I'm really not going to spend a lot more time on this. You have a problem with my electrical analogies when one of the first things taught in acoustics is that there are electrical analogies for acoustical components/circuits. All of Thiele and Small's work is based on this.
You speak of electrical to acoustical, are you aware that the electrical circuit is connected to the mechanical driver circuit (yes the mechanical circuit also has an electrical analogy) through a gyrator representing the motor, and to the acoustical circuit through a transformer representing the cone area which couples the mechanical circuit to the air. The most common models talked about in numerous publications are electrical analogies to speaker systems.
You also don't really understand what happens to the peak of a driver in free air as compared to being in a TL. Probably because you've never done the measurement and in fact your description above has a common error that was pointed out by my project advisor. You claim that the resonance moves up in frequency and is diminished in amplitude.
So tell me, why do you think it moves up?
And tell me how do you think you might get as much of a diminished amplitude in a vented system - do you know that it is possible?
Are you aware that there is a way to alter the driver mechanical parameters to get a lower peak? Which parameter would you change to lower the peak without changing the frequency?
Why would you expect a recent paper from WPI when I told you mine was finished in 1980?
You are not discussing facts, rather you are hand waving.
Last edited:
I am using Augspurger's software (version 3.1) and have tried a broad spectrum of configurations and tapers (positive through negative). Nowhere did I see SPL levels approaching a vented box's SPL. While I fully understand that this is simply software that is supposed to approximate (closely, one would hope) what happens in practice, I observed at least 3 to 6 dB less output.
I may be doing something wrong, too. So, if you know different, I would love to know what it is I'm doing wrong.
As far as impedance matching goes (I'll jump into this one), my understanding is that electrical and acoustic impedance matching are different things. In other words, you can match or smooth out the inductive impedance mismatch in the driver's motor, and that is one thing. However, acoustic impedance matching is another animal. For example, a horn is an acoustic impedance device. Not all configurations of a tapered TL would qualify as an acoustic impedance match between the driver and the room in that context. Would you agree?
I believe that it went into a bit more depth. I scanned one of them for Martin King some time ago - I'll have to look for it. The paper versions are filed - I'll see if I can find them but it has been a long time.
Ha! 😀
Well, when you compare my little amp to what typically is on tap for solid state amps you are talking a different story.
Are you using two different packages to simulate the vented vs. TL? I would not expect to see such a large difference. There are two possible causes for lower efficiency, first mass loading of the TL line and second highly resistive loading or low "box" Q. It is more likely that one simulator is assuming full space while the other half space. Or that one is using 1W or 1V in while the other something like 2.83V.
Are you all aware that vented, sealed and TL systems should all have the same passband efficiency well above the low bass range? It is a characteristic of the driver. Sealed and vented box input impedance is compliant above resonance, whereas with a TL there are regions where it is compliant others a mass, the mass region can lower efficiency and in that case would break the rule. The output from a high Q TL has large amplitude swings partly due to this mass loading.
Can't stand the smell of the stuff. Give me a packet of Jaffas any day.
I, too, originally thought he made a reference to power transmission lines, those wires on towers that criss-cross the countryside. If so, he's very wrong - they are designed with a low source impedance and a high load impedance, to minimise losses. There is no way any power company would accept half of their generated power being lost in transmission. In any case, for a circuit to be considered as a "transmission line" it usually has to be 1/4 wavelength or more in length. A few moments working out 1/4 wavelength of 50 or 60 Hz will show why "transmission line" means different things to electrical power distribution engineers versus RF engineers.
The peak resonance frequency (as indicated by both electrical impedance- see King's graph posted by Planet10 above, and as measured by increased acoustic output - see papers of Natkaniec, King, Shultz, and Augspurger) is shifted up in frequency because the distance from the exciting element in the damped or undamped pipe (same behavior with either) to the pipe opening corresponds to frequencies at or near the center of the driver's resonance distribution which is higher than the driver's free air resonance or Fs. The pipe in all TL's acts like a low pass filter whose corner frequency is set by pipe length and whose taper and stuffing determines the "order" of said filter's slope. The farther away the pipes opening to the first quarter wave peak of the wave pulse emitted by the transducer - the more attenuation. The closer the first quarter wave peak to the driver's opening - the less attenuation caused by the pipe's confinement. This analysis supports Natkaniec's finding that the peak sum of primary resonant modes in the pipe corresponds to 60 degrees phase distance from the point of excitation within the pipe - or a quarter wave frequency of 1.5 times the fundamental Fs.
