Hello all 
. I am interested in closed end TL designs, I don't think the topic has been covered a great deal with ported variants being discussed more.
As far as I understand it, the closed end TL is the "traditional" type, the type originally concieved. Features are a flattened impedence (mainly due to stuffing?) and as far as I've heard this can also improve the midrange response. I also believed the closed end TL provided bass reinforcement, beyond what say a sealed box might give, through quarter wave resonances?
A design that I find interesting is the B&W Nautilus:
http://www.bowers-wilkins.co.uk/display.aspx?infid=1729&terid=1583&sc=hf
whose specs quote a -6db point at 10hz . Is this due to the use of a "heavy" sub type driver in the curled up closed end TL, or are they perhaps quoting an EQ applied response with bass compensation which may be integrated into the active X-over?
Assuming its not the latter, what kind of bass response can one expect from a closed TL? Playing with the MJK sheets I find that virtually no matter what you do, you end up with a response practically identical to an IB response. Is this the norm for such a design or should there be more bass reinforcement when one is designed ideally?
I attach the response of the Hi-Vi M8a driver in a closed TL of 83inches, So=3, Sl=1.2, stuffing 0.5lb ft3, created with MJK worksheets. Would you say it were possible to achieve a "better" (ie, more bass reinforced) response in a closed TL than this? Would you perhaps say that is a good response when coupled with room gain? I'm just looking at various enclosure options for this driver and the closed TL has interested me for a while
Ultimately, why might you choose a closed end TL over a sealed box? Improved midrange performance? Lower group delay? Flatter impedence resulting in easier passive crossover design? Is there any bass extention advantages?
. I am interested in closed end TL designs, I don't think the topic has been covered a great deal with ported variants being discussed more. As far as I understand it, the closed end TL is the "traditional" type, the type originally concieved. Features are a flattened impedence (mainly due to stuffing?) and as far as I've heard this can also improve the midrange response. I also believed the closed end TL provided bass reinforcement, beyond what say a sealed box might give, through quarter wave resonances?
A design that I find interesting is the B&W Nautilus:
http://www.bowers-wilkins.co.uk/display.aspx?infid=1729&terid=1583&sc=hf
whose specs quote a -6db point at 10hz
. Is this due to the use of a "heavy" sub type driver in the curled up closed end TL, or are they perhaps quoting an EQ applied response with bass compensation which may be integrated into the active X-over? Assuming its not the latter, what kind of bass response can one expect from a closed TL? Playing with the MJK sheets I find that virtually no matter what you do, you end up with a response practically identical to an IB response. Is this the norm for such a design or should there be more bass reinforcement when one is designed ideally?
I attach the response of the Hi-Vi M8a driver in a closed TL of 83inches, So=3, Sl=1.2, stuffing 0.5lb ft3, created with MJK worksheets. Would you say it were possible to achieve a "better" (ie, more bass reinforced) response in a closed TL than this? Would you perhaps say that is a good response when coupled with room gain? I'm just looking at various enclosure options for this driver and the closed TL has interested me for a while
Ultimately, why might you choose a closed end TL over a sealed box? Improved midrange performance? Lower group delay? Flatter impedence resulting in easier passive crossover design? Is there any bass extention advantages?
Attachments
Dr.EM said:Hello all
Ultimately, why might you choose a closed end TL over a sealed box? Improved midrange performance? Lower group delay? Flatter impedence resulting in easier passive crossover design? Is there any bass extention advantages?
Impedance is flattened not by stuffing, but by calculating the length to cause a reflection from the tube end back to the driver that is 180 degrees out of phase when it arrives back at the cone, causing the driver to be extremely damped (controlled) at cone resonance. Wrong length could be bad news, as there is no port to help adjust tuning. There are practical limits to how much stuffing can do to correct the fundamental length. Stuffing is more useful to skew, and higher up, absorb, higher resonances of the tube - knock them off from being strong multiples as in a hollow pipe.
