AK Design Collaborative - TCB Subwoofer
- Thread starter Zilch
- Start date
A few observed "rules of thumb" that came out of my experience with these, in the Audio Engineering class I took at Georgia Tech under Dr. Wm. Marshall Leach, where designing one of these (on paper) was our class project (I wrote a little Q-BASIC program to calculate the parameters, since I got really tired of doing it on paper- I think I might have that program on a disk somewhere...), and out of building several of these in various configurations using that data (and other modeling programs later):
- For 6th-order bandpass (vented both sides), you want low-to-medium Qts (.15 to .35 or so), low-to-medium Fs (25-55 Hz) and whatever Vas you want, for the efficiency (and box volume) you want.
- These woofers do not have to have extremely high X-max... anything over 5-6mm is a reasonable candidate, for home sub use. Naturally, for us trying to bend people's spines in a PA environment, we go for massive X-max (and gigantic ports, as a consequence of the air-movement needed)... but, a good solid 1/4" (6.5mm) of X-max will give nice SPL levels at home, with a decent size woofer (10" or bigger).
-Having enough port area is pretty critical. While the SHAPE of the response at LOW volume won't change with a too-small port, the tuning DRASTICALLY starts to shift (usually, you start losing LF extension) when you start to over-run the ports.
-While I haven't gotten around to it yet, the idea of a PR-loaded 6th-order bandpass has ALWAYS been intriguing. The idea of NOT wasting lots of internal volume with vent diameter and length, and the easy, quick tune-ability of PRs are a big potential plus. It's just a matter of properly CALCULATING them. My take on this- I just model a vented box with the proper vent frequency, and look at where the impedance peaks are. Then, I model the same box with a PR, and try to make the dip between the peaks exactly match up, at the same frequency (which SHOULD be Fb, or very close). This should at least get you into the ballpark...
BTW: My take on the Tang-Band small woofers (6.5" sub and 5.25" sub) is that they'd probably be easier to tune in a 4th-order bandpass (one side sealed, one side ported) rather than a 6th order (both sides ported). This is due to the relatively-high Qts. However, it may be possible, with some significant tweaking of volumes and port frequencies on both chambers, to get them to behave well.
BTW^2": Most unlikely-seeming candidate for 6th-order bandpass- how about the ElectroVoice EVM12L?! That was Dr. Leach's class project example, that we used as a model! It would, in fact, give F3 of 30 and 120 Hz (or close thereabout) with 1ft^3 on the front chamber, and 2ft^3 on the rear. IIRC, tuning frequencies were around 60-80 Hz on the front chamber and 20-25 Hz on the rear...
Regards,
Gordon.
- For 6th-order bandpass (vented both sides), you want low-to-medium Qts (.15 to .35 or so), low-to-medium Fs (25-55 Hz) and whatever Vas you want, for the efficiency (and box volume) you want.
- These woofers do not have to have extremely high X-max... anything over 5-6mm is a reasonable candidate, for home sub use. Naturally, for us trying to bend people's spines in a PA environment, we go for massive X-max (and gigantic ports, as a consequence of the air-movement needed)... but, a good solid 1/4" (6.5mm) of X-max will give nice SPL levels at home, with a decent size woofer (10" or bigger).
-Having enough port area is pretty critical. While the SHAPE of the response at LOW volume won't change with a too-small port, the tuning DRASTICALLY starts to shift (usually, you start losing LF extension) when you start to over-run the ports.
-While I haven't gotten around to it yet, the idea of a PR-loaded 6th-order bandpass has ALWAYS been intriguing. The idea of NOT wasting lots of internal volume with vent diameter and length, and the easy, quick tune-ability of PRs are a big potential plus. It's just a matter of properly CALCULATING them. My take on this- I just model a vented box with the proper vent frequency, and look at where the impedance peaks are. Then, I model the same box with a PR, and try to make the dip between the peaks exactly match up, at the same frequency (which SHOULD be Fb, or very close). This should at least get you into the ballpark...
