It would be interesting to see data comparing coated with and without the pattern. I don¡¦t think I will be on that path since we have and will always be dealing with very stiff cones but it would be interesting to understand.BudP said:
Shipping processed drivers are probably more expensive than buying some simple equipment most DIYers use. I¡¦m not sure if I¡¦m getting the right understanding here. But exploring 3¡¨~2¡¨ wide range metal drivers show that weight and location is critical if the more significant issue is cone breakup. At this size allowable weight error is generally best controlled at no more than 25%. This changes the CSD and FR of the driver. The benefit of the patternized weight could be of benefit, which is something that I will look into a bit more, for metal cone only. Perhaps some other people are interested in cones of other material?BudP said:
I think this is actually best closely verified in simulation rather than a matrix of testing using FEA methods. Theoretically if we send an impulse through an edge and take data off different locations in the radiated pattern, we will probably find an altered impulse containing a different spectrum pattern. Hmm, looks like a good subject for a thesis. My instinct would be to make the reflected impulse look as flat as possible to spread the energy to the lower frequency range.BudP said:
Did you stick in the pattern at different speeds or did it occur at just one specific speed.BudP said:
I currently use SoundEasy, and it has been able to provide some pretty good information in the past year or so in both analysis and measurement capabilities. Hope it gets better year by year. If you don¡¦t have any measurement equipment, it¡¦s probably cheaper to get some kind of setup than ship the drivers around. But I¡¦d be glad to measure any ¡§metal¡¨ coned drivers, pay for return shipping, and provide data. Explanation of the data would be somewhat limited without authorization or some form of contractual relationship. The best system for the job I would suspect would be the Klippel system if someone can afford them. Every idea that I can think of they seem to have a solutions in the works, and cone vibration measurement and visualization is one of them. If I see wave radiation visualization, I think it would be perfect.BudP said:
I¡¦m sure these are not for metal cones. I¡¦ve tried a few that I could get my hands on, and none are good enough. The ones I haven¡¦t tried are either too expensive, or do not provide the necessary basic data.BudP said:
Well, I think objective measurements should be used on the type of cone subject for improvement, because the main purpose of the coating as you had mentioned before is to let the cone transmit the sound faster than air, which different material does differently due to the Young¡¦s modulus and the internal friction.BudP said:
Trying to understand
I might have an simpler analogy that might explain how this enabl concept works.
After some thinking it comes to me that this flexible conformal coating over the acrylic paint idea is similar to tennis racquet damper !!
It absorb and form a tunnel to trap vibration near the terminus of the cone, which result in reduced cone vibration/oscillation mode.
The conformal coating over the paper cone & the acrylic paint add the springiness and the enabl pattern add damping+tunnel trap that transforms cone oscillating(not well behaved) springiness to less oscillating(more well behaved) ones, more like a well behaved bending transducer.
If seen from this point of view, in the case of metal cone as BudP mentioned there's no need to apply the coating probably because the metal cone is already have certain degree of springiness that fits well in this application.
In case of extremely thin, light stiff cone ( diamond?? ) the effect of the acrylic paint is more subtle because the compliance of the cone and the acrylic paint and the conformal coating is very different, in this case the acrylic + conformal coating simply become double layer vibration damper/tunnel trap with slight compliance difference. but considering the lesser vibration that exist in this case. I think it still might provide audible benefit.
there might be a better analogy than this one, but this is the easiest one my brain presently can model after 2 days looking into this thread
does it work ? I was a bit skeptical when I first read it, but something interesting is kind of happening with this idea, so I try to give it thought.
If seen by analogy I presented above, it should work !! and should work especially well on relatively springy thin low mass cones. too heavy cone mass and the acrylic damper+trap will not work as well.
For speaker box/panel application where things is especially thick, it needs progressively larger (much larger) pattern toward the end/side and much more potent damper/trap material.
more potent absorbent than acrylic paint(damper/trap inner layer) should be used , the conformal coating+acrylic of the enabl pattern tcan be replaced with something else. 2 ,3 or more layer with increasing stiffness toward the outermost layer can be beneficial here, since mass is not an issue.
More potent damper+trap might also be needed with heavy stiff,not so springy cones.
Now is 3:25 a clock in the morning, I hope my brain still actually works and I just hope I didn't mess up with my babling .
I might have an simpler analogy that might explain how this enabl concept works.
After some thinking it comes to me that this flexible conformal coating over the acrylic paint idea is similar to tennis racquet damper !!
It absorb and form a tunnel to trap vibration near the terminus of the cone, which result in reduced cone vibration/oscillation mode.
The conformal coating over the paper cone & the acrylic paint add the springiness and the enabl pattern add damping+tunnel trap that transforms cone oscillating(not well behaved) springiness to less oscillating(more well behaved) ones, more like a well behaved bending transducer.
If seen from this point of view, in the case of metal cone as BudP mentioned there's no need to apply the coating probably because the metal cone is already have certain degree of springiness that fits well in this application.
