bi-pole bass bin - what would happen to polar response?

[brianpowers27]

I am considering building a bass bin with a woofer on both the front and the back. At the current time I believe that I will cross around 450hz.

This looks attractive for both cosmetic and opposed motor cancellation perspectives.

The two woofers would be wired out of phase for maximum motor distortion cancellation, correct?

I am having a hard time imagining what will happen to the polar response. I suppose that this would behave similarly to a dipole on a thick baffle.

These woofers would likely cross around 450hz lr2 to an open baffle mid and monopole tweeter.

[ThomasW]

Don't.....it's not a dipole on a thick baffle nor will it function as one.

[brianpowers27]

Don't.....it's not a dipole on a thick baffle nor will it function as one.

Why doesn't this behave as a dipole on a thick baffle? It makes sense to me that if both woofers were positioned at a equal height and centered on the baffle it would be similar. It would resemble a h-frame, minus the transmission line hframe resonance.

Perhaps a bipole configuration would work better?

[krips]

I might be wrong here but wouldn't the motors have to be wired in phase for cancellation? Otherwise they will be moving in the same direction at the same time and the cone movements would be additive...

[brianpowers27]

I might be wrong here but wouldn't the motors have to be wired in phase for cancellation? Otherwise they will be moving in the same direction at the same time and the cone movements would be additive...

I think you are right.

One other thing that I have come to realize is this: The frequency of the dipole cancellation will be much too high for this to work with my intended drivers.

I guess wiring them in phase would shift it from dipole to bi-pole. I haven't investigated bipole enough to know what will happen to the overall system. (started a new thread to investigate bipole.)

[JoshK]

I too have a hard time seeing why this doesn't work fairly close to an H frame.

[augerpro]

I might be wrong here but wouldn't the motors have to be wired in phase for cancellation? Otherwise they will be moving in the same direction at the same time and the cone movements would be additive...

I wouldn't look at it as direction, because you can orientate the drivers however you want. I would look at it as compressing or rarefying the outside air. In this case if you run both in phase they will be bipole, out of phase they will be dipole. I don't know how having an internal seperation affects the overall performance, but it won't change whether it is dipole or bipole. If you run them opposed but out of phase (dipole) I would think the box will be jumping all over the place.

Brian what do you mean the dipole cancellation is too high? Typically one would operate the driver below the first dipole peak and following null.

[Mazeroth]

I might be wrong here but wouldn't the motors have to be wired in phase for cancellation? Otherwise they will be moving in the same direction at the same time and the cone movements would be additive...

Out of phase for cancellation.

Thomas, could you clarify why this wouldn't work as a dipole? Wouldn't it be more optimal than a typical dipole we're used to as the front and rear output would be, in theory, identical, unlike a typical dipole where the rear of the cone is being obstructed by the basket and magnet?

[brianpowers27]

Out of phase for cancellation.

Thomas, could you clarify why this wouldn't work as a dipole? Wouldn't it be more optimal than a typical dipole we're used to as the front and rear output would be, in theory, identical, unlike a typical dipole where the rear of the cone is being obstructed by the basket and magnet?

I think the principle has more to do with the phase of the actual force vs the absolute phase of the driver. FOr example: 2 driver angled same direction the cancellation would be from a out of phase configuration. If the drivers face the opposite direction, the phase would need to be the same.

I think this application works better in a bippole than a monopole. As AJ pointed out you still have air spring non-linearity to deal with.

[augerpro]

Good point about the radiation being more identical compared to the radiation of a single driver with basket, etc. Two problems I see Maz: the box jumping all over the place, and what it means to share the same volume. The first is obviously bad, but I'm not sure what the second means? If you did run them with a separation and out phase they would act much a like an OB dipole, except the the AC's will not be in the same place, so the combing pattern will be different. But you could maybe work this to your advantage. It would be similar to making a wider baffle, but as we know going to wide impacts the stereo image, and flattens the main lobe, so that response actually climbs off axis until the dipole radiation can start to take over. So you could maybe try two drivers on the front and back of a speaker with a narrower baffle, and adjust the dipole effect by adjusting the the speaker depth. Maybe a cylindrical or elliptical speaker cabinet would work well?

[JoshK]

Does John K (of NaO) do something to this affect in one of his cardiod subs? He plays with a delay to the signal in the rear facing sub?

While I could see why it maybe not perfect, or has a different set of trade-offs, I can't see why it can't mimic the dipole behaviour.

The box dancing is a problem, but how much different would have been then an open baffle with same woofers?

[john k...]

