1.5 for the Tang Band W5-2143 or F.A.S.T.?
I was planning to build a two-way TL to replace a set of three-way floor-standers my wife found too big when I got sidetracked by my curiosity about FR speakers. So I bought a pair of Tang Band W5-2143 to play with. I put them in a 12 liter vented box tuned at 55 Hz. My wife asked for “something white that blends with the walls” so I made them as flat as possible intended to be hung of the wall (se attached images).
It was an experiment that turned out surprisingly well. They are far from a “perfect” speaker but what they do well they do really well. We use them a lot with low volume background music, mainly jazz and vocal and they are perfect for that. They are a bit bass shy and lack SPL for other material like rock, especially in our fairly big living room. I added a sealed sub tuned around 80 Hz and it helps but it’s far from perfect. I really like the sound from these elements so I have started to think about alternatives to counter the weaknesses of the current design and maybe even get rid of the sub. Two alternatives that comes to mind:
I would really need help on how to construct a XO for this speaker. I was hoping I could get along with just one inductor for one of the drivers to cut it around 300 to 500 Hz. But I got no experience building XOs so what do I know?
I would also like to know if a speaker design like the Eikona VTL could be used and if it would be better or worse? It’s not a ML TQWT, more of a -ve TL, so I guess I could get away with a SL of only 3 * Sd, 97.83 cm² * 3 = 293.49 cm², but I guess it would also require a bit longer line length. It would probably be a bit trickier to fit the two drivers on the baffle, I was planning on vertical with the one playing low on top but something like the Eikona VTL might force me to put them on a 45 degree axis or even horizontally! Any opinions on how that would affect the sound?
Then we have alternative 2. I have no idea of what woofer I could pair the Tang Band W5-2143 with in a F.A.S.T. design. I totally lack experience of F.A.S.T. and XO design and I would like to stay away from active XO in this case. So I’m stuck again, any help would be appreciated.
Regards / Rob
I was planning to build a two-way TL to replace a set of three-way floor-standers my wife found too big when I got sidetracked by my curiosity about FR speakers. So I bought a pair of Tang Band W5-2143 to play with. I put them in a 12 liter vented box tuned at 55 Hz. My wife asked for “something white that blends with the walls” so I made them as flat as possible intended to be hung of the wall (se attached images).
It was an experiment that turned out surprisingly well. They are far from a “perfect” speaker but what they do well they do really well. We use them a lot with low volume background music, mainly jazz and vocal and they are perfect for that. They are a bit bass shy and lack SPL for other material like rock, especially in our fairly big living room. I added a sealed sub tuned around 80 Hz and it helps but it’s far from perfect. I really like the sound from these elements so I have started to think about alternatives to counter the weaknesses of the current design and maybe even get rid of the sub. Two alternatives that comes to mind:
- Add another W5-2143 element and make it a 1.5 design.
- Add a woofer and make it a FAST design.
I would really need help on how to construct a XO for this speaker. I was hoping I could get along with just one inductor for one of the drivers to cut it around 300 to 500 Hz. But I got no experience building XOs so what do I know?
I would also like to know if a speaker design like the Eikona VTL could be used and if it would be better or worse? It’s not a ML TQWT, more of a -ve TL, so I guess I could get away with a SL of only 3 * Sd, 97.83 cm² * 3 = 293.49 cm², but I guess it would also require a bit longer line length. It would probably be a bit trickier to fit the two drivers on the baffle, I was planning on vertical with the one playing low on top but something like the Eikona VTL might force me to put them on a 45 degree axis or even horizontally! Any opinions on how that would affect the sound?
Then we have alternative 2. I have no idea of what woofer I could pair the Tang Band W5-2143 with in a F.A.S.T. design. I totally lack experience of F.A.S.T. and XO design and I would like to stay away from active XO in this case. So I’m stuck again, any help would be appreciated.
Regards / Rob
Attachments
Last edited:
RE: TB w5 2143
Hello EmuMannen, since nobody with more experience posted anything I'm going to give you a reply as someone very much interested in designing TL and this driver - the TB w5 2143.
The first thing I noticed about the thread is that you've build yourself a nice pair of speakers, really good looking... unfortunately you are not that familiar with designing enclosures and this is where the problem seems to have come from.
