Thanks for the suggestion but that looks like a site for those that are into the web and html and know what they are doing. All I need is to adapt a handful of lines of javascript from a how to get started with this package example to call a few functions in a large js package written by someone else and put up on the web. Although it is only a small amount of code it is larger than zero which is the amount allowed by many sites including this one.
I have opted for github, joined and created a repository. Won't have anything to post for a week or two but the project has started.
Thank you i wasn't aware of 'droop'. It makes sense especially with mdf.
Your comment as well as those from White Dragon and Planet10 about geometry is the core of my first question:
by using a 'stronger' material than wood there should be improvement. Intuitively i thought that a T profile will gives the most effect against flexion because there is 'matter' present in the middle of the strip to counteract the effect.
So i seeked a bit about it and discovered 'quadratic moment'.
While a read again link Fluid provided it occured to me the 'best' shape is an extruded 'hollow' rectangle profile. Something like a 50mmx30mm or 60x30mm.
It'll weight a bit more but being way more firm.
After discussion with my friend into boat it occured the glue he thoughts about won't do it as too flexible once cured: too much damping. It's a Silkaflex product iirc.
We came to conclusion epoxy is propably better in this case.
Check engineering texts for deflection of beams at point load, cantilever or continuously distributed load.
And torsion of shafts as well.
Then check the unit weight as well.
Mechanical and civil engineering, third and fourth year in my course of mechanical engineering.
It will give you a pretty good idea of what is the most suitable solution.
If you can, get steel window sections in T shape.
By the way, I agree with Dave, T is better than square section for stiffening, or as a brace.
And torsion of shafts as well.
Then check the unit weight as well.
Mechanical and civil engineering, third and fourth year in my course of mechanical engineering.
It will give you a pretty good idea of what is the most suitable solution.
If you can, get steel window sections in T shape.
By the way, I agree with Dave, T is better than square section for stiffening, or as a brace.
Use screws or bolts if you think glue will affect damping, depends how durable the screws will be in chip board, MDF and similar material, of course...
Bolts will pass through, so more durable, but looks may be affected.
Bolts will pass through, so more durable, but looks may be affected.
Hi,
I found a calculator to compare behavior of different profile under flexion:
http://www.mecatools.free.fr/inerties/mflexion.html
I found a calculator to compare behavior of different profile under flexion:
http://www.mecatools.free.fr/inerties/mflexion.html
It's trivially easy to reduce box radiation in the pressure zone with wood bracing. I doubt you will see any benefit with using aluminum. Probably the surface area of the end of the brace contacting the panel will have more impact on performance than changing the material.
Damping the radiation in the primary panel resonance region is much tougher, and more important given the SPL is much higher. Here, stiffening even more may actually be worse than using wood braces. At least wood has internal damping. Aluminum has very little.
Damping the radiation in the primary panel resonance region is much tougher, and more important given the SPL is much higher. Here, stiffening even more may actually be worse than using wood braces. At least wood has internal damping. Aluminum has very little.
Well it is related to the Audioholic's link Fluid published earlier.
In the simulation the 9 brace scenario with 4"x1" mdf brace is considered as the 'upper' limit wrt material used. The author suggest then to shift to other material ( aluminium) because the more 'simple' braces used the less efficient they become ( for a given material).
With 4"x1" mdf 'simple' brace you would say switch to 'shelve' or 'window' brace directly: this won't wast this much material and will connect 4 sides of the box together improving strength against flexion.
So let's imagine the same kind of bracing scheme for the panel as presented in the link ( 9 horizontal 'simple' brace, vertical being the longer panel length) but with 5 mdf 'window' brace and 4 aluminium tubing 'simple' brace. It should fullfil requirements displayed in the article?
about the article: in find highly unintuitive the way braces are oriented wrt panel's shape.
the sim reveal it makes sense but the longest dimension is the one in which i would have done it...
i get it's all a question of modes within a plate and to apply support to area in movement but still... i find it unintuitive.
