Ripol cavity resonance vs cross over design

[Bending]

I was reading about Ripol subwoofers for years and I am intrigued as any subwoofer I tried (mostly closed box design) are sounding bloated compared to my bending wave drivers. These are almost full range from around 150Hz to 30kHz. I am not aiming with that sub for very low frequencies, -6dB at 30HZ would be more than acceptable.

There a very few commercial subwoofers based on Ripol design like Voxativ, Modal and just announced Borrensen bass modules(no data yet). So likely I will be more interested in building my own ones.

Most Ripol designs are based on 12" drivers and have the cavity resonances often around 300Hz or a little lower. Would a design with smaller woofers and therefore smaller cavities significantly increase the cavity resonance frequency?

A crossover frequency of 150Hz would require a step slope but even that may be not quite enough. A notch filter would be useful but I am interested in using active crossovers and built in notch filters are not common. Presently I am using a Devialet amp with its built in digital crossover which allows 4th order slopes (my measurements show effectively more a third order slope). I am contemplating the use of an active analog crossover as digital ones induce a time delay of at least 20ms. I have not seen any ones with notch filters except the one from Alex Ridthaler (the inventor of the design) which are not anymore available.

 

[ppataki]

A notch filter would be useful but I am interested in using active crossovers and built in notch filters are not common.

If you use a computer as a DSP (like I do) then you can have notch filters and literally any kind of filters available. Basically you will have unlimited DSP capability.

 

[Heinrich]

There a very few commercial subwoofers based on Ripol design like Voxativ, ...

The so called Ripol tries to add mass to a driver's cone by coupling it to the air moved back and forth. It is a cumbersome attempt. In case it reduces distortion by summing up two drivers.The peak you are addressing is a regular occurance with every 'dipole' to begin with (see also Linkwitz' famous woofers).

If there were a real resonance in the folded enclosure, then you could alter it with techniques extensively used in so called meta materials - many possible pathlengths.

But first rethink how the resonance would form. Where in space would it be excited, center of cone, the circumference, elsewhere?

Anecdote: recently I tried to 'dipolse' a midrange. It faile misreably, because for one the dipole took away the fundamental's sound pressure by interference, which is intended by the dipole's design, second it generated a hump further above in frequency, which emphasized the distortion components of the fundamental. Combined the distortion was measured to be tenfold(!).

In case you do not need the mythological component of that dipole stuff, and other baffling story telling (Voxativ) then you may want to investigate the reasons of 'bloating' bass, which for sure are other than a fundamental problem with the concept of a sealed box.

 

[Arindal]

Would a design with smaller woofers and therefore smaller cavities significantly increase the cavity resonance frequency?

Possibly yes, but my experience with smaller drivers for Ripol and similar concepts (like Ecouton) is more than frustrating. Even a multiple-10" with sufficient power delivered somehow weak, hollow bass with a tendency of distorting quickly in the lower end of its band, lacking punch. Would always opt for a 12" or better 15".

 

[Heinrich]

Possibly yes, but my experience with smaller drivers for Ripol and similar concepts (like Ecouton) is more than frustrating. Even a multiple-10" with sufficient power delivered somehow weak, hollow bass with a tendency of distorting quickly in the lower end of its band, lacking punch. Would always opt for a 12" or better 15".

This may help; there's a little maths, or better to say there are calculations to support a sober plan:

Electro-acoustic models

Acoustical and electrical models for the design of a dipole loudspeaker with numerical examples for the PHOENIX project.

www.linkwitzlab.com

On downsizing, if the drivers get smaller, the cavities get shorter, which will lift the resonances and lower the efficiency proportionally.

All in all the Ripol is a pathetic concept at least in regards to distortion. Not only is the fundamental tone deliberately attenuated to form the dipole characteristics, but the distortion components get amplified in absolute terms proportionally, summing up to double effect (in calculated numbers: 3-fold HD in low registers, 10-fold in upper), and ... you've got the resonances which amplify the distortion further by an even bigger amount on top of that (aprox 3..10-fold, resulting in total 10 to 100 times the HD of a sealed box).

Better try to get a grip on the sealed woofer types.

 

[ppataki]

Better try to get a grip on the sealed woofer types.

+1 to that
Just use one or more sealed subs and use DSP to tame them - job done

 

[Bending]

If you use a computer as a DSP (like I do) then you can have notch filters and literally any kind of filters available. Basically you will have unlimited DSP capability.

I was using DSP for my headphones and via Roon for my speakers. The problem is that the filter is applied before the crossover but I require that notch filter only for low pass and I don't want any filtering for the the rest. Doing DSP filtering causes time delay and I want to avoid that the subs are delayed as my bending wave divers are perfectly time coherent. Analog active crossovers are supposed not to cause a time delay but I have not seen any with a notch filter for a single leg.