BTW: My take on the Tang-Band small woofers (6.5" sub and 5.25" sub) is that they'd probably be easier to tune in a 4th-order bandpass (one side sealed, one side ported) rather than a 6th order (both sides ported). This is due to the relatively-high Qts. However, it may be possible, with some significant tweaking of volumes and port frequencies on both chambers, to get them to behave well.
BTW^2": Most unlikely-seeming candidate for 6th-order bandpass- how about the ElectroVoice EVM12L?! That was Dr. Leach's class project example, that we used as a model! It would, in fact, give F3 of 30 and 120 Hz (or close thereabout) with 1ft^3 on the front chamber, and 2ft^3 on the rear. IIRC, tuning frequencies were around 60-80 Hz on the front chamber and 20-25 Hz on the rear...
Regards,
Gordon.
dnewma04
The Healer
I have found that the easiest way to get reasonably close on tuning a PR is to treat it like a port of the same effective diameter, determine the required length, calculate the mass of the air within the determined port volume and then mass load the PR. It should get you in the ball park, then you can swap in or take out a few grams here to tune.
bakes
Super Member
I built a fourth order bandpass sub using two Carbonneau 12" drivers, using Margerand's "Symmetrically Loaded" Speaker Builder article for the design parameters. The box is huge - 6 feet by 18" by 23" or something like that. The -3 dB corner points were (IIRC, design is at home somewhere) 17 Hz and 70 Hz within a Hz or two either way. It vents into the 28' long family room via a 6" or 8" Sonotube port. I have no way of measuring distortion levels, etc. but Analog Addict can confirm the the house-shaking (literally, see below) abilities of this beast. It makes electronic music and pipe organ music a truly memorable experience.
Why house shaking? I built it into the wall between the family room and workshop in the basement.
Don't do that.
The whole house shakes.
Worse WAF: Great big box in the middle of a room, or house rumbling when she's trying to get some sleep? Hmmm... :scratch2:
Why house shaking? I built it into the wall between the family room and workshop in the basement.
Don't do that.
The whole house shakes.
Worse WAF: Great big box in the middle of a room, or house rumbling when she's trying to get some sleep? Hmmm... :scratch2:
dnewma04
The Healer
Bakes, are the woofers opposing each other in the design? For instance, one down firing, one up firing into the same chamber? This usually does wonders for reducing vibrations that would be increased by directly coupling an enclosure to a house.
My friend and I installed 4 15" woofers in an IB in his ceiling. The 4 drivers faced one another and all of their motion opposed one another. The bass SPL would vibrate the house, but there was little structural noise.
My friend and I installed 4 15" woofers in an IB in his ceiling. The 4 drivers faced one another and all of their motion opposed one another. The bass SPL would vibrate the house, but there was little structural noise.
dnewma04
The Healer
I have a pair of the tangband woofers that have been brought up. I'd be willing to play around with some alignments. I'd be partial to a 4th order bandpass with a PR for a port. I suppose we might get better efficiency with the 6th order, however.
I did some quick modeling and don't mind the looks of a 8.7l front chamber tuned to 60hz and a 38.1L rear chamber tuned to 25hz. Gives a bandwidth of 25-87hz. We can even probably get away with a gentle tailing off of the low freqs and count on room gain to get us even deeper extension.
We could do something fun like use a port for the front chamber where it would require shorter port lengths and a PR for the rear chamber. Then build it so the rear chamber is in front and run the front chamber port through the rear and have it terminate under the PR. It would give it the look of a standard ported enclosure.
Now to spec out the required PR.
I did some quick modeling and don't mind the looks of a 8.7l front chamber tuned to 60hz and a 38.1L rear chamber tuned to 25hz. Gives a bandwidth of 25-87hz. We can even probably get away with a gentle tailing off of the low freqs and count on room gain to get us even deeper extension.