With adequate stuffing the Midrange (off a bass driver, at least) will effectively be in an infinite baffle condition - little of the reflection from the tube end survives the round trip back to the cone. (I dunno what they do for the short-tube on the midrange of the Nautilus....must just stuff the heck out of it)
That's some of the theory. MJK and others can better comment on why it's more popular to go open TL. I suspect that a well-tuned, well-damped open TL performs as well or better, possibly with more efficient deep bass for a given line length.
Dr.EM said:
A design that I find interesting is the B&W Nautilus:
http://www.bowers-wilkins.co.uk/display.aspx?infid=1729&terid=1583&sc=hf
whose specs quote a -6db point at 10hz
Assuming its not the latter, what kind of bass response can one expect from a closed TL? Playing with the MJK sheets I find that virtually no matter what you do, you end up with a response practically identical to an IB response. Is this the norm for such a design or should there be more bass reinforcement when one is designed ideally?
Hi,
I would assume it is the latter, i.e. active EQ.
As I understand it sealed TL cannot reinforce bass.
/sreten.
Thanks for your answers, that helps a lot
So in an application with the driver playing up to about 800-900hz I would likely be better off using an open end TL or more likely a MLTL to give bass extention. Using an active crossover and hence the bass drivers being connected directly to a high damping factor amp, the impedence flattening of a closed TL will have no advantage?
Makes me wonder why B&W chose to use one in thier design. The driver only plays up to 220hz so the advantages of possibly cleaner mids are fairly irrelevant and it is also actively driven.
Does the properly designed closed end TL possibly limit cone excursion better/differently to a typical sealed box making it more viable to EQ?
What is the design criteria for defining the length of a closed 1/2 wave TL; the driver resonance? Does fs change in the box, my 83 inch line corresponds to the 33hz resonance I think, but mabye I'm thinking of quarter wave
So in an application with the driver playing up to about 800-900hz I would likely be better off using an open end TL or more likely a MLTL to give bass extention. Using an active crossover and hence the bass drivers being connected directly to a high damping factor amp, the impedence flattening of a closed TL will have no advantage?
Makes me wonder why B&W chose to use one in thier design. The driver only plays up to 220hz so the advantages of possibly cleaner mids are fairly irrelevant and it is also actively driven.
Does the properly designed closed end TL possibly limit cone excursion better/differently to a typical sealed box making it more viable to EQ?
What is the design criteria for defining the length of a closed 1/2 wave TL; the driver resonance? Does fs change in the box, my 83 inch line corresponds to the 33hz resonance I think, but mabye I'm thinking of quarter wave
Ok, thought I'd try another sim. Made line length half wavelength of driver fs, which is a massive 204.7inches . S0 and Sl are both 1.4sd. It seems to "work" like this looking at the impedence plots (created with MJK worksheets):
Undamped:
Damped:
I think the impedence could still be flatter, is it possible to get it totally flat, as in 10 ohms max? This still rises to 20 ohms.
The other "big" problem is the size! If I built that it'd be enourmous yes? I could use a ratio in S0 to Sl which in an open TL can allow a decreased line length, but in this closed end sim it does't seem to have the same effect. The line dimensions at the start would need to be a certain size or the driver would have a wall right behind it, which really won't help!
Also, does using 2 drivers in the line make any difference to the parameters I'm using, should I double sd and vas driver parameters?
. S0 and Sl are both 1.4sd. It seems to "work" like this looking at the impedence plots (created with MJK worksheets):Undamped:
An externally hosted image should be here but it was not working when we last tested it.
Damped:
An externally hosted image should be here but it was not working when we last tested it.
I think the impedence could still be flatter, is it possible to get it totally flat, as in 10 ohms max? This still rises to 20 ohms.
The other "big" problem is the size! If I built that it'd be enourmous yes? I could use a ratio in S0 to Sl which in an open TL can allow a decreased line length, but in this closed end sim it does't seem to have the same effect. The line dimensions at the start would need to be a certain size or the driver would have a wall right behind it, which really won't help!
Also, does using 2 drivers in the line make any difference to the parameters I'm using, should I double sd and vas driver parameters?