In case of extremely thin, light stiff cone ( diamond?? ) the effect of the acrylic paint is more subtle because the compliance of the cone and the acrylic paint and the conformal coating is very different, in this case the acrylic + conformal coating simply become double layer vibration damper/tunnel trap with slight compliance difference. but considering the lesser vibration that exist in this case. I think it still might provide audible benefit.
there might be a better analogy than this one, but this is the easiest one my brain presently can model after 2 days looking into this thread
does it work ? I was a bit skeptical when I first read it, but something interesting is kind of happening with this idea, so I try to give it thought.
If seen by analogy I presented above, it should work !! and should work especially well on relatively springy thin low mass cones. too heavy cone mass and the acrylic damper+trap will not work as well.
For speaker box/panel application where things is especially thick, it needs progressively larger (much larger) pattern toward the end/side and much more potent damper/trap material.
more potent absorbent than acrylic paint(damper/trap inner layer) should be used , the conformal coating+acrylic of the enabl pattern tcan be replaced with something else. 2 ,3 or more layer with increasing stiffness toward the outermost layer can be beneficial here, since mass is not an issue.
More potent damper+trap might also be needed with heavy stiff,not so springy cones.
Now is 3:25 a clock in the morning, I hope my brain still actually works and I just hope I didn't mess up with my babling .
Hartono,
Your ideas are quite interesting. And one of them has allowed my own thinking to take another step.
I do not think this is quite correct, but it did cause me to recognize that the conformal coating does provide a more abrupt impedance change to the energy emitting from it into the air, as the energy travels across the cone, through the boundary layer.
The usual cone surface is actually a fairly gradual impedance change and the losses, while traveling through this gradient and expressing into the air, are probably the root cause of the local area ringing that would and does appear to cause corruptions across the untreated cone surface. Certainly you are correct about it being a "tunnel" but it is a tunnel that is emitting upward while allowing the energy to also traverse across the emitter within a much more uniform boundary layer.
If we pursue the tunnel analogy, modified slightly, it does make sense that a metal surface, with an already abrupt impedance change into the adjacent air, would not respond well to having the coating applied to it. In fact, for all but titanium, the coating only goes over the block zone, which does bolster your thought of it acting as a tunnel. Just add the third vector to your thought and I think your model hangs together quite well. On titanium there is no coating applied anywhere.
I think you need to reconsider your thoughts here. You are returning to an energy damping, by storage, in a trap mode. I think it will blind you to what does occur. Keep in mind that I am actually speaking from physical reality, events that always take place, not theory. The theory is lagging far behind the reality in this case.
The same, exact materials and in the same amounts, are used to control a hard surfaced box face.The treatment of a large soft material cone does require much more conformal coating, to actually create that abrupt impedance change boundary layer. Do try to view this pattern and coating as a method of flinging the energy off of the boundary layer surface, before it gets an opportunity to store in the boundary layer and ring as a standing wave.
The pattern and coating components, as I use them, are most definitely not a storage trap. Though, if it was only a two vector problem they certainly would be. But, there is that pesky third vector, out into the air, that is causing a differing set of thoughts to be needed.
Excellent thinking so far. Keep modeling it in your mind, no other tools are available. And, I wouldn't give up on the trap thoughts either, as they may lead you off to something different from the EnABL process and quite unique and effective too. You need only look at the Mamboni process to realize there are plenty of ways to do this job, that have not yet been thought up.
Bud
Your ideas are quite interesting. And one of them has allowed my own thinking to take another step.
I do not think this is quite correct, but it did cause me to recognize that the conformal coating does provide a more abrupt impedance change to the energy emitting from it into the air, as the energy travels across the cone, through the boundary layer.
The usual cone surface is actually a fairly gradual impedance change and the losses, while traveling through this gradient and expressing into the air, are probably the root cause of the local area ringing that would and does appear to cause corruptions across the untreated cone surface. Certainly you are correct about it being a "tunnel" but it is a tunnel that is emitting upward while allowing the energy to also traverse across the emitter within a much more uniform boundary layer.
If we pursue the tunnel analogy, modified slightly, it does make sense that a metal surface, with an already abrupt impedance change into the adjacent air, would not respond well to having the coating applied to it. In fact, for all but titanium, the coating only goes over the block zone, which does bolster your thought of it acting as a tunnel. Just add the third vector to your thought and I think your model hangs together quite well. On titanium there is no coating applied anywhere.
I think you need to reconsider your thoughts here. You are returning to an energy damping, by storage, in a trap mode. I think it will blind you to what does occur. Keep in mind that I am actually speaking from physical reality, events that always take place, not theory. The theory is lagging far behind the reality in this case.
The same, exact materials and in the same amounts, are used to control a hard surfaced box face.The treatment of a large soft material cone does require much more conformal coating, to actually create that abrupt impedance change boundary layer. Do try to view this pattern and coating as a method of flinging the energy off of the boundary layer surface, before it gets an opportunity to store in the boundary layer and ring as a standing wave.
The pattern and coating components, as I use them, are most definitely not a storage trap. Though, if it was only a two vector problem they certainly would be. But, there is that pesky third vector, out into the air, that is causing a differing set of thoughts to be needed.
Excellent thinking so far. Keep modeling it in your mind, no other tools are available. And, I wouldn't give up on the trap thoughts either, as they may lead you off to something different from the EnABL process and quite unique and effective too. You need only look at the Mamboni process to realize there are plenty of ways to do this job, that have not yet been thought up.
