.

 

[pooge]

I was thnking about that very same thing as i was considering John's offer to build 2 otherwise identical boxes... the box i had in mind is very much optimized for mostly 1/2" plywood (but very specifically uses some 3/8" MDF in a critical location to provide a dissimiilar material interface)... executing the same box in MDF would be like shooting an animal in a trap (ie MDF is just not stiff enuff to make a decent brace)

dave

My point, not expressed, is that every material is "transparent" at some frequency. Building enclosures to compare, where the design minimizes transparency using one material vs. not doing so with the other, is an unfair comparison regarding the criticism of MDF transparency. Unless the material characteristics are utilized to their best advantage in the design of the overall system, it is not a fair comparison.

Personally, I always wanted to try round panels made out of light weight, high-tensile-strength material to reduce th ability to flex, the light weight mimimizing any energy storage. I haven't done any testing, but I always wanted to try a sandwich of two metal or hardboard skins foamed in between with a stiff foam.

 

[tinitus]

A small bag of sand(1kg) tied to the middle of the baffle is effective too...and it dont transfer energy to other parts of the box

 

[gedlee]

Personally, I always wanted to try round panels made out of light weight, high-tensile-strength material to reduce th ability to flex, the light weight mimimizing any energy storage. I haven't done any testing, but I always wanted to try a sandwich of two metal or hardboard skins foamed in between with a stiff foam.

If the inner layer is well damped glue then you have the classic Constrained Layer Damping which I use on my baffles. This is very effective at reducing vibrations. Nothing that I have seen ever even comes close to CLD for effectivness with the panel vibration problem. (this comes from 20 years in automotive noise control.)

 

[MJL21193]

I agree with gedlee on the effectiveness of using the rod. For enclosure surfaces to resonate together as you have mentioned, the mass distribution of the total enclosure must meet quite specific conditions so that the net force will be zero to keep the box stationary.

??

Do a simple, cheap experiment. Take a cardboard box and cut the top and bottom out of it. Now use a dowel to join the sides together. How strong is this structure?

Now take another piece of cardboard, one big enough to fit inside the box so it touches all sides. glue this in so it bisects the space.
How strong is this structure?

If you actually do the experiment, you will find that the full panel makes the structure MUCH stronger.

Forget BS about mass distribution and net force. It's simple building science.

 

[pooge]

If the inner layer is well damped glue then you have the classic Constrained Layer Damping which I use on my baffles. This is very effective at reducing vibrations. Nothing that I have seen ever even comes close to CLD for effectivness with the panel vibration problem. (this comes from 20 years in automotive noise control.)

I agree that CLD is the way to go for flat panels, of which the baffle almost has to be for a large woofer. But if the side and back panels are curved, preferably made by one panel, material bending stiffness is not a big deal, as high tensile strength material prevents stretching of the material. Since such a material can also be light weight, it could also be easily damped, if not already inherent in the material itself. Not sure what material property affects transparency, since I haven't looked at this for a while, but I guess that an injected foam like you would use to foam around your doors may reduce this, or perhaps some type of poured "rubber" that hardens. Never got around to experimenting with this to find out how thick to make the sandwich, or whether to use stiff expanding foam, or the newer low expansion type that doesn't stiffen as much.

If the front baffle is laminated, the inside layer could be made slightly smaller to accept attachment of the inside curved layer to its side edge, or a groove routed into its inside face to accept the edge of the inside curved panel for gluing, with the outside curved layer attached to the outside baffle layer.

The top and bottom laminated panels would be similarly attached to the curved panels.

The bottom panel could not be laminated until the "foam" is injected or poured.

The foam would be the biggest cost. It would perhaps be easier to go to an insulating company to have it done, rather than try the finicky spray cans with their obstinate plastic tubes that may not reach far enough into the enclosure panels.

A further advantage of a curved wall is that it minimizes the available surface area for a given enclosed volume.

 

[graaf]

the rod idea that I talked about will actually work a little better if you put the rod in tension or compression (either way is about the same) - this is called pre-loading.

this was MGD (Michael Green Designs) approach to the issue
they used steel (or brass?) threaded rods so that the tension could be adjusted

where are they now?