Dr.EM said:
FYI, only straight tapered pipes are TLs and the Nautilus is an exponentially? expanding horn with a very high BW throat, so with the right driver specs in its mouth it will have a highly damped response in its pass band with a ~flat acoustic and electrical phase response as well as a critically damped alignment, i.e. all the 'perks' of a FLH, an excellent reason for using it IMO.
GM
I am intrigued
So, the Nautilus is not a tapered closed end TL, but rather a sort of horn in "reverse"? Does this concept only work with this ideal style of construction (ie, with all dimensions expanding) or could a simplified version still work made of wood (so curves possible, but flat sides for instance, as in most BLH designs)?
Can this style of enclosure be simulated with any MJK worksheet? Assuming it is possible to implement without needing fibreglass etc. It may be the enclosure I was seeking really, but I didn't know what it was . Any DIY projects doing this design, I know of Nautilus clones but anybody know any made with wood or with specs/calculations published?
So, the Nautilus is not a tapered closed end TL, but rather a sort of horn in "reverse"? Does this concept only work with this ideal style of construction (ie, with all dimensions expanding) or could a simplified version still work made of wood (so curves possible, but flat sides for instance, as in most BLH designs)?
Can this style of enclosure be simulated with any MJK worksheet? Assuming it is possible to implement without needing fibreglass etc. It may be the enclosure I was seeking really, but I didn't know what it was
. Any DIY projects doing this design, I know of Nautilus clones but anybody know any made with wood or with specs/calculations published?
Well, it would be designed like an ideal BLH, but with the driver sealing off the mouth instead of being in the filter chamber, so no, I don't know for sure of any program that will sim it, though Akabak is so flexible it might can, albeit with a steep learning curve.
I'm not aware of any DIY variants and yes, down where the WLs are long the expansion can be approximated with numerous flat panels as Olson proved with his late '30s BL corner horn that was the inspiration for this BVR: http://homepage.mac.com/tlinespeakers/FH/images/Sachico.gif
GM
I'm not aware of any DIY variants and yes, down where the WLs are long the expansion can be approximated with numerous flat panels as Olson proved with his late '30s BL corner horn that was the inspiration for this BVR: http://homepage.mac.com/tlinespeakers/FH/images/Sachico.gif
GM
Hmm, sounds near impossible to sim then
Is it a lot different acoustically to the tapered TL? Visibly the difference could be simplified to the rate of expansion but your saying this has effects acoustically?
I know virtually nothing about horns but presume this particular design won't increase efficiency as horns typically do? Also, is bass extension gained in this design over the closed end TL or sealed? Perhaps excursion is reduced as per horns typically, this would be a significant advantage.
It'd be nice to persue this design for a bass enclosure in a DIY system but it looks like it may be too difficult, unknown territory sort of
Is it a lot different acoustically to the tapered TL? Visibly the difference could be simplified to the rate of expansion but your saying this has effects acoustically?
I know virtually nothing about horns but presume this particular design won't increase efficiency as horns typically do? Also, is bass extension gained in this design over the closed end TL or sealed? Perhaps excursion is reduced as per horns typically, this would be a significant advantage.
It'd be nice to persue this design for a bass enclosure in a DIY system but it looks like it may be too difficult, unknown territory sort of
Dr.EM said:
Yes, you can flatten it to near Re if there's enough acoustic mass resistance.
Yes, for a given alignment, double the drivers, double Sd, Vas. If parallel wired, Re, Le is halved and BL is unchanged while in series they are all doubled.
Like any alignment, it's size will be based on its Vas, Qts for a given gain BW, so it need not be overly large unless reasonably high efficiency down low is required. For instance I used a theoretical 12" driver to make a simple zero throat reverse horn flat to 40 Hz in < 7 ft^3 with the trade-off being abysmal efficiency.
GM
Attachments
Thanks for your reply, that clears up driver multiples nicely
So, this response that you've simmed you say is a reverse horn; that being the same as used in the Nautilus? It looks like its simmed on the MJK sheets, is it just that it uses a linear expansion and hence behaves as a closed end TL? Or is there some other worksheet; I haven't looked at the horn ones yet.