Bud
I was playing around with some LoudSoft FineCone software once, and it does present some interesting insight on what happens to the cone at different frequencies. Ted Jordan had published some nice articles about cones possibly in the 80's in WirelessWorld which also gives insight to stiffness of cones etc. If the surround is already properly designed, then the energy is properly dissipated through the surround, and the pattern will further enhance the performance. If the surround is not properly designed, the pattern might help, but not as much as in combination with the right mass distribution.
soongsc
I agree. I have never done this experiment. I have used other coating materials, including epoxies with a very high solids content. All of them just exaggerated either a fault or a lack in the drivers and this dissuaded me from pursuing coating only activities. Just providing a better place for the energy to ring within, as a boundary layer supported standing wave seemed unnecessary.
Please keep in mind that these blocks, actually round dots, with patterns this small, are very tiny, no more than two degrees long by one degree wide. In reality they are about 1 degree in diameter as my hands will not work a technical pen more precisely and round has always been the best I could do. In addition I must cut this high solids content paint by 50%, just to get it to pass through the pen tip. Then, if there is a need for conformal coating it is only over the pattern dots, not ever out onto the rest of the dome or cone surface. With Titanium, no coating is used over the pattern.
I think I need to stress that this is not a mass damping procedure, in the classical sense. What I am doing is damping the emitter surface and just up into the boundary layer, where the energy is emitting and forming a traveling pressure wave that is continuously fed by the energy moving, as a transverse wave across the dome or cone surface. I am certainly mass damping that tenuous boundary layer, and rather severely too, by forcing the energy within it to conform to an unnatural pattern to traverse, where the EnABL pattern blocks are. But I am not materially affecting the whole mass balance on the driver emitter itself. Even if the driver was 0.002"aluminum foil. I can see where this type of energy control might affect a very light driver panel, just through the movement of the energy itself and the sort of constraints to free surface excitation this sort of control applies.
A somewhat confusing periodicity was noted. There were perhaps three speeds that were most illustrative of the described wave action and there were dipping rates that produced nothing but chop, though that was diminished pretty severely beyond the pattern.
I was not overly concerned with this as the tank being used was a pretty poor excuse for a real wave tank and I was not modeling the energy loss into the air with anything like the accuracy needed to make this an even remotely scientific experiment. I was just attempting to get some kind of information, so that I could begin to think about what was already happening on treated speaker cones. This EnABL process did not arise through a coolly reasoned application of scientifically derived procedures, at all.
I do have a test set up based upon the Lincoln Audio Dos based test program. A calibrated microphone, instrumentation grade amplifier and a portable with an ISA card slot capable docking station that I can use. I doubt that this will work with the newer test software and the Lincoln suite has no CSD capabilities that I can discover. Also, I haven't used it in a number of years and the battery in the portable has died, so, there is that expense to overcome too. Perhaps someone has an old MLSSA system they will sell me for a price I can afford.
Bud
I agree. I have never done this experiment. I have used other coating materials, including epoxies with a very high solids content. All of them just exaggerated either a fault or a lack in the drivers and this dissuaded me from pursuing coating only activities. Just providing a better place for the energy to ring within, as a boundary layer supported standing wave seemed unnecessary.
Please keep in mind that these blocks, actually round dots, with patterns this small, are very tiny, no more than two degrees long by one degree wide. In reality they are about 1 degree in diameter as my hands will not work a technical pen more precisely and round has always been the best I could do. In addition I must cut this high solids content paint by 50%, just to get it to pass through the pen tip. Then, if there is a need for conformal coating it is only over the pattern dots, not ever out onto the rest of the dome or cone surface. With Titanium, no coating is used over the pattern.
I think I need to stress that this is not a mass damping procedure, in the classical sense. What I am doing is damping the emitter surface and just up into the boundary layer, where the energy is emitting and forming a traveling pressure wave that is continuously fed by the energy moving, as a transverse wave across the dome or cone surface. I am certainly mass damping that tenuous boundary layer, and rather severely too, by forcing the energy within it to conform to an unnatural pattern to traverse, where the EnABL pattern blocks are. But I am not materially affecting the whole mass balance on the driver emitter itself. Even if the driver was 0.002"aluminum foil. I can see where this type of energy control might affect a very light driver panel, just through the movement of the energy itself and the sort of constraints to free surface excitation this sort of control applies.
A somewhat confusing periodicity was noted. There were perhaps three speeds that were most illustrative of the described wave action and there were dipping rates that produced nothing but chop, though that was diminished pretty severely beyond the pattern.
I was not overly concerned with this as the tank being used was a pretty poor excuse for a real wave tank and I was not modeling the energy loss into the air with anything like the accuracy needed to make this an even remotely scientific experiment. I was just attempting to get some kind of information, so that I could begin to think about what was already happening on treated speaker cones. This EnABL process did not arise through a coolly reasoned application of scientifically derived procedures, at all.
