• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Small Subwoofer Drivers in Tower Speaker

OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
No, but while I was digging yesterday I found this distributor in europe:
https://en.toutlehautparleur.com/ha...-pouce.html?___store=tlhp_en&___from_store=v2
https://en.toutlehautparleur.com/sp...9-inch.html?___store=tlhp_en&___from_store=v2

I also did a rough outline of a TB 3-way crossover with perceval's coaxial xover, a 500hz 1st order high pass filter on the woofer, and a digital LPF on the woofers. It's not pretty, even when the gain is adjusted per-channel.
Untitled.png
Screen Shot 2021-09-02 at 12.45.59.png
 
Last edited:

Wolf

Addicted to Fun and Learning
Joined
Nov 21, 2020
Messages
557
Likes
589
Location
Indiana
I think I'd notch out below 3kHz on the tweeter to remove too much output at below 1kHz as shown. Just an LCR across it should suffice.
 
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
The active version looks much better in every respect. It's a goal here, but not a requirement. Certainly helps get the tweeter under control. I'm not sure I understand time adjustments, but a few tweaks made a massive difference in phase and 2nd order response from the tweeter. I'll have to read up, but in the meantime, is there a way to time align without separate amplification?
Frequency response for the subs includes a 104L enclosure with dual 4x12" flared vents. Not sure I saved/uploaded it right into VCAD. I measured for all four subs in 2||2 config without crossover data, and applied the vent's frequency response to the first sub. Not sure if it's the right way to do it, but I'm done messing with this for one day.
For the Coax, a 2.2L (200x200x100mm) sealed box appears to be adequate for keeping down excursion below 2mm at 500hz (peak) @100W*3.6ohm. Also included in the FR.
Screen Shot 2021-09-02 at 21.21.40.png

Screen Shot 2021-09-02 at 21.21.51.png



The part I don't get here is how little the tweeter and woofer seem to want to play together. There are a lot of components in these passive crossovers from the diyaudio forums. The SEAS and SB Acoustics Coax drivers have really easy, inexpensive crossovers and on paper, have a clear range they're intended to work with. The fact that Tang Band's own crossover in the SUG2-25 is so different from the variety I've seen it's a little disconcerting.
 
Last edited:
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
Current box model has two flared 3x6.5" ports, 110L sub box (with a significant need for bracing design) and a 1.8L(200x200x100mm) sealed chamber for the coaxial. VCAD's estimated F3 is 25Hz, F10 is 22. With some EQ that should have some very deep, flat bass.
Screen Shot 2021-09-03 at 10.18.54.png
Screen Shot 2021-09-03 at 10.14.43.png
 

Wolf

Addicted to Fun and Learning
Joined
Nov 21, 2020
Messages
557
Likes
589
Location
Indiana
I would advise you to not go that small on the midrange, as it can make the lower mids sound 'stuffy' or 'congested'. I would recommend at least 5 ltrs for a 6.5" driver, but favor more like 10-15 ltrs. You don't want the stiffness of the air-spring to have an impact on the free motion of the driver.
Another thing to consider is the inside cavity reflection from the wall behind the driver. The closer this is the worse it can wreak havoc. I prefer to have the midrange chamber be that of a triangular format with the edge at the rear-most point so that there is no reflection.

IME, the 1139 likes about 1.25ft^3 tuned to 25Hz with 2x 2" diameter ports. It really plumbs the depths in this volume, and does not complain at higher output. Of all my uses (and I've done a lot!), I consider this the optimum cabinet for the 1139. At the same time, you can place it in a much smaller cab of 0.4-0.5ft^3 and get output into the lower 30s.
 
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
The closer this is the worse it can wreak havoc.
Wouldn’t polyfill reduce the reflections in a sealed box?
I prefer to have the midrange chamber be that of a triangular format with the edge at the rear-most point so that there is no reflection.
Basically like the B&W midrange, or the tapered cone on the HiVi dome midrange. I could maybe 3D print a cone.
I would recommend at least 5 ltrs for a 6.5" driver, but favor more like 10-15 ltrs.
The thing is, when I plugged in 50w to vituixCAD, I originally had 10 liters set aside for the midrange. When I applied the passive crossover from the midrange, the excursion dropped from 4mm@120 to zero. Even with the power turned to 100w, the excursion measured nothing. I kept lowering the box size until I hit parts express’ recommended sealed enclosure size, 1.8L. I had to use a calculated Mms T/S parameter, so I may have to do more digging and redo it. Otherwise, I’m happy to give it a larger box. I read it was the right thing to do.
IME, the 1139 likes about 1.25ft^3 tuned to 25Hz with 2x 2" diameter ports
If I combine these volumes to 5ft^3(141.5L) you’re saying I’m 1.1ft^3(32L!) short of optimal volume? I’ll try some calculations later, but would you say it’s better to enclose and vent them individually or just separate the bass/mids?
Another spitball idea I had from the Statements was open back mids, but I have a lot of relevant reading to do and some consideration to pay to how it will affect box design.
 

