"Folded" Sonotube TL Sub Complete!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I just put the finishing touches on my sub project and wanted to share and get some feedback.

It's a "folded" TL design, using a 12" sonotube nested concentrically (well, eccentrially, I suppose) inside a 16" sonotube. There's a 4 1/4" gap at the bottom of the smaller tube, which makes the "bend".

I had initially wanted to use 14" and 20" tubes, since the CSA's of both sections are nearly identical this way, but found that these sizes are uncommon and hard to find (and really big). I was skeptical of having the area of the first and second sections of the line differ, but after modeling the line in MJK's Tl Sections sheet, I found that the response actually looks quite good, with a broad flat fundamental peak from around 22 Hz to 65 Hz.

The design is intended for the Ascendant Audio Atlas 12, which I don't have yet, so sitting in for it now is my old Phoenix Gold car sub, which has a higher Q but nearly the same Fs and Vas specs.

So far, I'm quite stunned with the performance, especially considering the low cost. Bass is clean, powerful, and smooth, and extension in my relatively small listening room seems limitless!

The whole thing stands about 75" high; line length is 147".
 

Attachments

  • sono_tl_full.jpg
    sono_tl_full.jpg
    9 KB · Views: 2,264
Here's a shot looking down the line from the top. The first half of the line is stuffed with about .5 lb / ft^3 of Dacron fill. I cut and wedged some 5/16" hardwood dowels sideways across the tube to keep the density even and prevent the fluff from all settling to the bottom.
 

Attachments

  • sono_tl_top.jpg
    sono_tl_top.jpg
    23.6 KB · Views: 2,303
This is my favorite part of the whole thing... the inner tube is just bolted to the outer one with some big old binding posts from Parts Express. The Atlas has two voice coils... one for driving and the other for adjusting the Q via resistive damping, so I put two posts at the top (for the RDO coil) and two at the bottom for the driven coil.

The finishing touch was some nice neoprene-backed sealing washers from Home Depot. Airtight, and solid to boot!
 

Attachments

  • sono_tl_posts.jpg
    sono_tl_posts.jpg
    15.8 KB · Views: 1,876
Tweeker said:
$#@%! you beat me to it! Very nice sub. Im building a concentric sonotube tl too, but have a few too many other to dos to do on my to do list. Im having trouble finding the tubes I want too.

4 out 5 five people polled cant say the name of your city right:D
Thanks!

What sizes of tubes are you considering? I'd recommend the heavy-walled versions if you can get them, as mine is a bit more flexy than I'd like... no rattles or offensive resonances, you can just feel quite a bit by putting your hand on it.

And people around here, against all logic, say "pew-AL-lup". That's where I work; I actually live in Olympia, which is much easier to pronounce. :)
 
My grandfather lives in Shelton, fairly near Olympia. Was thinking about something crazy big before (20, 28sleeved30, 36) but it'd be an insane pain:dead: . Looking at 16,22,28 now. Plan to run bracers along the tube lengths tieing the tubes together for much more rigidity. Driver is mounted ~1/3 way along the line at the junction. The cone/motor forms part of the first turn. Driver is Avalanche 18.

______
| |....| |
|||....|||
|||....|||
|||....|||
|| /^\ ||


You might be able to make your sub much stiffer by bisecting or trisecting the outer tube with a couple boards and glueing/nailing/poswting/screwing them on each side to the tube face.
 
Tweeker said:
You might be able to make your sub much stiffer by bisecting or trisecting the outer tube with a couple boards and glueing/nailing/poswting/screwing them on each side to the tube face.
Meh, it doesn't really bother me that much. I don't think the vibration is caused by pressure in the line, it's more the mechanical motion of the driver making the inner tube want to oscillate around the point where it's joined to the outer wall at the top.

I may get inventive and try to knock a wedge-shaped block under the edge of the small tube in the bend, bracing it against the floor. Have to find a way to do this on the end of a 7-foot pole, though! :)
Driver is Avalanche 18.
An Avalanche 18?!? Woah. :bigeyes:

That's got a 16 Hz Fs, right? How long is your line going to be? My math says 17.5' or so...
 
Originally posted by Tweeker
You might be able to make your sub much stiffer by bisecting or trisecting the outer tube with a couple boards and glueing/nailing/poswting/screwing them on each side to the tube face.
Sorry to keep replying to my own thread, I'm just thinking out loud...

You know, adding a board as a stiffening brace is a pretty good idea. If I just do one at the widest point of the crescent, maybe a foot long, I could wedge it in there and attach it with screws without having to take anything apart. And narrowing the line at the end a tiny bit with the brace can only lower tuning and smooth out the response a smidge, both of which are good things.

Thanks for the great idea! :)
 
I'M IN!!!!!

I'd like to build one as well. Can you guys help me sift through the endless list of subs?

I'd like a freq response between 15 and 20 on the low side and extends to 40 or 50 on the high side. Okay, maybe 60Hz on the high sides. Of course staying relatively flat.

I have a 9 foot ceiling so, I was thinking two tubes, one inside the other, to create an 18 foot overall transmission line. Now, I could use an 18 and come close to hitting the ceiling (may not be a bad idea....I could use the ceiling as a brace and wedge the whole thing in like a pillar. Or I could play it safe and use a 15 inch sub and shorten as well as narrow my tubes a bit.

The monstrosity will sit back in an odd shaped corner of my 23 x21 foot living room. I have up to 30 inches diameter of footprint to play with here, so size of tube is irrelevant, only the height has limitations.

Again, I'm not looking for 10 Hz, just something between 15 and 20Hz.

Thanks

Chris
 
I think the 12" in mine is quite sufficient for low-end extension IMHO. My listening room is longer than yours (it is open to the dining room), but with corner loading and room gain low-end extension seems to have no limit.

