Most aluminum and mag cones exhibit severe resonances, the first of which typically occurs at 4 - 7 kHz. Below, waterfalls usually look way better than any paper, polyprop, hard paper, kevlar, hexacone whatever material.
Aluminum will have strong secondary resonances, whereas magnesium only has one dominant one, and higher orders are pretty benign. Why?
Does anybody know whether the dominant resonance is tangential or radial, i.e. whether the standing wave pattern is made up of concentric rings or rays originating from the center?
This would help to decide what the optimum shape and orientation bits of material (paper, fibreglass) should be that one could glue to the back side of the cone to suppres the dominant resonance. Anybody know why this is not being done by the manufacturers?
Regards,
Eric
Aluminum will have strong secondary resonances, whereas magnesium only has one dominant one, and higher orders are pretty benign. Why?
Does anybody know whether the dominant resonance is tangential or radial, i.e. whether the standing wave pattern is made up of concentric rings or rays originating from the center?
This would help to decide what the optimum shape and orientation bits of material (paper, fibreglass) should be that one could glue to the back side of the cone to suppres the dominant resonance. Anybody know why this is not being done by the manufacturers?
Regards,
Eric
Hi
see this tread, someone talk about damping rubber band.
http://www.diyaudio.com/forums/showthread.php?threadid=20067&highlight=
F
see this tread, someone talk about damping rubber band.
http://www.diyaudio.com/forums/showthread.php?threadid=20067&highlight=
F
I think it is because:
1- Magnesium has better internal damping (if i remember well my engineering courses)
2- Since mag is lighter than alu, Seas made the cone thicker and it is less resonant (from the Thor article by Joe D'appolito)
F
The first series of Excel magnesium drivers had a break up pattern similar to the alu cones with multiple peaks.
/Peter
/Peter
Doc B. from bottlehead fame ofcourse coated his drivers with shellac.
That is what I shall be doing.
These days he coats with a butyl coating.
Cheers,
Bas
Someone on this forum also suggested "puzzlecoat" not a familiar term over here in europe..
That is what I shall be doing.
These days he coats with a butyl coating.
Cheers,
Bas
Someone on this forum also suggested "puzzlecoat" not a familiar term over here in europe..
Yes, because the cure is usually worse than the disease. If it were just a matter of putting some goop on the cone, smart manufacturers like Seas would do that.
The best way to use these drivers is to have a steep rolloff well below the first big breakup, and use trap filters if necessary.
Does it lower efficiency or something? I'm guessing it would.
What about coating the rear of an already coated paper midbass driver, would that make it sound softer and warmer, even if there are no large resonance peaks? (Seas H571/W17PPI)
I think I may be punished for saying that... 🙄
-Simon
What about coating the rear of an already coated paper midbass driver, would that make it sound softer and warmer, even if there are no large resonance peaks? (Seas H571/W17PPI)
I think I may be punished for saying that... 🙄
-Simon
With metal cones, it probably won't have much effect until it's really glopped on. With enough glopping, the Q will be higher, fs will be lower, efficiency will be lower, and a whole new set of resonances will "replace" the old one.
Goop is more effective, in general, for cones that are made of pourous materials like paper or cloth.
Goop is more effective, in general, for cones that are made of pourous materials like paper or cloth.
Right this may get flamed but its what it sounds like aint it.
Anyway if you have read any of my previous posts here you know I use the KX drivers on the SB cards ie DSP xovers.
Now right at this moment (im listening to it 🙂) I have a three way going using the KX.
Tweet SSd2905/95
Mid Excel w15cy001
Bass Peerless 850146
Right Xovers centred at 3500hz and 150 hz
Now. I have the EQ peaking filter set to remove the res peak of the SEAS, I am also using 24db electrical on it, cant do 18 🙁 as I wanted anyway. So technically at about 8000hz the peak is already down by about 30 dB anyway due to the xover. BUT if I remove the peaking, by clicking bypass (this is all in real time so I can switch it in and out at will as im listening) I cant hear a difference at all, just on a quick AB.
