Hello X,
Me again, please indulge me. I drew out the TL on nominally 3/4" Ply. That way with the 6 " internal width I end up with the 7.5" speaker width.
I measured the length of the labyrinth using the centre of the path and have 53" which with the front baffle will be 53.75".
The internal panels are 15mm ply.
Your description says a 48" line.
How much of a difference will the 5.75" make and is it significant ? ( I tell myself yes, it is over 10% longer but I really don't know)
If yes, it is a significant difference, is there anything I can do to mitigate?
I attached a picture for your perusal.
The circled numbers are the labyrinth segment lengths.
Also, in general, given Vas, Qts, Re and Fs, how much difference with these values would be okay as a substitute?
e.g if all 4 mentioned T/S parameters being within +/- 10% be okay to substitute or a different percentage, maybe +/- 5% ?
Many thanks for any assistance you can offer.
Andrewbee
Me again, please indulge me. I drew out the TL on nominally 3/4" Ply. That way with the 6 " internal width I end up with the 7.5" speaker width.
I measured the length of the labyrinth using the centre of the path and have 53" which with the front baffle will be 53.75".
The internal panels are 15mm ply.
Your description says a 48" line.
How much of a difference will the 5.75" make and is it significant ? ( I tell myself yes, it is over 10% longer but I really don't know)
If yes, it is a significant difference, is there anything I can do to mitigate?
I attached a picture for your perusal.
The circled numbers are the labyrinth segment lengths.
Also, in general, given Vas, Qts, Re and Fs, how much difference with these values would be okay as a substitute?
e.g if all 4 mentioned T/S parameters being within +/- 10% be okay to substitute or a different percentage, maybe +/- 5% ?
Many thanks for any assistance you can offer.
Andrewbee
Last edited:
A TL’s length is sort of like a woodwind instrument’s length. For a closed end instrument like a clarinet the lowest note is the 1/4-wave length. To calculate that, take length in inches and multiply by 0.0254m per inch. So 48in x 0.0254m is 1.22m. Since this is a 1/4-wave device, this length is 1/4 the length of a full wave or 4x1.22m or 4.88m long is the fundamental low note. The speed of sound is about 342m/sec. The lowest frequency is (342m/sec) / 4.88m or 70 cycles/sec or 70 Hz. This refers to the -3dB point. This number is approximate and assumes a straight TL. Tapering it narrow at the exit makes this value lower. It can only be accurately arrived at by a test build (as in your case) or by comprehensive simulation which accounts for the taper and the stuffing such as Akabak or HornResp etc. I think I designed it for 55Hz and this represents the -3dB point. To estimate what the extra 5in does it simply scales. The tuning frequency is the minimum saddle point between the two impedance peaks for vented alignment.
48/53x55 is 49Hz. This is 5Hz lower but may come at a slight cost of a little less bass amplitude vs the 55Hz tuning.
Why did I design for 55Hz? It is close to kick drum fundamental and still gives more bass amplitude va trying to extract a very low note and having too low of amplitude.
But a lot is adjustable via stuffing tuning. Your ears are good or use a mic and REW software.
48/53x55 is 49Hz. This is 5Hz lower but may come at a slight cost of a little less bass amplitude vs the 55Hz tuning.
Why did I design for 55Hz? It is close to kick drum fundamental and still gives more bass amplitude va trying to extract a very low note and having too low of amplitude.
But a lot is adjustable via stuffing tuning. Your ears are good or use a mic and REW software.
Thanks very much for the explanation X. I have a better understanding of what happens with a TL now. I think I will probably go with it as is and play with the stuffing as suggested.
Andrewbee
That looks very solid. Did you choose bamboo to create a stiffer baffle?
Also, I looked at a video with PMC legend Peter and he explained how they approach the damping.
He mentioned 2 artifacts, where the mentioned 100 Hz was referred to as "around 100 Hz depending on the driver";
1 - They aim at damping everything out of the line above around 100 Hz to prevent the resonance of the mid-bass in the line to impact the backside of the driver, thus reducing distortion. That way, only the driver will produce the mid-bass part and there is NO contribution of the line itself. Damping should occur in such a way that there is still unobstructed airflow through the line.
