Actually one can visualize the effects if an appropriate Finite Element Analysis tool is used. The technology is there, it's just a matter of whether the people that use it feel it worth the effort to go in that direction or not. Since it is already common knowlege that all efforts should be made to minimize diffraction effects, it's easier to just apply it based on the current simple tools available. These tools are already sufficient to make diffraction less dominant than driver performance.johninCR said:
JohninCR,
This is a clear if somewhat abbreviated and non mathematical summation of boundary layers and what affects them. I am sure that a translation of Prandtl's original paper would also help with the OB questions raised so far..... wanna bet there are even more energy functions occurring than those already argued over in such a friendly and sane fashion?
http://units.aps.org/units/dfd/prandtl_vol58no12p42_48.pdf
Panomaniac,
Actually I think it would work, the hairy cones... but I do suspect that the efficiency would suffer a bit. But you might be able to see the waves in the grass and surely hawaiian dance music would get the stuff swaying...
Bud
This is a clear if somewhat abbreviated and non mathematical summation of boundary layers and what affects them. I am sure that a translation of Prandtl's original paper would also help with the OB questions raised so far..... wanna bet there are even more energy functions occurring than those already argued over in such a friendly and sane fashion?
http://units.aps.org/units/dfd/prandtl_vol58no12p42_48.pdf
Panomaniac,
Actually I think it would work, the hairy cones... but I do suspect that the efficiency would suffer a bit. But you might be able to see the waves in the grass and surely hawaiian dance music would get the stuff swaying...
Bud
soongsc said:
I don't need a tool, though an animation might be nice. Just an explanation is fine, especially if it's down to the level where Huygen's principle leaves off. I'll figure my own way around it, and have little interest in simulating it only to get bogged down in trial and error simulations. Those got old long ago.
BudP said:
Bud,
Interesting stuff. Now how about linkage with sound, since it propagates by back and forth movement instead of a flow. I'm seeing my shoe moving back and forth on a smooth floor that is covered lightly with sand, and the sand is my boundary layer. Also assuming there's linkage to sound, is it significant enough to become audible? I was visualizing your treatment's effect on driver cones to be something different, instead diffusing the waves propagating within the cone itself as opposed to something going on in a boundary layer.
Regarding the hairy cones, with enough micro fibers they may even increase sensitivity due to the greater surface area offering a better impedance match with air.
JohninCR,
Now move your foot really fast and really far, about 13k inches in a second.
The voice coil is a pump, the emitter surface is a passive device that is being acted upon by lateral energy displacement and this lateral displacement creates a compression wave in the adjacent air, that is attached to the emitter at the moving peak of the lateral energy displacement.
This is the classic Walsh description of a Bending Wave transducer. All transducers act like this and as Beranek and Corrington showed us 50 years ago or so, the piston model functions only at one frequency, all other frequencies use the wave ripple model, both above and below that perfect piston point. The wave ripple model uses the boundary layer to transform energy from a lateral wave to a compression wave.
That boundary layer is extremely fragile and can be disturbed by very minute events, like paint that alters the surface height or Reynolds number, or both.
Once the emitter has been passed by, the compression wave and that lateral connection to the boundary layer of the next surface, move across that surface. When the surface terminates, a multitude of events occur. All have been described in this argument to date.
One of these events is, that what energy cannot leave that boundary layer, reflects, creating a new pressure wave as it re traverses the surface, attached to the boundary layer. Control of that boundary layer reflection is what this EnABL thing is all about.
It works because the boundary layer interface is fragile and the blocks of paint are just enough to raise the impedance to that back wave reflection, to cause that energy to exit in one of the other modes available. The least expensive being in phase with the original compression wave.
Bud
Now move your foot really fast and really far, about 13k inches in a second.
The voice coil is a pump, the emitter surface is a passive device that is being acted upon by lateral energy displacement and this lateral displacement creates a compression wave in the adjacent air, that is attached to the emitter at the moving peak of the lateral energy displacement.
This is the classic Walsh description of a Bending Wave transducer. All transducers act like this and as Beranek and Corrington showed us 50 years ago or so, the piston model functions only at one frequency, all other frequencies use the wave ripple model, both above and below that perfect piston point. The wave ripple model uses the boundary layer to transform energy from a lateral wave to a compression wave.
That boundary layer is extremely fragile and can be disturbed by very minute events, like paint that alters the surface height or Reynolds number, or both.
Once the emitter has been passed by, the compression wave and that lateral connection to the boundary layer of the next surface, move across that surface. When the surface terminates, a multitude of events occur. All have been described in this argument to date.
One of these events is, that what energy cannot leave that boundary layer, reflects, creating a new pressure wave as it re traverses the surface, attached to the boundary layer. Control of that boundary layer reflection is what this EnABL thing is all about.
