DIY Infinite Baffle subwoofer - advice needed

[Boomzilla]

Being too lazy to construct a box, I'd like to play with making an infinite-baffle, dipolar subwoofer. From what I've read, I think I want the driver in close proximity to the floor to take advantage of the boundary effect. I'd also like to have sufficient cone area to play moderately loudly (95 dB peaks @ 20 Hz.) without reaching the excursion limit of the drivers. I'm also willing to experiment with isobaric loadings where two drivers are working in tandem. I can use DSP equalization to extend the lowest frequencies of the design and to flatten the response, but if I do, the drivers will have to have some power handling capability.

Unfortunately, most of the infinite baffle designs I've come across on the internet are for car audio, and this is to be a home audio application (music first - HT only occasionally). I'm not sure how parameters would change (if at all) when considering the driver to use.

For lots of watts at low frequencies, I think that a "pro" Class-D amp is probably the least expensive option, although I do have conventional Class-AB amps already in house.

I know from previous postings that some of you have used infinite baffle subs before. Suggestions?

Thanks kindly - Boomzilla (aka Glenn Young)

[millst]

You're not making an IB if you aren't sealing the rear driver's radiation off from the front. The area behind the driver doesn't have to be infinite, but should be rather large to avoid coloration. An attic is commonly used.

Ideally, you would use a driver that has good parameters for the use. There are purpose built ones out there. The idea is to use many of them, keeping distortion low by spreading the load. Without any box limiting the driver excursion, it doesn't take much power so any amp you have is probably good enough to start.

I read quite a bit on Cult of the Infinitely Baffled back when I built mine years ago. Believe this is it:

ibsubwoofers.proboards.com/

[DYohn]

1. Don't waste time with isobaric loading. The result is building a smaller enclosure, not more output.
2. Read the Cult posted above. Excellent resource.
3. IB does not really care where you place it in the room. However if you think it means less construction than building an enclosure, you would likely be mistaken unless you purchase something off the shelf, like a Sonance in-wall sub. Plus you will likely need a LF EQ.
4. My favorite IB subs are from Acoustic Elegance. I have four of their IB 15's in my HT ceiling now. aespeakers.com/ I have built successful IB installations using other drivers - look for a woofer with a large Xmax and a Qts around 0.7

[AudioHTIT]

I’m interested in what you learn here, I’ve always envisioned adding a third subwoofer for LFE, and placing it in the ceiling somewhat in a VOG position (though probably more forward). After adding height speakers, I know what I’m in for cutting a hole up there, but feel I’d need to strengthen the ‘baffle’, and would reinforce the sheetrock with plywood or particle board between the 24” centered TJ ceiling joists. Even then my main worry would be exciting resonance in the ceiling, or rattle from the can lights. That said it would be a fairly ‘infinite’ baffle, and with the right driver might sound pretty good.

[millst]

Mine is kind of like that. 4 Fi IB drivers that are a bit front and right of the first row. Ideally, I would have centered it, but wasn't going to happen given the roof line and that I wanted to have the drivers opposing each other to cancel out vibration. I don't have any issues with my can lights and they still use incandescent bulbs.

I'd agree that you better be ready to put some work in when building. However, it will be mostly hidden and doesn't have to be pretty, so that helps.

[DYohn]

Nov 8, 2021 19:28:35 GMT -5 millst said:

Mine is kind of like that. 4 Fi IB drivers that are a bit front and right of the first row. Ideally, I would have centered it, but wasn't going to happen given the roof line and that I wanted to have the drivers opposing each other to cancel out vibration. I don't have any issues with my can lights and they still use incandescent bulbs.

I'd agree that you better be ready to put some work in when building. However, it will be mostly hidden and doesn't have to be pretty, so that helps.

Fi makes excellent woofers.

[Boomzilla]

True IB is probably not in the cards, near term. I expect to sell this house in the next 15 months, and structural modifications won't appeal to potential buyers. The idea I had was more like what this guy did:

[DYohn]

Open baffle is very different from infinite baffle. Here's a few resources to peruse:

www.diysubwoofers.org/dipole/
www.quarter-wave.com/OBs/U_and_H_Frames.pdf
www.linkwitzlab.com/models.htm

[Boomzilla]

Nov 9, 2021 10:00:35 GMT -5 DYohn said:

Open baffle is very different from infinite baffle. Here's a few resources to peruse:

www.diysubwoofers.org/dipole/
www.quarter-wave.com/OBs/U_and_H_Frames.pdf
www.linkwitzlab.com/models.htm

WHOA... Not such a simple thing as one might think. That back-wave cancellation is absolutely brutal at lower frequencies. The infinite baffle design just throws away the back wave, lowering sensitivity, but allowing for really low frequency response.

