The Continuum....Subwoofer

[Jeff B.]

Somthing I have been planning for quite some time was to build a dedicated subwoofer designed to follow the philosophy of the Continuum monitors.

So, my idea was to create a small sealed subwoofer that was still fairly high in performance and would be able to offer the owner of Continuums (me, in this case) or anyone who would benefit from a small sub, a small full-range low distortion system that was very easy to incorporate into a room, espcially if you have a free corner somewhere.

I ended up using a TC Sounds Epic12 in a Dayton 1 cubic foot sealed box. Although this woofer is shoehorned into this 14" cube (Carver style) the measured Qtc was .775. I used the maple veneered sub cabinet and to my delight it was an excellent match to the solid maple enclosures of my Continuunms.

In my case I used an Oaudio 500W plate amp, because I already had one and it is an excellent choice for a sealed sub with the combination of parametric EQ and selectable highpass filter. However, there are quite a few plate amps that would work well with this design if someone wanted to use them instead. (I do really like the Oaudio though.)

Pure modeling of the driver in this small box at first indicates a fairly high F3. However, once you introduce a dialed in lowpass filter and the small amount of EQ from the 20Hz setting on the amp it becomes fairly flat with an F3 around 28Hz, and an F6 of 23Hz. You give up some sensitivity to do this, but that's what the 500 Watts are for. Since subs almost never play into a half-space environment in the last two octaves anyway, it was important to incorporate some room modeling into the design. It is easy to dial this little sub in so that it is essentially flat to 18Hz in a typical room and capable of 112 dB. I really don't need more than this for anything I do. I could be adjusted to work great with home theater too, but I use it almost exclusively for music. I run it from my sub-out jack using a combination of the receiver's 80Hz crossover and the adjustable lowpass on the plate amp. This gives me a fairly decent integration with the Continuums at an 80Hz crossover point.

The TC Epic12 really is a nice value driver for its performance. I was orignally going to use the TC Sounds LMS-R12, but realized that due to the stiff box compliance even the Epic wouldn't be pushed to anywhere near its limits in this sub and the low end extension was remarkably similar for both drivers. As a result the two LMS-R12's are still sitting on the shelf and this sub is being introduced with the Epic at half the price. One final comment on these drivers, the TC woofers work wonderfully in small packages due to their low compliance Vas parameters. As a result they allow you to build a smaller sub than you can with most any other line and you still get high Xmax and an advanced motor design too.

Here's a pic of the Continuum System as I took it to the MWAF:



Here's a pic of the sub itself:


and here's one for Thilo:



If anyone would like to duplicate the sub let me know. The details are pretty simple but there are a few tips for using it in this box. And it goes lower and has more output than you would expect for such a small subwoofer.

Jeff B.

[AR3A]

Nice job, setup looks awesome !

[arlis_1957@yahoo.com]

nice. they did sound very good together.

[craigk]

like i said before, thought they were the best sound of the show. amazed standing in the back of the room the sound stage thrown, the largest of any speaker there. when 2 ways are designed right there is just something about them that is magical. very good base also. like goldie L and the three bears, just right. great design. was reading on salk audio website and curious, would you let me know what other 5.25s you looked it before deciding on this design ? thanks

[fdieck]

That's it... park the trophy on top of it and rub it in. You know I am kidding, great job as always.

[Pete Basel]

I did a very similar design about 2 years ago.
500W, XBL 12", considered the O-Audio and Epic-12.
Mine needs 1000W to get the most out of the driver and
I'd expect the same with the Epic. I also use the subs in
pairs.

Going to disagree again, high Vas is better with the rest of the
driver parameters being correct in closed box woofers. Still the
Epic does fine in small boxes.

[kmibb]

I've got a 15" cube box from the M&K liquidation and a BASH 500W plate amp. May have to pick up a TC Epic and build this to match my Continuums.

[ryanbouma]

112db at 18hz?!? Wow. Three of these for HT would be a lot of fun. Any idea what would happen in a larger box considering the effeciency at 18hz has gotta be way down with a q over 0.7.

Only using 1 while 2 other good subs sit unused? Have any room measurements for how 1 handles modes?

[Jeff B.]

112db at 18hz?!? Wow. Three of these for HT would be a lot of fun. Any idea what would happen in a larger box considering the effeciency at 18hz has gotta be way down with a q over 0.7.

Only using 1 while 2 other good subs sit unused? Have any room measurements for how 1 handles modes?

Let me clarify those numbers, so I don't mislead. The 18hz is an in-room F3 based on room modeling. The 112dB is a one meter midband (roughly 30-80Hz region) at 500 Watts in the room as well. Both however, are very realistic values.

When I was modeling with larger box volumes I would have gone with a passive radiator, probably an VMP-12, since I have several here. Moving up to about 2.5 cu ft and using the PR increases output by about 6dB at 20Hz, going sealed with the same settings but with a 2.5 cu ft box increases output by approximately 3dB at 20hz.