It works because the boundary layer interface is fragile and the blocks of paint are just enough to raise the impedance to that back wave reflection, to cause that energy to exit in one of the other modes available. The least expensive being in phase with the original compression wave.
Bud
There are two things that take place with the pattern.johninCR said:
1. When used on structures, It stimulates a turbulant boundary layer, and delays the boundary layer separation, due to this delay in separation, the diffraction effects are reduced in magnitude. I believe you will can measure the differences. Effects can be visualized via analysis. The theory is similar to why golf balls have those dimples.
2. When used on thin membranes such as cones, the patterns will breakup the vribration wave. Wavelength of effectiveness depends on material size, material properties, and pattern layout and pattern material.
Across a structure, there isn't a flow, just very small back and forth movement of the same molecules, so my shoe would make very small fast movements. That's why I asked for the linkage that a similar analysis applies to sound. Due to the type of movement involved, boundary separation must be a very different thing.
With the cone treatment, I'm having real trouble getting past that EnABL isn't a strategic geometric damping/diffusion of waves in the cone itself, just like the blocks in the water tank (which doesn't look like something related boundary layers to me).
I guess I'll just have to try blocks of paint on one baffle and something thicker like a foam tape on another, and see what happens.
With the cone treatment, I'm having real trouble getting past that EnABL isn't a strategic geometric damping/diffusion of waves in the cone itself, just like the blocks in the water tank (which doesn't look like something related boundary layers to me).
I guess I'll just have to try blocks of paint on one baffle and something thicker like a foam tape on another, and see what happens.
Can't resist, sorry.
All ... Please keep up this endless BS of diffraction academics and allow the actual building of many more important aspects of building an OB slowly die.
At my age I will expire before anything gets accomplished. Forgive my impatience, and I do expect a little wrath over my butting in to this purely classroom boring rhetoric.
Zene
All ... Please keep up this endless BS of diffraction academics and allow the actual building of many more important aspects of building an OB slowly die.
At my age I will expire before anything gets accomplished. Forgive my impatience, and I do expect a little wrath over my butting in to this purely classroom boring rhetoric.
Zene
Possibly this thread has strayed away from its original direction.
I have not heard any more from Lynn Olson since this diffraction discussion began.
Maybe a new thread about diffration and its effects on an OB should be started.
I am interested in the latest OB design imformation that Lynn has
thought of.
Anything new Lynn?
I have not heard any more from Lynn Olson since this diffraction discussion began.
Maybe a new thread about diffration and its effects on an OB should be started.
I am interested in the latest OB design imformation that Lynn has
thought of.
Anything new Lynn?
Hi Bratislav,
Re your Post#1400.
Whilst I was getting my beauty sleep you have addressed me as if 'I' stated that simulation does not work.
I have not said this ! You have though put into words the mechanism I had in mind.
JohninCR made a point to JohnK that the posted simulations did not correlate with real life observations, I merely suggested that airflow (laminar/turbulent) rather than wavefronts could be simulated; much like we simulate voltage and current in circuits or consider voltage and magnetic fields in electromagnetic radiation.
Of course both aspects for each are directly inter-related, but not always as expected if *impedances* are not constant, and thus the results are not always intuitively cross-predictable. This is where simulation so often falls flat on its face, though it is not the simulation per-se which is wrong, but weaknesses related to its implementation.
Come to think of it, the Edge software shows a change of response as a baffle board is moved about beside a LS mounted in free space, but there is not any allowance for changing the shape of the edge of that board, even though the shape of any nearby sharp edge (in the null plane) becomes significantly audible. Of course the null plane itself is not constant with frequency either.
We could measure this re-radation, but our own ears already tell us what sounds better, besides, Lynn has already stated his intention to place the ends of his smoothly curved baffle away from a forward listening position.
Cheers ......... Graham.
Re your Post#1400.
Whilst I was getting my beauty sleep you have addressed me as if 'I' stated that simulation does not work.
I have not said this ! You have though put into words the mechanism I had in mind.
JohninCR made a point to JohnK that the posted simulations did not correlate with real life observations, I merely suggested that airflow (laminar/turbulent) rather than wavefronts could be simulated; much like we simulate voltage and current in circuits or consider voltage and magnetic fields in electromagnetic radiation.
Of course both aspects for each are directly inter-related, but not always as expected if *impedances* are not constant, and thus the results are not always intuitively cross-predictable. This is where simulation so often falls flat on its face, though it is not the simulation per-se which is wrong, but weaknesses related to its implementation.
Come to think of it, the Edge software shows a change of response as a baffle board is moved about beside a LS mounted in free space, but there is not any allowance for changing the shape of the edge of that board, even though the shape of any nearby sharp edge (in the null plane) becomes significantly audible. Of course the null plane itself is not constant with frequency either.
