If I were to build a classic aperiodic TL with these units, personally, I'd do something on these lines, derived mostly from Martin's alignment tables.
Line 76.3in long with a ~3:1 taper. So = 74.5in^3, Sl=23.25in^3. Driver centre mounted 25.5in down from So. Should allow for the vent to be near the floor too, so you get more lift from the room. Lynn Olson noted this when he built his Ariels -the room lights up when you get the vent close to the floor boundary condition. Stuff 0.54lbs ft^3 from the top 68in down & adjust to taste / the room requirements. The density will depend on what you use for material, though differences are usually very subtle. Augsperger found some differences, though nothing I'd loose any sleep over. FWIW, I'd use an internal width of ~7.5in.
Line 76.3in long with a ~3:1 taper. So = 74.5in^3, Sl=23.25in^3. Driver centre mounted 25.5in down from So. Should allow for the vent to be near the floor too, so you get more lift from the room. Lynn Olson noted this when he built his Ariels -the room lights up when you get the vent close to the floor boundary condition. Stuff 0.54lbs ft^3 from the top 68in down & adjust to taste / the room requirements. The density will depend on what you use for material, though differences are usually very subtle. Augsperger found some differences, though nothing I'd loose any sleep over. FWIW, I'd use an internal width of ~7.5in.
Attachments
dave wrote:
Tony knows that the Optimo design isn't fully neutral. At the moment he wants something really neutral he uses closed boxes, both in the mid and bass-ranges as in his "project-x". However his conclusion about a relatively large constant x-section for a two-way isn't fully wrong vs BR or closed enclosures: low-mid and midrange escaping from the mouth sure gives this special listening and at the same time a nice and tight bass-range...confirmed by anatole51.
I agree with your comments about the shortcomings of the different tunings. He simply didn't want to build several enclosures too much time consuming. So his conclusion has not to be regarded as a rule of laws but as a result of these six designs with these drivers. But what I can say is that my little TB TL was tuned 10Hz bellow fs and that this special (but not neutral) midrange presentation remained.
Tony knows that the Optimo design isn't fully neutral. At the moment he wants something really neutral he uses closed boxes, both in the mid and bass-ranges as in his "project-x". However his conclusion about a relatively large constant x-section for a two-way isn't fully wrong vs BR or closed enclosures: low-mid and midrange escaping from the mouth sure gives this special listening and at the same time a nice and tight bass-range...confirmed by anatole51.
Some speaker companies like PMC use a lot of damping material in the TL line to absorb midrange frequencies before exiting the port.
The Optimo design has port on top, so how troublesome is the placement of this speaker near back wall, since midrange frequencies are not totally absorbed inside speaker?
The Optimo design has port on top, so how troublesome is the placement of this speaker near back wall, since midrange frequencies are not totally absorbed inside speaker?
Scottmoose,
Thanks for your simulations and your personal idea about the loading of this scan-speak driver. But then:
1) the present midrange character of the Optimo will probably less or more disappear.
2) the filter will have to be redesigned.
Future Optimo builders have to know these consequences!
Thanks for your simulations and your personal idea about the loading of this scan-speak driver. But then:
1) the present midrange character of the Optimo will probably less or more disappear.
2) the filter will have to be redesigned.
Future Optimo builders have to know these consequences!
Re the above two points,
1) The midrange response between the cabinets should theoretically change very little, the tapered line being a little cleaner if anything as higher pipe harmonic modes are reduced naturally rather than needing extra damping to do it. Output from the terminus should be done by ~200Hz. In both cabinets, most of the midrange will be damped out by the stuffing / lining used in the cabinet, which is in the first half of the line, where the velocity of such frequencies is highest. Leaving the 2nd half of the line (relatively) undamped will preserve the cabinet 1st mode, and a little (but not much) of the lower mids.