I do have a test set up based upon the Lincoln Audio Dos based test program. A calibrated microphone, instrumentation grade amplifier and a portable with an ISA card slot capable docking station that I can use. I doubt that this will work with the newer test software and the Lincoln suite has no CSD capabilities that I can discover. Also, I haven't used it in a number of years and the battery in the portable has died, so, there is that expense to overcome too. Perhaps someone has an old MLSSA system they will sell me for a price I can afford.
Bud
Bud,
I think if your computer has a sound card, and can run Windows, perhaps the best way to get a feeling whether it will work or not is getting some freeware or demo software (can't save data). A few that come to mind are SpeakerWorkshop (freeware), ARTA (demo) this will at least get you started. If you've been doing some tests before, then you probably have the interface and hardware it takes.
I've done many tests on metal cones using toothpaste mainly because I could do multiple rounds of tests in a short time to figure out what worked, what not, and what the trends are. Some patterns were dotted as well. As a matter of fact, I'm going back through my data to see if there are trends I missed in light of the discussion here. This is going to be interesting.
I think if your computer has a sound card, and can run Windows, perhaps the best way to get a feeling whether it will work or not is getting some freeware or demo software (can't save data). A few that come to mind are SpeakerWorkshop (freeware), ARTA (demo) this will at least get you started. If you've been doing some tests before, then you probably have the interface and hardware it takes.
I've done many tests on metal cones using toothpaste mainly because I could do multiple rounds of tests in a short time to figure out what worked, what not, and what the trends are. Some patterns were dotted as well. As a matter of fact, I'm going back through my data to see if there are trends I missed in light of the discussion here. This is going to be interesting.
Just been talking on the phone with BudP, and the conversation got some ideas going.
Most people don't understand that direct-radiator drivers are constant-acceleration devices, at least in the piston-band where they are inherently flat. This means excursion increases at a rate of 12 dB/octave as frequency decreases, or to re-state that, 40 dB/decade.
That why you don't see the cone move at higher frequencies - it would take a microscope, or more accurately, a laser holography setup. Not only that, it isn't velocity you're interested in - it's acceleration. Acceleration of the cone is what produces sound, not velocity.
So the actual mechanism of sound production is a shock wave that propagates along the length of the voice-coil former, and this shock wave is then propagated to the base of the cone, where it creates a shear wave that propagates outward. All nice in theory, but there are reflections at the VC former/cone junction, the dust cap (if it isn't at the VC junction), and most important of all, the edge of cone where it meets the surround.
The speed of sound in the cone is much faster than air and the cone is much denser than air; this results in very inefficient energy transfer to the surrounding medium. The energy pulse moving though the cone tends to stay in the cone - it is important to remember the efficiency of a direct-radiator loudspeaker is between 0.3% (audiophile) and 5% (prosound).
Our problem is the other 95 to 99% of the rest of the energy put into the system. Most of it heats the voice-coil, but the cone itself is a very inefficient mechano-acoustic coupler.
I see the EnABL pattern (and I surmise there are many others possible) as de-cohering the wavefront that strikes the edge of the cone, randomizing the reflected energy of the shockwave before and after it strikes the edge. The edge itself reflects almost all of the energy - although a well-chosen rubber surround obviously helps.
This reflected energy, if left alone, causes lots of trouble. The cone itself is not a linear transmission medium, and being symmetric, is prone to strong standing waves. Since the cone is such an inefficient radiator, the air load is quite ineffective as a damping medium.
This is different than a stretched-film electrostat, or a freely hanging ribbon tweeter. With diaphragms as light as these (relative to the air-load), the coupling to the air is good enough to provide a fair amount of damping - thus the clean transient response these transducers are known for. The reason these transducers are inefficient isn't the air coupling, but the lossy drive method - capacitive coupling, or a very low BL product.
So with a cone radiator, almost all of the damping is provided by internal losses within the cone, not the air load. If EnABL provides even a slight difference in coupling efficiency over a small part of the cone, that will have a major effect on a system that is deficient in losses - just as a small amount of acoustic absorber has a large effect in a resonant room.
BudP also mentioned the interesting point that acoustic emission is mostly confined to the surface of the cone, while the shear wave is a bulk property of the entire thickness of the cone. So there are questions about how this shear wave propagates not just outward towards the edge of cone, but how how it moves from the interior of cone to the emissive surface, where the radiating happens.
If the surface is a skin of some kind, all the emission will be from that surface. If the surface is a fibrous tangle, then emission will happen in depth, from all of the fibers that are exposed to air. Unlike a skin that is a contiguous surface (and probably has a different speed of sound from the underlying substrate), a fibrous tangle is only loosely coupled to the rest of the cone, with every fiber free to move on its own.
This is why it is an old idea to dip the center region of the cone in a lacquer, forming a skin on the surface - but only in the area close to the voice coil. This acts as a mechanical crossover, with the center region having a semi-rigid surface with high emission (but low damping), and the outer region made from loosely woven paper, with high losses and less efficient emission.
So even if you measure the cone with a laser holography system that displays acceleration on the surface of the cone, a simple reflective treatment to assist the LH system get a good optical reflection could also, and probably would, affect the emissive properties of the cone.
