• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily 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

sarumbear

Major Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
3,256
Likes
2,929
Location
Southampton, 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.
 

tuga

Major Contributor
Forum Donor
Joined
Feb 5, 2020
Messages
3,202
Likes
3,218
Location
Oxford, England
Speaking of prototypes:

OlqLuen.jpg

source: https://www.bowerswilkins.com/en-gb/blog/products/history-of-nautilus
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
2,397
Likes
1,786
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: 49
  • 51645237880_70312c61b2_c.jpg
    51645237880_70312c61b2_c.jpg
    234 KB · Views: 50
  • 51660573954_a6fd692174_c.jpg
    51660573954_a6fd692174_c.jpg
    138.6 KB · Views: 56
  • 51658551407_1d249774af_c.jpg
    51658551407_1d249774af_c.jpg
    168.4 KB · Views: 62
  • 51687056333_bf80f346b4_c.jpg
    51687056333_bf80f346b4_c.jpg
    74.5 KB · Views: 70
  • 51687477709_14565850c7_c.jpg
    51687477709_14565850c7_c.jpg
    128.5 KB · Views: 69
Last edited:
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Manufacturer
Joined
Apr 25, 2019
Messages
688
Likes
1,263
Location
Norway
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.
Matlab is probably a good idea. I have used Matlab for dynamic systems simulations, back when I was working with cybernetics/control systems.

Nice speaker, and I observe the bass is moved out to separate boxes, like it should be.
 
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Manufacturer
Joined
Apr 25, 2019
Messages
688
Likes
1,263
Location
Norway
I think you should really like to do TL, Horn and so on if you want to do well. See it as a challenge,
I like the challenge of everything that is difficult. And today, with the computers and software available, it is a very different world from back when trying to calculate of a horn from simple formulas using a calculator.
 

sarumbear

Major Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
3,256
Likes
2,929
Location
Southampton, UK
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Manufacturer
Joined
Apr 25, 2019
Messages
688
Likes
1,263
Location
Norway
Was that before or after B&W Natilius?
I haven no idea what speaker this is. Does not look like a b&W - oh, wait, went back to the post by @tuga - there is a link to a b&w page, so I guess it is a b&w.
 

Vladimir Filevski

Senior Member
Joined
Mar 22, 2020
Messages
330
Likes
376
I'm a little late to the party, but here is my practical and theoretical experience with Transmission Line (TL) loudspeakers:
1. Almost all TL loudspeakers which have been designed without George L. Augspurger's TLworks, or Martin J. King's TL MathCad worksheets, or David McBean's HornResp, are plain wrong.
2. All TL loudspeakers measured so far in Stereophile magazine and SoundStage are plain wrong.
3. How can you recognize if the TL design is wrong? By two tell-tales signs:
a) Frequency response has big notch (or depression) at about 150 - 200 Hz (lower in bigger boxes). Also, frequency response in smaller enclosures is not extended toward low bass, in contrast to vented enclosures with similar volume and efficiency.
b) Impedance has two prominent peaks in the bass (like bass-reflex box).
4. Although "theoretical" definition for TL is "infinite long pipe which absorbs woofer's back wave", the practical definition for TL is something like this: "vented enclosure with quarter wavelength resonance (long channel/line) at, or lower than woofer own resonant frequency". This means that the output from the TL port should contribute the woofer output. TL channel/line length is about ¼ wavelength of cut-off frequency.
5. Mass-loaded TL (ML TL) is a different beast - it is bass-reflex box with TL quarter wavelength resonance, or TL box with restricted terminus. We can skip it for now.
6. Optimal TL design has woofer placed not in the beginning of TL channel, but with some offset. Classic Bailey TL (from 1965) and modern PMC TLs have woofers at the beginning of the line and therefore are not optimal designs.
7. The optimal TL impedance should have depressed lower (first) impedance peak (or completely flat), leaving only the second impedance peak higher in the bass. It is mostly stuffing/damping issue.
8. Optimally designed TL enclosure has tight, punchy bass – even with high Qts woofers. Enclosure volume is moderate to very big. Low cut-off frequency is higher than a vented box with the same driver, but lower than in closed box. All-in-all, not worth the pain, except for moderate to high Qts woofers.
 
Last edited:
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Manufacturer
Joined
Apr 25, 2019
Messages
688
Likes
1,263
Location
Norway
2. All TL loudspeakers measured so far in Stereophile magazine and SoundStage are plain wrong.
Hard to argue against, though I may not have read all the reviews. Makes me wonder what kind of - if any - engineering goes into hifi products.

But - as I already mentioned - perhaps it was intentional. To create a unique sound.
 