Can the efficiency become lower than basic driver efficiency when loaded like this? I might be interpreting this wrong but isn't 7ft^3 enourmous?! That infers 7ft in each dimension, L,H,W?
So, this response that you've simmed you say is a reverse horn; that being the same as used in the Nautilus? It looks like its simmed on the MJK sheets, is it just that it uses a linear expansion and hence behaves as a closed end TL? Or is there some other worksheet; I haven't looked at the horn ones yet.
Can the efficiency become lower than basic driver efficiency when loaded like this? I might be interpreting this wrong but isn't 7ft^3 enourmous?! That infers 7ft in each dimension, L,H,W?
The sim is either a simple linear or conic expansion AFAIK (not sure which in MJK's CLOSED that I used, etc., worksheets), not presumably the Nautilius's hyperbolic/expo/whatever expansion nor its filter chamber (the big cavity at its center), so no comparison, plus the driver specs required for each would be somewhat different.
Yes it can and why this theoretical driver has such a low Fs/high Qts, which the Nautilus surely doesn't.
7 ft^3 is its total cubic volume (Vb), so for a cube it would be ~1.44225 ft a side, not very big by my standards and being tapered/sealed it can be folded up to fit a more svelte/eye pleasing rectangular/column/tapered shape.
GM
Yes it can and why this theoretical driver has such a low Fs/high Qts, which the Nautilus surely doesn't.
7 ft^3 is its total cubic volume (Vb), so for a cube it would be ~1.44225 ft a side, not very big by my standards and being tapered/sealed it can be folded up to fit a more svelte/eye pleasing rectangular/column/tapered shape.
GM
The response seems bad to me for a TL -3db 50Hz I see in your plot.
You shout look for a high QTS and xmax low fs speaker.
Like this peerless great price quality ratio
sls315
Or hivi research d10
d10
even better tangband TIP
speaker TB
An externally hosted image should be here but it was not working when we last tested it.
complete TL-kit (227 euro !!)
An externally hosted image should be here but it was not working when we last tested it.
- Frequenzbereich (-8dB): 32 - 40 000 Hz
- Trennfrequenzen: 150 und 3000 Hz
tb
The tang band TL is really super.
Why? Because with a high x-max you will get Doppler effect distortion off the mid frequencies.
So I will advise to XO under 300Hz. The TB TL is super because it is XO on the right frequencies and their special full range chassis does the big deal and the ribbon makes the high frequencies makes it complete. And super price I think.
Nice alu chassis dayton
Dayton got a lot of candidates they look cool to.
An externally hosted image should be here but it was not working when we last tested it.
Dayton got a lot of candidates they look cool to.
An externally hosted image should be here but it was not working when we last tested it.
Thanks for your replies. I was looking specifically at applications for the M8a driver though you suggest it may not be optimal. I actually have a pair of D10G drivers! I was going to use them in a different 3 way in sealed boxes but am considering changing my design to a hard coned setup with M8a bass/mid (RS52 mid, RT2-EA tweet).
Is it generally bad practice to have the woofer of a 3 way cover up to 800hz due to the doppler effect? If I were using 2 a side I guess in some ways it'd be best to have one for below 80hz or so and one for above, so only 1 has high excursion. Disadvantages being reduced max spl of low frequency and requirement for a 4th stereo amplifier
Still interested in the Nautilus style enclosure if there is any chance of making something effective for the M8a. Experimenting with a test setup like in the photo planet10 posted would be good but you'd need at least some starting point. Once you have it roughly right from sims you can experiment with your test "box" until it achieves flat measured impedence?
Is it generally bad practice to have the woofer of a 3 way cover up to 800hz due to the doppler effect? If I were using 2 a side I guess in some ways it'd be best to have one for below 80hz or so and one for above, so only 1 has high excursion. Disadvantages being reduced max spl of low frequency and requirement for a 4th stereo amplifier
Still interested in the Nautilus style enclosure if there is any chance of making something effective for the M8a. Experimenting with a test setup like in the photo planet10 posted would be good but you'd need at least some starting point. Once you have it roughly right from sims you can experiment with your test "box" until it achieves flat measured impedence?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.