Sonotube TL for 21Hz

[Jeff20B]

I want to do a Sonotube TL with two woofers in a push-pull arrangement where one of the woofers is hooked out of phase and installed backwards in the enclosure. Basically, it'll be sort of like this. http://www.t-linespeakers.org/projects/steve/index.html

I have 8 foot ceilings in my listening area so I suppose I could probably get away with a stuffed 6 footer, but what I'd really like to do is build a TL long enough to take advantage of the 21Hz free air resonance of my 12" drivers.

Woofer specs:

Dual voice coil 8ohm/4ohm
FS 21Hz
QTS .38
QMS 4.2/4.4
QES 0.44/0.42
VAS 13.3 ft³ (376.6L)
SD 0.0532 m²
XMAX 0.460 (11.7mm)
(let me know if I forgot any important ones)

A little basic math tells me...

1128 ÷ 21 Hz = 53.71 feet full wave length
53.71 x 0.25 = 13.428 feet ¼ wave length

13.428 feet = 161.14 inches TL length

Assuming my math is correct, this 13+ foot length presents a problem in my listening area.

I came up with an idea that includes just one 90° bend. The top will only reach as high as about 7'2" leaving some breathing room between it and the ceiling. The rest will be on the floor run along a wall. I have easy access to some 14" x 14" x 14" (1 ft³ internal volume) 3/4" MDF boxes meant for 12" woofers and can probably find some 6' long x 12" diameter Sonotube nearby. I bet I could trim the tubes a little (if necessary) and cut a couple more holes in the boxes and have this thing together pretty quickly.

What do you guys think of this idea? Is there any chance of comb filter effects or any other undesirable operation possible from this design? Should I add something like a 45° reflector to the box in the middle? Thanks.

[GM]

Greets!

A single 12" pipe doesn't have enough Vb to perform well with one of these drivers, much less two. The math is mostly correct, an end correction of (0.613*pipe radius) needs to be subtracted from its length and the SoS will probably be somewhat higher, so worst case is it will be a little too long.

Yes, there will be comb filtering. All things considered, best IMO to build an optimum one fold cab out of 19mm no-void plywood with both the drivers and vent at floor level. Use this little calculator to find optimum length/CSA (don't forget to double Vas): http://www.diyaudio.com/forums/attac...amp=1117991086

BTW, for my records, what make/model drivers are these?

GM

[Jeff20B]

A few more questions for you guys. Is it totally necessary to build a TL exactly to the ¼ wave length of my driver FS? Or can I fudge it a little and make it shorter so it's tuned closer to 30Hz or something? What if I stuff it full of poly-fil?

Or how about two 6' tubes with one 1ft³ box and one woofer for each tube? This will make a stereo bass pair of TLs and each TL will be 7' long (about 1 foot below my ceiling).

[Jeff20B]

Hey, thanks for the post. It seems I'm still being moderated since I'm a new member, so that second post I made was actually from a few hours before yours.

Ah, so Sonotube of 12" diameter is too narrow for my driver(s)? Interesting. Yeah, I'll post the full specs.

Radio Shack 40-1350 Realistic Subwoofer (vintage 1989). It was the first dual voice coil woofer I'd ever seen.

19 ounce ceramic magnet
frequency response (-6 dB) 26Hz to 2.7kHz
8 ohm/4 ohm
flux denisty 7.8Kg
Sensitivity 89dB/95dB
FS 21Hz
RE 6.9 ohm/3.5 ohm
QTS .38
QMS 4.2/4.4
QES 0.44/0.42
VAS 13.3 ft³ (376.6L)
SD 0.0532 m²
XMAX 0.460 (11.7mm)
Vd 37.9 in²
Le 2.24 mH/0.56 mH (@ 1 kHz)
No 0.78%/0.79%
Pe 120/60 Watts

I'm not following your math. I barely understood enough about transmission lines to get this far. Should I go with a 14" or 16" tube for better Vb? I think in order to keep this from getting too complicated (as this is my first time), what if I went ahead and kept it simple to start with? How about something like the reconstructed El Pipe-O with only one woofer each? http://www.passdiy.com/projects/el-pipe-o-9.htm

I don't think it would be appropriate to go with 6' length because it would require a bunch of stuffing which would lead to poorer performance. If only I had a listening room with higher ceilings. Or lower floors.