I am guessing the painted-on EnABL treatment will have the greatest effect on metal cones and lossy-paper cones. With both types, the paint surface adds a thin surface with substantially different emissive properties than rigid metal or loose and fibrous paper. I'm sure it makes a difference on plastic cones, but the change in emission is probably smaller.
What is an emissive surface? It has two properties of interest: it is free to accelerate while in contact with the acoustic medium (air), while also passing on the mechanical shear wave to adjacent emissive surfaces.
Most people don't understand that direct-radiator drivers are constant-acceleration devices, at least in the piston-band where they are inherently flat. This means excursion increases at a rate of 12 dB/octave as frequency decreases, or to re-state that, 40 dB/decade.
That why you don't see the cone move at higher frequencies - it would take a microscope, or more accurately, a laser holography setup. Not only that, it isn't velocity you're interested in - it's acceleration. Acceleration of the cone is what produces sound, not velocity.
So the actual mechanism of sound production is a shock wave that propagates along the length of the voice-coil former, and this shock wave is then propagated to the base of the cone, where it creates a shear wave that propagates outward. All nice in theory, but there are reflections at the VC former/cone junction, the dust cap (if it isn't at the VC junction), and most important of all, the edge of cone where it meets the surround.
The speed of sound in the cone is much faster than air and the cone is much denser than air; this results in very inefficient energy transfer to the surrounding medium. The energy pulse moving though the cone tends to stay in the cone - it is important to remember the efficiency of a direct-radiator loudspeaker is between 0.3% (audiophile) and 5% (prosound).
Our problem is the other 95 to 99% of the rest of the energy put into the system. Most of it heats the voice-coil, but the cone itself is a very inefficient mechano-acoustic coupler.
I see the EnABL pattern (and I surmise there are many others possible) as de-cohering the wavefront that strikes the edge of the cone, randomizing the reflected energy of the shockwave before and after it strikes the edge. The edge itself reflects almost all of the energy - although a well-chosen rubber surround obviously helps.
This reflected energy, if left alone, causes lots of trouble. The cone itself is not a linear transmission medium, and being symmetric, is prone to strong standing waves. Since the cone is such an inefficient radiator, the air load is quite ineffective as a damping medium.
This is different than a stretched-film electrostat, or a freely hanging ribbon tweeter. With diaphragms as light as these (relative to the air-load), the coupling to the air is good enough to provide a fair amount of damping - thus the clean transient response these transducers are known for. The reason these transducers are inefficient isn't the air coupling, but the lossy drive method - capacitive coupling, or a very low BL product.
So with a cone radiator, almost all of the damping is provided by internal losses within the cone, not the air load. If EnABL provides even a slight difference in coupling efficiency over a small part of the cone, that will have a major effect on a system that is deficient in losses - just as a small amount of acoustic absorber has a large effect in a resonant room.
BudP also mentioned the interesting point that acoustic emission is mostly confined to the surface of the cone, while the shear wave is a bulk property of the entire thickness of the cone. So there are questions about how this shear wave propagates not just outward towards the edge of cone, but how how it moves from the interior of cone to the emissive surface, where the radiating happens.
If the surface is a skin of some kind, all the emission will be from that surface. If the surface is a fibrous tangle, then emission will happen in depth, from all of the fibers that are exposed to air. Unlike a skin that is a contiguous surface (and probably has a different speed of sound from the underlying substrate), a fibrous tangle is only loosely coupled to the rest of the cone, with every fiber free to move on its own.
This is why it is an old idea to dip the center region of the cone in a lacquer, forming a skin on the surface - but only in the area close to the voice coil. This acts as a mechanical crossover, with the center region having a semi-rigid surface with high emission (but low damping), and the outer region made from loosely woven paper, with high losses and less efficient emission.
So even if you measure the cone with a laser holography system that displays acceleration on the surface of the cone, a simple reflective treatment to assist the LH system get a good optical reflection could also, and probably would, affect the emissive properties of the cone.
I am guessing the painted-on EnABL treatment will have the greatest effect on metal cones and lossy-paper cones. With both types, the paint surface adds a thin surface with substantially different emissive properties than rigid metal or loose and fibrous paper. I'm sure it makes a difference on plastic cones, but the change in emission is probably smaller.
What is an emissive surface? It has two properties of interest: it is free to accelerate while in contact with the acoustic medium (air), while also passing on the mechanical shear wave to adjacent emissive surfaces.
This thread is getting close to driver cone design. But I think most of the effeciency loss is in the motor, which is beyond the subject of this thread.
I do notice some things through the impulse response due to additional patterns added to the cone. But since my system had gone through various hardware changes in the last year, it's a little difficult to compare some data. But it does seem that different thicknesses do have a slight different effect at different time locations in an impulse.
I also recall a picture somewhere in this forum that shows a MANGER driver monted on a round disk with a convex->concave ring pattern also showing data now it corrects a square wave response. So I guess it is clear that certain patterns will have some positive effects when implemented appropriately.
One should also realize that the cone is not a ridgid body, and any curves or deformation in the cone will cause the effective mass to change dynamically at different frequencies. Thus acceleration capability will be different. Ted Jordan had gone through the basics of this in his articles. I think I'll take a JX92S driver for a run the next time I get in the lab.
