• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Transmission-line speakers

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA
I think the practical application for transmission lines in hifi speakers is a somewhat niche thing. You have to have a bigger and more complex box, and you get less bass. However, in applications where the bass driver is also reproducing midrange, the transmission line does an unusually good job of damping the rearward radiation of the speaker than other enclosure types. This is the idea with the Nautilus, and it makes sense.

Consider that a wall with two layers of sheetrock and studs with some insulation will decrease sound transmission by say, 40db. This is frequency dependent, and at low frequencies, performance is much worse. Now look at your speaker. At best it's 1" of mdf, and then there's a port opening...and the woofer, which has a fairly significant area, is made out of a thin sheet of paper or aluminum or plastic. When you look at a speaker that way, you realize that the sound in the box really needs to be dealt with carefully, because as loud as the speaker is outside the box, it's just as loud inside.

There's a few solutions to this problem. The least elegant, and least effective, is to line the box with sound absorbing materials. Well, that's great, except sound absorbers don't do much below 100hz. Stuffing the box with fiberglass might be a bit better in the bass, but then your port doesn't work as well.

The next best solution is to put the speaker at the end of a long pipe, so the sound has to travel through a ton of insulation before it reflects back through the woofer. That works great, and since the insulation doesn't work at bass frequencies, you can leave the end of the pipe open, and the bass comes out! If the length of the pipe is sufficient, you well get a bass boost AND your rearward radiation is diminished.

The most elegant solution...no box at all. Dipoles such as the LX521, or Planar speakers, have their own issues relating to distortion, bandwidth, complexity and chaotic directivity, but the lack of any box radiation coming through the woofer is pretty great.
 

Kal Rubinson

Master Contributor
Industry Insider
Forum Donor
Joined
Mar 23, 2016
Messages
5,272
Likes
9,785
Location
NYC
When you look at a speaker that way, you realize that the sound in the box really needs to be dealt with carefully, because as loud as the speaker is outside the box, it's just as loud inside.
Classic TL speaker designers emphasized that one needed a generous volume behind the driver (with some loose damping material) and a generous opening from that into the initial length of the damped line.
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
I think the practical application for transmission lines in hifi speakers is a somewhat niche thing. You have to have a bigger and more complex box, and you get less bass. However, in applications where the bass driver is also reproducing midrange, the transmission line does an unusually good job of damping the rearward radiation of the speaker than other enclosure types. This is the idea with the Nautilus, and it makes sense.

Consider that a wall with two layers of sheetrock and studs with some insulation will decrease sound transmission by say, 40db. This is frequency dependent, and at low frequencies, performance is much worse. Now look at your speaker. At best it's 1" of mdf, and then there's a port opening...and the woofer, which has a fairly significant area, is made out of a thin sheet of paper or aluminum or plastic. When you look at a speaker that way, you realize that the sound in the box really needs to be dealt with carefully, because as loud as the speaker is outside the box, it's just as loud inside.

There's a few solutions to this problem. The least elegant, and least effective, is to line the box with sound absorbing materials. Well, that's great, except sound absorbers don't do much below 100hz. Stuffing the box with fiberglass might be a bit better in the bass, but then your port doesn't work as well.

The next best solution is to put the speaker at the end of a long pipe, so the sound has to travel through a ton of insulation before it reflects back through the woofer. That works great, and since the insulation doesn't work at bass frequencies, you can leave the end of the pipe open, and the bass comes out! If the length of the pipe is sufficient, you well get a bass boost AND your rearward radiation is diminished.

The most elegant solution...no box at all. Dipoles such as the LX521, or Planar speakers, have their own issues relating to distortion, bandwidth, complexity and chaotic directivity, but the lack of any box radiation coming through the woofer is pretty great.
Or you brace up subwoofers really good and thus postpone resonances higher up in frequency, then steep filters around 80-100 Hz and you get rid of that rubbish. No stuffing in the subwoofer. :D

See:

 
  • Like
Reactions: 617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA
Classic TL speaker designers emphasized that one needed a generous volume behind the driver (with some loose damping material) and a generous opening from that into the initial length of the damped line.
Yes and I think this is still best practice. Sometimes you see designs where the line goes down straight from the driver, but ideally you want the line to go back from the driver as far as possible, and then change direction with 45 degree facets.