I played some Mozart and Anastacia's freak of nature to see if differnt music would highlight the peak. None did.
If you deactivate the tweeter and then swtich in and out the peak you can hear the difference, buts its only quiet small.
This is with the W15 tho, small driver 24db xover and peak high.
If this were the W18 however with its peak at 5000hz the story mite be different.
If you have a soundblaster with the EMU10k chip on it I would strongly suggest using the KX with a coupla power amps and trying it for yourself to see what you can and cannot hear.
At the end of the day if you cant hear it why suppress it. Now Ok I just said I couldnt hear it and I am building an acitve xover to replace the KX as its not the best in hifidelity but its not half as bad as ppl would make it out to be. And I am including a res trap.
Buts its very easy to sort out with an active so I added it.
Its just trial and error, which unfortunatly us DIYers cant do easily.
Making ten different passive xovers for one speaker to see what sounds the best isnt practical, but it sure would help.
I think trial and error would help alot of ppl if it could be done. I see loads of posts that say, what xover freq should I choose, is the res peak audible, is baffle step nescessary. Now we want to get it right 1st time so we only have to purchase a minimum of parts, but experimentation is the key. For no logical reason one frquency mite sound better then others even tho two drivers would work well at five others. The KX helps a lot of this, as do some loudspeaker design packages that include real time simulation of what the xover will sound like, but the software costs, but the KX is free!
Anyway hope that helps
Matt
I ranted again oh well.
Anyway if you have read any of my previous posts here you know I use the KX drivers on the SB cards ie DSP xovers.
Now right at this moment (im listening to it 🙂) I have a three way going using the KX.
Tweet SSd2905/95
Mid Excel w15cy001
Bass Peerless 850146
Right Xovers centred at 3500hz and 150 hz
Now. I have the EQ peaking filter set to remove the res peak of the SEAS, I am also using 24db electrical on it, cant do 18 🙁 as I wanted anyway. So technically at about 8000hz the peak is already down by about 30 dB anyway due to the xover. BUT if I remove the peaking, by clicking bypass (this is all in real time so I can switch it in and out at will as im listening) I cant hear a difference at all, just on a quick AB.
I played some Mozart and Anastacia's freak of nature to see if differnt music would highlight the peak. None did.
If you deactivate the tweeter and then swtich in and out the peak you can hear the difference, buts its only quiet small.
This is with the W15 tho, small driver 24db xover and peak high.
If this were the W18 however with its peak at 5000hz the story mite be different.
If you have a soundblaster with the EMU10k chip on it I would strongly suggest using the KX with a coupla power amps and trying it for yourself to see what you can and cannot hear.
At the end of the day if you cant hear it why suppress it. Now Ok I just said I couldnt hear it and I am building an acitve xover to replace the KX as its not the best in hifidelity but its not half as bad as ppl would make it out to be. And I am including a res trap.
Buts its very easy to sort out with an active so I added it.
Its just trial and error, which unfortunatly us DIYers cant do easily.
Making ten different passive xovers for one speaker to see what sounds the best isnt practical, but it sure would help.
I think trial and error would help alot of ppl if it could be done. I see loads of posts that say, what xover freq should I choose, is the res peak audible, is baffle step nescessary. Now we want to get it right 1st time so we only have to purchase a minimum of parts, but experimentation is the key. For no logical reason one frquency mite sound better then others even tho two drivers would work well at five others. The KX helps a lot of this, as do some loudspeaker design packages that include real time simulation of what the xover will sound like, but the software costs, but the KX is free!
Anyway hope that helps
Matt
I ranted again oh well.
Bas Horneman said:
Puzzlecoat is a paint that is designed to (I assume) shrink slightly when it dries leaving a randomly cracked surface appearance, like what happens to normal paint after years of exposure.
I'm pretty sure puzzlecoat is a flexible clear coating that you paint on top of a jigsaw puzzle to preserve it once you have finished that is available where you buy these puzzles.
Thay way the little pieces stay together. Then you can hang it on the wall and fool your friends into thing that you have the original "Last Supper" fresco by Leonardo.