Since open baffle won't give what I want, and infinite baffle is not practical until after I move, maybe a feasible option would be to purchase a commercial sealed box, glue it together myself, and cut an extra hole for a second driver (effectively doubling the volume of the box without physically increasing its size)?

[DYohn]

Nov 9, 2021 16:55:32 GMT -5 Boomzilla said:

Nov 9, 2021 10:00:35 GMT -5 DYohn said:

Open baffle is very different from infinite baffle. Here's a few resources to peruse:

www.diysubwoofers.org/dipole/
www.quarter-wave.com/OBs/U_and_H_Frames.pdf
www.linkwitzlab.com/models.htm

WHOA... Not such a simple thing as one might think. That back-wave cancellation is absolutely brutal at lower frequencies. The infinite baffle design just throws away the back wave, lowering sensitivity, but allowing for really low frequency response.

Since open baffle won't give what I want, and infinite baffle is not practical until after I move, maybe a feasible option would be to purchase a commercial sealed box, glue it together myself, and cut an extra hole for a second driver (effectively doubling the volume of the box without physically increasing its size)?

Yea, OB is a difficult nut to crack. I use two DIY OB subs in H-frames and the dipole frequency is about 45 Hz. They are pretty large. Then the IB in the ceiling covers the bottom octiave plus.

As to your idea, you can't just randomly "cut a hole in the box" and add a second driver. A system is designed with the proper size enclosure, port, etc. for the driver. Adding or subtracting something will change performance in often very bad ways.

If you want DIY and don't want to design yourself, buy a kit from Parts Express. Here is a good one that is in stock currently:

www.parts-express.com/Dayton-Audio-15-Reference-Series-HO-Subwoofer-and-Cabinet-P-300-7093

[Boomzilla]

Well, the price is certainly right - I can't buy a decent commercial sub for twice that.

And since I'm overdue for a rant... The HT craze has ruined loudspeakers. Yes, you can get great-sounding speakers for ever lower prices, but (and it's a BIG but), they ALL require subwoofers! Even tower speakers in the $4,000 to $6,000 range have -3dB points of 40 Hz. or thereabouts. One would think that at that price, the manufacturers would at least attempt to offer "full-range speakers." And that's ever and always been defined as a speaker with 20 to 20,000 Hz. response. Now years ago, only a few speakers actually succeeded in offering that low a bass, but it was at least the widespread stated goal. Speakers then were evaluated on how closely they could come to "full-range" response. But any more, nobody even tries. The manufacturers realize that bass costs money, that the majority of their customers prefer a huge boom instead of accurate and extended bass (much less with low distortion), and so they assume that all customers will have a subwoofer already. Freed from having to spend a penny on decent bass response, the manufacturers just up their profits by selling cheap 40-Hz. boxes. "Want some bass? We'll be happy to sell you a that bottom octave - Buy our matching sub for another $4-6 thousand!"

Bah, Humbug.

[foggy1956]

Nov 9, 2021 19:34:10 GMT -5 Boomzilla said:

Well, the price is certainly right - I can't buy a decent commercial sub for twice that.

And since I'm overdue for a rant... The HT craze has ruined loudspeakers. Yes, you can get great-sounding speakers for ever lower prices, but (and it's a BIG but), they ALL require subwoofers! Even tower speakers in the $4,000 to $6,000 range have -3dB points of 40 Hz. or thereabouts. One would think that at that price, the manufacturers would at least attempt to offer "full-range speakers." And that's ever and always been defined as a speaker with 20 to 20,000 Hz. response. Now years ago, only a few speakers actually succeeded in offering that low a bass, but it was at least the widespread stated goal. Speakers then were evaluated on how closely they could come to "full-range" response. But any more, nobody even tries. The manufacturers realize that bass costs money, that the majority of their customers prefer a huge boom instead of accurate and extended bass (much less with low distortion), and so they assume that all customers will have a subwoofer already. Freed from having to spend a penny on decent bass response, the manufacturers just up their profits by selling cheap 40-Hz. boxes. "Want some bass? We'll be happy to sell you a that bottom octave - Buy our matching sub for another $4-6 thousand!"

Bah, Humbug.

I don't use a sub with my Tektoñ DIs

[KeithL]

It's worth noting that you cannot simply "throw away the back wave"...

With any dipole radiator, which describes ALL moving cone drivers, the back wave is equal and out of phase to the front wave.

Therefore, at the point where the baffle is not sufficiently large, the back wave "just goes around the sides and cancels with the front wave".

In order to have a true infinite baffle design the baffle must be infinitely large - with an infinitely large space behind it.

(You can get pretty close to this by mounting a driver in a wall - with a large room behind it.)

In either of those cases, there is no connection between front and back, so the back wave has no opportunity to cancel the front wave.

