Hi everybody 
I'm building a speaker. Problem is, I'm trying to decide between a Ported or TL, but I have never heard, or can’t recall hearing, a good ported design. And a TL I have heard once, if the Bose Wave radio is a TL
The ported design I have modeled has ~6 dB more output between 25-30 Hz than the tapered TL of the same size because I can use 2 woofers per speaker instead of 1 in the TL. I could use 2 woofers per speaker in the TL of the same size, but I would hardly get any more output with the same power and the 2 woofers would have about 2/3rds of the excursion of the design with 1 driver, at that same output level. It’s not really worth it to stuff an extra woofer in the TL to have 33% less excursion and gain maybe 1 dB, I figure. I can post some graphs if it’s clearer.
Is it worth it to build the ported design, if I've never recalled hearing a ported design I liked? The best ported design I can remember hearing was at a club; it was a JBL pro sub, probably the 18” version of the 2226. I was most likely slightly intoxicated though
Some body on this forum wrote that a TLs' sound is close to the sound a sealed enclosure with a Q of .5. Is that true? The reasoning I believe was that both have a rather low group delay compared to a ported enclosure. I'm kind of torn because I love bass, but I live in an apartment and might not even be able to play much bass So I will just have some bad ported speakers that I play at a low volume with clothes stuffed in the ports. Or the better scenario, of the possible negative outcomes, some ported speakers that I will play loud but sound sloppy.
I'm building a speaker. Problem is, I'm trying to decide between a Ported or TL, but I have never heard, or can’t recall hearing, a good ported design. And a TL I have heard once, if the Bose Wave radio is a TL
The ported design I have modeled has ~6 dB more output between 25-30 Hz than the tapered TL of the same size because I can use 2 woofers per speaker instead of 1 in the TL. I could use 2 woofers per speaker in the TL of the same size, but I would hardly get any more output with the same power and the 2 woofers would have about 2/3rds of the excursion of the design with 1 driver, at that same output level. It’s not really worth it to stuff an extra woofer in the TL to have 33% less excursion and gain maybe 1 dB, I figure. I can post some graphs if it’s clearer.Is it worth it to build the ported design, if I've never recalled hearing a ported design I liked? The best ported design I can remember hearing was at a club; it was a JBL pro sub, probably the 18” version of the 2226. I was most likely slightly intoxicated though
Some body on this forum wrote that a TLs' sound is close to the sound a sealed enclosure with a Q of .5. Is that true? The reasoning I believe was that both have a rather low group delay compared to a ported enclosure. I'm kind of torn because I love bass, but I live in an apartment and might not even be able to play much bass
So I will just have some bad ported speakers that I play at a low volume with clothes stuffed in the ports. Or the better scenario, of the possible negative outcomes, some ported speakers that I will play loud but sound sloppy.
Last edited:
I am always a bit wary when people make statements like I never heard a bass refex that at I liked.
I know this will probably ellicit howls of outrage from the TL set but, using a periodic transmission line and then attempting to damp out most of its resonanses is far from an optimum way of making a speaker system, the optimum way is to simple do without the resonances you dont want in the first place and only have the one you do want, i.e. the optimum TL IS A BASS REFLEX.
Reflex boxes in clubs are usually arranged for "punch region" bass, that is they emphasise the 60-120Hz. region, the illusion of "tightness" and "fastness" is simply this.
Floyd Toole long ago pointed out that the so called slow flabby reflex sound can be magically transformed into a fast tight one simply by moving the speaker system and or equalising the room resonance causing the "flabiness".
rcw.
I know this will probably ellicit howls of outrage from the TL set but, using a periodic transmission line and then attempting to damp out most of its resonanses is far from an optimum way of making a speaker system, the optimum way is to simple do without the resonances you dont want in the first place and only have the one you do want, i.e. the optimum TL IS A BASS REFLEX.
Reflex boxes in clubs are usually arranged for "punch region" bass, that is they emphasise the 60-120Hz. region, the illusion of "tightness" and "fastness" is simply this.
Floyd Toole long ago pointed out that the so called slow flabby reflex sound can be magically transformed into a fast tight one simply by moving the speaker system and or equalising the room resonance causing the "flabiness".
rcw.
RCW
With all due respect, your generalizations are almost as bad as the OP's !