About painting of the cone, well there are many things involved, but again as Ted had mentioned about his own drivers, the coating is used to control the cone vibration modes. Based on my own experiments, on metal cones, the coating plays more roll in dissipating energy by using material with a more internally resistive nature, this can be done on the whole cone, or at the terminating edge pretty close to where the EnAble pattern is located, or in proper combination. The location depends on the bandwidth of the driver.
But I think most of the effeciency loss is in the motor, which is beyond the subject of this thread.I do notice some things through the impulse response due to additional patterns added to the cone. But since my system had gone through various hardware changes in the last year, it's a little difficult to compare some data. But it does seem that different thicknesses do have a slight different effect at different time locations in an impulse.
I also recall a picture somewhere in this forum that shows a MANGER driver monted on a round disk with a convex->concave ring pattern also showing data now it corrects a square wave response. So I guess it is clear that certain patterns will have some positive effects when implemented appropriately.
One should also realize that the cone is not a ridgid body, and any curves or deformation in the cone will cause the effective mass to change dynamically at different frequencies. Thus acceleration capability will be different. Ted Jordan had gone through the basics of this in his articles. I think I'll take a JX92S driver for a run the next time I get in the lab.
About painting of the cone, well there are many things involved, but again as Ted had mentioned about his own drivers, the coating is used to control the cone vibration modes. Based on my own experiments, on metal cones, the coating plays more roll in dissipating energy by using material with a more internally resistive nature, this can be done on the whole cone, or at the terminating edge pretty close to where the EnAble pattern is located, or in proper combination. The location depends on the bandwidth of the driver.
Lynn, soongsc and all,
Boy, it would be nice to be able see how that happens. From the two very short interferometry looks I had, a number of years ago, it does appear that most of the energy has excurted from the cone, before the EnABL pattern is reached. Almost as if the inner and outer pattern ring were creating a large scale quantum signaling event that always provides only enough energy left, after the EnABL process, to exit properly into the surround and whatever other sinks are available. I realize this is a foolish notion, but I was quite serious about igniting voice coils, Kapton former voice coils with Nomex insert no less, and not really witnessing anything more than a flattening of correlatable perspective information, as a signal of the drivers distress.
It provided up to 9 dB of improvement in the tweeter under test and displayed on the EnABL paper at Positive Feedback Online. This was a pretty egregious example of a cone radiator, so quantitative improvement will likely be less in modern drivers, but the insulting thing ,to me, is that the qualitative improvement is still about the same.
I hope this thread gets into a huge debate about drivers and their weaknesses. Not so that anyone is belittled or embarrassed but so that everyone participating goes off with some new things to investigate. Maybe someone out there will get that flash that takes a step beyond the mental trench we are currently mired in for driver design and performance.
Certainly companies like RAAL, Manger, the radiostraller weirdness, Jordan and even some major players are beginning to accept that what is current status quo has about run it's course.
Nothing would please me more than to never have to EnABL another driver, just so I can stand to listen to it and have intelligible music come from it, that pleases my taste for clarity, performance space and the musicianship, that works that space, to entangle my emotions and provide what music provides in real live performances, connection to ART.
Boy, it would be nice to be able see how that happens. From the two very short interferometry looks I had, a number of years ago, it does appear that most of the energy has excurted from the cone, before the EnABL pattern is reached. Almost as if the inner and outer pattern ring were creating a large scale quantum signaling event that always provides only enough energy left, after the EnABL process, to exit properly into the surround and whatever other sinks are available. I realize this is a foolish notion, but I was quite serious about igniting voice coils, Kapton former voice coils with Nomex insert no less, and not really witnessing anything more than a flattening of correlatable perspective information, as a signal of the drivers distress.
It provided up to 9 dB of improvement in the tweeter under test and displayed on the EnABL paper at Positive Feedback Online. This was a pretty egregious example of a cone radiator, so quantitative improvement will likely be less in modern drivers, but the insulting thing ,to me, is that the qualitative improvement is still about the same.
I hope this thread gets into a huge debate about drivers and their weaknesses. Not so that anyone is belittled or embarrassed but so that everyone participating goes off with some new things to investigate. Maybe someone out there will get that flash that takes a step beyond the mental trench we are currently mired in for driver design and performance.
Certainly companies like RAAL, Manger, the radiostraller weirdness, Jordan and even some major players are beginning to accept that what is current status quo has about run it's course.
Nothing would please me more than to never have to EnABL another driver, just so I can stand to listen to it and have intelligible music come from it, that pleases my taste for clarity, performance space and the musicianship, that works that space, to entangle my emotions and provide what music provides in real live performances, connection to ART.
soongsc said:This thread is getting close to driver cone design.
OK...been following this thread with interest...but now that someone has breached the subject, I will offer the following:
French DIY Linaeum driver
An externally hosted image should be here but it was not working when we last tested it.
This could also go in Lynn's thread, given its high efficiency and dipole configuration. I'm going to build this thing sometime this year (I swear), but was wondering what BudP's thoughts would be on this design. My current thinking is to go with a 25% cotton paper membrane that has been impregnated with Dammar varnish - or something similar.
Mr. Paddock's patents are out there for guidance.