2) Why would the filter have to be redesigned? I see nothing in the XO that would make any dramatic difference in the response between the two cabinets below 1KHz. Above that, the enclosure ceases to have any real function, aside from providing a solid mounting point for the driver.
Going onto general terms, a proper TL (and you can count on the fingers of one hand how many real TLs there've been) should be just as neutral as a sealed box, but with a far easier impedence curve (so for those without monster amps, it'll probably beat the sealed box in the neutrality stakes as the amp isn't having to work as hard). Most, like this, or the rough suggestion I made above, are hybrid, aperiodic lines, damping most of the driver's BW & presenting as flat an impedence to the amp as practicable, response, while preserving a little of the QW fundamental to prop the bass up fractionally -a pretty good compromise. Bailey (ignoring his misleading title) did this with his original design, and the Radfords developed from it.
Sticking the vent on top of a speaker is interesting in a way. I've done a lot of work over the past couple of years with pipe-horns (the BIB) and it can work very well when pushed against a rear wall, or corner loaded. For short cabinets, I like to get it close to the floor, because then you're actually using a very large, expanding air-mass (the room corner) to load the pipe & you end up with a degree of horn-loading. You want to make sure nothing over about 300-350Hz is being released from the cabinet though, or phase and imaging get badly messed up.
1) The midrange response between the cabinets should theoretically change very little, the tapered line being a little cleaner if anything as higher pipe harmonic modes are reduced naturally rather than needing extra damping to do it. Output from the terminus should be done by ~200Hz. In both cabinets, most of the midrange will be damped out by the stuffing / lining used in the cabinet, which is in the first half of the line, where the velocity of such frequencies is highest. Leaving the 2nd half of the line (relatively) undamped will preserve the cabinet 1st mode, and a little (but not much) of the lower mids.
2) Why would the filter have to be redesigned? I see nothing in the XO that would make any dramatic difference in the response between the two cabinets below 1KHz. Above that, the enclosure ceases to have any real function, aside from providing a solid mounting point for the driver.
Going onto general terms, a proper TL (and you can count on the fingers of one hand how many real TLs there've been) should be just as neutral as a sealed box, but with a far easier impedence curve (so for those without monster amps, it'll probably beat the sealed box in the neutrality stakes as the amp isn't having to work as hard). Most, like this, or the rough suggestion I made above, are hybrid, aperiodic lines, damping most of the driver's BW & presenting as flat an impedence to the amp as practicable, response, while preserving a little of the QW fundamental to prop the bass up fractionally -a pretty good compromise. Bailey (ignoring his misleading title) did this with his original design, and the Radfords developed from it.
Sticking the vent on top of a speaker is interesting in a way. I've done a lot of work over the past couple of years with pipe-horns (the BIB) and it can work very well when pushed against a rear wall, or corner loaded. For short cabinets, I like to get it close to the floor, because then you're actually using a very large, expanding air-mass (the room corner) to load the pipe & you end up with a degree of horn-loading. You want to make sure nothing over about 300-350Hz is being released from the cabinet though, or phase and imaging get badly messed up.
I must second the sense of 'ease' when the impedance is benign in the nether regions. A trick I have used and worked nicely in my sealed box bass 3 way, was to stagger the peaks of the sealed woofers with the aperiodic midbass, and in a manner parallel them by gradually cutting the midbass unit, letting it go low but slow. That resulted in a very 'driveable' under 100Hz impedance with minute peaking, and a bass quality, that subjectively sounds almost as free as when a TL is being driven.
Attachments
Scottmoose,
What about the bass and low-mid ranges? Your tapered line will change enough of output energy in these areas and the baffle step compensation (so L2 and therefore almost all the other components too...it's a series filter!) will very probably have to be redesigned. I guess that the amount of damping material will not be suffisant to get the proper balanced sound. Personaly I would not change this loading design without preparing me to change the filter.