We could do something fun like use a port for the front chamber where it would require shorter port lengths and a PR for the rear chamber. Then build it so the rear chamber is in front and run the front chamber port through the rear and have it terminate under the PR. It would give it the look of a standard ported enclosure.
Now to spec out the required PR.
Gordon, Ray, and I all have it. You tell us what you want to use, and we'll do the appropriate button clicks. Three-chamber is conceptually a pair, belly to belly.
Gordon's done some sub designs with slot ports for us on LHF before, 123A-1 and LE14A, plus isobarik. Once we close in on something, maybe he'll take a shot at it here. :yes:
http://audioheritage.org/vbulletin/showthread.php?t=3922
[Nearly four years ago! Why do I remember this stuff and how to find it?
]
Russellc
Addicted Member
VOTT, maybe?
J/K....
Yeah they could, but I assume that would have to be huge to keep up. I meant the other two smaller ones next to the A7s in the picture. Added by edit later.
Maybe the 4647 w/2235H could keep up with A7s?russellc
Guaranteed.... :thmbsp:
dnewma04
The Healer
with my preliminary design, PR mass is going to be a big issue. To tune to 25 hz in that small of an enclosure requires 1500g of mass on a PR with an effective SD of 250cm^2. Back to the tuning phase. 
If one were to find a PR suitable for this app, you'd have a 25.5x21.75x8.5 enclosure, 105dB at rated power from 25-90hz.
If I bump up the rear cabinet volume just a hair, i can make a pair of ports work adequately. One 3" port 13.1" long, one 3" port 16.5" long. The dimensions would increase to 26.75x21.75x8.5 with just about the same bandwidth.
If one were to find a PR suitable for this app, you'd have a 25.5x21.75x8.5 enclosure, 105dB at rated power from 25-90hz.
If I bump up the rear cabinet volume just a hair, i can make a pair of ports work adequately. One 3" port 13.1" long, one 3" port 16.5" long. The dimensions would increase to 26.75x21.75x8.5 with just about the same bandwidth.
Russellc
Addicted Member
I'll give that a go then...my area JBL center had to order 2235H kits....they only had one, and I wanted to make dam sure they matched in color!
russellc
Something's up with your images, Dave. The one at #34 is showing blank on my screen.
We're talking TB W6-1139SI here, right?
I had to enter the T/S parameters into BB6P manually. It did not like spec Mms = 34.91g and recalced it to 41.45, with which value it found all of the other published parameters acceptable.
In the BB6P TCB suggested alignment (36.46 & 72 Hz,) it works in a neat 12 x 12 x 24" package. It doesn't play quite as low as the Dayton (F6 = 35.58 Hz,) but it can certainly take the power. At full 140W, however, it's only delivering 101.8 dB at F6.
I confirm Ray's finding of inordinately long duct lengths, but observe that the vent velocities are quite low, <6 M/s. BB6P must apply different criteria to bandpass vents than it does to ports in conventional alignments.
We're talking TB W6-1139SI here, right?
I had to enter the T/S parameters into BB6P manually. It did not like spec Mms = 34.91g and recalced it to 41.45, with which value it found all of the other published parameters acceptable.
In the BB6P TCB suggested alignment (36.46 & 72 Hz,) it works in a neat 12 x 12 x 24" package. It doesn't play quite as low as the Dayton (F6 = 35.58 Hz,) but it can certainly take the power. At full 140W, however, it's only delivering 101.8 dB at F6.
I confirm Ray's finding of inordinately long duct lengths, but observe that the vent velocities are quite low, <6 M/s. BB6P must apply different criteria to bandpass vents than it does to ports in conventional alignments.
Here it is back at you. Properties say it's at Photobucket, 58.898 K. Two grey bars and a box, right, or is the image here?
Upload to the AK server, maybe....
Attachments
Duffinator
Spin The Black Circle
Zilch, are you going to have a mock up by next Saturday?
I'm disappointed this thread didn't hit 100 posts in the first day.
I'm disappointed this thread didn't hit 100 posts in the first day.