Wolf

Addicted to Fun and Learning
Joined
Nov 21, 2020
Messages
557
Likes
589
Location
Indiana
Wouldn’t polyfill reduce the reflections in a sealed box?

Why give it the opportunity to reflect in the first place?

Basically like the B&W midrange, or the tapered cone on the HiVi dome midrange. I could maybe 3D print a cone.

You could do that for sure. I've seen triangular, 3d-printed, pentagonal-dagger/cone, etc to reduce this possibility.

The thing is, when I plugged in 50w to vituixCAD, I originally had 10 liters set aside for the midrange. When I applied the passive crossover from the midrange, the excursion dropped from 4mm@120 to zero. Even with the power turned to 100w, the excursion measured nothing. I kept lowering the box size until I hit parts express’ recommended sealed enclosure size, 1.8L. I had to use a calculated Mms T/S parameter, so I may have to do more digging and redo it. Otherwise, I’m happy to give it a larger box. I read it was the right thing to do.

I don't know where PE gets those or how BassBox calculates what is ideal. Fact is, the volume in midrange chambers goes a long way back to the studies of someone famous (and if only I could recollect his name), to where the larger volume has less of a damping effect on the cone's motion. It really has nothing to do with the motion required at output, but the fact that motion should be unhindered if it occurs.

If I combine these volumes to 5ft^3(141.5L) you’re saying I’m 1.1ft^3(32L!) short of optimal volume? I’ll try some calculations later, but would you say it’s better to enclose and vent them individually or just separate the bass/mids?
Another spitball idea I had from the Statements was open back mids, but I have a lot of relevant reading to do and some consideration to pay to how it will affect box design.

Never combine different model drivers in the same volume or chamber as one will inflict upon the other. It will be easier to vent once for 4 woofers in one volume, or vent twice for 2 woofers in 2 separate volumes. Since you are looking at WWMWW, this could keep things more symmetrical and easier to build in the 2 volume case. Anywhere from 0.75-1.1 is likely best, and keeps drivers under control. I would recommend at least a 2" diameter port per driver, but doubling that will improve vent mach speeds.

Best regards,
 
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
Thanks for all the replies. It's been really helpful.
I'm getting wildly different responses from different speaker box programs, but for now I'm settling on four 3"x20"(76x500mm). It's very similar to running eight 2"x2" ports. I may decide to scale it down to WMW, but I like the shape, honestly.
Final x-over and cab for the weekend:
Screen Shot 2021-09-03 at 19.18.51.pngScreen Shot 2021-09-03 at 19.19.03.pngScreen Shot 2021-09-03 at 19.19.28.png
Screen Shot 2021-09-03 at 19.23.41.png
 
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
Does anyone here have experience with the W6-1139SIF, the ferrite magnet version of the 6.5" sub? I heard the magnet size is bad for venting, but changing over lowers the final build price by like, $350. I'm nowhere near full excursion except at alarming wattages, like 350 or so, maybe a few mm less won't be so bad.
 

Wolf

Addicted to Fun and Learning
Joined
Nov 21, 2020
Messages
557
Likes
589
Location
Indiana
They are very similar, and I wouldn't worry about any differences coming to light.
I thought you were using the SIF, honestly. Swap and be happy.
 
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
Been doing some thinking since I last saw this project. The depth is insane. There's absolutely no space behind for the ports to breathe while the drivers sit on the same plane as my current bookshelves. 140L is way too much. I think I'll have to cut back the number of bass drivers to reduce the enclosure size.
If I take some liberties with the height by going WCW instead of CWW, I can squeeze 74L (2.6ft^3) out of a 1300x475x200mm box with a pair of 76x400mm (3"x15.5") ports. This also brings down the cost significantly, but I'm not sure how I feel about running the 1139SIF's in parallel yet. I really like the idea of running each tower with a MiniDSP board and a pair of ICEPower 50ASX. In my head, it seems perfect. Each speaker gets 50W, each is controlled by the miniDSP board individually. If I parallel the two subs, I can't put the full 170W through them, as it falls below the rated 3 ohms in BTL configuration. Sad, but not the end. Individually powered may be better, and there are plenty of sub amp alternatives.