Could you explain more about Vb when increasing tube diameter to say 14"? Also could you explain what SoS is? I hadn't seen anything about needing an end correction before (I'm sure I've still got a lot to learn). I couldn't view the xls file. Lastly, could you explain what comb filtering is with regard to audio? I had only heard of it in televisions.

Hey, I just had a brainstorm. How far apart can the driver and the open end be before comb filtering takes place? Better yet, how close together can the driver and open end be before problems occour? Can they be side by side? How about downward firing? I've seen ported home theater woofers where the port and the driver are both down-firing. What about a down-firing TL where both the open end and driver are facing the floor? Will 'floor gain' occour? Is it even desirable? I've got a carpeted floor. If this works, it'll save tons of time and effort (and money) since I could just cut the tubes at 45° angles and liquid nails them back together. It'd just be one fold and vented at floor level, just like you suggested. They'd get a couple bolts through both adjacent edges for greater stability and a couple legs to raise it an acceptible height from the floor (whatever that may be). They'd even probably be portable, to some extent.

I could fit two of these in the listening room without too much trouble. If necessary, the driver/open end could even be rotated up so as to aim at the ceiling if that would provide better SQ.

I whipped this up in like two minutes so pardon the sloppyness.

[GM]

Greets!

(your posts rearranged to suit me)

====

>Hey, thanks for the post. It seems I'm still being moderated since I'm a new member, so that second post I made was actually from a few hours before yours.

>Radio Shack 40-1350 Realistic Subwoofer (vintage 1989). It was the first dual voice coil woofer I'd ever seen.

====


You're welcome! A 'golden oldie! I thought those specs looked familiar! I designed/built quite a few different designs using them, though most were a tower design (now AKA ML-TL).

====

>A few more questions for you guys. Is it totally necessary to build a TL exactly to the ¼ wave length of my driver FS? Or can I fudge it a little and make it shorter so it's tuned closer to 30Hz or something? What if I stuff it full of poly-fil?

====

No, and in some cases a good trade-off between size/performance is to tune it to as high as ~(SQRT(2)*Fs), or ~29-30 Hz with this driver. This driver's Fs is low enough to use the higher tuning since room gain will boost it down low and the higher tuning will increase its midbass efficiency, a win-win alignment.

=====

>I don't think it would be appropriate to go with 6' length because it would require a bunch of stuffing which would lead to poorer performance. If only I had a listening room with higher ceilings. Or lower floors.

====

Agreed, at this length an ML-TL is preferred.

====

>Or how about two 6' tubes with one 1ft³ box and one woofer for each tube? This will make a stereo bass pair of TLs and each TL will be 7' long (about 1 foot below my ceiling).

>Ah, so Sonotube of 12" diameter is too narrow for my driver(s)? Interesting. Yeah, I'll post the full specs.

>Should I go with a 14" or 16" tube for better Vb? I think in order to keep this from getting too complicated (as this is my first time), what if I went ahead and kept it simple to start with? How about something like the reconstructed El Pipe-O with only one woofer each? http://www.passdiy.com/projects/el-pipe-o-9.htm

====

You're going to use a high stuffing density to lower the pipe's strong harmonic modes, but contrary to what's been published until recently, the SoS doesn't slow down much with increasing density. Really, the 'classic' TL rule-of-thumb (ROT) to make the TL's cross sectional area (CSA) ~= Sd rolls off/damps most drivers so much that once the pipe is lined, the ~completely damped Fs peak implies a much longer effective pipe length, ergo the false impression of a lower SoS.

For instance, your 12" sonotube design will overdamp the driver to below its Qts. Fine if it's a too high Q driver or you want a ~flat impedance for best power transfer when driven by an impedance matching amp, but the former doesn't apply in this case, and being a wide BW sub driver, I doubt the latter does either, and why I don't consider it a good design for you.