I could be wrong but I'd suggest that you find out before attempting a substitute!!! There's gonna be a big difference between a flexible coating and one the breaks into a bunch of cracks!
Thay way the little pieces stay together. Then you can hang it on the wall and fool your friends into thing that you have the original "Last Supper" fresco by Leonardo.
I could be wrong but I'd suggest that you find out before attempting a substitute!!! There's gonna be a big difference between a flexible coating and one the breaks into a bunch of cracks!
Variac said:
Exactly... puzzlecoat is actually a brand name that i 1st started using in the late 70s. I haven't seen it around for some time.
It is a version of PVA glue (white glue) and people will often use that with good results. It drys clear & flexible. I have a page with pictures and links.
http://www.t-linespeakers.org/design/tweeks.html
I've used this since almost forever on paper cone drivers -- particularily bass drivers, but also had success killing the (fairly small) resonances in the MCM55-1855 (shielded version of Doc Bottlehead driver).
http://www.t-linespeakers.org/drivers/MCM55-1855.html
Another fellow did a series of studies with treating the same driver, and found that "puzzlecoat" produced the best results.
dave
Rubber cement. As it comes from the bottle, it's too thick to spread evenly, so you need to dilute it. It's dead. It doesn't crack. It's cheap. It's readily available. Good stuff for experimentation purposes.
No, I've never tried it on metal cones. The only one I've got is the aluminum one in my Hartke cabinet, and I like it just fine the way it is.
Keep in mind that no matter what you put on the cone, you're adding mass. The T-S parameters will change accordingly.
Grey
No, I've never tried it on metal cones. The only one I've got is the aluminum one in my Hartke cabinet, and I like it just fine the way it is.
Keep in mind that no matter what you put on the cone, you're adding mass. The T-S parameters will change accordingly.
Grey
I want to encourage “diy”ers to experiment with canceling, altering, and damping loudspeaker cone resonances. This is a problem we are capable of addressing. I will admit that some problems are more difficult than others and that loudspeaker driver diaphragm resonance control is one of the more difficult problems. We have known about it for a long time. We have been working on it for a long time. We have had some successes and we have had some failures, and we have a way yet to go. It is an area of design that is still ripe for discoveries.
The “diy” world has been concerned with this problem for a long time. The second issue of Audio Amateur led off with a loudspeaker construction article by Peter Baxandall featuring two parallel resonant filters in series with the driver. The next year, 2 (3), featured an interview with Henry Kloss, where he briefly talks about loudspeaker driver diaphragm resonance and comments on work being done to structure the resonant response through shape and material selection. Kloss also talked about using resonant filters to control a resonance around 1 kHz in 5 inch full-range loudspeaker drivers.
The commercial world has also been concerned and has sometimes committed large sums of capital to address the problem. In the early 1980s Celestion developed a laser interferometry system to show (in animation) the complex movements of diaphragms at fixed frequencies. The first product produced with the technology was the SL-6. The SL-6 tweeter was a 1-¼ inch copper alloy dome tweeter that they claimed was without resonant break-up. I measured an early version of the loudspeaker and found high Q resonant structures. By the time J. Peter Moncrieff of IAR reviewed the loudspeaker, a resonant filter was added to the crossover to tame the major dome break-up resonant. That, however, is all it did. It tamed it but did not eliminate it. Just goes to show that even commercial concerns sometime resort to techniques available to us “diy”ers.
Celestion and others are still interested in controlling loudspeaker driver diaphragm and cabinet resonances. Instead of adding things on the outside of the cones, they are doping the materials internally. For example, Celestion is now using mica flake filler in their polypropylene cones. Another company is filling their polypropylene cones with shredded aluminum.
This little story of a long quest and commercial failure may sound pessimistic, but I do not mean it to be. There are several reasons why a “diy”er might have success where a trained engineer working for a commercial concern might fail. First, there is an esthetic that must be followed in the high capital commercial world. We equate visual perfection with quality. Having all sorts of things stuck onto a nice smooth and polished metal cone just looks bad.