Since the wavelength of a 20 Hz wave is 50 feet, in order to have an open baffle design able to reproduce 20 Hz well, the baffle would have to extend 25 feet in all directions from the driver.

And, in order to do so with a "pseudo infinite baffle", the room behind the speaker would have to be large enough to not have any significant resonance or other effects at that frequency.

I've heard of this being done reasonably well with drivers mounted in a wall whose back side was open to a garage or attic.

(So the room is both relatively large, somewhat apsorbtive, and unsealed...)

You also end up with a less effective approximation when you mount a small subwoofer in a wall panel.)

And, with an open baffle design, once you reach the frequency where the rear wave can cancel the front wave, for a given size baffle, the output drops off at 12 dB/octave below that.

And, in order to correct that, you must apply an equivalent amount of EQ, to boost the bass output to compensate for the roll off.

This means that you need a lot of amplifier power.

It also means that you need large drivers, capable of a large excursion, to move the required amount of air.

With an acoustic suspension speaker, the excursion required from the driver drops off as the same rate as the acoustic output, because the two are directly related.

With an open baffle speaker, the opposite is true.

At lower frequencies, the drivers must have MORE excursion, because they're required to move more air, even though the air they're moving is contributing less to making acoustic output.

And, since there is no damping provided by any cabinet, the driver must also provide it's own damping.

Because of this only certain drivers work well in that application.

Incidentally, if you start with a box, with a driver in it...

And then cut another hole for a second identical driver...

Driving them with identical signals...

You will be HALVING the effective volume (it's exactly the same as if each driver was mounted in a separate box - each HALF the size of the original).

So efficiency will go down, box resonance will go up, and the frequency where your woofer starts to roll off will he higher.

And, if you're using a ported design, all the calculations there will change as well.

(Basically you need to treat the sum of the areas and masses of your two drivers as if they were a single larger driver - in the same box.)

Nov 9, 2021 16:55:32 GMT -5 Boomzilla said:

Nov 9, 2021 10:00:35 GMT -5 DYohn said:

Open baffle is very different from infinite baffle. Here's a few resources to peruse:

www.diysubwoofers.org/dipole/
www.quarter-wave.com/OBs/U_and_H_Frames.pdf
www.linkwitzlab.com/models.htm

WHOA... Not such a simple thing as one might think. That back-wave cancellation is absolutely brutal at lower frequencies. The infinite baffle design just throws away the back wave, lowering sensitivity, but allowing for really low frequency response.

Since open baffle won't give what I want, and infinite baffle is not practical until after I move, maybe a feasible option would be to purchase a commercial sealed box, glue it together myself, and cut an extra hole for a second driver (effectively doubling the volume of the box without physically increasing its size)?

[Boomzilla]

Nov 10, 2021 15:04:20 GMT -5 KeithL said:

...

Incidentally, if you start with a box, with a driver in it...

And then cut another hole for a second identical driver...

Driving them with identical signals...

You will be HALVING the effective volume (it's exactly the same as if each driver was mounted in a separate box - each HALF the size of the original).

So efficiency will go down, box resonance will go up, and the frequency where your woofer starts to roll off will he higher.

And, if you're using a ported design, all the calculations there will change as well.

(Basically you need to treat the sum of the areas and masses of your two drivers as if they were a single larger driver - in the same box.)

OK, correct me if I'm wrong, but your statements above assume that the two drivers are facing each other (physically out of phase, but not electrically). If the drivers are working in tandem (both moving together) then the opposite should happen, yes?

In the following graphic, the top two drivers, wired in phase will (as you say) see half the volume of the box.

But in the bottom graphic, wired in phase, they will (ignoring the space that the right driver occupies in the box) see TWICE the actual volume. Yes?

[Boomzilla]

The more I think about it, the less enthused I am about building anything. One year from now, the question I'll face is "is this item worth moving across the country, or should I sell it / donate it and just deal with the loss?" Anything I spent money on, I can live without or replace. But things I spent TIME on will be harder to part with.

Emotiva has some nice looking subs for not much $$$.

[KeithL]

You're sort of right... but not the way you think... and I don't think that's what you want...

Let's assume that you set something up EXACTLY as in the second illustration.

The drivers are on opposite sides of a reasonably sized cabinet - and wired electrically in phase.

But one driver is facing in while the other is facing out.

Yes, whenever one driver is moving in, the other will be moving out.

And, because of this, each driver will see, NOT a cabinet of twice the size, but an effectively infinite cabinet.

One driver moves in, the other moves out, the net change in the volume of air in the cabinet is zero, so the cabinet is effectively "not there"

The catch is that, since the drivers are moving out of phase, relative to both the cabinet interior and to the room, you will also get virtually no change in the air pressure in the room...