If I understand correctly your problem is the fact that the bass is too thin / shy... I personally I'm considering buying a pair of w52143 but have postponed it for two reasons - bass too little & dispersion might be too bad for me (and there aren't that many choices for this driver size). You are not the first to build a bass reflex enclosure for this driver and realize it's not good enough - Baffle Step compensate for TB W5-2143 Full range - Techtalk Speaker Building, Audio, Video Discussion Forum - and if you read the whole thread the BSC did not solve the problem only made the sound less musical. I have made extensive research on the internet and the general consensus of people who used the driver is that is lacking in bass and several independent measurements confirm it - up to 7 db under nominal SPL depending on tuning and enclosure type. No need to spend 5 hour for this search - it was in the specs - Qts = 0.38 + Xmax = 2.5 mm ( https://www.parts-express.com/pedocs/specs/264-958--w5-2143-spec-sheet.pdf ) - those numbers generally point to a driver with (bellow) average capability in the bass.
The general hierarchy of enclosure bass output is - min: closed - bass reflex - TL/cone - TL straight - TL expanding - horn (full size) - max. The size of the enclosure is larger for more bass output (see "bigger is better" enclosure bass reviews). Second of all you chosen a 12 liter alignment for your bass reflex - without looking at a simulation I think this is a compromise sized enclosure (smaller than recommended for 55 hz tuning)... not to mention that you have a large room.
Another mistake (lack of knowledge) you made there was to consider that the size of a transmission line is related to Sd - it's not, it's related to Vas - Vas gives you internal volume, Fs & Qts gives you tuning frequency (line length...). And yes - adding a second driver / enclosure will improve bass output but in the case of a Mileva you can't just double the enclosure, you need to redesign it... so just start from zero for another enclosure optimized for your needs.
As far as your question is concerned I would probably add a subwoofer made with the same aesthetics and keep the nice looking and already finished and installed enclosures. But! if this would have been my build I would not consider any subwoofers of tiny bass reflexes... I would build a large floorstander (TL) with the terminus placed at floor level (for even more gain)... from simulations I know this will be at least 36 liter in volume, tuned around 50 hz, and build one test enclosure to see if it works in reality before any approval stamps.
The best bass output enclosure on the internet currently was at 0 db (in line with the rest of the audio spectrum) down to 45 hz, a large transmission line with contracting geometry (PartsExpress).
By the way - can you say a few words about the high frequency dispersion?
Hello EmuMannen, since nobody with more experience posted anything I'm going to give you a reply as someone very much interested in designing TL and this driver - the TB w5 2143.
The first thing I noticed about the thread is that you've build yourself a nice pair of speakers, really good looking... unfortunately you are not that familiar with designing enclosures and this is where the problem seems to have come from.
If I understand correctly your problem is the fact that the bass is too thin / shy... I personally I'm considering buying a pair of w52143 but have postponed it for two reasons - bass too little & dispersion might be too bad for me (and there aren't that many choices for this driver size). You are not the first to build a bass reflex enclosure for this driver and realize it's not good enough - Baffle Step compensate for TB W5-2143 Full range - Techtalk Speaker Building, Audio, Video Discussion Forum - and if you read the whole thread the BSC did not solve the problem only made the sound less musical. I have made extensive research on the internet and the general consensus of people who used the driver is that is lacking in bass and several independent measurements confirm it - up to 7 db under nominal SPL depending on tuning and enclosure type. No need to spend 5 hour for this search - it was in the specs - Qts = 0.38 + Xmax = 2.5 mm ( https://www.parts-express.com/pedocs/specs/264-958--w5-2143-spec-sheet.pdf ) - those numbers generally point to a driver with (bellow) average capability in the bass.
The general hierarchy of enclosure bass output is - min: closed - bass reflex - TL/cone - TL straight - TL expanding - horn (full size) - max. The size of the enclosure is larger for more bass output (see "bigger is better" enclosure bass reviews). Second of all you chosen a 12 liter alignment for your bass reflex - without looking at a simulation I think this is a compromise sized enclosure (smaller than recommended for 55 hz tuning)... not to mention that you have a large room.