I decided to target the smaller box as a design choice and also to try to target a very flat in-room response (modes not include here, just boundary reinforcement). Since these are primarily for music I didn't want any deep bass over-influence.

You can model all of this using my design tools. My measurments appear to validate those simulations to be pretty accurate.

[Jeff B.]

I did a very similar design about 2 years ago.
500W, XBL 12", considered the O-Audio and Epic-12.
Mine needs 1000W to get the most out of the driver and
I'd expect the same with the Epic. I also use the subs in
pairs.

Going to disagree again, high Vas is better with the rest of the
driver parameters being correct in closed box woofers. Still the
Epic does fine in small boxes.

All other things being equal a higher compliance woofer would generally be a more efficient driver, but this comes at the cost of requiring a larger box to achieve the same F3. It all depends on your design goals. The Iron Law is at work here.

For me, the smaller box was a design goal, so a lower Vas driver was ideal. When designing for sealed boxes the full lumped parameter of Qtc is the best indicator of what you have. In my case a Qtc in the .7 range was close to ideal in the box.

A 1000 Watt amp would give me a 3dB increase in overall output and the Epic would still have well-controlled excursions due to the box compliance, but I chose the Oaudio for the unique flexibility it offered. Some may prefer the 1000 Watt amp, and I'm fine with that.

Jeff

[Pete Basel]

I understand your choice with the O-Audio amp, I liked it very much myself. I just
wanted to point out that the driver has a lot more to give if anyone needs more peak
output. I would be concerned about abuse at 1000W - it has to be used with caution.

I'll give it one more shot, then perhaps when I have some time offer some simulations
or some sort of proof. If all we did was loosen the Epic 12 suspension say by 2-4X,
you could obtain exactly the same response that you have now in a smaller box.
The suspension and box compliance are effectively in parallel and all that matters is
the total compliance, make one less stiff the other can be more stiff. Everything else
including efficiency remains the same. Your point about Qtc does not apply to the case
I just described because in system Qtc does not change if the in system total
compliance remains the same.

[Jeff B.]

I understand your choice with the O-Audio amp, I liked it very much myself. I just
wanted to point out that the driver has a lot more to give if anyone needs more peak
output. I would be concerned about abuse at 1000W - it has to be used with caution.

I'll give it one more shot, then perhaps when I have some time offer some simulations
or some sort of proof. If all we did was loosen the Epic 12 suspension say by 2-4X,
you could obtain exactly the same response that you have now in a smaller box.
The suspension and box compliance are effectively in parallel and all that matters is
the total compliance, make one less stiff the other can be more stiff. Everything else
including efficiency remains the same. Your point about Qtc does not apply to the case
I just described because in system Qtc does not change if the in system total
compliance remains the same.

I will await your simulation then. If I follow what you are saying - you cannot increase Vas and then get the same Qtc and response by making Vab smaller. High compliance (large Vas) drivers require larger boxes.

[Pete Basel]

No, you are not following me if this is what you believe that I said:
"you cannot increase Vas and then get the same Qtc and response by making Vab smaller."

I said the opposite, as long as the total compliance remains the same, by increasing
Vas and decreasing Vb, all other final system parameters stay the same. Have you
ever read Villchur's articles? His profound insight was to make Vas high and Vb small,
it might not seem profound today but back then everyone was building systems with
huge boxes - or small boxes with no bass. It is before my time, but I have read the
history. My project advisor, when I studied audio and loudspeaker design, liked to
refer to Villchur's original design as "crippled" without a box. Fs was very low below
18 Hz, Vas very high, and Qts very low, it needed the small box to bring Fc, and Qtc
back up to an optimal range.

Let me offer this as food for thought:
In a theoretically ideal closed box system, Vas and Qms would equal infinity.
This would not be a physically realizable system but understanding it offers
food for thought.

Based on your response, I'm not sure I could convince you through simulation.
It is there in Villchur's, Small's and most others' papers.

[killa]

Would a old acoustic suspension type speaker be similat to what you are talking about? I once took a old marantz 12in. woofer out of a speaker and stuck it in a fairly small sealed box in my truck. It dug suprisingly deep. People at work thought I had a real sub That driver had very loose suspension with a cloth surround.

[Jeff B.]

No, you are not following me if this is what you believe that I said:
"you cannot increase Vas and then get the same Qtc and response by making Vab smaller."

I said the opposite, as long as the total compliance remains the same, by increasing
Vas and decreasing Vb, all other final system parameters stay the same. Have you
ever read Villchur's articles? His profound insight was to make Vas high and Vb small,
it might not seem profound today but back then everyone was building systems with
huge boxes - or small boxes with no bass. It is before my time, but I have read the
history. My project advisor, when I studied audio and loudspeaker design, liked to
refer to Villchur's original design as "crippled" without a box. Fs was very low below
18 Hz, Vas very high, and Qts very low, it needed the small box to bring Fc, and Qtc
back up to an optimal range.