I am not spouting my own theories here. Although I have an extensive background in laser design, electrical engineering, acoustical engineering, linear control theory, and quantum mechanics - the conclusions and theories I espouse herein are those of other accredited individuals with published work and scientific data to support said work. I am simply conveying my agreement with their findings and offering support thereto.
And yes, I'm aware you can increase driver diaphragm mass to lower its fundamental resonance....yada yada yada - can we put an end to the peeing match and just stick with the facts as they are presented? I'm really not interested in your background or Martin King's background for that matter - just trying to put some facts on the table.🙂
This is blatantly wrong that a line has to be 1/4 wave or more to be a TL. The fact is that a lumped approximation is good enough when the wavelength is long compared to the length of the line. It is still a TL. I taught this material. Transmission lines all have the same underlying equations doesn't matter how they are used.
I need to revisit my simulations. I did them awhile ago using a woofer with a 93 dB/W/m and I just remember that the overall efficiency of the system in the TL software was below that number. Could be something I was doing wrong and I never had the opportunity to actually build the TL to test if the theory was correct, although, if I can come up with a tantalizing theoretical design I would be tempted to cut lumber and see.
The high power transmission lines I was referring to were the undersea telegraph cables referenced here:
Transmission line - Wikipedia, the free encyclopedia
Which predates the acoustic "transmission line".
I just gave the Natkaniec paper a quick read and I wonder how or why you've placed so much value on this paper. He provides some real data for sealed box systems but nothing new, the plots look like they're right out of Thiele and Small's papers. His TL analysis where he claims the use of vector calculus and draws a picture of the wave? You have got to be kidding? Sorry I'm LOL here!
There is not one equation in his TL section. His picture in Figure 5 is his wishful thinking of how it works - he completely misses the l/4 resonance. The air is neither compressed or uncompressed ... please this is nonsense.
Ok, yes this thread is a waste of time.
Let me just mention that I find flaws in most papers that I read concerning TLs. The Olney papers look solid and I also found Martin Kings work to be excellent as far as a quick read goes. I did note one error in his interpretation of his measured data, still don't know if I convinced him that his measurements were better than he thought.
TLs are a difficult subject.
There is not one equation in his TL section. His picture in Figure 5 is his wishful thinking of how it works - he completely misses the l/4 resonance. The air is neither compressed or uncompressed ... please this is nonsense.
Ok, yes this thread is a waste of time.
Let me just mention that I find flaws in most papers that I read concerning TLs. The Olney papers look solid and I also found Martin Kings work to be excellent as far as a quick read goes. I did note one error in his interpretation of his measured data, still don't know if I convinced him that his measurements were better than he thought.
TLs are a difficult subject.
know its not helpful to the discussion but I liked Earl's "...be done with it part"
outdoors
tapered line response grabbing an Eminence woofer having ~19hz fs and 0.46 qts.
Sander's article line Eclipse W1038R (Qts ~0.3, fs ~20) box size 48"x18"x12" both cabinets ~same size /6CF -
hacked a big stupid slot in it to see what that would do
sealed = green, blue = Fulmer opening, violet = regular 50sq.in. rectangular slot
W1038R even with adequate stuffing had poor upper bass and a nasty thud - good BP4 or K15 like Exemplar was better for me.
owned only one commercial "t-line" ESS transtatic, and built huge Weems thing back in ~68
I'd like to hear one done right
outdoors
tapered line response grabbing an Eminence woofer having ~19hz fs and 0.46 qts.
An externally hosted image should be here but it was not working when we last tested it.
