Hi everyone,
I'm new to speaker design and I just started learning some theory and trying to design a set of bass-reflex speakers. However, I stumbled upon some information about transmission lines and now I have some questions that I can't really find answers to online.
First and most important question I have: I don't really understand where ports end and transmission lines begin. I've prepared a drawing with ports of different lengths and shapes (in the attachment). Can you tell me at which point on my picture the port becomes a transmission line?
Second question: if I decide that my port is so long that it qualifies as a transmission line does this mean that I have to stuff it with damping material?
Final question: Why do all the transmission line designs that I see use a full-range driver? I really haven't seen many tweeters on those designs.
Thanks in advance!
I'm new to speaker design and I just started learning some theory and trying to design a set of bass-reflex speakers. However, I stumbled upon some information about transmission lines and now I have some questions that I can't really find answers to online.
First and most important question I have: I don't really understand where ports end and transmission lines begin. I've prepared a drawing with ports of different lengths and shapes (in the attachment). Can you tell me at which point on my picture the port becomes a transmission line?
Second question: if I decide that my port is so long that it qualifies as a transmission line does this mean that I have to stuff it with damping material?
Final question: Why do all the transmission line designs that I see use a full-range driver? I really haven't seen many tweeters on those designs.
Thanks in advance!
Attachments
I think I see where you are coming from but let's say (f) can still be a classic bass reflex.
But we also have to keep in mind that every port itself is also a transmission line. That is why in your favourite box software, the "1st port resonance" (or whatever else the wording might be) is prominently displayed when you configure your port dimensions.
But we also have to keep in mind that every port itself is also a transmission line. That is why in your favourite box software, the "1st port resonance" (or whatever else the wording might be) is prominently displayed when you configure your port dimensions.
1, on "where ports end and transmission lines begin", I will state what I think. If we start out with a transmission line design it will never become a bass reflex, unless we expressly add a port to it as in an ML-TL. This is because for a Helmholtz resonator to work (as a port) an abrupt cross section change is required to form a boundary which is what transmission line designs want to avoid. However for the other way around, if we start out with a bass reflex, every port is a transmission line regardless of port length. By that I did not mean every port is a textbook orthodox transmission line, but that every port is an unintended transmission line amendable to the same analysis as intentionally designed transmission lines.
2, If you actually stuff the vent it becomes a resistive vent or an aperiodic membrane. Dynaudio used to make something called a Variovent.
3, TL can be applied to 2-way, 3-way or 4-way designs, not just full range drivers. You see TLs with full range drivers a lot because they are typically small diameter drivers in need of some bass extension. It is between TL and bass reflex and the vibrant community here is leaning towards the former. I can name advantages like lower group delay and better control of cone excursion below tuning.
2, If you actually stuff the vent it becomes a resistive vent or an aperiodic membrane. Dynaudio used to make something called a Variovent.
3, TL can be applied to 2-way, 3-way or 4-way designs, not just full range drivers. You see TLs with full range drivers a lot because they are typically small diameter drivers in need of some bass extension. It is between TL and bass reflex and the vibrant community here is leaning towards the former. I can name advantages like lower group delay and better control of cone excursion below tuning.
cyberstudio, thanks for you reply! It is really helping and I think I'm now getting close to understanding what's what. I'm currently trying to process what you said with the limited knowledge that I have. I'm going to take some time to digest these ideas and going to reply to you in more detail once I had a chance to give it a good though.
All are transmission line (but longer it gets, more different than bass reflex is).
ML-TL simulators can simulate any bass reflex enclosure more precisely.
But there are many variants of transmission line.
Last edited:
Have a look at the PMC speakers, commercially available transmission line based speaker, not a true 1/4 wave but their site gives a description. If you search pictures of the OB1 model you will see a 3 way TL with the mid driver in a plastic tube enclosure, pretty neat. TDL also made TL based speakers.
Some of the more modern ways to look at loudspeaker enclosure theory do not distinguish between closed / bass reflex / transmission line / horn anymore, at least not in the way they did back in the last millennium. The "acoustic impedance" approach (take a look at the Martin King website) can be used to model all kinds of loudspeaker enclosures, not just quarter wave tubes. There are a number of software packages that work on these principles, and they provide accurate physical models of the different loudspeaker enclosures. You might want to take a look at Chapter 2 of the AJ Horn manual*, which describes how things go from a horns to transmission lines to bass reflex and closed boxes (and how they are covered by the same physics). I'd guess that if speaker enclosures would be re-invented today, the distinction between the different enclosure types would turn out differently.
*: it's in German, and I am not aware of an English equivalent. Maybe try an online translator tool if needed.
*: it's in German, and I am not aware of an English equivalent. Maybe try an online translator tool if needed.
Thanks everyone for your replies! I'm currently trying to process all the information and I'm going to address the points that you all mentioned individually once I had a chance to go through all the material that you provided. Will probably take a few days to finish all the reading.