 

[ppataki]

I was using DSP for my headphones and via Roon for my speakers. The problem is that the filter is applied before the crossover but I require that notch filter only for low pass and I don't want any filtering for the the rest. Doing DSP filtering causes time delay and I want to avoid that the subs are delayed as my bending wave divers are perfectly time coherent. Analog active crossovers are supposed not to cause a time delay but I have not seen any with a notch filter for a single leg.

There are zero delay DSP plugins, those will not introduce any delay whatsoever
And you can place the notch filter wherever you want in the signal chain or on whichever channel

 

[Heinrich]

I was using DSP for my headphones and via Roon for my speakers. The problem is that the filter is applied before the crossover but I require that notch filter only for low pass and I don't want any filtering for the the rest. Doing DSP filtering causes time delay and I want to avoid that the subs are delayed as my bending wave divers are perfectly time coherent. Analog active crossovers are supposed not to cause a time delay but I have not seen any with a notch filter for a single leg.

Bending wave, Manger or DML? There is no time delay with x/overs, actually. A notch is simple made in the analog signal domain, line level or post power amplification.
Can you solder some lumped pieces together?

 

[Bending]

This may help; there's a little maths, or better to say there are calculations to support a sober plan:

Electro-acoustic models

Acoustical and electrical models for the design of a dipole loudspeaker with numerical examples for the PHOENIX project.

www.linkwitzlab.com

On downsizing, if the drivers get smaller, the cavities get shorter, which will lift the resonances and lower the efficiency proportionally.

All in all the Ripol is a pathetic concept at least in regards to distortion. Not only is the fundamental tone deliberately attenuated to form the dipole characteristics, but the distortion components get amplified in absolute terms proportionally, summing up to double effect (in calculated numbers: 3-fold HD in low registers, 10-fold in upper), and ... you've got the resonances which amplify the distortion further by an even bigger amount on top of that (aprox 3..10-fold, resulting in total 10 to 100 times the HD of a sealed box).

Better try to get a grip on the sealed woofer types.

I am using presently a sealed woofer but sealed woofers have two fundamental issues: pressure is asymmetric (+ vs -) which more an issue at high volume and radiation pattern is omni directional. That is the reason I am curious about dipole and Ripol woofer/subs. My first real speaker was an Apogee and I was quite satisfied with its low frequency performance even the lows did not play very loud.

 

[Bending]

Bending wave, Manger or DML? There is no time delay with x/overs, actually. A notch is simple made in the analog signal domain, line level or post power amplification.
Can you solder some lumped pieces together?

Yes I can solder! Since 2 decades I avoid passive filters post amp. About my bending wave driver: https://www.diyaudio.com/community/threads/improving-bending-wave-drivers.420495/

 

[Heinrich]

Yes I can solder! Since 2 decades I avoid passive filters post amp. About my bending wave driver: https://www.diyaudio.com/community/threads/improving-bending-wave-drivers.420495/

With all due respect, great achievements!! This shouldn‘t hinder me to comment on your referring project. Regarding the gas state equation pV = nkT, it is nonlinear. Only that its contribution is nelegable compred to the motor‘s. I recently built a successful box comprising a 7“ woofer in 8 liters going down to 20Hz in-room.

Regardless of my huge respect for S.Linkwitz, the dipole sub is a dead end, and even more so the so called ripole. Distortion gets out of hand, always. Better to have several sealed boxes, positioned reasonably in the room, and equalize to taste.

Regarding line level equalizing, I thought, for my own use, of combining real tiny inductors combined with capacitors very much like ‚solderdude‘ does for headphone equalization. Didn‘t try yet.

 

[M240i]

@Heinrich
Are you saying Linkwitz didn't understand the trade-offs of his dipole woofer configurations? If so, you'd be mistaken.
In fact, he has an entire section of his website dedicated to a closed-box solution for subwoofer duties.

THOR - subwoofer

Design of a sealed box subwoofer with electronic equalization

www.linkwitzlab.com

 

[Sashoir]

I know just enough to know that I don't know the answers to your questions (at all), but there's a gentleman called "Jazzman" on this forum who has built several of these: it might be worthwhile sending him a direct message. Even if he doesn't know, he might know someone who does.

 

[charlielaub]

I was reading about Ripol subwoofers for years and I am intrigued as any subwoofer I tried (mostly closed box design) are sounding bloated compared to my bending wave drivers. These are almost full range from around 150Hz to 30kHz. I am not aiming with that sub for very low frequencies, -6dB at 30HZ would be more than acceptable.

There a very few commercial subwoofers based on Ripol design like Voxativ, Modal and just announced Borrensen bass modules(no data yet). So likely I will be more interested in building my own ones.

Most Ripol designs are based on 12" drivers and have the cavity resonances often around 300Hz or a little lower. Would a design with smaller woofers and therefore smaller cavities significantly increase the cavity resonance frequency?

