Here's what the experts have to say:
"An expert is one who knows more and more about less and less until he knows absolutely everything about nothing." Nicholas Butler
"A theory is something nobody believes, except the person who made it. An experiment is something everybody believes, except the person who made it."" Albert Einstein
"Nothing is more practical than a good theory." Kurt Lewin
😀
Perhaps mwmkravchenko is trying to say something about thinkers vs. doers
To which I would say:
"Thinkers think and doers do. But until the thinkers do and the doers think, progress will be just another word in the already overburdened vocabulary of the talkers who talk." - unknown
...which pretty much sums up this thread.
Well did I stir up a hornets nets?
My statement was made in jest. Anybody who has read my comments over the past few years will know I have an evil sense of humor.
Jan. Your statement is true. Any comment such as I made will turn people off. It was not my intention and you have my apology for that.
Earl. Let me start with this. 22 years ago I read your paper in the JAES on bandpass designs. What I know I read from you. What you know you learned and read came from others as well. We all stand on the shoulders of those that came before us. My barb was definitely not aimed at you. Please don't take it seriously.
Pete. If you want to comment about someones work have the decency to let him know please. I have nothing evil to say about you because you seem to be a knowledgeable guy. The barb was not aimed at you either. In fact you are one person who has made me think constructively.
Every body.
We have all worked with people in some capacity or the other who could pontificate about all and sundry but couldn't do didlee. It is those who I have always dreaded working with.
Now bear this in mind. We all have strengths and weaknesses. I wish my foundation in math was a lot better. It would be oh so useful in some of the things I do. But it is a weakness. I have a computer and gifted programmers who can help me out in that respect.
So we work with what we have. We post on this forum to show and tell and get help. And I think that is the best reason to do anything worth while. Sharing. More joy in giving then in receiving. Right?
Mark
My statement was made in jest. Anybody who has read my comments over the past few years will know I have an evil sense of humor.
Jan. Your statement is true. Any comment such as I made will turn people off. It was not my intention and you have my apology for that.
Earl. Let me start with this. 22 years ago I read your paper in the JAES on bandpass designs. What I know I read from you. What you know you learned and read came from others as well. We all stand on the shoulders of those that came before us. My barb was definitely not aimed at you. Please don't take it seriously.
Pete. If you want to comment about someones work have the decency to let him know please. I have nothing evil to say about you because you seem to be a knowledgeable guy. The barb was not aimed at you either. In fact you are one person who has made me think constructively.
Every body.
We have all worked with people in some capacity or the other who could pontificate about all and sundry but couldn't do didlee. It is those who I have always dreaded working with.
Now bear this in mind. We all have strengths and weaknesses. I wish my foundation in math was a lot better. It would be oh so useful in some of the things I do. But it is a weakness. I have a computer and gifted programmers who can help me out in that respect.
So we work with what we have. We post on this forum to show and tell and get help. And I think that is the best reason to do anything worth while. Sharing. More joy in giving then in receiving. Right?
Mark
Mark
I would be careful that's all. The person who started this thread slandered me up and down on another site, so its hard to tell a joke from an insult.
Hi Mark,
I have to say that I think you're easy to talk to about this and have an open mind, I like that. I am interested in the TRIO drivers, more the 12" and that led me to take a look at your thread. When I saw your UNHORN concept it got me thinking as to how to simplify it enough so that some simple analysis might apply, and so far I've not come to a firm conclusion about it. I could also say that you work has led me to some constructive thinking.
As far as actually doing goes, I've built plenty of speakers, I do the woodworking myself for prototypes. I showed one concept to a fellow engineer at work and he, being an excellent carpenter, acted as if it would be too difficult for me to build without asking about my ability. They're built now, no problem. Working with my hands has always come naturally for me. I used to work on cars when I had more spare time and have not only rebuilt engines, but also transmissions, and set up a differential with new gears one time. I also did a full auto paint job myself. I have restored a 1967 Firebird, A 67 Alfa Romeo 2600, and a 1958 Alfa Romeo Giulietta Spider Veloce. My point is to not make assumptions. I don't do as much building these days because I am busy with my family and I concentrate more on business.