In the meanwhile I thought I would just summarize what I gathered so far, and please correct me where I'm wrong. So:
a) Any kind of vent in enclosure can be seen as a kind of transmission line. Different kinds of transmission lines are: TQWP, ML-TL, Back-loaded horn, bass-reflex etc.
b) Main purpose of any transmission line (including bass reflex) is to act like a mass-spring system that is being driven by the back of the woofer above the system's resonance frequency and therefore inverting the phase of the waves. Waves then come out of the vent and interfere constructively with the waves emitted from the front of the speaker.
c) Additional functionality of transmission line is that its resonance frequency can be matched to the woofer's resonance frequency (quarter length principle), so the transmission line takes energy out of the woofer at resonance and limits the woofer's excursion.
d) Main difference of bass reflex is that it uses Helmholtz principle (abrupt change of surface area in the transmission line) to separate the air into the "spring" part (main volume of the enclosure) and the "mass" part (air inside the port). This Helmholtz resonator is cheap to manufacture and helps to achieve reasonably low resonant frequency in small enclosure. But it doesn't work as nicely as a proper long transmission line without abrupt changes in area.
e) By adding a port to the end of transmission line we can produce ML-TL which is somewhere in between bass-reflex and proper transmission line.
Is any of this right?
Thanks!
In the meanwhile I thought I would just summarize what I gathered so far, and please correct me where I'm wrong. So:
a) Any kind of vent in enclosure can be seen as a kind of transmission line. Different kinds of transmission lines are: TQWP, ML-TL, Back-loaded horn, bass-reflex etc.
b) Main purpose of any transmission line (including bass reflex) is to act like a mass-spring system that is being driven by the back of the woofer above the system's resonance frequency and therefore inverting the phase of the waves. Waves then come out of the vent and interfere constructively with the waves emitted from the front of the speaker.
c) Additional functionality of transmission line is that its resonance frequency can be matched to the woofer's resonance frequency (quarter length principle), so the transmission line takes energy out of the woofer at resonance and limits the woofer's excursion.
d) Main difference of bass reflex is that it uses Helmholtz principle (abrupt change of surface area in the transmission line) to separate the air into the "spring" part (main volume of the enclosure) and the "mass" part (air inside the port). This Helmholtz resonator is cheap to manufacture and helps to achieve reasonably low resonant frequency in small enclosure. But it doesn't work as nicely as a proper long transmission line without abrupt changes in area.
e) By adding a port to the end of transmission line we can produce ML-TL which is somewhere in between bass-reflex and proper transmission line.
Is any of this right?
Thanks!
The difference follows from the type of resonance being used to enhance the bass response.
In a bass reflex speaker the mass of air in the port/line moves as an incompressible slug and "bounces" on the spring created by compressing the volume of air trapped within the enclosure. This is commonly referred to as a Helmholtz resonator. In a port/line you also have compressible resonances due to sound waves travelling up and down the line and being reflected at the ends. These are present in a bass reflex port and are a problem if they fall within the passband of the driver.
In a transmission line design the idea is to use a small amount of damping to let through the lowest frequency compressible resonance to enhance the bass response and to absorb the higher frequency resonances. If there is a chamber then there will also be an incompressible Helmholtz resonance present but it will almost certainly be at too a low frequency to be within the passband of the speaker.
There is also a resonance present from the mass and stiffness of the driver but this is not what distinguishes the operation of the two different types of cabinet loading.
The shape of the line, the location of the driver, the location and amount of the stuffing, chambers at various locations,... can have an influence on the performance but it doesn't change the dominant physical mechanisms which is either a Helmholtz resonance or a pipe resonance.
A vent tube or duct is a 1/2 wave resonator as it's a pipe open at both ends. So in a sense it's a TL because 'transmission line' has unfortunately become a catch-all term that is used broadly to refer to pretty much anything that happens to involve pipe resonance (eigenmodes / standing waves) rather than cavity (Helmholtz) resonance which assumes a uniform internal air-particle density & no standing waves. Note the caveat: I'm talking here about the vent / duct only.
Not really. A bass reflex enclosure designed as such is certainly not a TL variation, nor is its T/S derived vented box descendent. As noted, a duct nominally has its own resonant modes, but these are not typically a functional part of a box designed on Helmholtz assumptions and only rarely impact upon tuning, usually in relatively extreme cases where an excessively large / long duct is present. Onkens for example can suffer from unwanted vent harmonics due to their size / length, which may explain why some people like them. They aren't strictly speaking supposed to be there however, nor are they accounted for in the design so are not a functional aspect of it.
TQWP is sometimes written as TQWT, and is often used to describe an expanding (toward the terminus) pipe. The latter is often best described as a conical or parabolic horn depending on configuration, albeit one that isn't necessarily impedance matched down to the QW cutoff frequency. A back [loaded] horn is likewise a more or less extreme variation depending on the design specifics. Ditto for MLTLs which are more or less extreme QW / TL variations, depending on design.
The best message here is: try not to get too hung up on definitions which are not universally applied in the same way. 'Transmission Line' itself as noted is a term that is often used by people to describe boxes that are functional opposites of each other. What is more important is that you understand what it is you are doing & the operating physics thereof.