OP
Kvalsvoll

Kvalsvoll

Addicted to Fun and Learning
Manufacturer
Joined
Apr 25, 2019
Messages
688
Likes
1,263
Location
Norway
@Vladimir Filevski I am glad that someone mentioned Martin J. King and his work at http://www.quarter-wave.com

He recently posted an article about TLs and the impact of taper and stuffing : http://www.quarter-wave.com/TLs/TL_Alignments.pdf
Nice to see M.J. King still publishing, I had a brief look at this new article, highly recommended, here he also points out the similarities and differences between bass-reflex and TL, and talks about what benefits these more complicated designs can offer.
 

beneix

Member
Joined
Dec 25, 2021
Messages
14
Likes
5
6. Optimal TL design has woofer placed not in the beginning of TL channel, but with some offset. Classic Bailey TL (from 1965) and modern PMC TLs have woofers at the beginning of the line and therefore are not optimal designs.
That's incorrect for my PMC FB1s:
"The bass driver is positioned around a third of the way along the line and, through skilful balance of the numerous parameters involved, the FB1 has been blessed with a frequency response which extends between 22Hz and 20kHz at the ‑3dB points. In fact, it is substantially flat above 27Hz" (https://www.soundonsound.com/reviews/pmc-fb1)

8. Optimally designed TL enclosure has tight, punchy bass – even with high Qts woofers.
That's precisely what I love about my FB1s!
Low cut-off frequency is higher than a vented box with the same driver, but lower than in closed box.
Well, I'm quite happy with -3dB at 22 Hz.
 

Vladimir Filevski

Senior Member
Joined
Mar 22, 2020
Messages
330
Likes
376
That's incorrect for my PMC FB1s:
"The bass driver is positioned around a third of the way along the line and, through skilful balance of the numerous parameters involved, the FB1 has been blessed with a frequency response which extends between 22Hz and 20kHz at the ‑3dB points. In fact, it is substantially flat above 27Hz" (https://www.soundonsound.com/reviews/pmc-fb1)
...
Well, I'm quite happy with -3dB at 22 Hz.
Regarding woofers at the beginning of the TL, I stand corrected, but only for models FB1, OB1 and Fact.8. AFAIK, all other PMC models have woofers at the beginning of the TL line (or close).
Unfortunately, although PMC Fact.8 woofer position is not at the beginning of the line, measured frequency response (from Stereophile review https://www.stereophile.com/content/pmc-fact8-signature-loudspeaker) is bad:
420PMC8fig4.jpg

That is far from optimally designed TL.
I could not find measurements for PMC FB1.

Anyway, quoted -3dB at 22Hz for FB1 is PMC's marketing wishful thinking. For example, according to PMC brochure, Fact.8 has frequency response form 28Hz to 20kHz (no +/- dB tolerance), but measured response above clearly show -3dB is at about 50Hz.

Another example: PMC Twenty.24i has about the same enclosure size as FB1, same 6.5" woofer size, same "effective ATL line length of 3m", same "28Hz to 25kHz" frequency response, but the ugly truth is this:
fr_on1530.gif

As we can see, -3dB is higher than 50Hz. (Measurents above are from SoundStage review https://www.soundstagehifi.com/index.php/equipment-reviews/425-pmc-twenty24-loudspeakers ).
 

beneix

Member
Joined
Dec 25, 2021
Messages
14
Likes
5
My listening room is untreated and leaves much to be desired, but a quick SPL measurement in REW of my two FB1s together from my listening position looks like this:
FB1.png

While there are many issues there to sort out (which is why I am just getting into room correction), to me it looks like the in-room frequency response from the FB1 does not drop off like the measurement of the Twenty.24i above.
 

Vladimir Filevski

Senior Member
Joined
Mar 22, 2020
Messages
330
Likes
376
Of course in-room responses may have better low bass than anechoic measurements (as in SoundStage). Or worse, depending on the particular room acoustic. That is exactly why all manufacturers (should) give anechoic -3dB tolerances: to compare apples to apples. PMC FB1 frequency response spec is a rotten apple - no way to reach -3dB at 22Hz anechoically, or in a number of rooms different of yours.
Speaking of in-room response, my little B&W standmount speaker has -3dB at 25Hz in my room (left speaker near to corner) and the right speaker is -3dB at 45Hz. But that is completely irrelevant for any other room, including yours.
Your in-room measurement has dip between 150 and 200Hz - it may be room acoustic related, or typical to a badly designed TL. I hope it is former.
3. How can you recognize if the TL design is wrong? By two tell-tales signs:
a) Frequency response has big notch (or depression) at about 150 - 200 Hz ...
b) Impedance has two prominent peaks in the bass (like bass-reflex box).
 
Last edited:
Top Bottom