When I look at the UNHORN it doesn't look like a bandpass to me at all, rather I see the rear chamber as the box in a vented system and the slot where the drivers are mounted as a slot loaded vent. What I'm not sure of is how mounting the drivers in the slot alters the response. It seems to me that the final horn there probably boosts up the midbass more than anything. I wonder what the simulation would show if you chopped off that horn section just to better understand what each part is doing. The input impedance curve would also be interesting. I would load up your model into the simulator if I had more time and do some experiments, but it will be a long while before I have the time for that.
I appreciate that you've been open to my comments at the other site and hope to see the results of your work soon.
Last edited:
That is my experience as well. More than half of "experts" that I have met are so focused and specialized that they lose the view of the how it fits into the whole project.
Hi Pete
You have good taste in cars!
And it's nice to see a real engineer get his fingers dirty.
Ok
What happens is quite interesting. I wish I could figure out exactly in mathematical terms what is going on. But in pure empirical modeling terms here is what happens.
Lets start out by calling the UNHORN a bow tie. For all intense purposes that is what the shape resembles. One end is a closed resonator. The knot section of the bow tie is port from the resonator to the wide open world. The position of the drivers and the width of the knot section make a huge difference on the response of the box. Just as in a conventional bandpass when you try to tune the sections to far apart you get a large dip in the response. If you change the size of either section of the box you get the same big dip. They need balancing between air mass and port section between the open end and the closed end. You can tailor the response quite a bit by changing the size. I think I posted a much larger version in hometheatershack. It could hit 18hz if I remember correctly.
I am embarrassed to say that I cannot explain in any conventional useful terms what the two sections are really doing. There is a definite relationship between vented box and flared port design.
As you questioned what happens if you get rid of the flared port? Poof the response drops off to a useless hump. Hornresp is actually quite accurate in modeling this. So what exactly is this weird little box doing? It makes music. That about it. I am currently working on some rather large front loaded horns. They are far easier to explain in their means of functioning. The UNHORN is a heck if I know. I think it is the latest reincarnation of the Karlson coupler. A box design poo pooed back when he did it to. I'm at a loss to put some rigorous explanation to it. But this is what I have learned.
Like a horn or any other type of speaker alignment there are drivers that will work well with it and drivers that will not. Short answer look at the Thiele Small parameters of the TRIO8 and the ANARCHY drivers and you have something that will work. Deviate largely from this and you get pathetic response. Is this a variant type of alignment? Maybe. Did I invent it? I hoped so for a while. But in the loudspeaker field I think there is nothing really new. Just variations on a theme.
Here is what you asked for.
The example is with the ANARCHY driver. Good driver by the way.
Here is an example of the same drivers with a different size enclosure Notice the peaks that occur when you try to tune the sections to far apart:
Now these are not the TRIO8 driver this is a badly aligned Anarachy driver. Not the drivers fault it's a bad box a very bad box.
You can take a good driver and do this if you care to:
This is two TRIO8's in an enclosure that looks more like a tapped horn than anything else. Actually somewhat like a transmission line. One thing to take note of is that the red and green circle are the front and back of the driver as modeled in Hornresp. So placement of the drivers within this type of enclosure is also important. The large peak in the top end does not show up when you actually measure the box.
I'll return to the original rather flat Anarchy design as an example of moving drivers to non-optimal locations:
Move the woofers around and this is what can happen. Not very pretty. The box is exactly the same design. The location of the driver has changed. Again Hornresp models this quite accurately as well. My second fold attempt at the TRIO8 unhorn displayed a large sag in the response due to poor driver placement. Live and learn. A case of a driver telling me what is best!
So a bit of show and tell. All the input info is available on diyaudio.com in the old tapped horn for car thread. I posted the input screen for the ANARCHY version there this morning.