What about the bass and low-mid ranges? Your tapered line will change enough of output energy in these areas and the baffle step compensation (so L2 and therefore almost all the other components too...it's a series filter!) will very probably have to be redesigned. I guess that the amount of damping material will not be suffisant to get the proper balanced sound. Personaly I would not change this loading design without preparing me to change the filter.
If you noticed, although I suggested a 7.5in internal width for ease, all I did was give cross sectional areas. You can use any width you want -all you have to do is adjust the depth dimension to keep the line CSA correct. Hardly a big deal. The original for those who don't like fooling with XOs (me included, mainly because I don't like most XOs) or a different width to the original for those who do enjoy XO tweaks. That's up to the builder. You'd have to make a change of several inches in diameter though before things became noticable. A slightly wider baffle than a circuit is designed for isn't usually audible. It's when a baffle is narrower than said circuit is designed for that things start to go pear-shaped in a big way. 
Beyond that, gain shouldn't be any different (just with a considerable reduction of harmonic nulls), you just get more extension at the bottom end.
Beyond that, gain shouldn't be any different (just with a considerable reduction of harmonic nulls), you just get more extension at the bottom end.
Scottmoose,
I wrote BSC, I thought spl balance, nothing related to baffle diffraction, sorry (in English I don't use correct words sometimes).
IMO unsmoothed ripples can be heard in the bass and low-mid ranges: Besides other parameters, it is mainly these peaks and valleys that give the different sounding character of such speakers, not the impedance peak at fb or higher (except with valves amps or poor solid states which furthermore can be compensated with a LCR-network); dampening material cannot help in this matter except if very heavy.
So frequencies balance is different vs a speaker with smoothed ripples. And in a filter this needs to be corrected in an almost identical manner than BSC.
I wrote BSC, I thought spl balance, nothing related to baffle diffraction, sorry (in English I don't use correct words sometimes).
IMO unsmoothed ripples can be heard in the bass and low-mid ranges: Besides other parameters, it is mainly these peaks and valleys that give the different sounding character of such speakers, not the impedance peak at fb or higher (except with valves amps or poor solid states which furthermore can be compensated with a LCR-network); dampening material cannot help in this matter except if very heavy.
So frequencies balance is different vs a speaker with smoothed ripples. And in a filter this needs to be corrected in an almost identical manner than BSC.
Ah, I see what you're driving at now. Believe me though, your English is good. My language skills are ~non existant.
Actually, IMO, harmonic ripple should not be treated in the same way as baffle step loss. A BSC circuit works by attenuating everything above the transition from 4pi radiation space to 2pi radiation space to rebalance the sound. Harmonic ripple however are actually nulls in the response. There shouldn't be anything to attenuate, unless you're hoping to reduce the gain BW -a rather odd thing to do, becuase the damping will have brought it back into line with the drivers IB response anyway (that's the entire point of using it in a semi-resonant / aperiodic line). Whatever, the XO should work fine in a different line geometry, assuming its been designed to provide the same gain as the original. The basic balance of the speaker will be unaffected. Some might miss the nulls in the response, personally, I prefer something a bit more accurate.
Actually, IMO, harmonic ripple should not be treated in the same way as baffle step loss. A BSC circuit works by attenuating everything above the transition from 4pi radiation space to 2pi radiation space to rebalance the sound. Harmonic ripple however are actually nulls in the response. There shouldn't be anything to attenuate, unless you're hoping to reduce the gain BW -a rather odd thing to do, becuase the damping will have brought it back into line with the drivers IB response anyway (that's the entire point of using it in a semi-resonant / aperiodic line). Whatever, the XO should work fine in a different line geometry, assuming its been designed to provide the same gain as the original. The basic balance of the speaker will be unaffected. Some might miss the nulls in the response, personally, I prefer something a bit more accurate.
Scott:
Over the closed box of equal volume, (dotted blue line), your solution seems to yield 2 dB ripple in the midbass/midrange in exchange or a 2 dB extra sensitivity in the bass range.