I don't post much here or anywhere these days but I though I could add something. First, my old CRAW woofer system was such an exercise. Two woofers were mounted in a box (internally isolated from each other), both with the magnets in the box. There were 3 configurations possible. 1) both drivers wired in phase = monopole response. 2) One driver wired out of phase = dipole, 3) one driver wired out of phase and delayed an appropriate amount = cardioid.


The advantage of using two woofers is that they both radiate the same response so the response is a true monopole, dipole or cardioid (at low frequency). Additionally, there will be not "1/4 wave resonances" to deal with as there is with an H or U frame enclosure. The disadvantage is, of course, the cost. However, if the enclosure is internally separated so each driver has its own chamber there is not going to be distortion cancellation regardless of how the drivers are mounted or wired. This is due to the nonlinearity of the air spring created by the enclosure. If the drivers are mounted one motor in and one motor out and they share the same enclosure, then at low frequency there is a potential for distortion cancellation. Non the less, if the argument is kept at low frequency then there is more to be had by building a 2 driver H frame such as used in the Linkwitz Orion.

[brianpowers27]

I don't post much here or anywhere these days but I though I could add something. First, my old CRAW woofer system was such an exercise. Two woofers were mounted in a box (internally isolated from each other), both with the magnets in the box. There were 3 configurations possible. 1) both drivers wired in phase = monopole response. 2) One driver wired out of phase = dipole, 3) one driver wired out of phase and delayed an appropriate amount = cardioid.


The advantage of using two woofers is that they both radiate the same response so the response is a true monopole, dipole or cardioid (at low frequency). Additionally, there will be not "1/4 wave resonances" to deal with as there is with an H or U frame enclosure. The disadvantage is, of course, the cost. However, if the enclosure is internally separated so each driver has its own chamber there is not going to be distortion cancellation regardless of how the drivers are mounted or wired. This is due to the nonlinearity of the air spring created by the enclosure. If the drivers are mounted one motor in and one motor out and they share the same enclosure, then at low frequency there is a potential for distortion cancellation. Non the less, if the argument is kept at low frequency then there is more to be had by building a 2 driver H frame such as used in the Linkwitz Orion.

John. I now realize the q of the woofers is inappropriate for a dipole. I am considering what might happen to create a bipole (2 woofers on opposite sides of the box, magnets in and normal polarity.) configuration in another thread that I spun off today. If you have the opportunity, take a look.

http://www.htguide.com/forum/showthread.php4?t=34154

[john k...]

Let's consider two woofers separated by some distance, d. This can be two woofers mounted on the front and rear of a box of width d. If the woofers radiate out of phase and are omni directional at low frequency, where the wave length is greater than 2d, the response is a dipole. Starting at DC the response rises at 6dB/octave to a peak where the wave length is 2d. If the drivers radiate in phase the response at low frequency is flat and has a monopole response up to where the wave length is 2d. AT that frequency (WL = 2d) there will be a null in the response. Above the WL = 2d frequency both formats will have a daisy like polar response, IF the radiation form each driver remains omni directional, with more and more pedals as the frequency increases.

In reality, at some point the drivers will be come directional and the front and rear radiation will become independent of each other. The frequency at which this happens obviously depends on driver and baffle size. Once the directionality takes over there is little difference between in and out of phase connections other than the difference in phase between the front and rear sources. Since directionality will typically take over at frequencies above the modal region of your room the effect of connection the rear woofer in or out of phase will have little effect on the sound at higher frequencies (above WL = 2d).


The major differences are in what happens below the WL = 2d frequency and limiting the use of the response to about an octave below that frequency, assuming directionality is not an issue.

[brianpowers27]

Let's consider two woofers separated by some distance, d. This can be two woofers mounted on the front and rear of a box of width d. If the woofers radiate out of phase and are omni directional at low frequency, where the wave length is greater than 2d, the response is a dipole. Starting at DC the response rises at 6dB/octave to a peak where the wave length is 2d. If the drivers radiate in phase the response at low frequency is flat and has a monopole response up to where the wave length is 2d. AT that frequency (WL = 2d) there will be a null in the response. Above the WL = 2d frequency both formats will have a daisy like polar response, IF the radiation form each driver remains omni directional, with more and more pedals as the frequency increases.

In reality, at some point the drivers will be come directional and the front and rear radiation will become independent of each other. The frequency at which this happens obviously depends on driver and baffle size. Once the directionality takes over there is little difference between in and out of phase connections other than the difference in phase between the front and rear sources. Since directionality will typically take over at frequencies above the modal region of your room the effect of connection the rear woofer in or out of phase will have little effect on the sound at higher frequencies (above WL = 2d).