TL's ARE NOT over-complexificated Helmholtz resonators! It follows that pejoratively analogizing an optimized TL as a bass reflex is therefore also erroneous.
Without going into the details of why this is so just now, let's just look at how the bass response differs in a TL, and thus why one might consider building a TL rather than a ported system.
Helmholtz resonators only work in a relatively narrow band of frequencies around resonance. A couple of octaves above and they are out of the picture and it's all up to box shape and stuffing. At the lower end of their operating envelope, where they transition from 'Helmholtz resonator' to 'sealed box with big hole' you can expect a very steep rolloff in frequency response, at least 24 dB/Octave. This means that below their port frequency, they are basically useless.
The TL OTOH, exhibits quite a different lower cutoff characteristic, much more like a sealed box or infinite baffle. In fact, it is quite easy to make an 'IB mimic' response with just about any driver using a TL, and the resulting box is usually smaller that an actual IB for the same driver, provided you start with a reasonably suitable driver for a TL. You can't really get this kind of response with a ported box without doing a lot of EQ and you still lose the very bottom due to complete unloading. Heavy EQ means a lot less dynamic range, too. Eventually the TL is completely unloaded also, but this happens at like <10Hz or so. Yes it's really that low.
Then there's the whole 'exact tuning' issue; you really need to tune a ported box using impedance plots, something not everyone is equipped to do. This is unnecessary with a TL
I don't really care for the IB sound myself and like to tune for more 'fatness' in the bottom than the typical IB setup. This results in a smaller TL, so good all around.
With an appropriate driver and well-designed TL, the box will be 'competitive' in size to a ported box, too for a given F3. If you look at F10, the TL will usually win.
Jimbo
With all due respect, your generalizations are almost as bad as the OP's !
TL's ARE NOT over-complexificated Helmholtz resonators! It follows that pejoratively analogizing an optimized TL as a bass reflex is therefore also erroneous.
Without going into the details of why this is so just now, let's just look at how the bass response differs in a TL, and thus why one might consider building a TL rather than a ported system.
Helmholtz resonators only work in a relatively narrow band of frequencies around resonance. A couple of octaves above and they are out of the picture and it's all up to box shape and stuffing. At the lower end of their operating envelope, where they transition from 'Helmholtz resonator' to 'sealed box with big hole' you can expect a very steep rolloff in frequency response, at least 24 dB/Octave. This means that below their port frequency, they are basically useless.
The TL OTOH, exhibits quite a different lower cutoff characteristic, much more like a sealed box or infinite baffle. In fact, it is quite easy to make an 'IB mimic' response with just about any driver using a TL, and the resulting box is usually smaller that an actual IB for the same driver, provided you start with a reasonably suitable driver for a TL. You can't really get this kind of response with a ported box without doing a lot of EQ and you still lose the very bottom due to complete unloading. Heavy EQ means a lot less dynamic range, too. Eventually the TL is completely unloaded also, but this happens at like <10Hz or so. Yes it's really that low.
Then there's the whole 'exact tuning' issue; you really need to tune a ported box using impedance plots, something not everyone is equipped to do. This is unnecessary with a TL
I don't really care for the IB sound myself and like to tune for more 'fatness' in the bottom than the typical IB setup. This results in a smaller TL, so good all around.
With an appropriate driver and well-designed TL, the box will be 'competitive' in size to a ported box, too for a given F3. If you look at F10, the TL will usually win.
Jimbo
Jimmy,
It is not nearly as simple as you'd like. Different drivers prefer certain boxes... and different alignments give differing results.
Sealed is fairly straightforward. Bigger box, lower Q.
With ported boxes there are a many more ways to tune things (and screw them up too). As well a bass reflex is tuned... using T/S parameters. One thing most people forget is that as you change the drive the T/S parameters move... T/S is a curve not a fixed set of scaler numbers. This means that the bass in a bass reflex can change as the volume changes.