I do model SPL figures in a 1 Pi acoustical environment because I find that this most closely models a normally constructed drywall and stud room corner placement. If you are in a car or a poured cement basement against exterior cement walls then you can expect a 1/8th Pi response.
Mark
You have good taste in cars!
And it's nice to see a real engineer get his fingers dirty.
Ok
What happens is quite interesting. I wish I could figure out exactly in mathematical terms what is going on. But in pure empirical modeling terms here is what happens.
Lets start out by calling the UNHORN a bow tie. For all intense purposes that is what the shape resembles. One end is a closed resonator. The knot section of the bow tie is port from the resonator to the wide open world. The position of the drivers and the width of the knot section make a huge difference on the response of the box. Just as in a conventional bandpass when you try to tune the sections to far apart you get a large dip in the response. If you change the size of either section of the box you get the same big dip. They need balancing between air mass and port section between the open end and the closed end. You can tailor the response quite a bit by changing the size. I think I posted a much larger version in hometheatershack. It could hit 18hz if I remember correctly.
I am embarrassed to say that I cannot explain in any conventional useful terms what the two sections are really doing. There is a definite relationship between vented box and flared port design.
As you questioned what happens if you get rid of the flared port? Poof the response drops off to a useless hump. Hornresp is actually quite accurate in modeling this. So what exactly is this weird little box doing? It makes music. That about it. I am currently working on some rather large front loaded horns. They are far easier to explain in their means of functioning. The UNHORN is a heck if I know. I think it is the latest reincarnation of the Karlson coupler. A box design poo pooed back when he did it to. I'm at a loss to put some rigorous explanation to it. But this is what I have learned.
Like a horn or any other type of speaker alignment there are drivers that will work well with it and drivers that will not. Short answer look at the Thiele Small parameters of the TRIO8 and the ANARCHY drivers and you have something that will work. Deviate largely from this and you get pathetic response. Is this a variant type of alignment? Maybe. Did I invent it? I hoped so for a while. But in the loudspeaker field I think there is nothing really new. Just variations on a theme.
Here is what you asked for.
The example is with the ANARCHY driver. Good driver by the way.
An externally hosted image should be here but it was not working when we last tested it.
Here is an example of the same drivers with a different size enclosure Notice the peaks that occur when you try to tune the sections to far apart:
An externally hosted image should be here but it was not working when we last tested it.
Now these are not the TRIO8 driver this is a badly aligned Anarachy driver. Not the drivers fault it's a bad box a very bad box.
You can take a good driver and do this if you care to:
An externally hosted image should be here but it was not working when we last tested it.
This is two TRIO8's in an enclosure that looks more like a tapped horn than anything else. Actually somewhat like a transmission line. One thing to take note of is that the red and green circle are the front and back of the driver as modeled in Hornresp. So placement of the drivers within this type of enclosure is also important. The large peak in the top end does not show up when you actually measure the box.
I'll return to the original rather flat Anarchy design as an example of moving drivers to non-optimal locations:
An externally hosted image should be here but it was not working when we last tested it.
Move the woofers around and this is what can happen. Not very pretty. The box is exactly the same design. The location of the driver has changed. Again Hornresp models this quite accurately as well. My second fold attempt at the TRIO8 unhorn displayed a large sag in the response due to poor driver placement. Live and learn. A case of a driver telling me what is best!
So a bit of show and tell. All the input info is available on diyaudio.com in the old tapped horn for car thread. I posted the input screen for the ANARCHY version there this morning.
I do model SPL figures in a 1 Pi acoustical environment because I find that this most closely models a normally constructed drywall and stud room corner placement. If you are in a car or a poured cement basement against exterior cement walls then you can expect a 1/8th Pi response.
Mark
Regarding how well Hornresp models the "UNHORN":
Back when Mark originally developed the UNHORN, I modeled one of the final iterations in AkAbak and found that the results were almost identical to those in Hornresp.