Sander's article line Eclipse W1038R (Qts ~0.3, fs ~20) box size 48"x18"x12" both cabinets ~same size /6CF -
hacked a big stupid slot in it to see what that would do
sealed = green, blue = Fulmer opening, violet = regular 50sq.in. rectangular slot
An externally hosted image should be here but it was not working when we last tested it.
W1038R even with adequate stuffing had poor upper bass and a nasty thud - good BP4 or K15 like Exemplar was better for me.
owned only one commercial "t-line" ESS transtatic, and built huge Weems thing back in ~68
I'd like to hear one done right
Last edited:
The above quotes tell me that like Martin King, you really don't have a clue what the optimal operating principals are for a proper transmission line speaker loading. For your information, in a lot of TL designs, the first "undesireable" harmonic multiple of the driver's peak or free air resonance is actually desireable. Furthermore, the only undesireable backwave outputs are the odd ordered multiples of the driver's resonant peak. Even ordered multiples are standing waves which by definition do not produce pressure levels at the driver that are different from the system at rest (at equilibrium with the environment). They are standing waves within the line that cancel themselves out. As I've posted elsewhere, even Natkaniec confuses the "anti node's" importance.
Getting back to the premise of "desireability" - if a driver had an Fs of 15hz and was coupled to a 23hz pipe, there is little doubt in my mind that the third harmonic of the fundamental (69hz) would be considered "useful" or "desireable" output by most designers. The fifth through ninth harmonic are what the 1/5 wavelength offset seek to suppress for your information. A simple calculation would bear that out.
For example, an offset design of pipe length 85 inches corresponding to pipe frequency of 40hz optimized for a driver Fs of 27hz would place the driver at 17 inches from the closed end of the pipe. This would create primary out of phase cancellation of frequency about 160hz at the beginning of the line behind the driver. This tuning corresponds to a resonant frequency that is exactly half way between the 5th and 7th multiple - achieving maximal cancellation of both unwanted resonances.
The above is the real explanation for why offsets have been used long before Martin King started playing with other people's differential equations. And I've yet to hear one of Martin's "disciples" produce such an explanation - including you - most likely because you don't have the knowledge or background in acoustics and haven't done the research into what many real acoustical engineers and scientists have done since the first papers on transmission line theory were published back in the 1930's.
Your quarrel seems to be with Martin. He's an easy person to contact - if you wish to argue with him, please do so somewhere else.
Your posts in this thread are of a persistently hostile nature - if you have insights to share, please, document a project or write up a set of design guidelines.
In the meantime, the rest of us shall continue using Martin King's models as what they are - useful tools. Regardless of whether you agree with his methods of development, his models give accurate results.
Your posts in this thread are of a persistently hostile nature - if you have insights to share, please, document a project or write up a set of design guidelines.
In the meantime, the rest of us shall continue using Martin King's models as what they are - useful tools. Regardless of whether you agree with his methods of development, his models give accurate results.
My posts are factual - based on actual measurements which have been recorded by most everyone on here - including those who use King's spreadsheets. The quarrel I have, if any "quarrel" should exist, is with the general claims of the real and perceived differences between transmission line loading, bass reflex loading, and aperiodic loading. In point of fact, it is others who have entered into this thread scoffing at the work of an accredited university professor without giving the slightest cursory examination into the facts that are presented therein - others who use terminology like "B......S..." and "tarted up". It is quite evident that most of the guilty parties are friends or associates of Martin King. While that may be a significant number of individuals, I suspect that is not "everyone" - and you sir, do not speak for "everyone".
I have to agree with TheSeeker that your posts appear antagonistic and ad hominem in their nature.
If you are trying to make a case for your argument, you are not achieving that goal.
HI Scott
I missed most of the bantere, but its nice to see an accurate comment like this once in awhile.
The electro-acoustical systems "couple" and affect each other. This is NOT "impedance matching" at all, its just coupling. Because the acoustical impedance of the TL, can be very high, it couples quite strongly to the driver. But only a horn could be considered as an impedance matching device, except that an Acoustic Lever is also an impedance matching device. A TL is not - its simply a loading device with delay.
