Transmission line DIYs
Jan Karlsson
from Capells "Kappelmeister" design.
The basic design is this:
-The transmission line is folded three times with the opening on rear
bottom of the speaker.
- There are ceramic tiles embedded in concrete in the bends in order to
reflect the prapagating soundwave to the end of the line.
- The length of the line is chosen to be slightly longer than than the
corresponding quarter wavelength of the fs och the woofer.
- The woofer is monted so that the front faces the litstener (as in a
"normal" speaker) and the back fires into the beginning of the line
- A single full range speaker is used (I used a Fostex 6.5" unit with an
fs of 45Hz).
The speakers I built using the above scheme sounded fine in many ways but
lacked a deep bass response. In my current projekt I want to use the same
paradigm for building a pair of subwoofers. Therefore I have the
following questions:
1. When choosing woofers for a TL, what should I look for in terms of
Qts, Vas, fs and so on?
2. Should the length of the line be longer, shorter or just at the
quarter wavelenth of the fs of the woofer?
3. Can/should I use multiple woofers in this design, problems withs
interference, phase and so on?
4. If I use multiple woofers
-what happens to the effiency?
-should they be in series or in parallell?
5. Given a woofer with fs=22 Hz and a line length = 4.25 meters, what
sort of bass respons can I expect (-3 dB point? Roll of below this point?)
6. Another thought was to expand theese into full range speakers by
inkluding a separate line for midrange/high frequencies. Maybe a could
save som filter componets by using the "natural" roll of from this
shorter line (a little more than one meter for a 100 Hz tuning)?
Lionel Tun
Jan Karlsson <ja...@csd.uu.se> wrote:
>The speakers I built using the above scheme sounded fine in many ways but
>lacked a deep bass response.
I have studied Capel's book and I also thought the design would
sound great except for the lowest bass. I think it is the problem
of using a full range driver for everything.
I cannot answer your questions, but I would like to add another.
Capel says that bass output is limited by the distance separating
the driver from the transmission line port.
Q: How does this statement bear up in practice. For example the
TDLs with better bass do not necessarily maximise this distance.
--
________ Lionel Tun, lio...@cs.city.ac.uk ________
/ /_ __/\ Computer Vision Group /\ \__ _\
/___/_/_/\/ City University, London EC1V 0HB \ \___\_\_\
\___\_\_\/ 071-477 8000 ext 3889 \/___/_/_/
Uli Bodenhausen
|> About a year ago I built a pair of transomision line speakers using ideas
|> from Capells "Kappelmeister" design.
|>
|>
|> Qts, Vas, fs and so on?
I've read that the total q (both mechanical and electrical) should be > 0.6.
This was recommended by someone who designed a TML that I build and I'm
quite satisfied with the result. Many Morel drivers seem to be very well
suited for TMLs.
|> 2. Should the length of the line be longer, shorter or just at the
|> quarter wavelenth of the fs of the woofer?
|> 3. Can/should I use multiple woofers in this design, problems withs
|> interference, phase and so on?
I've only build TMLs with one woofer per box, but I've read that two
woofers can be arranged such that the frequency balance is indeed improved.
|> 4. If I use multiple woofers
|> -what happens to the effiency?
|> -should they be in series or in parallell?
depends on the lowest impedaance your amplifier can handle.
|> 5. Given a woofer with fs=22 Hz and a line length = 4.25 meters, what
^^^^^^^^^^^
This seems very long according to my experience with two TMLs. Both have
a lenght of about 2 - 2.5 m. Take into account that the damping inside the
line will slow down the the sound wave (at least that's what I'm told).
With a length of 2 - 2.5 m both TMLs I've build produce very deep bass.
One of them (TDL) seems not really challenged by the low frequencies
on most CDs.
Uli
Thomas Dunker
> In article <2sfdut$1h...@columba.udac.uu.se>, Jan Karlsson <ja...@csd.uu.se> writes:
> |> About a year ago I built a pair of transomision line speakers using ideas
> |> from Capells "Kappelmeister" design.
> |>
> |>
> |> 1. When choosing woofers for a TL, what should I look for in terms of
> |> Qts, Vas, fs and so on?