Of course if this is a prime performance goal, then by all means use one tube for both drivers if max acoustic damping is desired or one tube/driver if a moderate amount, but if a ~IB or slight gain response is desired, a much larger CSA will be required. IOW, the pipe's Vb needs to be at least = the driver's Vas, and in many cases 1.5-4x Vas for max performance, so to KISS it (keep it simple stupid) with good performance, design/build it as I already recommended.

====

>I'm not following your math. I barely understood enough about transmission lines to get this far.

>Could you explain more about Vb when increasing tube diameter to say 14"? Also could you explain what SoS is?

=====

CSA for a tube = (dia.^2*pi)/4

Vb = volume of box, or pipe in this case, so assuming dims in inches: (CSA*length)/1728 = ft^3

SoS = speed of sound. It books along at ~1130 ft/sec on average, but varies with the local pressure/temperature. Altitude per se has a negligible impact on it, so not a factor in speaker design.

=====

>I hadn't seen anything about needing an end correction before (I'm sure I've still got a lot to learn). I couldn't view the xls file. Lastly, could you explain what comb filtering is with regard to audio? I had only heard of it in televisions.

====

You haven't been reading the right docs. Do a search on pipe harmonics, it should bring up some
that elaborate on the finer details of pipe action/design.

Why not? Even if you have a Mac, won't Open Office or some emulator allow you to view/use an Excel spreadsheet?

Comb filtering is where the same frequencies from multiple sources, but offset in time (space), add/subtract to create a jagged summed response.

====

>Hey, I just had a brainstorm. How far apart can the driver and the open end be before comb filtering takes place? Better yet, how close together can the driver and open end be before problems occour?

====

There's always going to be comb filtering, but it can be minimized by placing the multiple sources as close together as practical.

====

>Can they be side by side? How about downward firing? I've seen ported home theater woofers where the port and the driver are both down-firing. What about a down-firing TL where both the open end and driver are facing the floor? Will 'floor gain' occour?

====

This driver is way too compliant for bottom firing. The calc'd sag is 5.63 mm, almost its Xmax! For max boundary gain, terminate ~at the floor in a perpendicular plane. Of course the TL's terminus can be floor loaded either way though.

====

>Is it even desirable? I've got a carpeted floor.

====

IMO it is. The way I look at it, there's no such thing as too much acoustic efficiency since it lowers the amp power required/distortion. IOW I'd rather EQ it down than boost it up to make it ~flat in-room. I recommend using spikes to anchor it to the floor rather than just sitting on the carpet.

====

>I could fit two of these in the listening room without too much trouble. If necessary, the driver/open end could even be rotated up so as to aim at the ceiling if that would provide better SQ.

====

While there's some advantages to using 'stereo' subs, there's something to be said for using dual drivers in a bipolar push-pull layout to cancel out their rocking motions/any non-linear excursion distortion WRT build simplicity/SQ, so I prefer the latter since these have enough excursion to make them an issue.

Since ceilings tend to be acoustically 'weaker' than a floor unless only the exterior roof is on the other side, I recommend floor loading.

GM

[Bill Fitzpatrick]

Scrap woofers. Get 4 real ones. Build sealed enclosures. Bi-amp. Ta-da!

[Jeff20B]

Thanks for taking some time to answer my questions. I appreciate the effort.

You're welcome! A 'golden oldie! I thought those specs looked familiar! I designed/built quite a few different designs using them, though most were a tower design (now AKA ML-TL).

So they can indeed be used in a TL. That's great to know because I wasn't sure if anyone had ever attempted a TL with these oldie-but-goodies.

No, and in some cases a good trade-off between size/performance is to tune it to as high as ~(SQRT(2)*Fs), or ~29-30 Hz with this driver. This driver's Fs is low enough to use the higher tuning since room gain will boost it down low and the higher tuning will increase its midbass efficiency, a win-win alignment.