This is a really interesting thread, with links to very intense experiments. Good work.
After reading it all, one can come to two different conclusions:
First: it is incredible complicated to construct a box suited to serve as a loudspeker case.
The second one, which I prefer: Loudspeaker boxes follow simple rules of physics which have been used for thousands of years by engineers and builders.
If a Roman civil engineer had to build a bridge over a river in Germania and a flat planc of wood was to weak, he braced it.
When Ceasar wanted to have an undisturbed night with Cleopatra, he pulled some heavy curtain over the windows.
You can take these two simple ideas and build very good loudspeaker cases with them.
If you do not get the picture, braces are for high load which equals low frequency. Soft, heavy media stops high frequncy.
Some years ago there was a very practical test in the German Hobby HIFI magazine, using commonly available materials for speaker building. In the end braces 3-4 times as high as the basic boards strengh in distances about 6-8 times the board strength proved to be the best low end solution. Nothing new, the T-bar.
Heavy, self adhesive elastic matereal (like Resonix, some aluminum foil with bitumen or other heavy elastic stuff ) was the winner in the higher frequency range, considering the price.
Lead balls (like bird shot) in some elastic material where best, but incredible expensive and ecologicaly questionabel. Just like massive tiles glued to the inside, which have a good effect, but the weight and problems of fixing them permanently make them impractical.
In practice, the simple method of building a box and bracing it where the knock test sounds "boxy" works fine. This is for the low end, preventing flexing of walls.
If you build a full range housing, after bracing, glue additional dampening to the inside walls. Material like Resonix is fine, but if you use a fair amount of elastic glue, carpet leftovers or thick felt are fine. The idea is the same: To let a material take up sound energy and convert it into thermal energy. Hard matter transports, while soft stuff absorbs.
If you want to find problems in your finished box or while building, simply make an impedance sweep. You should easily seey when irregular peaks disturb the ideal cuve. This is much simpler than complicated microphone measurements and takes only a few minutes. If you do not know how, you only need a 5 Cent resistor and a tutorial.
This does not contradict the BBC approach, Wood outside and some kind of tar inside. I personally think, if wheigt, size and cost would not have been soo important, the wood of the BBC monitors would have been a bit thicker, like todays examples show.
After reading it all, one can come to two different conclusions:
First: it is incredible complicated to construct a box suited to serve as a loudspeker case.
The second one, which I prefer: Loudspeaker boxes follow simple rules of physics which have been used for thousands of years by engineers and builders.
If a Roman civil engineer had to build a bridge over a river in Germania and a flat planc of wood was to weak, he braced it.
When Ceasar wanted to have an undisturbed night with Cleopatra, he pulled some heavy curtain over the windows.
You can take these two simple ideas and build very good loudspeaker cases with them.
If you do not get the picture, braces are for high load which equals low frequency. Soft, heavy media stops high frequncy.
Some years ago there was a very practical test in the German Hobby HIFI magazine, using commonly available materials for speaker building. In the end braces 3-4 times as high as the basic boards strengh in distances about 6-8 times the board strength proved to be the best low end solution. Nothing new, the T-bar.
Heavy, self adhesive elastic matereal (like Resonix, some aluminum foil with bitumen or other heavy elastic stuff ) was the winner in the higher frequency range, considering the price.
Lead balls (like bird shot) in some elastic material where best, but incredible expensive and ecologicaly questionabel. Just like massive tiles glued to the inside, which have a good effect, but the weight and problems of fixing them permanently make them impractical.
In practice, the simple method of building a box and bracing it where the knock test sounds "boxy" works fine. This is for the low end, preventing flexing of walls.
If you build a full range housing, after bracing, glue additional dampening to the inside walls. Material like Resonix is fine, but if you use a fair amount of elastic glue, carpet leftovers or thick felt are fine. The idea is the same: To let a material take up sound energy and convert it into thermal energy. Hard matter transports, while soft stuff absorbs.