>
> I've read that the total q (both mechanical and electrical) should be > 0.6.
> This was recommended by someone who designed a TML that I build and I'm
> quite satisfied with the result. Many Morel drivers seem to be very well
> suited for TMLs.
>
some questions regarding woofer Qts vs. TMLs. The reply was that woofers with
a real low Qts (0.4-0.6) seemed to work better with double folded lines, and that
the ones with Qts values above that would work better in single folded lines.
This, according to Timmermans, was reached by empirical work and no theory
really explains it.
> |> 2. Should the length of the line be longer, shorter or just at the
> |> quarter wavelenth of the fs of the woofer?
> |> 3. Can/should I use multiple woofers in this design, problems withs
> |> interference, phase and so on?
>
> I've only build TMLs with one woofer per box, but I've read that two
> woofers can be arranged such that the frequency balance is indeed improved.
But what about the wavelength? Can you direct either of us to any literature
that explains or describes the use of two woofers?
>
> |> 4. If I use multiple woofers
> |> -what happens to the effiency?
> |> -should they be in series or in parallell?
>
> depends on the lowest impedaance your amplifier can handle.
Anybody know anything about the effects of series connection in a TML? What
happens to the Q-values???? I actually want a high impedance speaker, because I
use tube amps....
> |> 5. Given a woofer with fs=22 Hz and a line length = 4.25 meters, what
> ^^^^^^^^^^^
> This seems very long according to my experience with two TMLs. Both have
> a lenght of about 2 - 2.5 m. Take into account that the damping inside the
> line will slow down the the sound wave (at least that's what I'm told).
> With a length of 2 - 2.5 m both TMLs I've build produce very deep bass.
> One of them (TDL) seems not really challenged by the low frequencies
> on most CDs.
>
My woofer (the speaker is an old K&T design called Cheap Trick 077) has an
fs=36Hz, and my line is about 2,20 m long....works fine! Rolls off around 30Hz.
Keep posting more about TMLs!!!!!!
Tom
> Uli
>
>
Jan Karlsson
writes:
> > In article <2sfdut$1h...@columba.udac.uu.se>, Jan Karlsson
<ja...@csd.uu.se> writes:
> > |> About a year ago I built a pair of transomision line speakers
using ideas
> > |> from Capells "Kappelmeister" design.
> > |>
> > |>
> > |> 1. When choosing woofers for a TL, what should I look for in terms
of
> > |> Qts, Vas, fs and so on?
> >
> > I've read that the total q (both mechanical and electrical) should be
> 0.6.
Is this what they call "Qts"?
Yes, please!
As I understand it, according to theory, the length of the line should be
a quarter of the wavelength corresponding to the lowest frequency the
speaker is designed to reproduce. Damping reduces the necessary length,
but not by much (again according to theory). This means 4.25 meters for
20 Hz, 2.125 for 40 Hz and so on (minus the effects of damping, depending
on how dense it is)
Jan Karlsson.
Francis Vaughan
I would have though Dick Pierce would have weighed in by now,
perhaps he is saving his strength, or perhaps my seriously
non-causal newsfeed has yet to deliver his response.
Failing that I though I would say a coule of things. A lot is
paraphrasing things Dick has said in the past. Needless to
say the mistakes are mine.
Bottom line is that all of these transmission line designs for bass
speakers are not transmission lines. The entire basis for the
term and much of the logic of the design has nothing to do with
real transmission lines and much to do with wishful thinking.
Vance Dickason notes that the design has its genesis in a Stromborg
Carlson desgined speaker that placed a labrynth behind the driver
and acheived a bass boost out of what was essentially a tuned pipe.
The pipe was about 1/4 of the speaker resonant frequency. From here
it seems began the basis for the design mystique.
Bailey published work which sought to adapt the idea. Filling the
line with absorbent material he hoped to create a speaker where
no energy was reflected back from the line, modelling essentially
a true transission line. However at bass frequencies this simply
is not the case. Claims about the effects of damping material
on propogation velocity are also not true. It seems that measurments
of velocity depended upon relative phase between the driver and the
opening at the end of the line. The measurements assumed that the
enclosure was a real TL and they got the results they wanted, ie
40% velocity drop. However since it is not a TL the inference from
the phase was false. A case of assuming what you want to prove.