A transmission line has even more possibilities (using the term TL to encompass all quarterwave resonantors). One of these possibilities looks a lot like a bass reflex... if you increase the height to cross-section ratio of a BR you get a transition to a quarter-wave behavior and your BR becomes an ML-TL. A lot of BR that get stretched to become ML-TLs and the designers are either at a loss to explain why they didn't turn out like planned or are oblivious. More damping than in a BR typically helps with the varying T/S issue... a traditional TL is a constant or tapering (larger-to-smaller), about as long as a quarter wavelength of its Fs, and then damped fairly heavily. Voigt, ML-Voigt, pipes, one can even stretch to include horns (particularily back-loaded-horns)
Driver, box choice, and execution all play a role. It is possible to make a loudspeaker with very good bass using any of these kinds of boxes. Drivers, execution, and room all figure in.
The more possible alignments the easier it is to design & build a screw-up. Modern tools can really trim the failures, and good examples abound.
If you haven't build speakers before you are best to build a proven example...
dave
It is not nearly as simple as you'd like. Different drivers prefer certain boxes... and different alignments give differing results.
Sealed is fairly straightforward. Bigger box, lower Q.
With ported boxes there are a many more ways to tune things (and screw them up too). As well a bass reflex is tuned... using T/S parameters. One thing most people forget is that as you change the drive the T/S parameters move... T/S is a curve not a fixed set of scaler numbers. This means that the bass in a bass reflex can change as the volume changes.
A transmission line has even more possibilities (using the term TL to encompass all quarterwave resonantors). One of these possibilities looks a lot like a bass reflex... if you increase the height to cross-section ratio of a BR you get a transition to a quarter-wave behavior and your BR becomes an ML-TL. A lot of BR that get stretched to become ML-TLs and the designers are either at a loss to explain why they didn't turn out like planned or are oblivious. More damping than in a BR typically helps with the varying T/S issue... a traditional TL is a constant or tapering (larger-to-smaller), about as long as a quarter wavelength of its Fs, and then damped fairly heavily. Voigt, ML-Voigt, pipes, one can even stretch to include horns (particularily back-loaded-horns)
Driver, box choice, and execution all play a role. It is possible to make a loudspeaker with very good bass using any of these kinds of boxes. Drivers, execution, and room all figure in.
The more possible alignments the easier it is to design & build a screw-up. Modern tools can really trim the failures, and good examples abound.
If you haven't build speakers before you are best to build a proven example...
dave
I've touted looking at F10 over F3 for a long time... it was encouraging to read in Floyd Toole's new book, the same thing... his assertion backed up with solid research.
dave
The point is that it is a linear phase system and any concantination of filters will give the same result, but the reflex will always have lower cone excursion for a given spl. and therefore lower distortion.
The sound is also not coloured by the series of pipe resonances that are still apparent in a TL, if you do manage to damp them sufficiently the open end of the line has no really useable output so then you might as well use a reflex box.
I have shown all of this in forums several times and I know that the TL fraternity will yet claim some mystical property is possed by the TL and no other scheme.
So be it but if you claim technical superiority for the scheme that is something else because there is non, and no amount of hand waving and obfuscation will change it.
rcw
The sound is also not coloured by the series of pipe resonances that are still apparent in a TL, if you do manage to damp them sufficiently the open end of the line has no really useable output so then you might as well use a reflex box.
I have shown all of this in forums several times and I know that the TL fraternity will yet claim some mystical property is possed by the TL and no other scheme.
So be it but if you claim technical superiority for the scheme that is something else because there is non, and no amount of hand waving and obfuscation will change it.
rcw
What Dave said.
Just to point out that there might be some misinterpretation of what a TL is lurking around. A Transmission Line by the strict sense of the word is inherently unresonant. That, and the fact that it presents as near to an 'perfect' load as you're likely to get in this life, is the entire point of them. They do not have the same goals as a vented box or trad. BR, and therefore trying to compare them is meaningless. Comparisons are only useful if the same design goals are applied.
Of course, as Dave points out, there are a lot of boxes that are lumped under the TL moniker, even though they do not fall under the strict definition of what a TL is. MLTLs (strictly speaker Mass Loaded Quarter Wave Resonators), Mass Loaded Horns, Reverse Taper Horns (TQWT)... &c. They all have their advantages & disadvantages. Superior to a BR? Not necessarily. Inferior? Again, not necessarily. Back to design goals.
Just to point out that there might be some misinterpretation of what a TL is lurking around. A Transmission Line by the strict sense of the word is inherently unresonant. That, and the fact that it presents as near to an 'perfect' load as you're likely to get in this life, is the entire point of them. They do not have the same goals as a vented box or trad. BR, and therefore trying to compare them is meaningless. Comparisons are only useful if the same design goals are applied.