Please elaborate. Is this a Linnaeum device? Some sort of live wall horn, which I have some pretty serious thoughts on?
I, being most ignorant, am not familiar with your reference, or even really how to do much of a patent search as Boolien logic just twists my brain into a knot.
What ever it is I am interested! I suspect Lynn will be also and perhaps Alex from RAAL as the two of them are off plotting in the shadows currently. There is also Moray James from Highwood? fame lurking about ready to contribute. And this doesn't even mention the other folks brave enough to step into this den of mad scie......fools. Perhaps the most needed here is C2CThomas, from over on the Walsh thread most of the time. Very clever man and quite skilled with tools of all shapes and sizes.
Please expand, I am quite interested.
Bud
BudP said:
Yes, it's a DIY Linnaeum full ranger(see link above the image for more details and images).
Here are a couple of Paul Paddock's patents relating to it:
original patent
later "improvement" patent.
Well, I'm almost positive that the 9db notch in the untreated driver posted in Positive Feedback really is due to the edge and the center moving in inversed directions. This does happen when the shape and stiffness of the cone is not well designed. There are lot more things involved in cone shaping and material selection than most may think. The pattern can smooth high frequency response as long as they are not caused by cone breakup.
Some drivers use concaved surrounds which make me think it possibly helps in similar ways as the EnABL Pattern. Surround material and weight effects the cone breakup modes and acts somewhat like an impedance transition and energy dissipation device between the cone and the basket.
Some drivers use concaved surrounds which make me think it possibly helps in similar ways as the EnABL Pattern. Surround material and weight effects the cone breakup modes and acts somewhat like an impedance transition and energy dissipation device between the cone and the basket.
pedroskova,
Ah yes Linnaeum's. I missed hearing the full range units that were built in the factory in Oregon, by a few months. We sold them the non satruable reactors they used in the cross overs for these drivers.
I have all of the left over Radio Shack (made in Japan by the folks who purchased the Linnaeum rights and spirited all of the drivers etc off before I could travel down there) upgrade tweeter kits left, west of the Mississippi, but could not work a deal for the ones east of the big river. I use these tweeters from 4K to their natural roll off at 13 k and once EnABL'ed I doubt there is a more musical and dynamic device made, though I am sure RAAL will have something to say about this region and lower, before the dust settles.
That French unit looks very seductively interesting, but I might try to find a polycarbonate film to use as a first choice. My second choice would be a woven silk fabric, with stiffener, as is used in the Vifa mid dome and many tweeters. Both of these materials will have enough self rigidity to eliminate the need for the fiber fill to help keep the leaf's from drooping. Likely the paper being made by Chinese over on the My 14" thread, in full range, would also work as he is using a very thin and rigid fiber based paper, possibly chopped silk fiber with a bonding agent.
The EnABL process, on the tweeters, is applied to both sides of each leaf, right at the clamping edge for the frame and mounting edge for the voice coil. The gloss coat is applied just over the pattern lines and a round, toothpick middle sized, dot of PVA is placed dead center, in both dimensions, on the outside, of each leaf, to expand the dispersion and eliminate the slight can resonance peaking there. This dot is also covered with one coat of gloss. The pattern dots are tiny here, definitely technical pen work only, and no dots are placed on the long edges of the leaf as they make the curved surface.
Zen Mod,
sSSHHHhhhhhhh! Wizards are best left to their own devices, until ready to spring them on us...... Alex assures me he will have something "interesting" very soon....
soongsc,
I would hope you are correct about the 9 dB fill.
The application of the Gloss is pretty critical on high frequency drivers. You can easily increase their output beyond what the other drivers can match, so I always use as little Gloss as possible. And this is actually true of domes on cone drivers too. Just one coat will be enough, if that much, even if three coats is correct for the cone portion.
My own take on an ideal driver would see a voice coil applied to the surface of two almost complete spherical tubes of polycarbonate, right at the point where they would touch and flatten against each other, in their center, like a Linnaeum without a separate voice coil slab. A little glue to hold them together there, some strong neo magnets up close and EnaBL out at the clamped ends of the almost circles.
Would eliminate one of the biggest driver problems of joining dissimilar materials with glue and transferring the shear wave from the voice "coil" to the emitter surface. The glue used to fix the two surfaces could be a damping material, just to keep the voice coil from ringing at it's natural frequencies. Maybe EnABL would not even be needed.....
Bud
Ah yes Linnaeum's. I missed hearing the full range units that were built in the factory in Oregon, by a few months. We sold them the non satruable reactors they used in the cross overs for these drivers.
I have all of the left over Radio Shack (made in Japan by the folks who purchased the Linnaeum rights and spirited all of the drivers etc off before I could travel down there) upgrade tweeter kits left, west of the Mississippi, but could not work a deal for the ones east of the big river. I use these tweeters from 4K to their natural roll off at 13 k and once EnABL'ed I doubt there is a more musical and dynamic device made, though I am sure RAAL will have something to say about this region and lower, before the dust settles.