So what do think a "loading device with a delay" constitutes? My friend, resonant cavities produce acoustic loading - by definition. This is fundamental to any understanding of acoustics. "Acoustic loading" is by definition acoustic impedance. The "matching" that's taking place in a properly designed transmission line speaker is the coupling of the driver to a resonant cavity whose acoustic impedance is the inverse of the driver's acoustical impedance - which has the pretty obvious effect of maximizing acoustic output and in doing so - minimizing losses to energy storage of the driver - as witnessed by the effective suppression of electrical impedance. These are fundamental concepts of acoustics.
By contrast, horns are not meant to be "resonant" cavities. Their purpose is to confine the spread of acoustical energy into a narrower beam to improve the transfer of acoustic energy in the direction of intended output. They operate most effectively when reflection - the principle upon which they are based - is kept to a minimum. Your own papers and treatises on this subject support this notion fully - multiple reflections in a horn are an anathema to fidelity and desired output. And generally the deeper the horn, the more opportunity for multiple reflections and more skewed temporal and spatial response. Proper horns have exponential or linear (or some combination thereof) tapers outward, transmission lines have inward tapers. Horns are methods for "unloading" a driver - not loading one. The acoustic loading of a compression driver in a horn results in reduced waveform linearity according to Baranek - not desireable. This is what exponential horn tapers are designed to reduce.
But we digress. Let's stick to understanding one theory at a time. transmission line theory and horn theory have opposite goals. They are in many respects - the reverse of one another. And trying to prove their equivalence either in design or in ultimate purpose is futile.
By contrast, horns are not meant to be "resonant" cavities. Their purpose is to confine the spread of acoustical energy into a narrower beam to improve the transfer of acoustic energy in the direction of intended output. They operate most effectively when reflection - the principle upon which they are based - is kept to a minimum. Your own papers and treatises on this subject support this notion fully - multiple reflections in a horn are an anathema to fidelity and desired output. And generally the deeper the horn, the more opportunity for multiple reflections and more skewed temporal and spatial response. Proper horns have exponential or linear (or some combination thereof) tapers outward, transmission lines have inward tapers. Horns are methods for "unloading" a driver - not loading one. The acoustic loading of a compression driver in a horn results in reduced waveform linearity according to Baranek - not desireable. This is what exponential horn tapers are designed to reduce.
But we digress. Let's stick to understanding one theory at a time. transmission line theory and horn theory have opposite goals. They are in many respects - the reverse of one another. And trying to prove their equivalence either in design or in ultimate purpose is futile.
Acoustic loading is half the story. You can effectively couple the driver to the box, but the other half is the box to the room.
Horns are an impedance matching device between the speaker and room. The speaker can be a simple driver or a driver and an enclosure.
While my knowledge on transmission lines is a little weak, to my understanding, transmission lines can have expanding or contracting tapers and they need not be linear.
Horns are an impedance matching device between the speaker and room. The speaker can be a simple driver or a driver and an enclosure.
While my knowledge on transmission lines is a little weak, to my understanding, transmission lines can have expanding or contracting tapers and they need not be linear.
Except that they are not correct. The mechanical impedance of a driver, even after it is reflected into the acoustical domain, is orders of magnitudes greater than the acoustical impedance of a damped TL (in air, acoustics is a very soft medium - not so in water). This large impedance level difference is not "matching", at least not by my deffinition, but you seem to be saying that it is. The rest of your claims are false by virtue of their relying on a false assumption of "matching".
A horn with an extremely low flare rate becomes a TL - they are certainly NOT the "reverse of one another". You stuff a TL, I stuff waveguides - not so different.
I am not clear on exactly what it is that you are claiming about TLs. They have been modeled and understood for decades as many people have pointed out. Is your claim that some Prof in Poland discovered a new concept that changes all of that? Or do you claim that there are errors in all the previous work?
SPEAK can model TLs and I'll done some work in that respect. The math in SPEAK is beyond reproach since hundreds to thousands of products have been designed with it for nearly thirty years. To say that it is wrong would be a big claim that you would have to substantiate with some real math and data - none of which have I seen as yet.
- Status
- Not open for further replies.