5% velocity drop seems more real.
The enclosure behaves exactly like what it really is, a vented enclosure
with a very big vent with a lot of loss in the vent.
So the question is: why when a very complete theory of vented designs
exists do you want to mess about with poorly understood and flawed
models of vented design? You will not achieve a better bass response,
the system is not in any way a 1st order system despite some absurd
potestations by some, and can only deliver worse bass performance than
a correctly designed system that takes advantage of the accepted
models of vented systems.
That said, the mid-bass performance and physics of some transmission
lines is both very very good and does operate with the charateristics
of a transmission line. Once the frequencies get high enough a long
enough line with appropriate damping material will essentially result
in what looks to the driver like a transmission line. Coupled with
the inherent bracing of the cabinet of a transmission line design
the results in the mid bass can be very good.
Personally I am about to try out a TL enclosure for a mid bass driver,
but the bass below 200Hz or thereabouts will be managed by an enclosure
sporting impecable Thiele-Small design credentials.
Francis Vaughan
Department of Computer Science _--_|\ Phone: +61 8 303 5592
University of Adelaide / \ Fax: +61 8 303 4366
Adelaide 5005 \_.--*_/ Home +61 8 364 2649
Australia v
Thomas Dunker
>
>
> So the question is: why when a very complete theory of vented designs
> exists do you want to mess about with poorly understood and flawed
> models of vented design? You will not achieve a better bass response,
> the system is not in any way a 1st order system despite some absurd
> potestations by some, and can only deliver worse bass performance than
> a correctly designed system that takes advantage of the accepted
> models of vented systems.
Wow, thanks! What about all us happy TML owners who get bass of subterranean
depth and who love it? I must admit that I am a bit confused about the concepts
myself, and this is a problem because I would like to design my own TML some day.
However, when I happen to have a pair of inexpensive TMLs that outclass most
speakers I have heard at two or three times the price, I'm not going to abandon
the concept just because I don't know the exact workings of it. The only right
thing to do would be to try and find literature about it and talk to other TML
owners/designers for impulses.
>
> That said, the mid-bass performance and physics of some transmission
> lines is both very very good and does operate with the charateristics
> of a transmission line. Once the frequencies get high enough a long
> enough line with appropriate damping material will essentially result
> in what looks to the driver like a transmission line. Coupled with
> the inherent bracing of the cabinet of a transmission line design
> the results in the mid bass can be very good.
I have terrific mid bass, but where my speakers impress the most is with the
clean deep bass. I have yet to hear a non-TML speaker with an 8" woofer of my
admittedly budget class produce such astounding subwoofer-like bass. Some people
rudely insist that TMLs have no particular quality with clean bottom octaves. I
am tempted to ask: Have you listened to any? Are all the happy TML owners a
bunch of deaf braggers?
To me it seems that most TML designs use a combination of the 1/4 wavelength
'organ pipe' thinking combined with proper damping of the line. Regardless of
whether you, I, or anyone else can computer simulate or otherwise make
theoretical models about it, it works well in practice. Even down below 20 Hz in
some cases (such as the TDL RSTL). I have not tried to design my own yet, and I
understand that it can be difficult without a full understanding of the theory.
Fortunately a lot of TMLs have been successfully designed in the past, so why not
make use of the experiences gained before? I, for one think it's a waste of
resources trying to reinvent the wheel every time one is doing something that has
already been done.
I also think it's healthy to explore different ways of doing things. That
something is 'easy' to design (bass reflex speakers) doesn't mean it's more
worthwhile. Who said audio design was easy?
> Personally I am about to try out a TL enclosure for a mid bass driver,
> but the bass below 200Hz or thereabouts will be managed by an enclosure
> sporting impecable Thiele-Small design credentials.
>
In case you haven't already noticed, Thiele-Small parameters are pretty crucial
in TML design as well. It's not as if all the great TMLs like Bailey and the RSTL
were just blind shots in the dark!
FYI, I am really keen on engaging in horn speakers! (I can see the head-shaking
already....)