If you want to find problems in your finished box or while building, simply make an impedance sweep. You should easily seey when irregular peaks disturb the ideal cuve. This is much simpler than complicated microphone measurements and takes only a few minutes. If you do not know how, you only need a 5 Cent resistor and a tutorial.
This does not contradict the BBC approach, Wood outside and some kind of tar inside. I personally think, if wheigt, size and cost would not have been soo important, the wood of the BBC monitors would have been a bit thicker, like todays examples show.
I had this same problem with the first MDF table top I built for my CNC. It was dead flat all around the edge but in the middle where it was unsupported it had sagged under it's own weight. The solution was an aluminium torsion box with a birch ply top, at a wildly increased price than two pieces of MDF glued and screwed together.
As you are speaking of having separate cabinets for different ranges you can choose the best strategy for each as when all the drivers are in the same cabinet there is inherently some sort of compromise between the different regions and what works best overall.
Aluminium T pieces would work well when glued or fixed on the panels to stiffen each 'plate' but practically they would be harder to use as cross braces. RHS or SHS would probably be easier to fix. As it's DIY there's not much downside in trying to over engineer. To me the hard part is I don't find many links back to how perceptible any of this is, which only seems to cause people to want to go to great lengths to try and make sue it is as low as possible, just to be safe 🙂
Here is some more from a B&W white paper on the 800D, what they say there is very much in line with augerpro's image depicting the different regions and it also includes the same image that b-force posted about Mass, stiffness and coincidence.
"In general, at low frequencies the stiffness of the walls dominates their behaviour, while at high frequencies it is their mass which rules. Between these two extremes they interact in a resonant manner which can grossly magnify and time smear the transmission at certain frequencies. This situation is rendered tolerable by resistive losses or damping. In general one is trying to maximise all these variables, although at times it can make more sense to ensure that a resonance is outside the frequency band to be used than it is to keep it subdued with the use of mass or damping. The stiffness of a panel for a given mass can be increased dramatically by curving it. Hence axially loaded tubes and spheres have long been recognised as the most efficient users of materials, though in our rectilinear society they have usually been relegated to more exotic designs. The mass of panels can be increased simply through the use of dense material. Bricks, lead and sand layers all offer increasing attenuation with frequency, but these are all definitely for the DIY enthusiast. In the real world, cabinets were made of wood with various degrees of panel bracing and damping, like bituminous mats and suchlike, which also helped increase the mass. By increasing both the mass and the stiffness, the lowest point in the curve corresponding to the minimum transmission loss may be brought up to reasonable levels. (figure AVI.1)"
At some point I read another paper about the latter version where the bracing was changed to plywood with aluminium inserts based on FEA modelling to find a more efficient use of materials. I have never been able to find it again, but there are some images in this video showing what they did and some screenshots of the sims.
A paper from B&W that I could find
https://www.comsol.com/paper/download/199487/cobianchi_paper.pdf
If it ever get's intuitive then you are probably going the wrong way 🙂
There was a thread here, a bitumen layer was used as a damping element, and it would act as a restraint as well.
So a tar felt sheet from a roofing supplies business would work, a heat gun may be needed.
A similar product is used in car doors as well.
Easy, no fasteners needed.
So a tar felt sheet from a roofing supplies business would work, a heat gun may be needed.
A similar product is used in car doors as well.
Easy, no fasteners needed.
@andy19191 If you want to show/demonstrate what happens in a simulation, wouldn't an animated gif be enough to display it on screen?
Or are you really going for an interactive sort of environment (seems a bit much to get the point across).
I'd say fire up your favorite program to run the sims, but capture it in pictures and create a short animation to get the message out.
This could easily be presented in blog form on Wordpress or even in a thread here.
Or do you want to share the entire simulation, including software, then Github would be a fine solution if the software is all open source.
On another note, what I don't get in that bracing example in the link from fluid is to have the braces equal distance from each other.
I'd choose an uneven number of braces, but have them vary just slightly in distance as to not break up the panel in equal sized parts.