See above: a BR is not a transmission line. The mass-spring analogy is a little too vague to be broadly accepted, particularly since not all TLs are designed to provide resonant action supporting the LF. Some are designed for the sole purpose of providing the flattest possible impedance with zero other considerations; these are typically aperiodic & have no active output from the terminus. Likewise, sealed TLs (pipes sealed at both ends) are TLs, but by definition contribute no active resonant support. So it depends what you mean by 'TL'.
Not especially. There is no requirement for that, and a pipe tuned to Fs does not necessarily provide optimal limitation of driver excursion.
A bass reflex & its vented box descendent as noted assume Helmholtz (cavity) resonance. They are designed on the assumption of a uniform internal air particle density & no standing waves are considered to be present. An MLTL deliberately stretches one physical dimension of an enclosure relative to the others in order to generate and use standing waves as an active part of the enclosure alignment. For practical purposes the transition from one to the other can be said to occur when for a given Vb and vent, one dimension of the enclosure has been stretched sufficiently for the alignment shape to change relative to what is predicted under Helmholtz assumptions, i.e. what you might expect for a given Vb & vent dimensions in something like WinISD, Bassbox or similar.
Last edited:
It is best to consider a transmission line as a kind of box design. The possibility of quarter-wave behaviour in the vent of of a bass-reflex is really a different thing, and one usually tries to supress that. Althou if the vent gets long enuff one gets to a TL design historically called a Daline, where one has a pre-chamber with a long skinny line/vent tapped into it.
I’d suggest that you are mixing up the purpose with the way it works. The purpose can be varied, but is most often associated with bass enhancement — using part of the back radiation to enhance the front radiation, but the design goal can be anywhere from there to a completely aperiodic line were the back wave is completely suppressed (and the impedance flattened).
That is one of the tenants of classical TL theory… something you can toss aside with modern design using one of the TL modelers.
A BR will transition to a TL as one increases one dimension such that it becomes considerably larger than the other 2.
A BR is short enuff that any 1/4 wave resonances are moot, as one increases one dimension the quarter wave resonance becomes significant and changes the behavious of the box.
By reducing the terminus of a pipe one adds an extra low-pass element which allows one to shorten the line for the same Fb, and simplify box construction. With the extra LP one can use less damping so that more of the desired fundemental is left to reinforce the low bass.
diyFEST SUnday over here on the island and i have 4 TLs currently in my living room (3 FRs and 1 a WAW with 2 TLs, an ML-TL for bass, and a 10:1 tapered aperiodic line for the midTweeter). You see so many FR TLs because a couple of us (at least) are FR fanatics and TL lovers. Bailey’s seminal TL was a 2-way, the Radfords that followed were 3-way, and many of the IMFs and TDLs were 4-way.
dave
Attachments
Ok, so the transmission line is something that makes use of pipe resonances and the bass-reflex is something that makes use of the Helmholtz resonance? And since port is also a pipe it will also have some pipe resonances but they are usually ignored in classic bass-reflex design approach?
Is there a ML-TL simulator that you can recommend? I saw that your wrote your own speaker simulator software. Does that support ML-TL calculations?
Thanks! I took a lot at the Martin King's website and there is definitely a lot of good information there. I'm planning to read his "Anatomy of the transmission line loudspeaker" as soon as I can find time. It's a pretty big document.
I also looked at the AJ Horn manual. Google did a decent job of translating it to English for me. It was quite insightful.
Are there any (preferably free) software packages that you personally would recommend?
I’d suggest that you are mixing up the purpose with the way it works. The purpose can be varied, but is most often associated with bass enhancement
A BR will transition to a TL as one increases one dimension such that it becomes considerably larger than the other 2.
A BR is short enuff that any 1/4 wave resonances are moot, as one increases one dimension the quarter wave resonance becomes significant and changes the behavious of the box.
Thanks, your reply provides a log of insight! What software did you use go generate these simulations?
Correct.
Mostly correct. They are not always 'ignored' per se; most reasonable simulation packages will note & display at least the first major vent resonant mode, largely so people can avoid or minimise them as far as practical. However, they are not a functional part of the underlying operating principle. In most cases they are of too high a frequency and too low an amplitude to have any impact on overall behaviour.
FWIW the graphic Dave provided was done by Martin [King] as part of an early project; it was done in an FEM package.
Free software = Hornrsp and AkAbaK, both of which are very accurate although you'll have to put some time into learning the interfaces. The former is the easier proposition for most people.
Free software = Hornrsp and AkAbaK, both of which are very accurate although you'll have to put some time into learning the interfaces. The former is the easier proposition for most people.
Leonard Audio's TL program is good and more flexible plus generally easier to learn than Hornresp plus the latter can only do end/bottom location vents, though it can be approximated along the [ML]TL's length by using the 'path' delay feature in the Loudspeaker Wizard's 'chamber' window.
http://www.diyaudio.com/forums/soft...n-line-modelling-software-17.html#post5160274
Help File | Leonard Audio
GM