I'm not writing this to irritate, but I just spent hours enjoying music from my
DIY system including some great organ pedal work in the lower octaves vibrating
my guts. I can't stand having people make erroneous statements going: 'TMLs can't
go deep(er), 'cause according to this book here they won't.' Or words to that
effect.
Tom
Tom
Richard D Pierce
> To me it seems that most TML designs use a combination of the 1/4 wavelength
>'organ pipe' thinking combined with proper damping of the line.
While the rest of this thread has some pretty seriously flawed assertions
on the part of transmission line advocates, not the least of which is the
completely boguis and unsopportable claim by at least one commercial
exponent of the technique, this one is the most glaringly obvious one
and, frankly, needs to be debunked once and for all.
An implementation of a labyrinth, a transmission line, or a straight tube
with the driver and one end and the other end open, damping in the line
or not, WILL NOT OPERATE AS A 1/4 WAVELENGTH RESONANT STRUCTURE. That
statement is false, so drop it once and for all.
Here is why. A tube that is open at both ends or closed at both ends
resonates at the 1/2 wavelenth frequency (and even multiples of that).
That is because in the former case, the particle velocity is maximum at
both open ends, and the pressure is 0, while in the latter case, the
pressure variations are at a max, and the velocity is 0. This means that
the phase difference of the wave at the ends of the tube at resonance must
be an even multiple of 180 degrees (1/2 wavelength) for these conditions
to occur.
A tube with one end open and one end COMPLETELY closed resonates at the
quarter wavelength frequency. This is because at the open end, the
particle velocity is at a maximum and pressure excursion are 0, while at
the closed end, the pressure excursions are maximum and the velocity is
0. This condition is met when the phase difference of the wave at the
ends of the tube must be an odd multiple of 90 degrees (1/4 wavelength).
Now, the typical implementation of a line is with the driver in one end
and the other end is open. This is completely and unambiguously equivalent
to the model where both ends are open. Why? Because, obviously, the far
end of the line is open, and the end where the driver is is completely
equivalent to being open because of the presence of the driver which
forces the particle velocity to be a maximum there.
The "organ pipe" analogy you cite is completely innaproriate in this
case. Indeed, stopped organ pipes, such as the gedackts, bourdons, and
the like, speak at 1/4 wavelenth, because one end (the top) is completely
closed, the other (at the mouth) is completely open. However, all other
organ pipes, (principles, open flutes, all strings or "geigenwerke", even
the "half-stopped" pipes, such as the rohrflote or "chimney flute", the
spitzflote, and so on), have both ends open to one extent or another, and
speak at 1/2 wavelength.
The ultimate issue here is not whether you like the results of a well
done transmission line or not, at least not as several of us have engaged
in the discussion. It's over wherther the claims made by many, including
some of the more ridiculous and strident claims made by some at companies
that specialize in this sort of product, have basis in physical fact or
not. And, for the most part, they provable do not. Most prominent among
those claims are:
1. Transmission lines have a first-order rolloff characteristic.
This is absolute, complete, and utter hogwash, and I challenge
each and every person making such a claim to advance ANY proof
of ANY kind supporting this claim.
There does not exist a single physical mechanism that will permit
a loudspeaker to act as a first order system. The mass of the
loudspeaker diaphragm and the compliance of the surround together
form a second order mechanical resonant system that provide the
ultimate limits of the system. There exists not a single acoustical
element or combination of elements that can reduce the system from
simplest possible operation (a second order high-pass) to a lower
order system. It is impossible. Additional mechanical or acoustical
elements are only capable of operating in parallel with one of the
mechanical elements of the system (like the compliance of the air
ine the enclosure will work in parallel with the compliance of the
suspension to ultimately raise the fundamental resonance, but it's
still a second order system), or more complex combinations (such
as the compliance of the air in the box and the mass of port) will
conspire to raise the order of the system, almost always by two
orders at a time. A transmission line, no matter what theory you
propose, is a higher order system than second order, whether you or
the hand-waving, overbearing manufacturers of same are willing or
able to realize that or not.