Second, there is a reliability and durability question. Anything stuck on can come off. A “diy”er can coat the metal and polypropylene woofer cones with a water-soluble glue and if it delaminates it is no big problem. Could be a reputation killer for a large cap commercial company.
Third, there is the prior learning problem. What you learned in the past can interfere with new learning. Engineers are intensely trained in specialties of engineering. Within engineering there are two main specialties, electrical engineering and mechanical engineering. Each thinks in precise patterns and have a language of discipline that does not always translate easily from one to the other. Sometimes that thinking can be so ingrained that it prevents trained engineers from considering different approaches. Yet loudspeakers are both electrical and mechanical devices that are tasked with reproducing music, a signal based upon natural mechanical resonances.
This is not a criticism of the person who wrote it, but in an earlier posting a resonant filter was referred to as a “trap” filter. This term comes out of electromagnetic radio or microwave work. It is part of the language of radio (and all of its derivations). It is most often used in transmission work involving carrier waves. While you can make a “trap” filter out of electrical components, you can also construct a physical trap in the electromagnetic microwave world. The use of “trap” here is metaphorical. Like all metaphors, it reveals some aspects while obscuring others. The word implicitly carries certain preconceptions that may or may not be productive in acoustical work.
Fourth, there is a culture of late 20th and early 21st century capitalist businesses that follows a deskilling model of labor division. Even the technical people in a company are given such a reductionist project scope that tunnel vision results. Think Dilbert here. Large cap audio manufacturers operate no differently than Dilbert’s software (or hardware) company. For an example, let us look at Magnesium. Magnesium is a “hot” commodity. Someone, maybe a marketing employee decides that since magnesium is selling golf balls that magnesium would be a good material to make a commercially successful driver diaphragm. An engineer is given the task to develop a process for fabricating the new cone material. That becomes their goal. They want to be able to consistently and affordably fabricate the material into the shape and thickness and mass that was the project of some other engineer or engineering team. Another engineer or team has responsibility for another aspect of the driver and so on until the driver is produced. Sometimes, these projects become so complex and subdivided that no one has a complete understanding or concern for the final product.
Once a company has invested heavily in a new product, what will they do if it is only slightly better, the same, or slightly worse than what came before it? Remember that magnesium is a hot commodity right now. The buying public has already been convinced that anything utilizing magnesium is a better product. Wouldn’t you put it out anyway?
If, however, they treat that cone to eliminate the problems, doesn’t that undermine the theory that the material made a better driver? Wouldn’t this make it harder to market and sell? Wouldn’t it make it harder to sell at a premium?
These are not problems for the “diy”er. The “diy”er should have few misconceptions and few limitations in the development of hypotheses for correcting loudspeaker driver diaphragm resonances.
The treating of loudspeaker driver diaphragm resonances is a task that should be operable by the application of science. Scientific method consists of no more than developing hypotheses and then testing them. Everyone reading this forum is capable of doing that. If you find pleasure in such a project, then I urge you to do it.
You will need a way to test your modifications. Not all analysis systems are equal. Some (whether because of how they are used or because of inherent limitations) have less resolution than others. I have three sets of graphs for the performance of the Tang Band W4-654S 4 inch paper cone driver. My own testing has the greatest resolution, Parts Express’ Clio tests come in second, and Tang Band’s LMS test has the least resolution. I have downloaded Speakerworks and am trying to learn how to operate the program. My intent is to calibrate it against my present system, and if I can duplicate my current resolution to provide this forum with the set-up criteria.
It will also help to have a known reference loudspeaker for calibrating the analysis system with your microphone. I am trying to work out a set of modifications to the W4-654S. If I am able to succeed with the modifications, then I will post them to this forum. If you know what the response should look like under two conditions (raw and modified) you should be able to “calibrate” your microphone to produce those response curves on Speakerworks.
Lastly, it will help to have an example of a resonant control modification. Driver diaphragm resonance is complex. It is a product of both shape and material. Tap test an unmounted loudspeaker cone. The sound will change as you tap at the outer or inner edge.