Of course, as Dave points out, there are a lot of boxes that are lumped under the TL moniker, even though they do not fall under the strict definition of what a TL is. MLTLs (strictly speaker Mass Loaded Quarter Wave Resonators), Mass Loaded Horns, Reverse Taper Horns (TQWT)... &c. They all have their advantages & disadvantages. Superior to a BR? Not necessarily. Inferior? Again, not necessarily. Back to design goals.
Last edited:
As Scottmoose points out, in the strict sense of the usage, a transmission line is aperiodic.
However in order for this to be true the impedance matching between its two ends must be conjugate, and this is simply not true of the transmission line speaker systems we see described, and indeed can only be achieved with a flaring duct, i.e. a horn.
rcw.
However in order for this to be true the impedance matching between its two ends must be conjugate, and this is simply not true of the transmission line speaker systems we see described, and indeed can only be achieved with a flaring duct, i.e. a horn.
rcw.
Scott, so if the design goals and not simply the frequency exstension are making the difference here what are the goals of the specific kinds? can you point out to a resource over the internet to understand it, and understand what to expect, unfortunately not all of us can manage to hear the real thing... for example i still did not ear any BLH design, over an year i've been interested in audio... no shop in rome have them... nor they do have any TL or what...
and specifically what would be the difference in presentation for a TL, or a reflex/BVR design given the saim driver (say an ALPAIR10?
and specifically what would be the difference in presentation for a TL, or a reflex/BVR design given the saim driver (say an ALPAIR10
?
Hmmm . . . sounds like I should go with ported Pun not intended
Here's what have. The one on the left is an Usher 8955a in a 10:1 tapered TL with a start area of 5 Sd and length of 90 inches. The middle one is a ScanSpeak 21W8555-01 in the same TL and the one on the right is a Usher in an enclosure half the volume of the TLs, so I can use two for my design. However, for the ported box, I get a different result using MJK's ML TQWT spreadsheet with same dimensions, I believe. The one in the picture is his standard "ported" spreadsheet.
The ported box has more output and less excursion above 20 Hz than the TLs. Also it has less output from the port above 100 Hz with less stuffing, which is good since I think I would like to have the crossover around 200 Hz.
I wasn't saying I didn't like ported enclosures, just that I've never heard one I liked. Which isn't saying much, since I've heard little.
I have read claims that the group delay of ported enclosures between 20-30 Hz is in the range of being audible by the human hear, around 23 ms vs 8 for a straight TL or sealed enclosure.
Pun not intended Here's what have. The one on the left is an Usher 8955a in a 10:1 tapered TL with a start area of 5 Sd and length of 90 inches. The middle one is a ScanSpeak 21W8555-01 in the same TL and the one on the right is a Usher in an enclosure half the volume of the TLs, so I can use two for my design. However, for the ported box, I get a different result using MJK's ML TQWT spreadsheet with same dimensions, I believe. The one in the picture is his standard "ported" spreadsheet.
The ported box has more output and less excursion above 20 Hz than the TLs. Also it has less output from the port above 100 Hz with less stuffing, which is good since I think I would like to have the crossover around 200 Hz.
I wasn't saying I didn't like ported enclosures, just that I've never heard one I liked. Which isn't saying much, since I've heard little.
I have read claims that the group delay of ported enclosures between 20-30 Hz is in the range of being audible by the human hear, around 23 ms vs 8 for a straight TL or sealed enclosure.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
True, but we're talking nominally infinite flared ducts / horns / whatever you wish to call them, & approaching the dead ground of mathematical semantics. In the real world, compromised horns are rarely employed solely to present as near to an optimal load to the amplifier as practical (unlike TLs, where this is the principal goal) but for gain & a measure of directivity control.
Re the question about what to expect from a TL-proper, the attached will give you an idea. It's not perfect, but it was to hand. This is, basically, what they do & are for. Damped LF, unreactive load & excellent behaviour in the time domain.
Re the question
Attachments
Notice Scott's TL clearly show the 'IB mimic' response I referred to earlier. The driver/TL response is virtually identical to the driver's IB response shown by the blue dotted line, and has a smooth rolloff below driver Fs. This would indeed be quite a 'hat trick' to accomplish with a Helmholtz box, but is 'old hat' for a TL.