Further, not a single, repeatable and verifiable measurement has been
advanced to support this claim (the result of the system being
order MUST be a an asymtotic response limit of 6 dB per octave BELOW
cutoff, not 6 dB for a little while, then 12, but 6 db per octave
forever. No speaker does this.) In fact, one of the most vocal
commercial proponents of this view had the balls, in the same phone
conversation, to make the claim that he regularily measures this
phenomenon, then turn around not ten minutes later and enquire if
I would be willing to make measurements for his company (i have
extebnsive facilities for doing so), because he had no measurement
equipment of his own.
2. Damping in the line can slow the wave propogation by 40% or more.
Nonsense. The absolute BEST that can be achieved, under theoretically
perfect conditions, is about 40%, and that's achieved with damping
materials that don't exist at frequencies much higher than what is
claimed in transmission lines. At 50 Hz and below, the effect is
verifiably more on the order of 10% or so. Don't take my word for
it, go measure it. I've done it so many times myself, I'm sick of
the excercise.
The argument is not over whether transmission lines cvan be made to sound
good, some can. The issue is over the outrageous physics claimed by some.
Does it make a difference. Well, if you don't care, no it doesn't. But it
does matter when claims are made that simply violate physics, then those
claims are then used in an attempt to build a veritable house-of-cards
theory, and the proponents of such then get pissed of when the slightest
breath of reality comes by and knocks there little house down. Instead of
trying to understand the issues, they start name calling and, in the case
of my friend on the phone, threaten law suits.
If you like the way your lines sound, fine. I like some lines, too. But I
and others in this industry don't particularly liked being lectured about
physics be newbies or threatened by lawyers-turned-carnival-barkers when
it comes to something we happen to know quite a bit about.
--
| Dick Pierce |
| Loudspeaker and Software Consulting |
| 17 Sartelle Street Pepperell, MA 01463 |
| (508) 433-9183 (Voice and FAX) |
Thomas Dunker
>
> While the rest of this thread has some pretty seriously flawed assertions
> on the part of transmission line advocates, not the least of which is the
> completely boguis and unsopportable claim by at least one commercial
> exponent of the technique, this one is the most glaringly obvious one
> and, frankly, needs to be debunked once and for all.
>
Thank you for the article, Richard. If you had come with this earlier, we could
have avoided the misleading posts that I and others have been guilty of writing.
Again, proof that it is unwise to be big-mouthed without knowing what one is
talking about. I wrote in anger because someone implied that TMLs can not sound
the way they some times do, clearly with no basis in either sound theory or
actual listening experience. Thank you for clarifying some of the mystery.
Tom
Richard D Pierce
No, the article that you so vociferously and erroneously responded to
said, in my reading NOTHING about whether or not transmission lines do or
do not SOUND any particular way. Rather that person specifically stated
that they could not WORK the way you and others claimed, regardless of
how you think they might sound.
A common, kee-jerk reaction on the part of many is to take a statement in
challenge to a particular theory and interpret it as a personal challenge
of one's hearing ability or manhood. The gentlement was specifically
addressing the flaws in one popular theory of transmission lines. You
immediately misinterpreted his statements, in a fairly snide fashion, as
an accusation that, in effect, all transmission line lovers are deaf. He
never said such, and I think that, regardless of the motivation to your
erroneous post, you owe the gentlemen at least a private apology.
>Thank you for clarifying some of the mystery.
Your quite welcome, despite the fact that I expect another 3 days of
telephone harrassment from my "friend" who will threaten me with yet more
stupid, pointless litigation will result.
Francis Vaughan
> Wow, thanks! What about all us happy TML owners who get bass of subterranean
> depth and who love it?
Thats great, you own a successful design and really like the result.
I well remember the first time I heard some IMF Mark IVs I was
pretty blown away. However that was a long time ago. There are
successful TL designs and many have a commited following of supporters.
None of this says that the design rationale used for the design
is correct. The basic problem addressed in this thread is not that
there have been good sounding designs but that the premise that
the basic design criterion used is fundamentally flawed.
What we will say is not that the speaker cannot work, but that
if you performed the conventional vented box analysis on the
final design the analysis would agree perfectly with the reality.