Driver diaphragm resonance is also chaotic. Small changes in beginning conditions can have a great influence on outcome. Sometimes very little needs to be modified to change the resonant structure of a driver. Most of the work involves identifying and isolating the mechanical causes of the material resonances.
If I succeed (and it is proving a much more challenging project than I envisioned) then the W4-654S can serve as one example of how to improve the acoustic performance of a driver through driver diaphragm resonance control. The W4-564S is a difficult driver to work with. The resonant structure is complex and the resonant structures show a high degree of coupling. The highest frequency resonance is centered at 12.32 kHz, is up 12 db from average output, and is only 1/5th of an octave in bandwidth. This resonance cannot be complementarily canceled with an electrical pre-filter. The axis of vibration is not in parallel with voice coil motion. The solution will be to either mechanically damp the vibration or to mechanically change its axis sufficiently to allow pre-filter cancellation (or a combination of both).
The material alone does not cause the resonant structure of this driver. Like many drivers, the cone is a flat profile over most of its area and then curves to parallel with the voice coil starting about 2/3 of the way in from the outside edge. The 12 kHz resonance appears to be caused by a shear bounded by this curve and the outer edge. The shape of the cone plays an important part in this resonance.
Mark
The “diy” world has been concerned with this problem for a long time. The second issue of Audio Amateur led off with a loudspeaker construction article by Peter Baxandall featuring two parallel resonant filters in series with the driver. The next year, 2 (3), featured an interview with Henry Kloss, where he briefly talks about loudspeaker driver diaphragm resonance and comments on work being done to structure the resonant response through shape and material selection. Kloss also talked about using resonant filters to control a resonance around 1 kHz in 5 inch full-range loudspeaker drivers.
The commercial world has also been concerned and has sometimes committed large sums of capital to address the problem. In the early 1980s Celestion developed a laser interferometry system to show (in animation) the complex movements of diaphragms at fixed frequencies. The first product produced with the technology was the SL-6. The SL-6 tweeter was a 1-¼ inch copper alloy dome tweeter that they claimed was without resonant break-up. I measured an early version of the loudspeaker and found high Q resonant structures. By the time J. Peter Moncrieff of IAR reviewed the loudspeaker, a resonant filter was added to the crossover to tame the major dome break-up resonant. That, however, is all it did. It tamed it but did not eliminate it. Just goes to show that even commercial concerns sometime resort to techniques available to us “diy”ers.
Celestion and others are still interested in controlling loudspeaker driver diaphragm and cabinet resonances. Instead of adding things on the outside of the cones, they are doping the materials internally. For example, Celestion is now using mica flake filler in their polypropylene cones. Another company is filling their polypropylene cones with shredded aluminum.
This little story of a long quest and commercial failure may sound pessimistic, but I do not mean it to be. There are several reasons why a “diy”er might have success where a trained engineer working for a commercial concern might fail. First, there is an esthetic that must be followed in the high capital commercial world. We equate visual perfection with quality. Having all sorts of things stuck onto a nice smooth and polished metal cone just looks bad.
Second, there is a reliability and durability question. Anything stuck on can come off. A “diy”er can coat the metal and polypropylene woofer cones with a water-soluble glue and if it delaminates it is no big problem. Could be a reputation killer for a large cap commercial company.
Third, there is the prior learning problem. What you learned in the past can interfere with new learning. Engineers are intensely trained in specialties of engineering. Within engineering there are two main specialties, electrical engineering and mechanical engineering. Each thinks in precise patterns and have a language of discipline that does not always translate easily from one to the other. Sometimes that thinking can be so ingrained that it prevents trained engineers from considering different approaches. Yet loudspeakers are both electrical and mechanical devices that are tasked with reproducing music, a signal based upon natural mechanical resonances.