Is it really worth the trouble to go through all that fine tuning the TL requires over the closed box?
http://www.diyaudio.com/forums/attachment.php?postid=1284252
Over the closed box of equal volume, (dotted blue line), your solution seems to yield 2 dB ripple in the midbass/midrange in exchange or a 2 dB extra sensitivity in the bass range.
Is it really worth the trouble to go through all that fine tuning the TL requires over the closed box?
http://www.diyaudio.com/forums/attachment.php?postid=1284252
Scottmoose,
Thanks to you to think that my English is relatively correct...and thanks to the translate machine
OK, I understand what you mean with the harmonic ripples only giving nulls and no peaks. So you are discussing about harmonics of the foundamental resonance of the system, OK.
For me a Transmission Line acts on this foundamental resonance (quarter or half wave depending of the lenght of the line, sealed or opened) and as you said provides gain over a large bandpass and flattens the impedance.
But for me the tapered line used in the mid enclosure of the B&W acts mainly on the standing waves( read: suppress their "birth") by progressively increasing the impedance during the waves travel (contrary of a horn): this enclosure damps the sound behind the cone so no back waves will return to the cone.
A different shape of the line on the Optimo will provide different standing waves and maybe this fact is the real reason of a different sound that will have to be dealed within the filter.
Do you think I'm wrong?
Thanks to you to think that my English is relatively correct...and thanks to the translate machine
OK, I understand what you mean with the harmonic ripples only giving nulls and no peaks. So you are discussing about harmonics of the foundamental resonance of the system, OK.
For me a Transmission Line acts on this foundamental resonance (quarter or half wave depending of the lenght of the line, sealed or opened) and as you said provides gain over a large bandpass and flattens the impedance.
But for me the tapered line used in the mid enclosure of the B&W acts mainly on the standing waves( read: suppress their "birth") by progressively increasing the impedance during the waves travel (contrary of a horn): this enclosure damps the sound behind the cone so no back waves will return to the cone.
A different shape of the line on the Optimo will provide different standing waves and maybe this fact is the real reason of a different sound that will have to be dealed within the filter.
Do you think I'm wrong?
Crazyhub:
I am not Scott, but I would like to hazard an answer.
It is true that the difference between bass reflex and transmission line will have repurcussions above the lowest bass octave. However, the bass loading, whether transmission line or bass reflex, is at a maximum an octave or so above the -3 dB down point. After that, the port output in either type of enclosure goes down, down, down. There might be a few peaks in the TL line a couple of octaves above that if you don't damp the enclosure correctly, but that is it. The more octaves you go above the octave where the bass loading is most active, the less you will be affected.
The -3 dB down point of this speaker is 40 Hz or so.
But the crossover is at 2500 Hz. The crossover point is fully six octaves above the octave where the bass loading is most active. Whatever differences exist between bass reflex and transmission line enclosures will have have diminished to nothing by the time you get up to the sixth octave, so the same crossover should be used in both types of enclosure.
As for the B&W midrange enclosure, remember that it is indeed a midrange enclosure. That is to say, the tuning will be far above 40 Hz, so the crossover point to the tweeter will be considerably less than six octaves above the point where the bass loading is most active.
I am not Scott, but I would like to hazard an answer.
It is true that the difference between bass reflex and transmission line will have repurcussions above the lowest bass octave. However, the bass loading, whether transmission line or bass reflex, is at a maximum an octave or so above the -3 dB down point. After that, the port output in either type of enclosure goes down, down, down. There might be a few peaks in the TL line a couple of octaves above that if you don't damp the enclosure correctly, but that is it. The more octaves you go above the octave where the bass loading is most active, the less you will be affected.
The -3 dB down point of this speaker is 40 Hz or so.
But the crossover is at 2500 Hz. The crossover point is fully six octaves above the octave where the bass loading is most active. Whatever differences exist between bass reflex and transmission line enclosures will have have diminished to nothing by the time you get up to the sixth octave, so the same crossover should be used in both types of enclosure.