That French unit looks very seductively interesting, but I might try to find a polycarbonate film to use as a first choice. My second choice would be a woven silk fabric, with stiffener, as is used in the Vifa mid dome and many tweeters. Both of these materials will have enough self rigidity to eliminate the need for the fiber fill to help keep the leaf's from drooping. Likely the paper being made by Chinese over on the My 14" thread, in full range, would also work as he is using a very thin and rigid fiber based paper, possibly chopped silk fiber with a bonding agent.
The EnABL process, on the tweeters, is applied to both sides of each leaf, right at the clamping edge for the frame and mounting edge for the voice coil. The gloss coat is applied just over the pattern lines and a round, toothpick middle sized, dot of PVA is placed dead center, in both dimensions, on the outside, of each leaf, to expand the dispersion and eliminate the slight can resonance peaking there. This dot is also covered with one coat of gloss. The pattern dots are tiny here, definitely technical pen work only, and no dots are placed on the long edges of the leaf as they make the curved surface.
Zen Mod,
sSSHHHhhhhhhh! Wizards are best left to their own devices, until ready to spring them on us...... Alex assures me he will have something "interesting" very soon....
soongsc,
I would hope you are correct about the 9 dB fill.
The application of the Gloss is pretty critical on high frequency drivers. You can easily increase their output beyond what the other drivers can match, so I always use as little Gloss as possible. And this is actually true of domes on cone drivers too. Just one coat will be enough, if that much, even if three coats is correct for the cone portion.
My own take on an ideal driver would see a voice coil applied to the surface of two almost complete spherical tubes of polycarbonate, right at the point where they would touch and flatten against each other, in their center, like a Linnaeum without a separate voice coil slab. A little glue to hold them together there, some strong neo magnets up close and EnaBL out at the clamped ends of the almost circles.
Would eliminate one of the biggest driver problems of joining dissimilar materials with glue and transferring the shear wave from the voice "coil" to the emitter surface. The glue used to fix the two surfaces could be a damping material, just to keep the voice coil from ringing at it's natural frequencies. Maybe EnABL would not even be needed.....
Bud
LOWTHER
Hi all. Here are the first set of treatments. The larger rings only as I find I am missing a pen point size and must go hunting. The pen used for all of these blocks is a Speed Ball A3 pen tip. The paint is a tan pigmented, uncut, Poly S flat acrylic paint. The upper whizzer ring will need an A4 and it looks like A5 will work for the lower whizzer ring. Then comes the high wire act.... the lower main cone ring.
By the way Jon Ver Halen from Lowther America emailed me to encourage me on and provide some operational insight into the drivers. His recommendation is to not Gloss coat the area at the very bottom of the whizzer, as that is the "ring radiator" working above 13 kHz and he is concerned the Gloss coat material will have an unpredictable effect here. I will approach this area with caution and apply the lower whizzer pattern up above the ends of the triangular peaks found on the whizzer surface. Thanks Jon!
Please note the wavery line control for the blocks. This is not the fault of the pen, rather that of the operator. My nervous system is not a steady one, most of you will do a better job than this. However, the end results will be the same, so if you are as jumpy as I, do not fret.
http://picasaweb.google.com/hpurvine/LowthwerDX4EnABLPatterns
Bud
Hi all. Here are the first set of treatments. The larger rings only as I find I am missing a pen point size and must go hunting. The pen used for all of these blocks is a Speed Ball A3 pen tip. The paint is a tan pigmented, uncut, Poly S flat acrylic paint. The upper whizzer ring will need an A4 and it looks like A5 will work for the lower whizzer ring. Then comes the high wire act.... the lower main cone ring.
By the way Jon Ver Halen from Lowther America emailed me to encourage me on and provide some operational insight into the drivers. His recommendation is to not Gloss coat the area at the very bottom of the whizzer, as that is the "ring radiator" working above 13 kHz and he is concerned the Gloss coat material will have an unpredictable effect here. I will approach this area with caution and apply the lower whizzer pattern up above the ends of the triangular peaks found on the whizzer surface. Thanks Jon!
Please note the wavery line control for the blocks. This is not the fault of the pen, rather that of the operator. My nervous system is not a steady one, most of you will do a better job than this. However, the end results will be the same, so if you are as jumpy as I, do not fret.
http://picasaweb.google.com/hpurvine/LowthwerDX4EnABLPatterns
Bud
easy to deal with ...
just place the damping in an acoustically open form inside the leaf of the driver, or make a cylinder of thin felt and fix that to the frame and stuff the inside of the felt cylinder with wool fluff. There are any number of ideas that will do the job.
I had thought about using silk screen mesh rather than mylar film for a driver like this. I saw the first patent in the early 1980's. A fine mesh diaphragm material would support significant internal loss of diaphragm resonances and be just as easy to work with as solid mylar film. Regards Moray James.
just place the damping in an acoustically open form inside the leaf of the driver, or make a cylinder of thin felt and fix that to the frame and stuff the inside of the felt cylinder with wool fluff. There are any number of ideas that will do the job.
I had thought about using silk screen mesh rather than mylar film for a driver like this. I saw the first patent in the early 1980's. A fine mesh diaphragm material would support significant internal loss of diaphragm resonances and be just as easy to work with as solid mylar film. Regards Moray James.
- Status
- Not open for further replies.