I’d like to use a force cancelling (magnet to magnet) bipole sub to go “under” some dipole speakers I’m planning. I anticipate a low pass for the sub at about 80hz.
In order for the force cancelling to work the rear facing woofer will obviously need to be electronically in phase with the forward firing woofer. This means the rear wave of the sub will be out of phase with the rear wave of the rest of the speaker. Am I designing in a problem or are the wavelengths so long that its just a chaotic system at those frequencies and so entirely unpredictable ?
In order for the force cancelling to work the rear facing woofer will obviously need to be electronically in phase with the forward firing woofer. This means the rear wave of the sub will be out of phase with the rear wave of the rest of the speaker. Am I designing in a problem or are the wavelengths so long that its just a chaotic system at those frequencies and so entirely unpredictable ?
In the ripple tank simulator I've set up four sources, two back to back, upper and lower.
At the start, I've put baffles above and below so you can compare the front and back. Here it is easier to see the way it is configured looking at a higher frequency..
Now, using a lower frequency you can clearly see that the null is becoming significant..
Next, I've removed the baffles so you can see the normal dipole null at the top, and the new null at the left (rear)..
And at lower frequencies..
At the start, I've put baffles above and below so you can compare the front and back. Here it is easier to see the way it is configured looking at a higher frequency..
Now, using a lower frequency you can clearly see that the null is becoming significant..
Next, I've removed the baffles so you can see the normal dipole null at the top, and the new null at the left (rear)..
And at lower frequencies..
At low frequencies the bipole is no longer bipole but really a monopole, as shown in the lowest pic above.
At the crossover point between a dipole and a monopole the response transitions through a cardioid over a very narrow band when the filter is steeper like LR4, etc. You will not hear this. But the power response will be 4.77dB higher for a monopole compared to a dipole when the on axis SPL is the same, so you might sense that the bass level is that much higher below the crossover point, which should be relatively low, eg. no higher than 100Hz.
Also, it might be uglier, but you will get some 2nd order distortion reduction if one driver is facing cone in while the other is facing cone out. The cones move inwards or outwards at the same time. You could put the magnet out driver to the rear to hide it.
At the crossover point between a dipole and a monopole the response transitions through a cardioid over a very narrow band when the filter is steeper like LR4, etc. You will not hear this. But the power response will be 4.77dB higher for a monopole compared to a dipole when the on axis SPL is the same, so you might sense that the bass level is that much higher below the crossover point, which should be relatively low, eg. no higher than 100Hz.
Also, it might be uglier, but you will get some 2nd order distortion reduction if one driver is facing cone in while the other is facing cone out. The cones move inwards or outwards at the same time. You could put the magnet out driver to the rear to hide it.
At those low freuencies (~ <500 Hz), the size of the boax is such that the waveform coming out the driver “do not see the box”.
A 2nd driver on the back (sides, top, or bottom) within a quarter-wavelength will act as a single woofer, push-push gives reaction force cancelation greatly reducing vibrational load on the box.
dave
Yes the force cancellation benefit was what prompted me to think about the bipole configuration.
So if I’ve understood correctly the net effect of doing this would be that I’d get the benefit of reduced vibration and it would behave like a front firing monopole sub.
I have heard for myself how well a monopole box sub can work below an OB mid / tweeter - this speaker sounds superb even in hi fi show conditions:
https://www.qualioaudio.com/
So if I’ve understood correctly the net effect of doing this would be that I’d get the benefit of reduced vibration and it would behave like a front firing monopole sub.
I have heard for myself how well a monopole box sub can work below an OB mid / tweeter - this speaker sounds superb even in hi fi show conditions:
https://www.qualioaudio.com/
In bipole, two woofers on oposite sides work against each other. Squeezing and stretching the air molecules inside. Box sides move in and out. Even with extensive bracing. Where is the force cancellation?
In dipole, woofers operate such, that there is constant pressure inside box. While one woofer moves out, the other moves in the box. Air molecules move back and force like in isobaric.
Side walls barely move. Dipole, in my measurements, has even lower fr extension, as woofers are not fighting agains each other, slowing each other cone movement.
I preffer sound of dipole. Bipole sounds just like monopole.
In dipole, woofers operate such, that there is constant pressure inside box. While one woofer moves out, the other moves in the box. Air molecules move back and force like in isobaric.
Side walls barely move. Dipole, in my measurements, has even lower fr extension, as woofers are not fighting agains each other, slowing each other cone movement.
I preffer sound of dipole. Bipole sounds just like monopole.
They work mechanically in opposition causing the active force cancelation. They work in harmony acoustically. You get% or more of the potential reaction energy making it to the box.
dave
Yes, but a dipole generates maximum vibration because both cones move together while the "bipole" has two drivers moving in opposite directions, thus cancelling vibration. The "side walls" are NOT the problem!
A bipole (sub)woofer radiates as omnipole (like a pulsating ball), until frequency gets too high for opposing waves to match, typically above 500Hz. And a monopole is actually omni too at low freq, before baffle step starts to happen (but we are more used to thinking of axial spl).