As for the B&W midrange enclosure, remember that it is indeed a midrange enclosure. That is to say, the tuning will be far above 40 Hz, so the crossover point to the tweeter will be considerably less than six octaves above the point where the bass loading is most active.
Hello kelticwizard,
Now I fully understand your thoughts (Scott also) about the effects and amount of resonances on TL and their (non)-repercussion on the x-over. So the Optimo x-over should work fine on the suggested loading design from Scott, assuming that the global gain of the band-pass remains the same and perhaps that the different standing waves will certainly be killed by the amount of dampening material.
Also I guess that in my Tang-Band TL (large constant x-section as in the Optimo) the uncommon spacial presentation could come from both:
- reduced standing-waves so no waves returning to the cone
- a gain in the bass and low-mid ranges (I perfectly heard high-bass/low-mid escaping from the mouth).
I never heard any other TL TQWT or rear loaded horns, so my question about this particular sounding character.
Now I fully understand your thoughts (Scott also) about the effects and amount of resonances on TL and their (non)-repercussion on the x-over. So the Optimo x-over should work fine on the suggested loading design from Scott, assuming that the global gain of the band-pass remains the same and perhaps that the different standing waves will certainly be killed by the amount of dampening material.
Also I guess that in my Tang-Band TL (large constant x-section as in the Optimo) the uncommon spacial presentation could come from both:
- reduced standing-waves so no waves returning to the cone
- a gain in the bass and low-mid ranges (I perfectly heard high-bass/low-mid escaping from the mouth).
I never heard any other TL TQWT or rear loaded horns, so my question about this particular sounding character.
I have just discovered too late this discussion. I hope that my contribution will be wellcomed ...
After many years of listening of a prototype now I would like to have a better, nicer (and more expensive) enclosure for my TL.
In the meanwhile I discovered the Tony Gee's work with the Optimo, and since we use the same mid-woofer (Scan-Speak 18W/4531G00) it was quite natural to me to make some comparisons, ... and have some doubt.
So: if I have to spend money for a better enclosure, what is better, keep my original design or use the Optimo design ?
My design has S0=0.037m^2, SL=0.013m^2, L=0.164m, woofer offset=36.7cm (sorry if I don't use Imperials). On the line there are 300g of suffing distributed over 4/5 of the line under the woofer (12kg/m^3).
The MJK worksheet predictions are in attach.
How it sounds ? different from a BR. Bass is present but without emphasis, there are recordings full of basses where they go deep, really deep, other recordings have less basses and the basses are not "invented", like it happens with other systems.
A BR aficionado can react with much perplexity.
Me, I could on the other side emphasize the speed of the basses, and the cleanliness of the midrange.
Any comment at this stage will be greatly appreciated.
Thanks
After many years of listening of a prototype now I would like to have a better, nicer (and more expensive) enclosure for my TL.
In the meanwhile I discovered the Tony Gee's work with the Optimo, and since we use the same mid-woofer (Scan-Speak 18W/4531G00) it was quite natural to me to make some comparisons, ... and have some doubt.
So: if I have to spend money for a better enclosure, what is better, keep my original design or use the Optimo design ?
My design has S0=0.037m^2, SL=0.013m^2, L=0.164m, woofer offset=36.7cm (sorry if I don't use Imperials). On the line there are 300g of suffing distributed over 4/5 of the line under the woofer (12kg/m^3).
The MJK worksheet predictions are in attach.
How it sounds ? different from a BR. Bass is present but without emphasis, there are recordings full of basses where they go deep, really deep, other recordings have less basses and the basses are not "invented", like it happens with other systems.
A BR aficionado can react with much perplexity.
Me, I could on the other side emphasize the speed of the basses, and the cleanliness of the midrange.
Any comment at this stage will be greatly appreciated.
Thanks
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.