Some new mockups:
Screen Shot 2021-09-11 at 10.36.34.png
Tang Band W6-1139SIF Six-pack.png
Tang Band 3-way Coaxial Tower XO-schema-5.png
Tang Band 3-way Coaxial Tower Six-pack.png
Untitled.png


EDIT: This also brings the BOM down to $2000 including lumber and some fudge money, from approximately $2400 earlier, and $2600 from when I was running the neodymium subs. Way more doable, still a "good" kind of excessive. The max SPL is down a bit, but in stereo I think it'll be fine.
 
Last edited:

Biblob

Addicted to Fun and Learning
Joined
Sep 13, 2018
Messages
633
Likes
600
It's nice to see you fine-tune between what's reasonable and what's possible. I recognize the same thing, when I did my diy project.

I would advice you add some cross bracing between the sidewalls, because of the depth of the speaker. For some inspiration, look at the Kef R5/7/9/11 crosscut's.
 
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
It's nice to see you fine-tune between what's reasonable and what's possible. I recognize the same thing, when I did my diy project.

I would advice you add some cross bracing between the sidewalls, because of the depth of the speaker. For some inspiration, look at the Kef R5/7/9/11 crosscut's.
Thanks! The R11 and now R7 are my inspirations for this design. I plan to model better cross bracing, but 3D modeling is hard enough for me right now. Gotta really put in the hours on this next model.

The number one detractor for this whole design is still the SEAS coaxial and woofers hanging over my head. A Hypex FA123, SEAS reed coax and woofers are well known and have proven designs behind them. It also costs approximately the same to go this route, either with two or four woofers. The sub depth may not be there, but the box will be smaller by a huge margin.
 

Wolf

Addicted to Fun and Learning
Joined
Nov 21, 2020
Messages
557
Likes
589
Location
Indiana
Would 2.5 cubes per channel then be doable now that you are not going with a quad?
 
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
Nah, not really. When I measured my living room there wasn't sufficient depth. I want the depth to be less than two feet, and without significantly increasing the baffle width, there isn't a lot of volume wiggle room. I already have my HiVi's on stands about 1.5ft from the wall, which isn't ideal. Nothing can really keep coming forward from there. It's one of the reasons I was so excited about cutting the midrange down to 2L, a 9L enclosure takes approximately 12L of internal volume, vs the 2 inside to 4 outside. Saving depth is important when I don't own a house.
 
OP
D

D!sco

Senior Member
Joined
Jun 16, 2021
Messages
483
Likes
381
Put a bit more work into the passive version today. I can maybe pull off a full blown 3-way passive if I balance these out decently. The tweeter is still unruly in the midrange, but switching the old perceval crossover up a bit yielded some decent results. A 3rd order is more effective at quelling the 1Khz bump, but that problem is why I wanted to try an active solution in the first place. The crossover is expensive, running around $120 for the pair. It wasn't a huge compromise adjusting some components to be duplicates, knocking a few bucks off the final price on the way. The most important change was definitely going series on the subs. That gives a maximum load of 125W with the crossover. It's so tempting to just go for POWER above all, but a nominal 7.2 ohms is so much easier to pair with amplifiers. I was struggling to justify a sub-2ohm impedance in my bass without some crazy expensive setup. I could still go FA123/UMIK on this build, and it would probably be my best choice being a newer designer. Going passive and getting decent amps would save about $200-500 in this build, though. It's nice to have the option.

Not sure why the DI (in red, under Power & DI) isn't showing up as anything but a flat line, but I don't trust the listening window either. It's funny, on the SEAS tower I have floating in the pile, it's reversed. DI is just like any proper measurement on this site, listening window is garbage. I don't get it.
Tang Band 3-way Coaxial Tower Passive XO-schema-6.png
Tang Band 3-way Coaxial Tower Six-pack.png
 
Last edited:

bigjacko

Addicted to Fun and Learning
Joined
Sep 18, 2019
Messages
721
Likes
359
From the griup delay graph the mid range and tweeter have 180 degree phase difference, is this your intended design? The power and DI graph looks alright though the phase difference is 180 degree.
 
Top Bottom