Another mistake (lack of knowledge) you made there was to consider that the size of a transmission line is related to Sd - it's not, it's related to Vas - Vas gives you internal volume, Fs & Qts gives you tuning frequency (line length...). And yes - adding a second driver / enclosure will improve bass output but in the case of a Mileva you can't just double the enclosure, you need to redesign it... so just start from zero for another enclosure optimized for your needs.
As far as your question is concerned I would probably add a subwoofer made with the same aesthetics and keep the nice looking and already finished and installed enclosures. But! if this would have been my build I would not consider any subwoofers of tiny bass reflexes... I would build a large floorstander (TL) with the terminus placed at floor level (for even more gain)... from simulations I know this will be at least 36 liter in volume, tuned around 50 hz, and build one test enclosure to see if it works in reality before any approval stamps.
The best bass output enclosure on the internet currently was at 0 db (in line with the rest of the audio spectrum) down to 45 hz, a large transmission line with contracting geometry (PartsExpress).
By the way - can you say a few words about the high frequency dispersion?
Thanks for your reply krakatoa. The 12 liter box was actually modelled and is "perfect" for the 55 Hz tuning, I just made it flatter than usual. I knew beforehand that it would be bass shy and initially played with the idea of making even smaller 6 liter sealed enclosures and use them as satellite speakers. But I wanted to know how "deep" I could make them go and what it would sound like in a bog standard BR enclosure (I choose BR because of the simplicity of modelling and predictable result). As it turns out, not enough SPL for the fairly large room and not enough bass by it's own in this configuration. Did not come as a surprise but the question is if I could do something to remedy these limitations? I do not expect them to challenge my BIC Soundspan TPR 600, that would defy the laws of physics, just make them a bit better, make them go a bit lower and add a bit more SPL down low.
I know I have to redesign something like the Minelva and I started to make some rough estimations. My first stab using Fs: 55 Hz, Sd: 196 cm2, a depth of 15 cm and SL 3 * Sd ended up something like (internal measures in cm):
Height: 80.9, Width: 42.4, Depth: 15
T1: 1, T2: 20, T3: 62-70, T4: 63
Port length: 12 and Port area: 49
Line length: 155.45 and F3: 40.9 Hz
It's a bit too big for my likings to I went back and forth a bit ending up with a depth of 13 cm and SL 2.25 * Sd, witch gave me:
Height: 80.3, Width: 37.3, Depth: 13
T1: 2, T2:18, T3: 62-70, T4: 64
Port length: 12 and Port area: 49 cm2
Line length: 155.45 and F3: 44.7 Hz
Simulations look fairly ok in Hornreps (the only modelling software I got for TL) but I'm not sure how to model the 1.5 part. I am now using the "Driver Arrangement" OD with two W5-2143 in parallel but there is no way to add a filter to one of the drivers.
I will probably build one of these boxes just to test it out with the two drivers I already got. But I really need some advice for the XO, hard to find information about 1.5 XO for two full range elements. I know active XO would be better but it would be nice with the simplicity of a passive one (hopefully just an inductor on one of the drivers) if possible. Don't know where to cross them, 250-500 Hz maybe? 250 Hz would make for a massive air core I guess.
I know I have to redesign something like the Minelva and I started to make some rough estimations. My first stab using Fs: 55 Hz, Sd: 196 cm2, a depth of 15 cm and SL 3 * Sd ended up something like (internal measures in cm):
Height: 80.9, Width: 42.4, Depth: 15
T1: 1, T2: 20, T3: 62-70, T4: 63
Port length: 12 and Port area: 49
Line length: 155.45 and F3: 40.9 Hz
It's a bit too big for my likings to I went back and forth a bit ending up with a depth of 13 cm and SL 2.25 * Sd, witch gave me:
Height: 80.3, Width: 37.3, Depth: 13
T1: 2, T2:18, T3: 62-70, T4: 64
Port length: 12 and Port area: 49 cm2
Line length: 155.45 and F3: 44.7 Hz
Simulations look fairly ok in Hornreps (the only modelling software I got for TL) but I'm not sure how to model the 1.5 part. I am now using the "Driver Arrangement" OD with two W5-2143 in parallel but there is no way to add a filter to one of the drivers.