The use of a 'chamber' behind the driver changes the performance of the line somewhat, and Martin King's calculators take this into account. The chamber is a feature often seen in back loaded horns as well.
 

pseudoid

Master Contributor
Forum Donor
Joined
Mar 23, 2021
Messages
5,116
Likes
3,416
Location
33.58 -117.88
At the lowest frequencies, yes, but when the frequency increases the sound from the port will eventually be delayed more than one cycle, and this sound adds to the much earlier sound from the front, creating a mess in the time domain. If the frequency response from the port output is designed to fall off sufficiently, this can be avoided.
This is exactly what I was trying to get an answer to; by asking the levels of harmonic 'coloration' contributed by the TL attempt...
[without sounding like/if I was criticizing your great post]
It had seemed to me like both time-domain and phase (Ø) shift issues would arise, unnecessarily complicating the original signal.
10Q
 
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
This is exactly what I was trying to get an answer to; by asking the levels of harmonic 'coloration' contributed by the TL attempt...
[without sounding like/if I was criticizing your great post]
It had seemed to me like both time-domain and phase (Ø) shift issues would arise, unnecessarily complicating the original signal.
10Q
And you can never avoid the phase issue, because the sound will be delayed approximately by the time it takes to travel the distance through the channel. When this delay approaches 1 wavelength and more, the summed sound output will be a mess. Only way to make it work is to attenuate the output from the pipe at higher frequencies. If this attenuation is achieved, it will work - over a limited frequency range.
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
It might have been a bit negative on my part in this thread. It should be fun with speakers.:)

So instead of highlighting negative examples can anyone point to some good TL speaker that measure well. Which costs around $ 1000.
 
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
To show show the port output and driver output sums up in the F2 speaker I used in the example in post #1 - it works, but only if the port output is limited in frequency range, and the port outpt must of course be free of severe resonances:

F2.png
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
To show show the port output and driver output sums up in the F2 speaker I used in the example in post #1 - it works, but only if the port output is limited in frequency range, and the port outpt must of course be free of severe resonances:

View attachment 166463
There seems to be potential there. :)

Your graph. Theoretical calculation or measured with a microphone?
Are there calculation programs for TL constructions?

What do you think the level of distortion will be?

Which drivers do you use in your example?

Note I do not question your results. I am just curious. I do not mind any design principle when it comes to speakers. As long as they measure well and thus sound good.:)
 
Last edited:

pseudoid

Master Contributor
Forum Donor
Joined
Mar 23, 2021
Messages
5,116
Likes
3,416
Location
33.58 -117.88
... As long as they measure well and thus sound good.:)
That is a big ask and is a conundrum we know doesn't have one single answer!
Especially when that equation gets even more complex, since price is known to be a big 'driver' [in addition to the cost of the TL driver]
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
I only found 1 transmission line speaker that is not a kit or used. They are rare especially @ the $1000 price range.

$269.98
$334.98
Euro 532.00
$599.98
Euro 820.00
$1018.00 not a kit.
Euro 1230.00
Thanks! Interesting. :)

I will read more about them. Took the first one you tipped about:

Dayton Audio MK442T

I do not know. Maybe that's what you're expecting, isn't it? Then $ 270 for a couple should be weighed in, I think. Price vs performance that is.:)


Edit:

I looked a little quickly at the next one in the list. I'll watch more tomorrow: :)

TriTrix MTM
Frequency response: ± 6 dB from 40-20,000 Hz
 

Attachments

  • image_preview2 (4).jpeg
    image_preview2 (4).jpeg
    19.5 KB · Views: 90
  • image_preview2 (2).jpeg
    image_preview2 (2).jpeg
    15.6 KB · Views: 81
  • image_preview2 (1).jpeg
    image_preview2 (1).jpeg
    22.8 KB · Views: 98
Last edited:

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
That is a big ask and is a conundrum we know doesn't have one single answer!
Especially when that equation gets even more complex, since price is known to be a big 'driver' [in addition to the cost of the TL driver]
That's true, what does it mean to measure well? Relatively maybe but, well relative to what? :)
 
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
There seems to be potential there. :)

Your graph. Theoretical calculation or measured with a microphone?
Are there calculation programs for TL constructions?

What do you think the level of distortion will be?

Which drivers do you use in your example?

Note I do not question your results. I am just curious. I do not mind any design principle when it comes to speakers. As long as they measure well and thus sound good.:)
This is a simulation, the real speaker measures similar.

Distortion is quite low in this speaker, at any practical loudness level in small rooms. It features a 12" pro driver with strong motor and moderate moving mass. It is sort of the opposite of a traditional TL - this is high spl, high output, good efficiency. The bass is like solid concrete - very hard, totally dry, no low bass.

I used the F2 as example because it has been discontinued for some time now - so no commercial interest, and because it shares the basic challenges with all variations of back-loaded pipes or horns or TL.