As to how it works, I'm puzzled too. I consider it a hybrid of reflex and tapped horn. Port area seems to be more critical than that of a reflex enclosure, and driver position seems to be more critical than that of a typical tapped horn. Playing with the dimensions in Hornresp's enclosure wizard is a good way to pass a few hours or days.
Back when Mark originally developed the UNHORN, I modeled one of the final iterations in AkAbak and found that the results were almost identical to those in Hornresp.
As to how it works, I'm puzzled too. I consider it a hybrid of reflex and tapped horn. Port area seems to be more critical than that of a reflex enclosure, and driver position seems to be more critical than that of a typical tapped horn. Playing with the dimensions in Hornresp's enclosure wizard is a good way to pass a few hours or days.
Hi Don
It's actually a sixth sense! LOL
Kidding aside modeling those beasts is not for the faint of heart.
Mark
It's actually a sixth sense! LOL
Kidding aside modeling those beasts is not for the faint of heart.
Mark
I have a question for gedlee.
I was trying to understand what happens when one deals with a TL with bends. I was convinced that the effects of bending were small, until I discovered the work by Sven Tyrland (unfortunately he is dead, so I cannot ask him any question). Given his results
the effects of bending are not irrelevant. I could guess that the statements made by the acousticians studying wind instruments are not applicable in the TL case, but I’m unable to do the math (multimodal analysis).
So I tried to afford the problem using FEM/BEM tools. However they are too difficult for a person (me) that doesn’t want to become an expert of them.
Can you give your advice on that topic ?
Thanks
Teodoro
I was trying to understand what happens when one deals with a TL with bends. I was convinced that the effects of bending were small, until I discovered the work by Sven Tyrland (unfortunately he is dead, so I cannot ask him any question). Given his results
An externally hosted image should be here but it was not working when we last tested it.
the effects of bending are not irrelevant. I could guess that the statements made by the acousticians studying wind instruments are not applicable in the TL case, but I’m unable to do the math (multimodal analysis).
So I tried to afford the problem using FEM/BEM tools. However they are too difficult for a person (me) that doesn’t want to become an expert of them.
Can you give your advice on that topic ?
Thanks
Teodoro
Not being gedlee i nevertheless try an answer as
educated guess:
If a bend lies in a velocity maximim of the standing wave
- better: at the distance from the source where the velocity
maximum would have been in the stretched version -
i would expect that particular resonance or anti resonance to
be damped, since the velocity of the air molecules has a
direction in that region of the pipe. That direction of movement
cannot be turned 180 degrees without loss due to friction.
For those standing waves having minimum velocity and a pressure
maximum in the bend region, i expect the (impedance) difference
between stretched and bended version to be smaller.
Kind Regards
educated guess:
If a bend lies in a velocity maximim of the standing wave
- better: at the distance from the source where the velocity
maximum would have been in the stretched version -
i would expect that particular resonance or anti resonance to
be damped, since the velocity of the air molecules has a
direction in that region of the pipe. That direction of movement
cannot be turned 180 degrees without loss due to friction.
For those standing waves having minimum velocity and a pressure
maximum in the bend region, i expect the (impedance) difference
between stretched and bended version to be smaller.
Kind Regards
I feel a little bit discomforted when I read some comments and discussions still alive in the world of the speaker builders.
Speed of sound in the fiber ?
You are discussing about the validity of the “solar system model” of the atom a long time after Quantum Mechanics has given his answers.
You could eventually discuss if the appropriate model is a rigid one, a limp, or if you need (I hope not) to use the Biot model. If you are lucky (so the model is rigid), you could discuss if the Delany-Bazley model is good enough, or if the Miki model or the Champoux-Allard (both preserve minimum phase in presence of the fiber) are more appropriate.
A lot of work has been done, decades ago, by people studying sound propagation in fibrous materials.
Same for the solutions of the Bernoulli-Webster Horn Equation (one for all: Putland).
I feel discouraged when the work of MJK is still under discussion. It can be improved, it can be made more “official”, but it gives good enough solutions.