Jimbo
How are you measuring distortion (as we know typical THD measures are meaningless.
And what about when a BR unloads below its tuning? Its excursion goes thru the roof.
And how do you deal with shifting T/S parameters.
dave
PS: i avoid straight BR designs as i've usually found the character of the bass to be off...
Okay, so it's settled. I'll build the tapered TL 
But in all seriousness, I guess I would need to listen to both. I'm just guessing I would like the sound of the tapered TL better.
There seem to be some advantages TLs. Sounds like maybe the high group delay of a ported box between 20-30 Hz is what I don't like about the sound of a ported enclosure. But the article I read said it was right at the threshold that is audible to humans. I would also imagine this is what would create that sloppy ported sound I seem to remember not liking. But then again I'm using my imagination a lot here
But in all seriousness, I guess I would need to listen to both. I'm just guessing I would like the sound of the tapered TL better.
There seem to be some advantages TLs. Sounds like maybe the high group delay of a ported box between 20-30 Hz is what I don't like about the sound of a ported enclosure. But the article I read said it was right at the threshold that is audible to humans. I would also imagine this is what would create that sloppy ported sound I seem to remember not liking. But then again I'm using my imagination a lot here
Last edited:
Jimmy,
I was unable to duplicate anything near your results using the TL dimensions you posted and the Usher 8955 driver. Where did you get the 'complete' T/S parameters, as they are not published? I derived them from formulas, with a good 'best guess' for Re of 6.5 Ohms, which duplicated the Zres number (from the published Z plot) exactly, so I'm pretty confident in it. I got a BL of 8.1 and Le of .3mH. The rest is all obtained from their data sheet. How do those numbers compare with what you used ?
Jimbo
I was unable to duplicate anything near your results using the TL dimensions you posted and the Usher 8955 driver. Where did you get the 'complete' T/S parameters, as they are not published? I derived them from formulas, with a good 'best guess' for Re of 6.5 Ohms, which duplicated the Zres number (from the published Z plot) exactly, so I'm pretty confident in it. I got a BL of 8.1 and Le of .3mH. The rest is all obtained from their data sheet. How do those numbers compare with what you used ?
Jimbo
That is the point Jimmy154, on being told that the reflex is a "resonant" device and that this is bad people all of a sudden start to hear this "badness", the TL is also a resonant device and has a whole series of resonances in stead of just the one you need to increase the acoustic impedance seen by the driver and greatly increase its effective efficiency.
As the data posted by Scotmoose shows, by the time a TL is acceptably damped it is just an over large and complicated sealed box, so you might as well just make a sealed box in the first place.
But as I pointed out in the article I wrote that appears on Rod Elliots page, most small drivers used in this way are pushed into very audible and harsh sounding odd order harmonic distortion at even modest listening levels, not to mention the flux modulation and d.c. offset effects that cause intermodulation, (these being caused by the extra power and cone excursion needed).
In addition small two way TL systems that have crossovers in the 2-3kHz. region use the bass driver through a region where the inherant pipe resonances of a TL are very audible, and as I pointed out before by the time you get them damped sufficiently the end has no usable out put.
Volume filling a small reflex box eliminates most of the internally reflected sound and has no pipe resonances, and gives lower distortion from all causes for the same spl. it is also smaller and easier to make.
rcw.
As the data posted by Scotmoose shows, by the time a TL is acceptably damped it is just an over large and complicated sealed box, so you might as well just make a sealed box in the first place.
But as I pointed out in the article I wrote that appears on Rod Elliots page, most small drivers used in this way are pushed into very audible and harsh sounding odd order harmonic distortion at even modest listening levels, not to mention the flux modulation and d.c. offset effects that cause intermodulation, (these being caused by the extra power and cone excursion needed).
In addition small two way TL systems that have crossovers in the 2-3kHz. region use the bass driver through a region where the inherant pipe resonances of a TL are very audible, and as I pointed out before by the time you get them damped sufficiently the end has no usable out put.
Volume filling a small reflex box eliminates most of the internally reflected sound and has no pipe resonances, and gives lower distortion from all causes for the same spl. it is also smaller and easier to make.
rcw.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.