2 - Around the Fs of the driver, they want to start having a more smooth roll-off, where they found that this will provide a more natural low-end. The line should be at maximum efficiency from 100 Hz downwards. The line should be supporting the reduction of cone excursion, described as, "the driver should not be influenced by the line and stop after the signal applied has stopped".
As can be seen from various detailed pictures that fly around the internet, but also mentioned in the video, they use various types of damping, not just one type across the line. Also, you can clearly she the amount of damping is different in a few locations (please read the following with that in mind).
Given this, which sounds logical, I started a simple test using Leonard Audio transmission line software to simulate this in the following way:
b.) - The position of the driver (the offset location), contribution of the line / efficiency of the line in the low -end
c.) - The last part of the line, where after 25% inwards the line, the effect becomes less apparent, both smooth out the mid-high & high frequencies.
- After having these 3 positions in the line, I played a bit with the density in each location.,
As for point 1 mentioned above from Peter, it looks like adding foam with high density at 30 - 45 cm (at position c.) of the in my case 180 cm line did exactly what he is suggesting, minimize the TL contribution above around 100 Hz.
.
(Stuffing the complete line with high density resulted in the obvious, killing the line contribution as a whole.)
So, is position c.) the magical spot?? and what is the science behind that. Of-course this is just the modeling software, but that's what we have until we build and measure.
Adding high density foam to position A.) roll-off was more gradual and easy to taylor.
May be this is worth diving in to, to further improve the low-end of the TL. I have no deep technical knowledge in the mathematics and science behind this, so don't kill me if I don't make sense from a scientific point of view.
kind regards and keep up the good work sharing your experience and knowledge.
Also, I looked at a video with PMC legend Peter and he explained how they approach the damping.
He mentioned 2 artifacts, where the mentioned 100 Hz was referred to as "around 100 Hz depending on the driver";
1 - They aim at damping everything out of the line above around 100 Hz to prevent the resonance of the mid-bass in the line to impact the backside of the driver, thus reducing distortion. That way, only the driver will produce the mid-bass part and there is NO contribution of the line itself. Damping should occur in such a way that there is still unobstructed airflow through the line.
2 - Around the Fs of the driver, they want to start having a more smooth roll-off, where they found that this will provide a more natural low-end. The line should be at maximum efficiency from 100 Hz downwards. The line should be supporting the reduction of cone excursion, described as, "the driver should not be influenced by the line and stop after the signal applied has stopped".
As can be seen from various detailed pictures that fly around the internet, but also mentioned in the video, they use various types of damping, not just one type across the line. Also, you can clearly she the amount of damping is different in a few locations (please read the following with that in mind).
Given this, which sounds logical, I started a simple test using Leonard Audio transmission line software to simulate this in the following way:
- First design a standard TL, just straight 180 cm pipe with driver (Faital Pro 6FE 100, lying to do nothing) offset to set it up correctly.
- As a function of the LA software, I split up the complete TL in segments (elements called in the tool) of 15 cm.
- With the SPL window open, I started to fill the segments of the TL. Using a high damping rate of the filling at first, it is easier to see the impact.
- I noticed the obvious effects from locations where I expected to see the line behavior change, like the rimpeling at higher frequencies become less with more damping material applied. I turns out there are 3 (maybe not such a surprise) positions, the damping host most effect.
b.) - The position of the driver (the offset location), contribution of the line / efficiency of the line in the low -end
c.) - The last part of the line, where after 25% inwards the line, the effect becomes less apparent, both smooth out the mid-high & high frequencies.
- After having these 3 positions in the line, I played a bit with the density in each location.,
As for point 1 mentioned above from Peter, it looks like adding foam with high density at 30 - 45 cm (at position c.) of the in my case 180 cm line did exactly what he is suggesting, minimize the TL contribution above around 100 Hz.