 

[dlr]

this was MGD (Michael Green Designs) approach to the issue
they used steel (or brass?) threaded rods so that the tension could be adjusted

where are they now?

That was little more than a gimmick. It was supposed to be a "tuning device". Rather than eliminate wall vibration, it provided coloration with some degree of control if you believed their ad copy. Where are they now? Probably where many poorly conceived ideas go.

Dave

 

[gedlee]

I agree that CLD is the way to go for flat panels, of which the baffle almost has to be for a large woofer. But if the side and back panels are curved, preferably made by one panel, material bending stiffness is not a big deal, as high tensile strength material prevents stretching of the material. Since such a material can also be light weight, it could also be easily damped, if not already inherent in the material itself. Not sure what material property affects transparency, since I haven't looked at this for a while, but I guess that an injected foam like you would use to foam around your doors may reduce this, or perhaps some type of poured "rubber" that hardens. Never got around to experimenting with this to find out how thick to make the sandwich, or whether to use stiff expanding foam, or the newer low expansion type that doesn't stiffen as much.

If the front baffle is laminated, the inside layer could be made slightly smaller to accept attachment of the inside curved layer to its side edge, or a groove routed into its inside face to accept the edge of the inside curved panel for gluing, with the outside curved layer attached to the outside baffle layer.

The top and bottom laminated panels would be similarly attached to the curved panels.

The bottom panel could not be laminated until the "foam" is injected or poured.

The foam would be the biggest cost. It would perhaps be easier to go to an insulating company to have it done, rather than try the finicky spray cans with their obstinate plastic tubes that may not reach far enough into the enclosure panels.

A further advantage of a curved wall is that it minimizes the available surface area for a given enclosed volume.

I don't think that I follow since I don't see any significant difference in flat versus curved. The automotive panels that I dealt with weren't flat.

Foam is pretty easy to get - look up US Composites in Florida. I would think that the foam would be cheap compared to the high tensile strength sheet. I think of things like carbon fiber, etc. which are far from cheap. I can't think of a cheap material with these properties.

(Oh, and I've made carbon fiber cabinets too - no difference.)

 

[gedlee]

Do a simple, cheap experiment. Take a cardboard box and cut the top and bottom out of it. Now use a dowel to join the sides together. How strong is this structure?

Now take another piece of cardboard, one big enough to fit inside the box so it touches all sides. glue this in so it bisects the space.
How strong is this structure?

If you actually do the experiment, you will find that the full panel makes the structure MUCH stronger.

Forget BS about mass distribution and net force. It's simple building science.

This is not the correct experiment. Its biased to highlight the advantages of your approach. Put two rods across the box and glue them at the center. Now which one yields the most strength for weight and volume of the stiffener? And when you use very flimsy materials then the stiffness increase does not show up for the rod as well as when the materials are stronger. You need a fair test to make judgements from.

 

[pooge]

I don't think that I follow since I don't see any significant difference in flat versus curved. The automotive panels that I dealt with weren't flat.

Foam is pretty easy to get - look up US Composites in Florida. I would think that the foam would be cheap compared to the high tensile strength sheet. I think of things like carbon fiber, etc. which are far from cheap. I can't think of a cheap material with these properties.

(Oh, and I've made carbon fiber cabinets too - no difference.)

Hmmm. I didn't expect this response. Since it is pretty well known that bending stiffness is a lot better for a curved panel than a flat panel, so I'm not sure if we are on the same subject.

Also, I don't think that searching for an exotic or expensive material with the highest tensile strength is required here, as a modicum of tensile strength in a curved panel would probably fundamentally trump a flat panel for bending stiffness. (I'm winging it, here.) An aluminum skin on the inside may have quite enough tensile strength exceeding what is necessary, has pretty good self damping, but not be very cheap. On the other hand, 1/8 in hardboard would be much cheaper, easier to work with, and still might have very adequate tensile strength.