When omni and dipole radiation sum around crossover, the net is cardioid. The wideness of this range depends on type and steepness of the crossover.
Here my outdoor-measured dispersion, with 150Hz LR2 crossover, cardioid range is roughly one octave (-6dB)
How does this sound? Pretty good, because sound direction is practically impossible to detect below 300Hz. Major benefit is avoiding front-wall interference, but this happens only if distance and cardioid range match. One can use SBIR calculator to find this correlation. Eg. 500mm distance from the midpoint of the omni to wall gives 172Hz first null.
Tech studies of John Kreskowsky
https://musicanddesign.speakerdesign.net/craw_cross.html
https://musicanddesign.speakerdesign.net/Boundary_reflections.html
When omni and dipole radiation sum around crossover, the net is cardioid. The wideness of this range depends on type and steepness of the crossover.
Here my outdoor-measured dispersion, with 150Hz LR2 crossover, cardioid range is roughly one octave (-6dB)
How does this sound? Pretty good, because sound direction is practically impossible to detect below 300Hz. Major benefit is avoiding front-wall interference, but this happens only if distance and cardioid range match. One can use SBIR calculator to find this correlation. Eg. 500mm distance from the midpoint of the omni to wall gives 172Hz first null.
Tech studies of John Kreskowsky
https://musicanddesign.speakerdesign.net/craw_cross.html
https://musicanddesign.speakerdesign.net/Boundary_reflections.html
Attachments
Last edited:
How? Just like i said already.
Repeating again: when one woofer moves forward, making air molecules inside the box thinner, the other woofer is doing the same thinning the air molecules, just in other dirrection, fighting against each other. One woofer is slowing the other. Making it harder to move.
Quite easy to understand.
I have measured spl and fr response in both configurations. Its easy. Just flip the polarity of one woofer and measure.
In dipole config not only spl was higher, but fr response extension towards very low fr was noticable. This was reproducible in two different bipole vs dipole subs.
This post posted on my page.
Last edited:
Dipole generates zero pressure change inside the box.
In bipole, drivers may be tied together, but there are very big pressure changes inside, causing box vibration. And that IS a problem.
Imagine closed box with woofer...cone moving forward, against suspension, and air. Air acts like spring. Thats how its often presented in textbooks. Cone movement is restricted against this spring.
Now imagine that when cone moves forward, against the spring, you pull on that spring from inside, against the cone. Restricting its movement in crest.
Thats what oposite woofer is doing in bipole. Thats what both woofers are doing to each other.
I hate to say it, but even bose understood this. Hence separate enclosure for each woofer.
I did not fall off the turnip truck just today... Also a dipole makes for a pretty lame subwoofer, which is the application in question here.
Pressure change is no more of a problem for a "bipole" subwoofer as it is for any other closed box subwoofer. When you make the box small with respect to Sd*Xmax at high excursion there is some distortion increase from the different pressures seen by the cone when moving into the box and out of the box. This is well known. Linkwitz even has a spreadsheet to calculate the percent distortion contribution for this phenomena. It will still be way less than if you try to get the same in-room SPL at low frequency by using the same two drivers but configured as a dipole, even when given all the amplifier power in the world. The dipole would require much, much more displacement to reach the same SPL at LF, and the drivers' Xmax will eventually be exceeded. Combined with the high excursions and distortion penalty of the dipole type driver arrangement a dipole "subwoofer" is a losing proposition. Look, I am a dipole proponent and I am saying this emphatically.
The OP is simply making a closed box subwoofer and was confused that the driver arrangement would make it a "bipole".
If I understood the thread linked below correctly this might eventually be undesirable, because the (pleasing) 2nd order harmonics will not mask the (slightly more unpleasant) 3rd harmonics anymore.
https://www.diyaudio.com/community/threads/push-pull-vs-normal-distortion-compared.191833/
It is the vibration connected with the inertia of the moving mass of the cones that is being alleviated. This depends on how the cones move; it does not depend on how hard the motor needs to push to produce that movement.
Because there is no box. Using a forth & aft driver in a box wired as a dipole is to my mind a huge waste. AFAIC a diopole is an OB (including ELS and big planners).
Like any box. Just that 90% or more (est) of the energy in the box that could cause panel vibration are actively cancelled. It then becomes easy to push anything left way above the bandpass of the woofer. Hence NO box vibration.
Pressure can easily be shown to be only a small source of the energy that can cause panel vibration, and then - because of damping — only at the lowest frequencies where it has its own name — ballooning. Very easily avoided.
Most of the energy loaded into the box material is from the direct connection of the driver to the baffle. In the case of a dipole (OB) that all loads into a falt panel. It is going to be prone to vibrate much more than a wel designed box.
So your argument falls apart since building a vibration free OB is much more difficult than a vibration free box. And in the bass push-push is an easy way to actively eliminate most of energy that would cause box panels to ring.
And a dipole suffers from a particular kind of distrortion, unless outside far away from any surface, that i find particuarily annoying.
dave