This is not a criticism of the person who wrote it, but in an earlier posting a resonant filter was referred to as a “trap” filter. This term comes out of electromagnetic radio or microwave work. It is part of the language of radio (and all of its derivations). It is most often used in transmission work involving carrier waves. While you can make a “trap” filter out of electrical components, you can also construct a physical trap in the electromagnetic microwave world. The use of “trap” here is metaphorical. Like all metaphors, it reveals some aspects while obscuring others. The word implicitly carries certain preconceptions that may or may not be productive in acoustical work.
Fourth, there is a culture of late 20th and early 21st century capitalist businesses that follows a deskilling model of labor division. Even the technical people in a company are given such a reductionist project scope that tunnel vision results. Think Dilbert here. Large cap audio manufacturers operate no differently than Dilbert’s software (or hardware) company. For an example, let us look at Magnesium. Magnesium is a “hot” commodity. Someone, maybe a marketing employee decides that since magnesium is selling golf balls that magnesium would be a good material to make a commercially successful driver diaphragm. An engineer is given the task to develop a process for fabricating the new cone material. That becomes their goal. They want to be able to consistently and affordably fabricate the material into the shape and thickness and mass that was the project of some other engineer or engineering team. Another engineer or team has responsibility for another aspect of the driver and so on until the driver is produced. Sometimes, these projects become so complex and subdivided that no one has a complete understanding or concern for the final product.
Once a company has invested heavily in a new product, what will they do if it is only slightly better, the same, or slightly worse than what came before it? Remember that magnesium is a hot commodity right now. The buying public has already been convinced that anything utilizing magnesium is a better product. Wouldn’t you put it out anyway?
If, however, they treat that cone to eliminate the problems, doesn’t that undermine the theory that the material made a better driver? Wouldn’t this make it harder to market and sell? Wouldn’t it make it harder to sell at a premium?
These are not problems for the “diy”er. The “diy”er should have few misconceptions and few limitations in the development of hypotheses for correcting loudspeaker driver diaphragm resonances.
The treating of loudspeaker driver diaphragm resonances is a task that should be operable by the application of science. Scientific method consists of no more than developing hypotheses and then testing them. Everyone reading this forum is capable of doing that. If you find pleasure in such a project, then I urge you to do it.
You will need a way to test your modifications. Not all analysis systems are equal. Some (whether because of how they are used or because of inherent limitations) have less resolution than others. I have three sets of graphs for the performance of the Tang Band W4-654S 4 inch paper cone driver. My own testing has the greatest resolution, Parts Express’ Clio tests come in second, and Tang Band’s LMS test has the least resolution. I have downloaded Speakerworks and am trying to learn how to operate the program. My intent is to calibrate it against my present system, and if I can duplicate my current resolution to provide this forum with the set-up criteria.
It will also help to have a known reference loudspeaker for calibrating the analysis system with your microphone. I am trying to work out a set of modifications to the W4-654S. If I am able to succeed with the modifications, then I will post them to this forum. If you know what the response should look like under two conditions (raw and modified) you should be able to “calibrate” your microphone to produce those response curves on Speakerworks.
Lastly, it will help to have an example of a resonant control modification. Driver diaphragm resonance is complex. It is a product of both shape and material. Tap test an unmounted loudspeaker cone. The sound will change as you tap at the outer or inner edge.
Driver diaphragm resonance is also chaotic. Small changes in beginning conditions can have a great influence on outcome. Sometimes very little needs to be modified to change the resonant structure of a driver. Most of the work involves identifying and isolating the mechanical causes of the material resonances.
If I succeed (and it is proving a much more challenging project than I envisioned) then the W4-654S can serve as one example of how to improve the acoustic performance of a driver through driver diaphragm resonance control. The W4-564S is a difficult driver to work with. The resonant structure is complex and the resonant structures show a high degree of coupling. The highest frequency resonance is centered at 12.32 kHz, is up 12 db from average output, and is only 1/5th of an octave in bandwidth. This resonance cannot be complementarily canceled with an electrical pre-filter. The axis of vibration is not in parallel with voice coil motion. The solution will be to either mechanically damp the vibration or to mechanically change its axis sufficiently to allow pre-filter cancellation (or a combination of both).