We could measure this re-radation, but our own ears already tell us what sounds better, besides, Lynn has already stated his intention to place the ends of his smoothly curved baffle away from a forward listening position.
Cheers ......... Graham.
My 2 cents
I think it is amazing that a 'Vaporware' project has garnered 1415 posts. This isn't to say that it won't be built, and I read it with interest, but it does at times seem ridiculous and more of a soapbox deal than fleshing out the actual design based on specific criteria.
This isn't meant as a flame, or meant to offend either. I simply hope that the resultant 'cutting of wood' becomes something spectacular.
Chris
I think it is amazing that a 'Vaporware' project has garnered 1415 posts. This isn't to say that it won't be built, and I read it with interest, but it does at times seem ridiculous and more of a soapbox deal than fleshing out the actual design based on specific criteria.
This isn't meant as a flame, or meant to offend either. I simply hope that the resultant 'cutting of wood' becomes something spectacular.
Chris
Posturing
Academics to the Left and real hands-on doers to the Right. My measly 5 years in college was enough to get me out in the shop to do some DIYing. I don't think I need to go back.
Learnin' ain't all that bad as I have to every day as new problems arise. I use you knowlegible guys here to get down to the real solutons. I never call on one of my old profs.
I absolutely want to know more about wood and crossovers and speaker spacing and size and on and on. I believe we are still in the very old methods of OB design. Isn't that what Lynn said with his first post? A flat square panel with the speaker in the middle surely isn't correct for baffle step. Edge shows that very clearly. So, what is better? Waiting, waiting.
Peanut butter on the edges for diffraction may not be that important on the list of priorities.
Ask Lynn about glues. He will give you a lesson on what he went thou with the Ariels. That's worth discussing along with a hundred important things to do with design and building.
Lynn ... your last post was 60+ ago. Hope you are OK?
Zene
Academics to the Left and real hands-on doers to the Right. My measly 5 years in college was enough to get me out in the shop to do some DIYing. I don't think I need to go back.
Learnin' ain't all that bad as I have to every day as new problems arise. I use you knowlegible guys here to get down to the real solutons. I never call on one of my old profs.
I absolutely want to know more about wood and crossovers and speaker spacing and size and on and on. I believe we are still in the very old methods of OB design. Isn't that what Lynn said with his first post? A flat square panel with the speaker in the middle surely isn't correct for baffle step. Edge shows that very clearly. So, what is better? Waiting, waiting.
Peanut butter on the edges for diffraction may not be that important on the list of priorities.
Ask Lynn about glues. He will give you a lesson on what he went thou with the Ariels. That's worth discussing along with a hundred important things to do with design and building.
Lynn ... your last post was 60+ ago. Hope you are OK?
Zene
I'm convinced that driver selection is much more critical than the baffle design. Once the simple baffle diffraction simulators provide you with the lowest effect you can come up with. Then it's the drivers that count. No crossover design is going to compensate for bad drivers unless you are satisfied with less performance.
Judging from the MLSSA data shown at Lynn's web site, the impulse response looks funny, and I would expect instruments like cymbals to sound a bit suppressed. The waterfall data seems to extend a bit long. I would aim for a goal to have a drop withing the first 0.1 ms of 12db. This way the recorded room reflections and timbre come out more naturally.
The next critical point is XO design. Lynn has his drivers connected in phase with each other. I find that this is the best configuration. All XO design needs be be based on this configuration. If the drivers are out of phase, no matter what XO you have, when the cymbals sound right, the drums sound a bit muffled, and vice versa.
Component selection is the next critical point. The previous need to be done well before you can really hear the benefits of components. The first pass in components selection can be done by using a passive first order filter driven by an amp. The components that give the lower distortion and harmonic contents are generally the better components, and you can categorize them and see if listening results are the same as measured results.
The final is internal wiring. This can also be done in a similar way as component selection.
Don't wait on others to start cutting, why not everyone do something?
Judging from the MLSSA data shown at Lynn's web site, the impulse response looks funny, and I would expect instruments like cymbals to sound a bit suppressed. The waterfall data seems to extend a bit long. I would aim for a goal to have a drop withing the first 0.1 ms of 12db. This way the recorded room reflections and timbre come out more naturally.
The next critical point is XO design. Lynn has his drivers connected in phase with each other. I find that this is the best configuration. All XO design needs be be based on this configuration. If the drivers are out of phase, no matter what XO you have, when the cymbals sound right, the drums sound a bit muffled, and vice versa.
Component selection is the next critical point. The previous need to be done well before you can really hear the benefits of components. The first pass in components selection can be done by using a passive first order filter driven by an amp. The components that give the lower distortion and harmonic contents are generally the better components, and you can categorize them and see if listening results are the same as measured results.
The final is internal wiring. This can also be done in a similar way as component selection.
Don't wait on others to start cutting, why not everyone do something?