Bitumen is pretty nasty stuff and eventually will dry out over time. Most of that has been replaced with stuff like butyl. Butyl often is found in combination with an alu-foil layer for car door damping (aftermarket) but also to glue windshields. It stays flexible, pliable and sticky. For some goals bitumen might actually function better but it still is nasty stuff. For adding weight to panels mass loaded vinyl could be a valid option. Basically acting as dead weight.
Friendlier than using lead.
Or are you really going for an interactive sort of environment (seems a bit much to get the point across).
I'd say fire up your favorite program to run the sims, but capture it in pictures and create a short animation to get the message out.
This could easily be presented in blog form on Wordpress or even in a thread here.
Or do you want to share the entire simulation, including software, then Github would be a fine solution if the software is all open source.
On another note, what I don't get in that bracing example in the link from fluid is to have the braces equal distance from each other.
I'd choose an uneven number of braces, but have them vary just slightly in distance as to not break up the panel in equal sized parts.
Bitumen is pretty nasty stuff and eventually will dry out over time. Most of that has been replaced with stuff like butyl. Butyl often is found in combination with an alu-foil layer for car door damping (aftermarket) but also to glue windshields. It stays flexible, pliable and sticky. For some goals bitumen might actually function better but it still is nasty stuff. For adding weight to panels mass loaded vinyl could be a valid option. Basically acting as dead weight.
Friendlier than using lead.
Last edited:
I think it is done to try and show how the division of the panels changes the resonant frequency, if the sub panels were different sizes that would make it harder to compare. It seems to me it is meant to be instructive rather than be followed directly, in reality it would be better to mix up the sub panel sizes as you say.
The instructive value often is missed and people tend to just copy what they see. As we see the modes in their models, it should be obvious they travel with equal spaced nodes, hence we should place the braces slightly off (and not in the nodes) to break up that behavior. Not showing that effect is not looking that instructive to me. 🤔
I think I've seen quite a bit of symmetrically placed braces in loudspeaker projects. I don't think it's wrong to try and alter that trend. I guess strong symmetry is attractive to many (even though we might never see the braces again and they are there to do a job), but not the most logical choice in cases like this.
I think I've seen quite a bit of symmetrically placed braces in loudspeaker projects. I don't think it's wrong to try and alter that trend. I guess strong symmetry is attractive to many (even though we might never see the braces again and they are there to do a job), but not the most logical choice in cases like this.
Last edited:
That's true, it seems obvious to me that in a real build you wouldn't want every subpanel to be the same size and have the same resonant frequency, but maybe it isn't that obvious to everyone and that wasn't addressed in the article.
Depends on what one is trying to get across. If it is to make a point then yes selecting a few images is good. If you want to inform about something fairly complicated like the motion of a speaker cabinet then less so. To understand a speaker's mode shape including the motion of the drivers and internal bracing would require perhaps 3 or 4 images. Multiplied by 5 or so interesting modes. Plus their sound radiation patterns. Multiplied by ten or more parametric changes. Covering everything would require a lot of pictures and substantial effort (which I have done for internal reports in the past) or dumping a plot file for each build into a viewer with some notes on what is interesting.
The motivation is minimising my effort and getting a lot of information across efficiently (plus learning how to do stuff in the modern web based world).
Sharing the full results is the current short term objective. Sharing some software would be a longer term objective if interest develops. I opened a GitHub Pages account to create a website but that required opening a github software account first and then storing the website inside it which confused me a bit at first.
That would depend on what is trying to be achieved and why.
Setting aside the question of whether adding mass is a good, bad or unimportant one will usually require some structures to be effectively damped. Effective damping requires a significant damping force which is the product of the displacement and the loss modulus of the damping material. Compared to walls in rooms, computer casings, car doors,... the displacement of speaker cabinet wall is small suggesting it will need a material with a higher loss modulus to be effective. That is, a stiff lossy material rather than a soft lossy material.
Uneven is not necessary, but i do plce then such that none of the subpanels are equivalent.
dave