The material alone does not cause the resonant structure of this driver. Like many drivers, the cone is a flat profile over most of its area and then curves to parallel with the voice coil starting about 2/3 of the way in from the outside edge. The 12 kHz resonance appears to be caused by a shear bounded by this curve and the outer edge. The shape of the cone plays an important part in this resonance.
Mark
Interesting comments so far!
Maybe I didn't make myself clear enough because nobody's yet picked up on it: I don't want to put some sticky coat on the cone, I was thinking about laminating pieces of fibreglass (or paper) to the back side). I have very thin fibreglass mats that would add close to no weight. With some lightweight glue (wallpaper glue), the change in mms would be minimal. At the same time, I assume the bending stiffness would improve locally where a piece of fibreglass is attached. Therefore, it is important to know the shape of the resonant modes.
The gue thing has been done: Audio International have a special version of the W17 made for them which has a reddish paint rather than the gray stuff. Mms is almost unchanged (maybe + 1 g), but resonances are not as severe.
Maybe I didn't make myself clear enough because nobody's yet picked up on it: I don't want to put some sticky coat on the cone, I was thinking about laminating pieces of fibreglass (or paper) to the back side). I have very thin fibreglass mats that would add close to no weight. With some lightweight glue (wallpaper glue), the change in mms would be minimal. At the same time, I assume the bending stiffness would improve locally where a piece of fibreglass is attached. Therefore, it is important to know the shape of the resonant modes.
The gue thing has been done: Audio International have a special version of the W17 made for them which has a reddish paint rather than the gray stuff. Mms is almost unchanged (maybe + 1 g), but resonances are not as severe.
There are a couple of nice laser interferometry animations on the Optonor Web site.
Just go to:
http://www.optonor.no/loudspeaker2.htm
Look at the diaphragm flexure modes. Are they radial or are they tangential?
It does not mean they cannot be controlled, but I believe that the vibration modes are too complex to be constrained by just these two "metaphors."
Since they have more experience with the technique, I wish Celestion would publish more of their interferometry animations. They don't, however, so we have to make do with what we can access.
Still, if you can measure the acoustic output (frequency response) of the driver, then apply your mats and measure again, you can tell if they are achieving what you want. This is just basic empiricism. There is nothing wrong with empirical experimentation. It has a long and revered tradition.
Mark
Just go to:
http://www.optonor.no/loudspeaker2.htm
Look at the diaphragm flexure modes. Are they radial or are they tangential?
It does not mean they cannot be controlled, but I believe that the vibration modes are too complex to be constrained by just these two "metaphors."
Since they have more experience with the technique, I wish Celestion would publish more of their interferometry animations. They don't, however, so we have to make do with what we can access.
Still, if you can measure the acoustic output (frequency response) of the driver, then apply your mats and measure again, you can tell if they are achieving what you want. This is just basic empiricism. There is nothing wrong with empirical experimentation. It has a long and revered tradition.
Mark
MarkMcK said:
That is a pretty well defined breakup in the animation -- usually things are a bit more caotic.
dave
Chaos (as in chaos theory) is not easy to understand or to visualize. I am not going to claim expertise in chaos theory.
Still, Dave has posed an interesting question. Optonor has posted several other animations (not simulations) of real world objects vibrating at specific frequencies. Do any of these animations show vibration modes that would qualify as chaotic?
http://www.optonor.no/vibration.htm
My totally inadequate understanding is that chaos refers to phenomena difficult to predict, where small variations in conditions of origin can have great effect upon outcome. It does not mean poorly defined or inconsistent. Thus turbulence (chaos) in water flowing down a streambed or in non-laminar flow over an airfoil or the vibration of a diaphragm may be very well defined and still be chaotic. Think of the vortex oscillating over the trailing edge of De Havilland's wing before his plane broke up.
Whether the vibration mode(s) shown in loudspeaker2 is chaotic or not, does it tell us anything about possible techniques to control that vibration?
Just for fun and to check your understanding of interfermetric analyis, and without e-mailing Optonor, what type of driver is tested?
Mark
- Status
- Not open for further replies.