Dual Opposed Subwoofer Theory?

[gnarly]

I had one question that came to my mind.

Usually Dual-Opposed Subwoofers (DOS) are designed like this:
(view from the side)
View attachment 247496

Theoretically speaking, would this design also work fine?

View attachment 247497

I am asking as I am aiming for a slim design and the latter driver arrangement would allow that
I think it would work just fine but any comments would be welcome
Thank you

I think the force cancelation you can except is basically the cosine of the angle between the drivers.

In the top pict and horizontal with an angle of 0 degrees, the cosine is 1.0 and there's 100% cancellation.

In the bottom pict, let's just call the angle 60 degrees upward. Which has a cosine of 0.5. I'd expect a maximum of 50% force cancellation.

 

[Head_Unit]

I used to build isobaric subwoofers as a hobbyist. I made a direct comparision (using close-miked response curves) between a single-woofer sealed box system and an isobaric system half the size. Both used the same woofer. The published equation indicates a halving of effective Vas for isobaric loading and therefore the frequency responses should have been essentially identical. The actual measured response was very different; the isobaric system had much deeper bass than predicted, so I think the published equation is incomplete.

Do you still have the measurements? And cabinet dimensions? I'm quite curious.

 

[Head_Unit]

I would wire them in parallel (i.e. 8ohm chassis required) instead of serial since this should work better in case there are differences between the drivers.

Yes because (a) there are always differences and (b) Dr. Richard Greiner said so (AES.org but I can't find the paper I have in mind, I'll have to dig in my garage. Also from years working with amps my belief is never mind the specs, loading them down to very low impedances does not actually give more dynamic power at clipping.

 

[Head_Unit]

I had one question that came to my mind.

Usually Dual-Opposed Subwoofers (DOS) are designed like this:
(view from the side)
View attachment 247496

Theoretically speaking, would this design also work fine?

View attachment 247497

I am asking as I am aiming for a slim design and the latter driver arrangement would allow that
I think it would work just fine but any comments would be welcome
Thank you

As some others noted while there should be significant cancellation there will be more and more torquing kind of effect as they get farther and farther apart. Now if you had one facing left on top, two facing right in the middle, and another facing left at the bottom, that should cancel well. Otherwise get your woofers as close as possible to minimize torquing...although a little piece of my brain is thinking if you have two cone backs very close moving in opposite directions the airflow could be problematic.

 

[Duke]

Do you still have the measurements? And cabinet dimensions? I'm quite curious.

That experiment was almost forty years ago, so I'll be going by memory.

Both boxes used the Dynaudio 21W54 woofer. The sealed box was 2 cubic feet (net internal volume) and the isobaric was 1 cubic feet. Or at least that was the ratio; the internal volume of the sealed box was intentionally twice the internal volume of the isobaric. The woofers in the isobaric were wired in parallel.

Using close-miced measurements, the sealed box was -3 dB in the low 70's and the response fell off by about 9 dB per octave below that, indicative of a low-Q sealed box.

The isobaric was -3 dB in the high 40's and fell off at about 15 db per octave below that, indicative of a fairly high-Q sealed box.

I calculated the effect of the air mass in the "isochamber" coupling to the cones and thereby increasing the effective moving mass, and my recollection is that calculated effect of the added mass was insufficient to explain the measured difference.

 

[ppataki]

(Hopefully) last question:

If done this way, is there a minimum recommended distance between the two drivers? (basically between the two magnets)

Due to space constraints I would need to put them relatively close to each other (like 5cm max) - would that be an issue?
Thank you

 

[Keith_W]

Sure, a number of companies stack their drivers together like that. Here are a few examples:
B&W DB series:

SVS Micro Series:

KEF KF92:

 

[Keith_W]

Additionally, if slimness is a priority, you could arrange your drivers like this:

As mentioned above, you lose the vibration cancellation feature which may or may not be an issue for you.

 

[digitalfrost]

Sure, a number of companies stack their drivers together like that.