Can we go deeper ? and stop discussing about the validity of the Copernican system ? General Relativity is one century old ...
Speed of sound in the fiber ?
You are discussing about the validity of the “solar system model” of the atom a long time after Quantum Mechanics has given his answers.
You could eventually discuss if the appropriate model is a rigid one, a limp, or if you need (I hope not) to use the Biot model. If you are lucky (so the model is rigid), you could discuss if the Delany-Bazley model is good enough, or if the Miki model or the Champoux-Allard (both preserve minimum phase in presence of the fiber) are more appropriate.
A lot of work has been done, decades ago, by people studying sound propagation in fibrous materials.
Same for the solutions of the Bernoulli-Webster Horn Equation (one for all: Putland).
I feel discouraged when the work of MJK is still under discussion. It can be improved, it can be made more “official”, but it gives good enough solutions.
Can we go deeper ? and stop discussing about the validity of the Copernican system ? General Relativity is one century old ...
Teodoro,
Those results would be from an unstuffed TL. Once an appropriate amount of damping is added all the HF stuff should go away.
dave
Those results would be from an unstuffed TL. Once an appropriate amount of damping is added all the HF stuff should go away.
dave
As others have pointed out, the effects of bending are small for frequencies lower than those corresponding to the fundamental resonance of each line section. In short, bends don't affect the bass. They become important for fullrange use. The resonances of each line section cause colouration of the sound. But since the purpose of the line is to improve the bass response, high frequency effects are undesirable. In practice, they are normally suppressed by the "stuffing" of the line.
AkAbak is a good tool for this, for frequencies below those that will cause "side to side" resonances in the line. You can accurately model the corners. See this example, which I used to Akabak model the box in the post before it:
http://www.diyaudio.com/forums/subwoofers/145603-tapped-horn-car-19.html#post1901061
Note how each corner is modeled by several diverging/converging sections. You need up to 4 sections to accurately model a hairpin bend. But after going to all that trouble, you will find that the performance for bass frequencies is very similar to that of a straight pipe.
I guess that its my turn. First it is entiorely incorrect to say that a bend causes damping because the wave changes direction. Wave can bounce off of walls, i.e. change direction, without any damping at all. So completely disregard the comments that claim damping. Also any that tells you that the effects are predictable via anything but BEM or FEA are just plain wrong. The situation is very complex and there is no way to predict the effects short of these more complex analysis.
The effects will be greater the higher the frequency and the sharper the bend. If the bend is gradual compared to a wavelength then its basically not there. Thats why musical instrument horns have gradual bends rather than very sharp ones. Use www.Falstad.com for a non-analytical way to look at bends and what they do. You will see that its anything but simple.
The effects will be greater the higher the frequency and the sharper the bend. If the bend is gradual compared to a wavelength then its basically not there. Thats why musical instrument horns have gradual bends rather than very sharp ones. Use www.Falstad.com for a non-analytical way to look at bends and what they do. You will see that its anything but simple.
Oops ! I’m unable to find (it’s late, here in Italy, maybe tomorrow ...) the paper written by one of the acoustician studying the effects on the pitch caused by the bending in a wind instrument (e.g. a trombone).
Anyway the (approximate) conclusion is that the bending (a cylindrical bend) affects the calculation of the transfer matrix by a +10% of ZC(w) and -10% of ro(w) (or -10% and +10%, I don’t remember), where the usual mechanism of taking the characteristic impedance and the density as complex functions of w.
So, since the bend is only a portion of the line and the modification on the transfer matrix are of order +-10%, I don’t expect so much variation.
I could be more precise once I shall include those results in “my” version of the MJK worksheets (yes I have reworked them in such a way an acoustic expert could not argue about them).
My problem are then the results of (the late) Sven Tyrland: it could be that the calculations valid for a trombone are not valid for a TL . For that reason I asked if someone has done some experiment or some FEM/BEM calculation.