Let me offer this as food for thought:
In a theoretically ideal closed box system, Vas and Qms would equal infinity.
This would not be a physically realizable system but understanding it offers
food for thought.

Based on your response, I'm not sure I could convince you through simulation.
It is there in Villchur's, Small's and most others' papers.

OK, I follow what you are saying now. You are talking about the old "Acoustic Suspension" design where the box air spring is necessary to control the cone. But, as your already described that driver typically had a small Qts and a low Fs to accompany its high compliance. In other words, they have a very loose suspension.

There are still a few woofers made like this today, but not many. Driver design philosophy has changed quite a bit over the years, driven mostly by the availability of higher and higher powered amplifiers. The bottom line is that our mathematical understanding of box alignments has evolved some since that time, and people like us have access to modeling software that most engineers in Villchur's day did not.

I agree with what you are saying now, but it still works within the same mathematical framework as any other driver and the math is pretty clear regarding the relationships between Vas, Qts, and Vab. More drivers today have gone the direction of smaller Vas and higher Qts, rather than Villchur's approach, but in the right balance either woofer could work well in the same box. However, it much harder to find true acoustic suspension drivers these days and there isn't a real advantage it's just a matter of parameter trade-offs.

But... I see what you are saying now and I agree with it.

[Pete Basel]

OK, good. I agree the fundamentals and math behind the problem have not changed.
Acoustic suspension is just a marketing term, all closed box systems follow the same
theory. The only reason to go for a lower Vas in a driver intended for closed box
designs is to avoid cone sag, and it is a good reason with high mass cones. I believe
that the main reason we do not see a lot of very high Vas drivers is that there is a
point of diminishing returns in closed box systems and the lower Vas allows the driver
to also be used in more reasonably sized vented systems - most drivers are dual
purpose today.
I have always questioned how they chose which parameters (the standard minimal
set) to publish for T&S parameters. Cone mass is a much better choice when dealing
with closed box systems but instead it has to be calculated from Fs and Vas.

[Jeff B.]

OK, good. I agree the fundamentals and math behind the problem have not changed.
Acoustic suspension is just a marketing term, all closed box systems follow the same
theory. The only reason to go for a lower Vas in a driver intended for closed box
designs is to avoid cone sag, and it is a good reason with high mass cones. I believe
that the main reason we do not see a lot of very high Vas drivers is that there is a
point of diminishing returns in closed box systems and the lower Vas allows the driver
to also be used in more reasonably sized vented systems - most drivers are dual
purpose today.
I have always questioned how they chose which parameters (the standard minimal
set) to publish for T&S parameters. Cone mass is a much better choice when dealing
with closed box systems but instead it has to be calculated from Fs and Vas.

I would say the data chosen for publishing are those parameters most necessary for modeling in most software. Hey, it's come a long way. Remember the old catalogs where all you got was "200 Watt 12" woofer with 40 oz magnet. Frequency response: 35-3500 Hz"?

It would be remiss to not point out how Qts controls box size. As you pointed out the Villchur woofer had a very low Qts, most woofers today do not have Q's this low. As a result, designing to avoid peaking in the response turns into a box size constraint.

[Pallas]

I did a very similar design about 2 years ago.
500W, XBL 12", considered the O-Audio and Epic-12.
Mine needs 1000W to get the most out of the driver and
I'd expect the same with the Epic. I also use the subs in
pairs.

Two things you're not considering.

First, the extra power isn't going to give much real extra output. It's going to be wasted as compression. The Epic (nee TC1000, nee TC2+) has a fairly small voicecoil diameter, and nothing extravagant for thermal management. The same is likely true of your XBL^2 12 - Shiva-X? - incidentally. If anything, more so because the XBL^2 voicecoil is shorter, and thus has less thermal mass.

Second, TC drivers (by design) have basically no safety margin. The motor is strong enough to push the voicecoil fairly linearly until it collides hard with something metal, like a backplate. (That's one reason they often claim - and legitimately have - more linear throw than other drivers using the same baskets and/or soft parts.) So prudent use of such drivers requires limiting, either by an outboard processor or by an amp that's not quite enough to smash the coil into a wall. Now, here it could be that the line is something above 1kW. I didn't model it. I'm just pointing out a general concern that many ignore.

There is, of course, one problem with the amp Jeff used, though: it's out-of-stock, and I don't know if Todd's going to order another run of them. I hope he does, because it's all the subwoofer amp most people need.

[Pete Basel]

You might notice my many comments about thermal compression on the forum.
The additional power is for headroom, not to increase the average power so increased
compression should not be an issue. Overexcursion is always a potential problem however in
a very small box such as these more power is usually required. I did state in another post that
the power should be used with caution.

What is the basis of your statement "TC drivers (by design) have basically no safety margin"?
Have you tested them ... overdriven them? Have you discussed it with Thilo?

[dfg]

I just ordered the Continuum kit this week from Meniscus, and have been looking for a small sub to go with it. I have limited space. I'd definitely be interested in the details on this one.

Greg