This speaker was designed using simulation - the complete electroacoustic system was modelled, including all parts of the internal horn channel and ports and damping material inside. The crossover with quite substantial eq to achive flat response is also included. This is really the only way to make such a design, if you want a predictable and good result.

For any pipe or TL or horn, you can use Hornresp to experiment with different designs. It will show you a good enough analysis that you can use it to adjust and tune a bit, and then build a speaker. Build it, just for fun, to test and hear.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,313
Location
UK
This speaker was designed using simulation - the complete electroacoustic system was modelled, including all parts of the internal horn channel and ports and damping material inside. The crossover with quite substantial eq to achive flat response is also included. This is really the only way to make such a design, if you want a predictable and good result.
May I ask what software you use, especially for simulating the material inside?
 
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
It might have been a bit negative on my part in this thread. It should be fun with speakers.:)

So instead of highlighting negative examples can anyone point to some good TL speaker that measure well. Which costs around $ 1000.
But it is also fun to observe those not-so-fortunate designs. Here is one - the same F2, but now I decide to remove most of the damping inside, and this is the result - a speaker with a truly unique sound:

F2_NOD.png
 
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
May I ask what software you use, especially for simulating the material inside?
You can use software like Matlab, or something dedicated to acoustic systems, like Akabak. The real work and knowledge is in making the models. Damping material is tricky, as there is a difference between viscous and acoustic resistive damping, both properties must be modeled, and it can be difficult to determine the physical numbers for the material in use.

I should add that this is something I have been working on for several years, in my company, developing loudspeakers and sound solutions.

Like I mentioned above - you can use Hornresp to simulate such pipes/TL systems, it gives good information and it is not very complicated to get started.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,313
Location
UK
You can use software like Matlab, or something dedicated to acoustic systems, like Akabak. The real work and knowledge is in making the models. Damping material is tricky, as there is a difference between viscous and acoustic resistive damping, both properties must be modeled, and it can be difficult to determine the physical numbers for the material in use.

I should add that this is something I have been working on for several years, in my company, developing loudspeakers and sound solutions.

Like I mentioned above - you can use Hornresp to simulate such pipes/TL systems, it gives good information and it is not very complicated to get started.
These days I use MatLab for almost everything :) Like you, I also believe in simulation before prototyping. I designed Silver 5L the same way.

I was interested in damping material simulation but you have answered my question. Thank you.
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
You can use software like Matlab, or something dedicated to acoustic systems, like Akabak. The real work and knowledge is in making the models. Damping material is tricky, as there is a difference between viscous and acoustic resistive damping, both properties must be modeled, and it can be difficult to determine the physical numbers for the material in use.

I should add that this is something I have been working on for several years, in my company, developing loudspeakers and sound solutions.

Like I mentioned above - you can use Hornresp to simulate such pipes/TL systems, it gives good information and it is not very complicated to get started.
Exciting!

There seems to be potential. Try ploping the elements into boxes. Microphone,
then measure and see if theory and practice agree.

Fill the boxes and measure before each change. Measure. In with more or less moff, stopping, change, more less padding, measure, in ....

... transmission lines should be avoided as far as possible.

... studying the theory ..., taking part in various dissertations and trying out the principle practically, is that one should not even think the idea.


Edit:
I think you should really like to do TL, Horn and so on if you want to do well. See it as a challenge, love to carve, love to test with various fillings and padding and measure. Like this guy did with a couple of bass boxes. By the way, I have heard them. Good bass.
If you like it so absolutely.

If we are talking subwoofer.I'm too lazy. I would use the same cavity volume and insert two drivers . One with the engine outwards, the other inwards. Tighten together to reduce second ton distortion.Really braced. No stuffing/filling. Seald so I did not have to get a port Then microphone, measure and EQ.My guess it would take much shorter time that DIY, would have the potential to get better (depending on the choice of driver of course).:)

 

Attachments

  • 51645038279_4de39ed01e_c.jpg
    51645038279_4de39ed01e_c.jpg
    138.4 KB · Views: 146
  • 51645237880_70312c61b2_c.jpg
    51645237880_70312c61b2_c.jpg
    234 KB · Views: 124
  • 51660573954_a6fd692174_c.jpg
    51660573954_a6fd692174_c.jpg
    138.6 KB · Views: 137
  • 51658551407_1d249774af_c.jpg
    51658551407_1d249774af_c.jpg
    168.4 KB · Views: 166
  • 51687056333_bf80f346b4_c.jpg
    51687056333_bf80f346b4_c.jpg
    74.5 KB · Views: 180
  • 51687477709_14565850c7_c.jpg
    51687477709_14565850c7_c.jpg
    128.5 KB · Views: 230
Last edited:
Top Bottom