A crossover frequency of 150Hz would require a step slope but even that may be not quite enough. A notch filter would be useful but I am interested in using active crossovers and built in notch filters are not common. Presently I am using a Devialet amp with its built in digital crossover which allows 4th order slopes (my measurements show effectively more a third order slope). I am contemplating the use of an active analog crossover as digital ones induce a time delay of at least 20ms. I have not seen any ones with notch filters except the one from Alex Ridthaler (the inventor of the design) which are not anymore available.

A ripole sub will have tilted response all the way up to the cavity resonance. The resonance will be a sharp peak, and after that the response rapidly falls off into a null or dip. To get a flat response you have to add a LOWPASS filter with a very low corner frequency, e.g. 30 Hz or even lower. I think usually second order is used. This flattens the response, but will set the system level to whatever the level is of the system before the filter was applied, but down around 30 Hz where your filter corner frequency is located. This makes the sensitivity in the lower bass very, very low. For example under 80dB would not be out of the ordinary.

Here is a link to a 12" Ripole design. There is a plot of the frequency response after the filter has been applied to flatten the response:

RiPol Subwoofer

RiPol

www.lautsprechershop.de

You can see from the plot that the cavity resonance was around 200Hz. The sensitivity is given as 80dB at 2.83V input. The system can be used below 100Hz or so. 150Hz would be pushing it because you want that sharp transition at the resonance peak to be suppressed so there is not too much ringing in the time domain.

 

[sigbergaudio]

I was reading about Ripol subwoofers for years and I am intrigued as any subwoofer I tried (mostly closed box design) are sounding bloated compared to my bending wave drivers. These are almost full range from around 150Hz to 30kHz. I am not aiming with that sub for very low frequencies, -6dB at 30HZ would be more than acceptable.

I am not sure how you can compare the sound between two speakers that doesn't play the same frequencies? How do you know that one is bloated and the other is not when they're not reproducing the same sound?

I am using presently a sealed woofer but sealed woofers have two fundamental issues: pressure is asymmetric (+ vs -) which more an issue at high volume and radiation pattern is omni directional. That is the reason I am curious about dipole and Ripol woofer/subs. My first real speaker was an Apogee and I was quite satisfied with its low frequency performance even the lows did not play very loud.

You are presenting two "problems" with sealed subwoofers, but I am not sure it is clear from your post why this is a problem exactly.

Sealed subwoofers will play linear and with low distortion, so is asymmetric pressure actually an issue?
With regards to ripol/dipole, any perceived benefit will come at the cost of efficiency and lack of low end extension.

I suspect there are other and better solutions to your problems, typically employing at least two subs combined with EQ/DSP.

 

[KSTR]

Most Ripol designs are based on 12" drivers and have the cavity resonances often around 300Hz or a little lower. Would a design with smaller woofers and therefore smaller cavities significantly increase the cavity resonance frequency?

Yes. Please read about slot-loaded open baffle, for example here: https://www.firstwatt.com/wp-content/uploads/2023/12/art_slob.pdf
In the end you build a bunch of small-driver "Ripoles" (just a fancy name) and stack them on top of each other. With half the depth of the cavities the pipe frequency goes up one octave, but efficiency suffers.

There is also the possibility of making the cavity depth blurred so the resonances are spread out over a whole range. The design of these is extremely complex, you either go by trial and error via three dozens of prototypes or invest and delve into full 3D acoustic modeling.

 

[Heinrich]

@Heinrich
Are you saying Linkwitz didn't understand the trade-offs of his dipole woofer configurations? If so, you'd be mistaken.
In fact, he has an entire section of his website dedicated to a closed-box solution for subwoofer duties.

THOR - subwoofer

Design of a sealed box subwoofer with electronic equalization

www.linkwitzlab.com

Nope, my intention was to point to the downsides of the dipole attempt. S.Linkwitz is still an authority. For example his notions on Doppler induced distortion are ignored entirely. I mentioned them several times. No (!) reaction, go figure …

In regard to ripole the caveats were stated. It is for the o/p to infer his own conclusions. I personally would be cautious not the least because it is advertised by just dubious outfits.

 

[Heinrich]

Yes. Please read about slot-loaded open baffle, for example here: https://www.firstwatt.com/wp-content/uploads/2023/12/art_slob.pdf

O/k, on plausibility. If there was a 9dB gain in efficiency from the slot loading alone, why don‘t we slot-load every driver, especially bass drivers mounted to closed boxes? Or, for something in reach of an armlenght, why don‘t we use compression drivers without matching horns? (A compression driver does exactly that slot-loading, that‘s the compression, velocity transformation.)

That throws a deep shadow on Heil‘s ‚transformer‘ aka AMT, doesn’t it? As I said …

 

[M240i]

Nope, my intention was to point to the downsides of the dipole attempt. S.Linkwitz is still an authority. For example his notions on Doppler induced distortion are ignored entirely. I mentioned them several times. No (!) reaction, go figure …

Linkwitz was one of the few people to not play fast and loose with the term "subwoofer."