I will probably build one of these boxes just to test it out with the two drivers I already got. But I really need some advice for the XO, hard to find information about 1.5 XO for two full range elements. I know active XO would be better but it would be nice with the simplicity of a passive one (hopefully just an inductor on one of the drivers) if possible. Don't know where to cross them, 250-500 Hz maybe? 250 Hz would make for a massive air core I guess.
Interesting, could you please elaborate because every example and paper I have read on the topic uses Sd to calculate CSA one way or the other...
I have not experienced any problems related to high frequency dispersion but it's my personal preference and maybe it's because other shortcomings been in focus...
Last edited:
I guess that would go for a passive 1.5 XO too, less parts maybe but still a massive and expensive air core inductor, right?
Couple of options not mentioned.
You could biamp with a passive line level crossover.
TLS.org | Passive Line-Level Crossover
Several on this site have done this, much cheaper than normal passive.
You could also try a castle arrangement, nice and simple.
http://www.diyaudio.com/forums/full...ge-speaker-photo-gallery-199.html#post2932320
Spacious sound, more bass and dynamics.
You could biamp with a passive line level crossover.
TLS.org | Passive Line-Level Crossover
Several on this site have done this, much cheaper than normal passive.
You could also try a castle arrangement, nice and simple.
http://www.diyaudio.com/forums/full...ge-speaker-photo-gallery-199.html#post2932320
Spacious sound, more bass and dynamics.
I guess another alternative would be to drop the W5-2143 entirely, buy a pair of 8" W8-2145 and redesign the Minelva for them instead (I have only read good things about the W8-2145 but I haven't heard it myself). XO problem would "disappear" but an even larger Sd (220 cm2) calls for larger CSA and therefor larger enclosure, just like the lower Fs, but I wanted it to go lower, right? Seems like I could expect an F3 of around 40 Hz in a fairly compact TQWT. The limited X-max of only 3mm for a 8" feels a bit scary though...
Seems like I could expect an F3 of around 40 Hz in a fairly compact TQWT. The limited X-max of only 3mm for a 8" feels a bit scary though...
Thanks for your input ewollowe. I played with the idea of a "castle arrangement" initially (when I stumbled over Bert Doppenberg's ML-TQWT design) and I might have to go back and reevaluate it. Maybe bipolar is the way to go if I continue down the 4 * W5-2143 path, no XO right? Just let both drivers run full range. I wonder if it would work in something like the Minelva? It would at least prevent my wife from putting things on top of the speakers... Or rather, how could I prevent my wife from putting stuff on top of them?
I've only heard good things from those that have made castle style speakers. I bodged a very dodgy pair together myself with some small 3" tang bands and they didn't sound odd for that arrangement.
Just thinking outside the box but you could also try having the second driver reversed "chassis out", put even less extra treble into the room.
Just thinking outside the box but you could also try having the second driver reversed "chassis out", put even less extra treble into the room.
Anyone got a "formula" for how to calculate a "castle style" bipolar speaker?
I really like "outside the box" but "chassis out" is really tough to get pass the WAF, and divorce is not an option...
Ok, reading up on the BD-Pipes. It is claimed that the 45 degree rear facing driver compensates for the baffle diffraction step. True (could someone explain the theories behind it)? And what about a 90 degree up firing driver, would it also compensates for the baffle diffraction step? Better or worse?
Thread on full range castles.
http://www.diyaudio.com/forums/full-range/148901-microtower-bipolar-ml-tl-chr-70-el70.html
http://www.diyaudio.com/forums/full-range/148901-microtower-bipolar-ml-tl-chr-70-el70.html
I can't give you any links or formulas (not enough time to locate them) but from memory - one of the senior members of the forum (GM maybe... Planet10) had a similar comment on the issue to someone who was calculating TLs in Sds; another thing is that if you simulate different 4 inch drivers (almost all have Sd=50cm2) you get different volumes so Sd clearly has no effect on the size of the enclosure. And maybe there are some reference to the formulas in MJK pdfs... or some comments he made on the forum... unfortunately I lost about 7 month of daily research in 2012 from a hard drive malfunctioning. I just don't have the time to relocate some of the things I left in my memory as thumb rules. By the way - TL enclosure research is minimal... just go with what the simulations indicate; by the way - the alternative to Hornresp is LeonardAudio - about the same results but with much more design freedom from the mouse pointer.