. (Stuffing the complete line with high density resulted in the obvious, killing the line contribution as a whole.)
So, is position c.) the magical spot?? and what is the science behind that. Of-course this is just the modeling software, but that's what we have until we build and measure.
Adding high density foam to position A.) roll-off was more gradual and easy to taylor.
May be this is worth diving in to, to further improve the low-end of the TL. I have no deep technical knowledge in the mathematics and science behind this, so don't kill me if I don't make sense from a scientific point of view.
kind regards and keep up the good work sharing your experience and knowledge.
Bamboo is both stiff and takes on wood screws without splintering well. Mostly for the cosmetics though. Baltic birch is just as good.
Good analysis on stuffing/damping.
Stuffing density near the terminus (exit) part C, I have found in my simulations using Akabak to have the highest sensitivity to changing the overall bass extension in frequency and too much here will kill the efficiency. Foam lining (melamine foam is best) along first 2/3rds is important to stop the resonances above 100Hz, but doesn’t change bass frequency extension much. You definitely want polyfill or fiberglass etc fluffy stuffing in part A to reduce overall resonances to make the rolloff shallower. This can be seen as a flattening of the two impedance peaks in simulation and measurement. It is one way to tune the stuffing in practice as access to this part is easy by removing woofer. In my case, the whole baffle is removable, like an LS3/5A. One of the nicest things about a TL is that it really reduces the compressive loading behind the woofer and the sound is much more open in the mid range vs a sealed box.
Here are some details on the damping pads internally:
Backside with binding post plate.
Good analysis on stuffing/damping.
Stuffing density near the terminus (exit) part C, I have found in my simulations using Akabak to have the highest sensitivity to changing the overall bass extension in frequency and too much here will kill the efficiency. Foam lining (melamine foam is best) along first 2/3rds is important to stop the resonances above 100Hz, but doesn’t change bass frequency extension much. You definitely want polyfill or fiberglass etc fluffy stuffing in part A to reduce overall resonances to make the rolloff shallower. This can be seen as a flattening of the two impedance peaks in simulation and measurement. It is one way to tune the stuffing in practice as access to this part is easy by removing woofer. In my case, the whole baffle is removable, like an LS3/5A. One of the nicest things about a TL is that it really reduces the compressive loading behind the woofer and the sound is much more open in the mid range vs a sealed box.
Here are some details on the damping pads internally:
Backside with binding post plate.
Last edited:
Thanks for the quick reply. I have been building loudspeaker for over 40 years now, always good fun. Never dared to burn my fingers on a TL, because of this damping issue. Now I finally seem to have found a usable kind of best practice to get the first shot on damping close to the end goal, I think I need to have a stab at it.
Has there ever been a good detailed analyses on "how to stuff" as far as you know?
I don't want to show on my grave "He never got to design and build a TL, nice man overall though.." So I need to get cracking
Has there ever been a good detailed analyses on "how to stuff" as far as you know?
I don't want to show on my grave "He never got to design and build a TL, nice man overall though.."
So I need to get cracking
Was, you are welcome.
Good luck on your build. There is no risk. Just add the foam as shown and adjust fluffy stuffing to taste. Reticulated aquarium filter foam is an excellent alternative to stuffing at the terminus. The size can be precisely cut to fit opening. The length adjusted to suit and it keeps bugs and mice out of your speaker.
Yes, foil faced mass loaded butyl autosound damping sheets. Noico is one typical brand.
Good luck on your build. There is no risk. Just add the foam as shown and adjust fluffy stuffing to taste. Reticulated aquarium filter foam is an excellent alternative to stuffing at the terminus. The size can be precisely cut to fit opening. The length adjusted to suit and it keeps bugs and mice out of your speaker.
Yes, foil faced mass loaded butyl autosound damping sheets. Noico is one typical brand.
I did stuff it similarly, however didnt put in any poly fills will have to try it. But its sounding absolute fantastic I am keeping them 1meter from back wall and 31 inches from side walls still very deep base