This is extremely interesting. Hmm, speaking of midbass, I currently have too much midbass from an Obcon enclosure with two Sony tens (vintage 1992) in it (driver specs available upon request) that has a Qtc of around 1 or so with aproximately 1.269 ft³ volume. According to an internet sealed encloser calculator, my drivers want a 1.9 ft³ enclosure. I added a pillow and that helped quite a bit. It has a few dips in frequency response as observed with sine wave sweeps, but it falls flat on its face below 35Hz. The typical bass boost from 70 to about 120 is still there, as is common in most consumer bass products, so I'd like to avoid a repeat if possible on the TL. Incidentally, my bro has a KLH floor sub for home theater and its published frequency response is 40-180Hz. It has a single front firing 10" and a rear firing port. It has very similar performance to the Obcon box. Surprisingly so (it can get louder and has a smoother curve, but falls flat right around 35Hz as well). Heh, that KLH is supposed to be a subwoofer? Dubious. I supose it's ok for non critical listening, as I suspect most consumer bass products are, but it is a bit too slow, muddy, and port noise can be heard etc. This is the sort of thing I want to avoid and why I chose to start learning about transmission lines. I want the effortless, clean bass everyone is always talking about. I prefer SQ over SPL and will do whatever it takes, within my limited budget, to get it.

Agreed, at this length an ML-TL is preferred.

What is the ML in ML-TL? Mass loaded? Does mass refer to the stuffing density? Or just that it has stuffing in it as opposed to an empty one?

You're going to use a high stuffing density to lower the pipe's strong harmonic modes, but contrary to what's been published until recently, the SoS doesn't slow down much with increasing density. Really, the 'classic' TL rule-of-thumb (ROT) to make the TL's cross sectional area (CSA) ~= Sd rolls off/damps most drivers so much that once the pipe is lined, the ~completely damped Fs peak implies a much longer effective pipe length, ergo the false impression of a lower SoS.

Ah, I think I understand. So the driver is damped to the point where it will start to behave as if it were in a larger encloser, even though it isn't? That's sort of what I had imagined happening instead of actually slowing the speed of sound down. That's a relief. Heh, wouldn't slowing down the speed of sound lower octaves and generally make things sound too slow? Kinda like the opposite of trying to speak in a less dense atmosphere like talking after breathing in helium from a balloon?

For instance, your 12" sonotube design will overdamp the driver to below its Qts. Fine if it's a too high Q driver or you want a ~flat impedance for best power transfer when driven by an impedance matching amp, but the former doesn't apply in this case, and being a wide BW sub driver, I doubt the latter does either, and why I don't consider it a good design for you.

I think I see what you're saying. As far as I can tell, my driver's only Q that is too high is the QES at 0.44/0.42. I'm supposed to aim for QES of 0.3 to 0.4, right? In this case, wouldn't a tube of 14" diameter be better than 12"? If QES is electrical and QMS is mechanical, then QTS is the total, right? Anyway, the QTS of 0.38 is on the high side, but probably not enough to turn me away from using tubes instead of wood.

Of course if this is a prime performance goal, then by all means use one tube for both drivers if max acoustic damping is desired or one tube/driver if a moderate amount, but if a ~IB or slight gain response is desired, a much larger CSA will be required. IOW, the pipe's Vb needs to be at least = the driver's Vas, and in many cases 1.5-4x Vas for max performance, so to KISS it (keep it simple stupid) with good performance, design/build it as I already recommended.

Heh, at this point, I'm so confused I don't even know what I want anymore. 12" diameter, 14" diameter, one driver, two drivers...

I think I'd like to keep stereo bass which would require two TLs with one driver each. I have some material that has stereo bass, so this is important to me. Then again, SQ and accurate low frequencies are also important to me. Can I have both?

What is IB? If I was going to tackle woodworking, non-void plywood sounds good. I bet it'd be safer (less dusty) than working with MDF. I'd honestly like to avoid woodworking if I could, and the sonotube idea would allow it. Hence my stubborness to abandon it.