A 15" or 18" would work fine, though, as long as the line's not too tall for your room.

I would be careful about getting the end of the line too close to the ceiling, especially in a corner, lest the confined air up there "load" the driver and line.
 
Well Clarkcr, the AV15 is currently one of the best deals available.
I don't know much about TLs, but there's alot of people with alot of knowledge on these.

I guess your 18 feet long TL would be really really nice.

Adire Audio will restock soon, not much choice beside that.
If you want cheaper alternatives, nice drivers on the PartsExpress website, like the Daytons...
 
SteveCallas said:
Excuse my ignorance, but by going with an 18' transmission line, isn't group delay going to shoot up pretty high?

Greets!

Some info I saved from an early Basslist thread. TLs 'shine' where it counts:

GM

-----------

Group Delay and Time Alignment

For a driver in a specific box type:
By definition, group delay is the slope of the phase response (the negative
derivative of phase slope). It describes the relative delay of the spectral
components of a waveform.

Group Delay = -[(phase@f2 – phase@f1)/(f2 – f1)]

A flat group delay curve indicates that all frequencies are arriving
simultaneously at the test point, while a peak in the group delay shows that
some frequencies arriving late. Typically higher frequencies will arrive
before lower ones, but not always, it depends on the design of the
driver/box, and more importantly the crossover. A typical group delay curve
will have a number of small (less than 1ms) dips and peaks.

Since the Qs of the system comes from the shape and absolute level of the
enclosure group delay curve,
a separate characterization has to be done for each type of enclosure.

For a closed box: Qtc can range 0.5 - 0.707 - 1.2
Qtc = 0.5 critically damped - transient perfect.
Qtc = 0.577 Bessel response (D2), max flat delay
Qtc = 0.707 Butterworth response (B2), max flat amplitude.
Qtc > 0.707 Chebychev (elliptical - C2), equal ripple response
max power, somewhat degraded transients.

As Qtc increases from 0.5 the group delay becomes greater and transient
response is degraded. However, independent sources have verified that any
group delay lower than 2ms is inaudible, and above 2ms group delay can
typically only be head in critical listening with headphones. In such an
experiment the group delay was introduced via electronic manipulation and
not inherit in the headphones themselves (fear not headphone buffs).


From "Perception of Phase Distortion in All-Pass Filters", by J.A. Deer,
P.J. Bloom and D. Preis as printed in the JAES, Oct, 1985:

"Results from listening tests indicate that a statistically significant
perceptual threshold is reached when peak group delay distortion at 2 kHz is
in the neighborhood of 2 ms (for diotic presentation via earphones)."


Loudspeaker guru Ken Kantor had this say about group delay:

“There have been many careful studies of the audibility of group delay in
loudspeakers and filters. Blauert, Lipschitz,
and the "Tufts" studies, to name a few. You can, for example, start with
the best, most "coherent" loudspeaker you can find, and degrade its time
performance using group delay filters.

What has been found time and time again is that "time alignment" effects
NOTHING in the quality of the reproduction until delays of well over 2 mS
are introduced, 10 times this in the bass.

I know of no published research to the contrary, other than snake oil and
floobydust.

Degradation of impulsive or square waveforms due to group delay is audibly
meaningless. Degradation of impulse response due to resonances or early
reflections, or phase differences between L&R are quite audible, on the
other hand.

To claim otherwise is both hard to support with published data, and is
contradictory to current models of hearing neurophysiology.”


Representative Group Delay by Enclosure Type:

Frequency Sealed Vented TL
10 Hz. 7 ms. 12 ms. 10 ms.
20 8 15 10
30 8.5 23 8
40 8.5 12 8
50 6 7 3
100 2 2 0


For a system:
The driver group delay is meaningless without consideration of the crossover
group delay. The 2nd. and 4th. order crossovers have much larger associated
group delays than that of the driver/box type. However, careful crossover
design can reduce the system group delay to well below 2ms over most of the
spectrum. Most commercial loudspeakers’ group delay curves exceed the 2ms
range in the low hundreds of hertz range. For examples just pick up a copy
of Audio or Stereophile and look at the graphs of any number of speakers.

Representative Crossover Group Delay Characteristics

frequency 1st. 2nd. 3rd. 4th.
40 Hz. 32 us 230 us 320 us 450 us
80 32 us 230 us 320 us 450 us
100 32 us 230 us 320 us 450 us
200 32 us 230 us 320 us 450 us
400 30 us 250 us 500 us 500 us
800 28 us 230 us 700 us 550 us
1000 16 us 220 us 680 us 450 us
2000 8 us 60 us 120 us 120 us
4000 ~1 us ~20 us ~20 us ~20 us

Notes:
1st order filter data for a Butterworth type.
2nd order filter data for a Linkwitz-Riley type.
3rd order filter data for a Butterworth type.

By E. Jakulis with additional comments by Jonathan Atkins
 
Well, I see the Tumult is now available, or at least up on their website. Which of the two 15 inch subs would be best for a TL line? The 15D2 or the 15D4?

With a little figuring, the area taken away by the 14 inch sonotube inside a 20 inch sonotube is roughly (6.28 sq inches off) the same. I could go to a 26 inch sonotube on the outside, an 18 inch sub on the inside mounted into an 18 inch sonotube, which leaves 12.98 sq inches of extra room (on the cross section) for internal bracing (3/4 inch wooded ribs.)

Can anyone calculate what my resonant frequency would be? Can TLs work this way? Here's a drawing. So, if it can be wedged into the ceiling (the top is closed off) I would think the bracing would be superior.

Am I way way off here??? Someone stop the ranting.

Here is a drawing of what I'm talking about.

C
 

Attachments

  • scan1.jpg
    scan1.jpg
    4.4 KB · Views: 817
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.