Some drivers have vented pole pieces, I wonder what happens if you put them back to back.

 

[fineMen]

The dual opposed configuration (woofers on opposite sides of the cabinet to cancel vibrations) was patented, ...

Wow! That cancellation is about the inner workings of a smoothly running internal cumbustion engine, and they got it patented--again?!

Anyway, people find the 'isobaric' more intersting, mysterious.

I read hundreds of posts but I couldn't found exact theory and building guide about DOS.

Reason is, there is none, because not needed. Yes, the need for just another theory is painfully missing. We cannot do anything about it, so better we cope with the situation.

In reality the, as you name it, DOS is just about putting two drivers into one enclosure, so that the momentum (Newton's acceleration thing, physics class?) of the moving cones cancels out. It reduces the common rattle with subwoofers that usually have quite heavy cones.

If one doubles the (internal) volume of what would be chosen for a single driver, the twin team will act the same, especially in regard to f3, the low frequency limit. Because two are used, the maximum output will rise by 6dB compared to a single driver, the power consumption will also double. Only caveat is what you said, it can only be used for relatively low frequencies, say up to 200Hz.

 

[Prana Ferox]

Additionally, if slimness is a priority, you could arrange your drivers like this:

As mentioned above, you lose the vibration cancellation feature which may or may not be an issue for you.

DO subs tend to be sealed (although they don't have to be.) Sealed subs often need to be huge to get LF extension, but because they inherently protect the driver from unloading / violating Xmax, you can often make up for a too-small enlosure with EQ and extra wattage.

Generally by the time you get to the point of Keith's example above, you're probably fighting a losing battle and would be better off replacing one driver with a tuned passive radiator.

 

[ppataki]

I have already designed and built two DOS systems (this and this) and they both worked like a charm but now I might venture into a third one that is supposed to be a slim design hence I was asking the above questions
I will stick to the 0° angle version so vibration is fully eliminated (that's the main goal) but I will keep the drivers close to each other to keep the design as slim as possible

 

[digitalfrost]

DO subs tend to be sealed (although they don't have to be.) Sealed subs often need to be huge to get LF extension, but because they inherently protect the driver from unloading / violating Xmax, you can often make up for a too-small enlosure with EQ and extra wattage.

Generally by the time you get to the point of Keith's example above, you're probably fighting a losing battle and would be better off replacing one driver with a tuned passive radiator.

You could just as well do them with passive membranes.

 

[Keith_W]

Wow! That cancellation is about the inner workings of a smoothly running internal cumbustion engine, and they got it patented--again?!

Anyway, people find the 'isobaric' more intersting, mysterious.

I think that cancellation feature only applies for boxer engines, all other types of engine including V types and inlines have an angular momentum so have more vibration.

As for "isobaric", my understanding is that it refers only to the air space that is common to both drivers, which I have coloured in red in the diagrams below. The supposed advantage of an isobaric configuration is that it allows you to halve the Vas for the cone facing the sealed box, allowing you to make a smaller cabinet. However, there is no advantage in smaller cabin size with magnet to magnet configuration (third diagram on the right), if you are going to do that you may as well get one driver. Cone to magnet gives you a marginally smaller cabinet, and the most compact design is cone to cone.

You could just as well do them with passive membranes.

Passive membranes need to be tuned by adding mass or playing with tension. Not so easy for those of us who are DIY.

 

[digitalfrost]

Passive membranes need to be tuned by adding mass or playing with tension. Not so easy for those of us who are DIY.

In these regions the room takes over anyway. What makes you think any closed box speaker is gonna be better than any reflex tuned speaker? There is no perfection to be found there.

 

[Keith_W]

In these regions the room takes over anyway.

If you are going to build a subwoofer box with passive radiators, the passive radiators need to be tuned to give the proper Q. Yes, I know the room takes over, but you still want your sub to at least start off as linear as you possibly can before applying room correction, don't you?

What makes you think any closed box speaker is gonna be better than any reflex tuned speaker? There is no perfection to be found there.