In other words, at any frequency where the wavelength is significantly longer than the distance from one side of the cabinet to the other, NO ACOUSTIC OUTPUT.

(So, with something the size of a sub, all you'll be doing at 20 Hz is blowing air back and forth around the box.)

There are a few designs where using two drivers serves a purpose.

1.

Putting two drivers, in phase, on opposite sides of the cabinet can serve to cancel out external vibrations (becomes their momentum is opposite).

2.

If you use two drivers, one facing in and one facing out, wired OUT of phase, you still effectively have each driver in a box half the size, but delivering their expected outputs. 

However, with luck, some distortions that occur as a result of asymmetry of the movement of the drivers may cancel out, resulting in lower overall distortion.

(M&K and a few others have done this.)

3.

There is an arrangement where two drivers are mounted very close - one in front of the other.

The rear driver is essentially intended to "isolate the front driver from the box by doing the heavy work of pushing the air out of the way".

If you drive them with the same signal you basically end up with a driver that has the area of one of them, the mass of both, and the motor strength of both, in a normal sized cabinet.

This can give you better low frequency extension in a small cabinet - but at the cost of overall maximum displacement and efficiency.

(This is referred to as "an isobaric arrangement".)

Remember that, if you flip a driver around, then also flip the wires, you have not actually changed anything at all.

(Reversing both "the mechanical phase" and "the electrical phase" at the same time puts you right back where you started.)

At low frequencies there isn't much difference between the front and back of the driver... all that matters is if both move out at the same time (in phase), or out at opposite times (out of phase), relative to the inside of the box.

Nov 10, 2021 18:38:51 GMT -5 Boomzilla said:

Nov 10, 2021 15:04:20 GMT -5 KeithL said:

...

Incidentally, if you start with a box, with a driver in it...

And then cut another hole for a second identical driver...

Driving them with identical signals...

You will be HALVING the effective volume (it's exactly the same as if each driver was mounted in a separate box - each HALF the size of the original).

So efficiency will go down, box resonance will go up, and the frequency where your woofer starts to roll off will he higher.

And, if you're using a ported design, all the calculations there will change as well.

(Basically you need to treat the sum of the areas and masses of your two drivers as if they were a single larger driver - in the same box.)

OK, correct me if I'm wrong, but your statements above assume that the two drivers are facing each other (physically out of phase, but not electrically). If the drivers are working in tandem (both moving together) then the opposite should happen, yes?

In the following graphic, the top two drivers, wired in phase will (as you say) see half the volume of the box.

But in the bottom graphic, wired in phase, they will (ignoring the space that the right driver occupies in the box) see TWICE the actual volume. Yes?

<button disabled="" class="c-attachment-insert--linked o-btn--sm">Attachment Deleted</button>

[Boomzilla]

So in picture 2, I've essentially created a Dipole system?

[fbczar]

Nov 8, 2021 9:44:31 GMT -5 Boomzilla said:

Being too lazy to construct a box, I'd like to play with making an infinite-baffle, dipolar subwoofer. From what I've read, I think I want the driver in close proximity to the floor to take advantage of the boundary effect. I'd also like to have sufficient cone area to play moderately loudly (95 dB peaks @ 20 Hz.) without reaching the excursion limit of the drivers. I'm also willing to experiment with isobaric loadings where two drivers are working in tandem. I can use DSP equalization to extend the lowest frequencies of the design and to flatten the response, but if I do, the drivers will have to have some power handling capability.

Unfortunately, most of the infinite baffle designs I've come across on the internet are for car audio, and this is to be a home audio application (music first - HT only occasionally). I'm not sure how parameters would change (if at all) when considering the driver to use.

For lots of watts at low frequencies, I think that a "pro" Class-D amp is probably the least expensive option, although I do have conventional Class-AB amps already in house.

I know from previous postings that some of you have used infinite baffle subs before. Suggestions?

Thanks kindly - Boomzilla (aka Glenn Young)

Boom, Go to GR-Research. They sell OB subwoofer kits and OB speaker kits. Also go to SpatialAudioLab.com. They make OB speakers. 

[millst]

Nov 10, 2021 15:04:20 GMT -5 KeithL said:

In order to have a true infinite baffle design the baffle must be infinitely large - with an infinitely large space behind it.

(You can get pretty close to this by mounting a driver in a wall - with a large room behind it.)

In either of those cases, there is no connection between front and back, so the back wave has no opportunity to cancel the front wave.

 While technically true, having 10xVas of all the drivers in an IB array is basically "infinite" for all intents and purposes. Even 4x, while less ideal, is probably good enough for most cases.

[DYohn]

Also technically all “sealed” systems are infinite baffle, with smaller ones approaching acoustic suspension (see Villchur.)