The major differences are in what happens below the WL = 2d frequency and limiting the use of the response to about an octave below that frequency, assuming directionality is not an issue.

John,

Thanks for the detailed response. I spent some time trying to visualize the polar response of a bi-pole as described above. Does it make sense to assume that the 90 degrees to the drivers will exhibit a lobe (peak in polar pattern) where F < (WL = 2d)

[john k...]

John,

Thanks for the detailed response. I spent some time trying to visualize the polar response of a bi-pole as described above. Does it make sense to assume that the 90 degrees to the drivers will exhibit a lobe (peak in polar pattern) where F < (WL = 2d)

Well below the WL = 2d frequency the response will be uniform at all angles (omnidirectional). As the frequency rises towards WL= 2d the on axis response will begin to degenerate to a null but at 90 degrees the response will remain uneffected (assuming no driver directionality effects). In fact, as the frequency continued to rise the response at 90 degrees will always have a lobe maximum (again assuming no driver directionality effects). Compare this to a dipole where there is always a null at 90 degrees.

[brianpowers27]

Perhaps an application could be a cabinet that is rotated such that the drivers are facing sideways. If crossed low enough this would provide the advantage of sideways nulls, much like a monopole. This would then gain the advantage of using the box to support the low end.

The deepest of frequencies would still be omnidirectional while frequencies around wl=2d could be affected. If the baffle width were selected carefully, the null could be centered somewhere in the usual problem frequencies (80-200hz)

[Johnloudb]

I have some Nelson/Reed subs that are dipole. They have 2 12" woofers on the front and back (4 woofer total for each sub). The crossover they use has a 2nd order lowpass slope across the mid band (16 to 60Hz), to flatten the response of the woofers.

Are you using these as subs or just woofers? You likely won't get much low frequency output without equalization. I wouldn't turn them sideways either, especially if your crossing over at 450Hz.

This configuration does well as subs and it does help reduce box resonances at least for the sides of box, which would be the most troublesome. Kind of waste of a woofer from my perspective, though, since you're still limited to the displacement and Xmax of a single woofer. As a bi-pole, woofer wired in phase you'd get more output, I think. Well, I'm not entirely sure about that.

[brianpowers27]

I am thinking about running them up to about 200hz. Dipole is out of the question due to the woofer's characteristics.

The details I have floating around in my head are changing by the minute, as I gather more information.

[john k...]

Perhaps an application could be a cabinet that is rotated such that the drivers are facing sideways. If crossed low enough this would provide the advantage of sideways nulls, much like a monopole.

I think you mean dipole.

This would then gain the advantage of using the box to support the low end.

The deepest of frequencies would still be omnidirectional while frequencies around wl=2d could be affected. If the baffle width were selected carefully, the null could be centered somewhere in the usual problem frequencies (80-200hz)

The presence of the null does not mean the woofer system would behave as a dipole at that frequency with regards to room modes. The dipole's potential to limit excitation of certain room modes is a result of the drivers operation out of phase. This is not the case with a bipole. When room modes are considered it is really a case of multiple sources, their placement and there phase relationship. The biggest problem with low frequency reproduction in typical rooms is that positioning of the speakers for best presentation of the sound filed, or imaging is contrary to the best positioning for low frequency. You may want to look at this article addressing a novel way to set up a woofer system.

[brianpowers27]

I think you mean dipole.

Understandable, but I am still referring to bi-poles.

The presence of the null does not mean the woofer system would behave as a dipole at that frequency with regards to room modes. The dipole's potential to limit excitation of certain room modes is a result of the drivers operation out of phase. This is not the case with a bipole. When room modes are considered it is really a case of multiple sources, their placement and there phase relationship. The biggest problem with low frequency reproduction in typical rooms is that positioning of the speakers for best presentation of the sound filed, or imaging is contrary to the best positioning for low frequency. You may want to look at this article addressing a novel way to set up a woofer system.

Thanks for the link. I have read this one several times and have a very similar configuration in my living room.

It is my postulation that a dipole rotated sideways would behave like a monopole until WL = 2d Around this area there should be nulls (in the front and back?). THis could be potentially different than a monpole. The bipole listener might witness a reduction in radiated energy.

[john k...]

When I said I think mean dipole it was in reference your the statement, "Perhaps an application could be a cabinet that is rotated such that the drivers are facing sideways. If crossed low enough this would provide the advantage of sideways nulls, much like a monopole." It seemed to me that you shoud have said dipole instead of monopole. I.E. a monopole would have sideways nulls.