Are you sure you can move sideways from those deep cones and not experience a "shadow" of high frequencies? you said you have a large room - it's inevitable to get past 30 degrees offset.
I read every paper on TL I came across and I still got my notes, every reference to CSA I got is based on Sd. Thats why I am really interested if you got something else...
I would love to try the sim from LeonardAudio but it's .NET based and I am running Mac and Linux (I can run Hornresp on Linux under Wine).
Ok, I had to get up an try it out. Our living room is fairly large, 6 * 6 meters + an open floor plan into adjacent space. The sweet spot is within 30 degrees between the speakers and I don't experience much problem within 30 degrees. Going wider on the outer side is another story. 45 degrees still ok, 90 degrees not so much...
I stand corrected, I found the following statement in Pearls from Martin J. King Quarter Wave Design - "*(Editor: Note that Martin uses sd as a convienient unit of area and is not really directly related to the cross-section of the line -- Vas is really the determining parameter)".
I would love to know more so if someone could explain the relationship between Vas and CSA please...
Found one or those posts about Sd - Vas confusion: "Sd has no impact on the size of a TL (other than that there is a VERY loose correlation between driver size & Vas -- it is unfortunate that Martin choose Sd as a unit of area in his model, it confuses things). Just like any other box it is Vas (in combination with Qt and Fs)" - from http://www.diyaudio.com/forums/subwoofers/154182-el-pipo-using-jensen-15-a.html - post no. 10 ( by Planet10).
I don't focus too much on the cross section off a TL because the length gives you the functional aspect of the enclosure (unless it's horn / expanding geometry when the variation of the cross section is also functional). The cross section is resulted from the volume of air needed to have a flat output down to tuning frequency (larger volume in TL gives you more output). The only problem would be if the cross section of the TL would turn out to be too small to mount the driver with magnet inside (a subwoofer can be used with magnet outside, a fullranger no) - but this is almost unheard of.
Re: dispersion - thank you for checking things out... if you can compare it with other fullrange drivers you have (like w52143 is better than X driver) that would be even better; i have measurements from here and there but since I'm used to one of the best dispersion fullrange cones out there (almost BMR level) I can't relate to a driver that is so different.
I don't focus too much on the cross section off a TL because the length gives you the functional aspect of the enclosure (unless it's horn / expanding geometry when the variation of the cross section is also functional). The cross section is resulted from the volume of air needed to have a flat output down to tuning frequency (larger volume in TL gives you more output). The only problem would be if the cross section of the TL would turn out to be too small to mount the driver with magnet inside (a subwoofer can be used with magnet outside, a fullranger no) - but this is almost unheard of.
Re: dispersion - thank you for checking things out... if you can compare it with other fullrange drivers you have (like w52143 is better than X driver) that would be even better; i have measurements from here and there but since I'm used to one of the best dispersion fullrange cones out there (almost BMR level) I can't relate to a driver that is so different.
You are right and I already have a subwoofer, that I would like to get rid of...
I took a stab at this and made a quick drawing. I initially tried to make it square like the microTower but did not manage to fit the two drivers the way I wanted so I fell back on something like the BD-Pipe. Simulations in Hornresp gives an F3 around 45 Hz. I wish I had access to Martin J. King's MathCad models but I don't so this have to do. Any thoughts?
Ps. I was planning to do these in 15 mm ply with some additional bracings. It might take a while though since I am under a wife constraint, no more speakers in the house unless I manage to get rid of at least on of the existing pairs, oh my and the garage is already full... And why the 70 mm port diameter? Mostly because I already got suitable pipes with that inner diameter, no other reason...
Ps. I was planning to do these in 15 mm ply with some additional bracings. It might take a while though since I am under a wife constraint, no more speakers in the house unless I manage to get rid of at least on of the existing pairs, oh my and the garage is already full... And why the 70 mm port diameter? Mostly because I already got suitable pipes with that inner diameter, no other reason...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.