Easier to design and build. You look at the Vas, choose less if you want less Q, more if you want more Q, build a box, and that's it. No need to mess with tuning the ports, mass loading and tensioning passive radiators, and so on. For me, the ease of design was the biggest advantage but I understand if other people have different priorities, e.g. they might need more output, or fit the sub somewhere in the room, etc. Not to mention, they might have more skill than me.

I agree there is no such thing as perfection, but IMO a sealed subwoofer has advantages that I want - less ringing, less phase rotation, more linear frequency response. The disadvantage is lower output (doesn't matter to me because I have multiple subs) and poorer low frequency extension (actually undesirable for me because my walls resonate at about 15-25Hz). You pick and choose which solution fits your needs the best, and if it fits your needs perfectly, then it is perfect. I hope you note that I don't go around telling people that my way is the best, I explain the options and the advantages/disadvantages and let them decide for themselves.

 

[fineMen]

If you are going to build a subwoofer box with passive radiators, the passive radiators need to be tuned to give the proper Q. Yes, I know the room takes over, but you still want your sub to at least start off as linear as you possibly can before applying room correction, don't you?

Not quite. The room is going to amplify the deep bass considerably. For reference, I'm using a tiny, sealed box. Of course I have to correct for the high Fspeaker/box. But not as much as expected. In the near-filed the bass @30Hz is still about 8dB below mids. In room I get a -3dB at 25Hz. And it is pretty acceptable in regard to distortion.

I agree there is no such thing as perfection, but IMO a sealed subwoofer has advantages that I want - less ringing, less phase rotation, more linear frequency response.

As stated above, in some cases the bass of a sealed would happily mate with the room's amplification. That cannot be done as easily with ported or with passive radiators. With the latter the tuning would determine the lower frequency cut-off. To set the tuning lower has some limitations, e/g port noise or the moving mass of a passive radiator. It's a tricky compromise.

 

[audiofooled]

In these regions the room takes over anyway. What makes you think any closed box speaker is gonna be better than any reflex tuned speaker? There is no perfection to be found there.

It actually can be better https://www.audiosciencereview.com/forum/index.php?threads/analytical-analysis-room-gain.23211/

 

[Chromatischism]

Not quite. The room is going to amplify the deep bass considerably. For reference, I'm using a tiny, sealed box. Of course I have to correct for the high Fspeaker/box. But not as much as expected. In the near-filed the bass @30Hz is still about 8dB below mids. In room I get a -3dB at 25Hz. And it is pretty acceptable in regard to distortion.


As stated above, in some cases the bass of a sealed would happily mate with the room's amplification. That cannot be done as easily with ported or with passive radiators. With the latter the tuning would determine the lower frequency cut-off. To set the tuning lower has some limitations, e/g port noise or the moving mass of a passive radiator. It's a tricky compromise.

In my room, everything under 35 Hz is boosted. That means ported subs get the same coupling as sealed. Advantage: ported.

In a larger room, that frequency shifts downward (speed of sound/frequency=wavelength).

 

[fineMen]

In my room, everything under 35 Hz is boosted. That means ported subs get the same coupling as sealed. Advantage: ported.

In a larger room, that frequency shifts downward (speed of sound/frequency=wavelength).

Previously my way bigger ported speakers were tuned to about 40Hz. That was too high to take advantage of my (!) room to anywhere lower.

I came to use a sealed because with just 8 liters of internal volume a ported design is useless. If some reasonable limits for port noise are stated the max output is restricted by that alone, never the cone's excursion. I tried a damped bass reflex, but that elevated the 3rd order intermodulation distortion (!) despite half of the cone movement by a factor of two. I still don't know why that is ... (maybe a 'snapping' cone movement due to highly non-linear airflow).

I don't believe that the greater group delay, vulgo 'impulse' speaks against ported, but the port noise, the chuffing. Ironically most with high power applications. As said, a complicated, individual compromise.