"It is my postulation that a dipole rotated sideways would behave like a monopole until WL = 2d Around this area there should be nulls (in the front and back?). This could be potentially different than a monopole. The bipole listener might witness a reduction in radiated energy."

Again, above you appear to be typing something other than what you are thinking.

In any event it is difficult to understand what your point is without knowing what frequency range you are referring to.

Form the beginning, if we consider monopoles, bipoles and dipoles constructed from omni-directional source then the single source monopole with have an omni-directional radiation pattern over all frequencies. A bipole composed of two such sources would have an omni-directional radiation pattern at frequencies roughly an octave below WL = 2d. A dipole would have a figure 8 pattern below WL = 2d. But as WL = 2d is approached the radiation pattern becomes highly frequency dependent. ultimately sufficiently above WL = 2d both the dipole and bipole radiate the same total power and have many lobed polar patterns unsuitable for loudspeakers.


In any event, "It is my postulation that a dipole rotated sideways would behave like a monopole until WL = 2d." is certainly an unclear statement. Do you really mean monopole or perhaps you should have said bi-pole? "until WL = 2d" form above or below? What frequency range are you considering? But to make one thing clear, a dipole never behaves like a monopole.

[brianpowers27]

John,

Thanks for your patience.

"It is my postulation that a dipole rotated sideways would behave like a monopole until WL = 2d." is certainly an unclear statement. Do you really mean monopole or perhaps you should have said bi-pole?

I did mean bipole. I am considering this approach for a woofer that is crossed low <200hz lr4. (Possibly with a plate amp.) I haven't taken the time to model this yet but I was considering a 10" w,20"d,32"t box with two woofers.

[john k...]

Ok, if I understand you, you are trying to make a comparison between a dipole and a bipole rotated 90 degrees. Lets say that d = 50 cm (20"). That means the 2d frequency will be 344 Hz. At 344 Hz the dipole would have a bloated figure 8 pattern. The bipole rotated 90 degrees would also have a figure 8 pattern, but it would be narrower. That is, at 45 degree the SPL would be lower than that of the dipole. Now, as the frequency drops the dipole will narrow down to a figure 8 pattern which will remain more or less constant at the frequency continues to drop. However, the bipole nulls will fill in and the response will become omnipolar. AT 200 Hz and below the bipole would essentially be identical to a monopole. If you want dipole response at low frequency you need to build a dipole.

[Mark K]

Hi John,
Good to see you posting. What have you been up to-I don't see you posting much and not alot new on your site (of course, the same can be said about me!!!)
Mark

[john k...]

Hi John,
Good to see you posting. What have you been up to-I don't see you posting much and not alot new on your site (of course, the same can be said about me!!!)
Mark

Hi Mark. What have I been up to? Not much building new stuff. Pretty much listening, revising the SoundEasy guide to keep up with the updates, assembling an occasional crossover to the NaO or U-frame woofer system. Sailing, tennis and driving. The usual suspects. I picked up a used Z4 roadster in May. Nice car to drive on the NE back roads.

I am working on a pair of dipole woofers to mate with my Martin-Logan Monolith panels. Should be a big improvement.

Not posting much because, well, life is too short to sit in front of the computer, or is it because I'd have to think? Still lurking in the mornings though. Today is an exception.

[Mark K]

Sorry about the slight thread hijack

Hi Mark. What have I been up to? Not much building new stuff. Pretty much listening, revising the SoundEasy guide to keep up with the updates, assembling an occasional crossover to the NaO or U-frame woofer system. Sailing, tennis and driving. The usual suspects. I picked up a used Z4 roadster in May. Nice car to drive on the NE back roads.

I am working on a pair of dipole woofers to mate with my Martin-Logan Monolith panels. Should be a big improvement.

Not posting much because, well, life is too short to sit in front of the computer, or is it because I'd have to think? Still lurking in the mornings though. Today is an exception.

Those Z4's are cute. Hard to fit 4 kids and the wife in one!

I've been pretty busy with my kids, work, and a major sidetracking back into sports-splitting my time between running and biking. I've dropped over #50, and been running 5k/12k's. My new toy is a titanium road bike that I'm building up and am trying to get myself ready for a century in Oct, as well as try to break 24 minutes for 5k and 1 hour for a 12k by early next year.

I've been meaning to post a little blurb about this on my website to explain why I've been so constipated with regards to new projects and, er, finishing my old ones...but even putting up a webpage well takes 2-3 hours. I would rather ride 50 miles on the bike!!