More to come.

[Jeff20B]

CSA for a tube = (dia.^2*pi)/4

Eep! Sorry, I don't understand that formula. There's pi, which I can recognise, but I don't understand the ^ or * or /. Are they just easier ways of typing x and ÷?

Vb = volume of box, or pipe in this case, so assuming dims in inches: (CSA*length)/1728 = ft^3

Once I learn how to figure CAS, I know I can do the rest of the calculation. Then I can factor in the end correction and arrive at a good set of parameters (I hope).

SoS = speed of sound. It books along at ~1130 ft/sec on average, but varies with the local pressure/temperature. Altitude per se has a negligible impact on it, so not a factor in speaker design.

Thanks for the clarification. I suck at math and don't use abbreviations or acronyms too often so you'll have to forgive me every now and then.

You haven't been reading the right docs. Do a search on pipe harmonics, it should bring up some that elaborate on the finer details of pipe action/design.

Why not? Even if you have a Mac, won't Open Office or some emulator allow you to view/use an Excel spreadsheet?

Comb filtering is where the same frequencies from multiple sources, but offset in time (space), add/subtract to create a jagged summed response.

I don't have MS office. Maybe I can try that xls file on a computer that does.

That sounds like multi-path in radio wave propigation.

There's always going to be comb filtering, but it can be minimized by placing the multiple sources as close together as practical.

That's good to know.

This driver is way too compliant for bottom firing. The calc'd sag is 5.63 mm, almost its Xmax! For max boundary gain, terminate ~at the floor in a perpendicular plane. Of course the TL's terminus can be floor loaded either way though.

The suspension is definitely rather lightly sprung on these old woofers. I've seen many much stiffer modern drivers. I thought XMAX was measured for one direction? Maybe the way it's measured has changed since the original specs were printed? I've never measured the excursion of them myself. If the published spec of 11.7mm is accurate, it seems similar to modern drivers though.

So what if I just layed the single driver folded tube TL (as pictured above) on its side so both the terminus and driver are perpendicular to the floor? I could still fit two in my listening room, I think.

IMO it is. The way I look at it, there's no such thing as too much acoustic efficiency since it lowers the amp power required/distortion. IOW I'd rather EQ it down than boost it up to make it ~flat in-room. I recommend using spikes to anchor it to the floor rather than just sitting on the carpet.

Ah yes, I've seen spikes in the picture of the bottom-firing HT sub.

I'd prefer as flat a response down past the driver's Fs as humanly possible. Something like a drop-off at around 10.5Hz would make me happy. I don't know how much room gain I'll have, so I think I should design for an anechoic chamber and then EQ it for any room gain (if any). I'd feel better doing it that way. However from what I've read, it seems to be a difficult goal to reach without 12 (dancing) Shivas or something, but again I probably haven't been reading the right docs.

I'd like it to be musical and would rather have SQ over SPL any day, but does this, by definition, require a shorter TL? If the single driver folded TL is a decent idea, a total length of around 12' folded into two equal sections would fit in my listening room.

While there's some advantages to using 'stereo' subs, there's something to be said for using dual drivers in a bipolar push-pull layout to cancel out their rocking motions/any non-linear excursion distortion WRT build simplicity/SQ, so I prefer the latter since these have enough excursion to make them an issue.

So you're saying that since these drivers have a lot of excursion, it's best to limit it in some way such as through a push-pull arrangement (one hooked out of phase) like this guy did? http://www.teresaudio.com/haven/subs/subs.html

I think I'd prefer stereo subs since these have a wide bandwidth and I was thinking about using it up to about 200Hz or so. Maybe even higher. And now my newbieness is starting to show.

I've read through some of the debates for stereo subs vs mono, and think I'd prefer stereo although I understand that mono has its advantages.

Since ceilings tend to be acoustically 'weaker' than a floor unless only the exterior roof is on the other side, I recommend floor loading.

Cool.

[Jeff20B]

How about something like this?

[Jeff20B]

Side view.