Anyway the (approximate) conclusion is that the bending (a cylindrical bend) affects the calculation of the transfer matrix by a +10% of ZC(w) and -10% of ro(w) (or -10% and +10%, I don’t remember), where the usual mechanism of taking the characteristic impedance and the density as complex functions of w.
So, since the bend is only a portion of the line and the modification on the transfer matrix are of order +-10%, I don’t expect so much variation.
I could be more precise once I shall include those results in “my” version of the MJK worksheets (yes I have reworked them in such a way an acoustic expert could not argue about them).
My problem are then the results of (the late) Sven Tyrland: it could be that the calculations valid for a trombone are not valid for a TL . For that reason I asked if someone has done some experiment or some FEM/BEM calculation.
Does any one know why Fried stuck with open cell polyurethane foam and did not use Acousta Stuff in his designs. When I purchased some new Fried drivers from him in late 90's he told me the foam was better. I think it has to do with the way the lines were stuffed. Using Acousta Stuff all of the sound passes through it. The Fried lines were true llines Only part of the sound passed through the foam due to the vertical spacing. I would think there would be less of a time delay at line terminus. Time delay is critical in tweeter and mid bass enclosure and crosoover design. I know from purchasing many of his kits that he developed universal boxes. He changed stuffing density for each driver used in exactly the same t-line enclosure. Not all the drivers had similiar overall parameters. The size of the box was determined by driver Sd and F3. Many people will disagree . I have used wide variety of drivers in these encloures and easily get them to test flat by making a few minor changes in stuffing density. Drivers with Bl of 10-12 and large Vas, low qts., low Fs sound the best. I once tried Bose 6.5" woofer I bought from Martin Sound for 4$ a piece in the C3L cabinet and it tested ruler flat to cutoff as did an old Audax HD17B37R4C12 driver, and Fried Dalesford 6.5" I chose to use the Bose and Audax woofer in a simpler t-line cabinet since the C3L are complicated to build.
I think it was unfortunate Stragiato was banned . I have talked to the guy and he is highly educated. A lot of his opinions coincide with what I derived Fried did in his t-lines. Most forum users I think are buying software and thinking they are producing ruler flat t-lines. The types of t-lines vary from from the type of design they are derived from. There is more then one way to design a t-line. Look at all of the Speaker Builder articles dating back to Gary Galo in 81. Bailey's Radford designs. Whose t-line design is the most acousticaly correct . Are any forum users members of the Yahoo T-line Group. Many people I have come to know have patents and copyrights on their t-line designs. There is far to much hype over their design. I was fortunate enough to become a Fried speaker customer at age 17. I live 25 miles from where Bud sold his speakers on City Line Avenue. I had many conversations and visits with him over the years. It was pretty easy to know when I was correct on his theory designs by his change in temperment. I have been building designs based on his work since the early 80's. Fried sold the business and relocated to Norristown,PA . The first owner Bud sold the company to wants to remain anonymous. Later Bud sold it to Shaynet Tenace. before Dr. Raines became the new owner. I think it unfortunate for moderators to ban some one that might have a higher understanding then them. Most people would rather argue that can't see the trees for the forrest. Try the older Fried designs on imfelectroncis using half decent drivers adjusting foam and measure results the old fashioned way. Look at these designs and do the math. If any one asks what math to do it shows how little they really know. Does any one want to discuss series crossovers?
Attachments
Last edited:
The link Strangiato posted is correct.When the author states using low cost cheap drivers in t-lines he is correct. Not that I would buy one what do you think Bose did in the Wave Radio? Bose used cube drivers in 1/4 wave lenght enclosures. I have done the same thing with the same exact drivers. Maybe this forum isn't for helping each other out. I try to give my fellow man helpful information to improve his listening experience. Many forum users are trying to put themseleves on a pedastool. Martial arts, powerlifting, or becoming a skilled marksmen might be a bit more challenging experience for these people rather then speaker building. I try to learn from other people's postings.. Every one can learn from each other's experiences. Most of my friends in the business stay away from forums. I can see why.
Attachments
- Status
- Not open for further replies.