Predictions predicated upon the 1st order damped 1/4 wave myth
would not. Almost every time I have discussed TL design with
anyone or read of exploits in design a common feature crops up.
"Of course you need to spend a lot of time tweaking the design."
What this tweaking always seems to mean is considerable effort
expended playing with the stuffing in the line. In vented box
analysis terms this means changing the vent losses.
Why not admit that the vented box analysis is what is really
going on and use it to predict the ultimate performance? It
would avoid a lot of messing about and yield considerably more
reproducable results.
> I must admit that I am a bit confused about the concepts
> myself, and this is a problem because I would like to design
> my own TML some day. However, when I happen to have a pair of
> inexpensive TMLs that outclass most speakers I have heard at
> two or three times the price, I'm not going to abandon
> the concept just because I don't know the exact workings of it.
What is being proposed is just what you desire. A fully self consistent
model that does describe the working of the speakers you love.
> The only right thing to do would be to try and find literature
> about it and talk to other TML owners/designers for impulses.
Good idea, but don't assume that because someone has managed to build
a design you enjoy they have a firm grasp on the physics. The tweak
factor in TL realisations is so owerwhelming that what you are doing
in trusting the designer is not so much a vote of confidence in their
engineering prowess as an agreement on prefered sound. Nothing wrong
with that, just be careful.
> I have terrific mid bass, but where my speakers impress the most is with the
> clean deep bass. I have yet to hear a non-TML speaker with an 8" woofer of my
> admittedly budget class produce such astounding subwoofer-like bass.
> Some people rudely insist that TMLs have no particular quality with clean
> bottom octaves. I am tempted to ask: Have you listened to any? Are all
> the happy TML owners a bunch of deaf braggers?
No, some audiophiles I know and respect love their TL speakers. However we
need to be really careful when assertions of astounding bass performance
are aired. Arguing here invarably seems to lead to a sudden heating
of the argument, BUT
There is no reason to belive that these transmission lines really produce
deeper bass. There is considerable evidence to the contary.
I will add another observation. Very often the phrase deep bass is
either further compounded or replaced with "clean bass". Now this
is quite interesting and perhaps hearin lies a clue. Just why is it clean?
An improvement in the quality of those frequencies where harmonics
of the deep bass might exist?
A personal observation, one which I ascribe no particular merit too,
but maybe of value. I suspect (and have in my limited experience
seen some evidence) that many percieved improvments in bass performance
have has nothing whatsoever to do with the bass driver operating
in the region of real bass. Rather a perceived improvement in both
depth and quality of bass has occured when some element involved in
the reproduction of higher frequencies has been improved. (I once
replaced the tweeters in a system and the most obvious and commented
upon change in quality was a perceived improvement in the bass !)
The quality of mid-bass through midrange frequencies occuring in
TL designs resultant from both operation in the range where TL
mechanisms really do occur and improved box mechanics through
the inherent bracing could account for much of this.
> In case you haven't already noticed, Thiele-Small parameters
> are pretty crucial in TML design as well. It's not as if all the
> great TMLs like Bailey and the RSTL were just blind shots in the dark!
Parameters maybe, but these are a very minor part of the Thiele-Small
work. They chose the parameters that suited their analysis. You can't
claim that by using the particular parameter set but ignoring the
models for which the parameters were chosen, that you are using T-S
blessed physics.
Francis Vaughan
Lionel Tun
>going on and use it to predict the ultimate performance? It
>would avoid a lot of messing about and yield considerably more
>reproducable results.
This argument sounds attractive. But where are the manufacturers
who have taken VB design to its ultimate limits? Does _anyone_
know of any VB speakers _in practice_ which outperform a TL in
the bass?
>What is being proposed is just what you desire. A fully self consistent
>model that does describe the working of the speakers you love.
So why have VB designers not converged towards the TL?
>I will add another observation. Very often the phrase deep bass is
>either further compounded or replaced with "clean bass". Now this
>is quite interesting and perhaps hearin lies a clue. Just why is it clean?
>An improvement in the quality of those frequencies where harmonics
>of the deep bass might exist?
This may well be it. TLs have been described as `clean', `deep',
`rich', `full'. But I don't know how VB theory can predict all this.