Yet Another RiPol Sub Project

[jazzman53]

Hi All,

I'm pleased to share my latest RiPol sub project below. It's actually a matched pair of subs using Peerless SLS 12 woofers. Not everyone is familiar with the RiPol concept so I will first give my take on it, and why RiPol's are my favorite subs.

The configuration was patented by German speaker builder Axel Ridtahler, hence the name "RiPol" is short for a "Ridtahler dipole".

A RiPol is basically a compact, folded-baffle dipole with two woofers in opposing push/push arrangement. Its radiation pattern isn't the classic dipolar figure-8, but cardioid shaped with a pronounced frontal lobe, and off-axis nulls.

I think a the most interesting RiPol characteristic is it's effect on woofers' fundamental resonance. All woofers have a natural resonance frequency which is typically loud and quite undesirable because it's a distortion that isn't in the music.

Conventional sub enclosures force the woofer's resonance higher than it's resonance in free air (more so in a smaller box, less so in a bigger or ported box, but always up/never down).

A RiPol enclosure lowers the woofer's resonant frequency as much as 10Hz, which not only allows it to play lower but has a profoundly positive effect on tonal quality.

Let's say we are using woofers that resonate at 30Hz in free air:

- A conventional enclosure forces the resonance upward, within the audible bass band. And if the music contains any energy at that frequency, the resonance will be excited and its distortion will be heard.

- A RiPol forces the woofers' resonance downward about 10Hz (down to 20Hz in this case). Most music contains little or no energy down at 20Hz, in which case the resonance would not be excited so would not occur. And even if the music does contain energy that low, 20Hz is at the threshold of human hearing, so you might feel it but you wouldn't actually hear it as a tone.

Another advantage is the dipolar off-axis nulls, which tends not to excite room resonances that can render the bass sluggish and inarticulate, or even produce the dreaded "one note boom".

Dipolar bass isn't very efficient so I always recommend a pair of RiPol's rather than just one, and they don't pressurize the room in the same way as conventional subs. Not everyone likes them because they don't hit you in the chest like a sledge hammer... but I LOVE them. Their sound arises from nowhere and recedes back to nowhere-- wonderful for jazz, and simply the cleanest, most unobtrusive bass I've ever experienced.

Modak Akustik in Germany markets a similar RiPol sub which uses the same Peerless SLS woofers.
A Product Review of the Modal Akustik sub by 6-Moons Audio includes the following statement:

"For music-first listeners who prioritize speed, articulation and enunciated clarity, it's the long awaited messiah."
Srajan Ebaen, 6moons.com

I share that sentiment about mine too, and I will share my drawing & parts list with anyone who asks-- just PM me an email address.
************************************************** ************************************

Build info:

The woofers are Peerless SLS 12's and the cabs are 3/4 red oak plywood with oak edge members inserted and rounded over, which looks much better than exposed plywood edges but is a LOT MORE work. The center section is solid brown oak, indexed to the cabs with oak dowel pins. The cabs and center section are held together with all-thread rods and cap nuts.

The light-colored oak cabs were oil stained with a mixture of 1/3 golden oak, 2/3 natural, and a bit of powdered turmeric root added for a yellowish tint. The darker oak center section was oil stained with red oak, for contrast. The finish is sprayed-on coats of clear satin polyurethane with a #320 sanding step between coats.

Enjoy!
Charlie in Savannah, GA

Photo 1: Rabbited plywood cab pieces self-locate and bank together for gluing.

Below: Cab glued and clamped.

Below: Cab edges were notched on the table saw, to accept the red oak edge members.

Below: Mitered edge members installed and held in place with painter's tape while the glue sets.

Below: Cab with edge members installed and sanded flush.

Below: The 3-piece assembly details.

Below: Both subs temporarily assembled and edges rounded over to 3/8 radius on router table.

Below: Details stained and spray coated with clear satin polyurethane.

Below: Completed subs with woofers installed.

Below: Back view

 

[TLS Guy]

Enjoy!
Charlie in Savannah, GA



Below: Cab glued and clamped.
View attachment 63163

Below: Cab edges were notched on the table saw, to accept the red oak edge members.
View attachment 63164

Below: Mitered edge members installed and held in place with painter's tape while the glue sets.
View attachment 63165

Below: Cab with edge members installed and sanded flush.
View attachment 63166

Below: The 3-piece assembly details.
View attachment 63167

Below: Both subs temporarily assembled and edges rounded over to 3/8 radius on router table.
View attachment 63168

Below: Details stained and spray coated with clear satin polyurethane.
View attachment 63169

Below: Completed subs with woofers installed.
View attachment 63171

Below: Back view
View attachment 63172

What is the proof that it lowers driver Fs? Also what is the mathematical model that shows that it does? Count me skeptical.

 

[jazzman53]

What is the proof that it lowers driver Fs? Also what is the mathematical model that shows that it does? Count me skeptical.

I don't blame you for being skeptical. It's a bit hard to get one's head around at first, but if you think about if for a while it starts to makes sense.

Unfortunately I can't provide a model. I heard some years ago that Martin J. King (a well known math wizard / transmission-line speaker modeler) was working one but I don't know what became of that, or whether anyone has created a usable model. I'm betting one exists.

I also heard that the patent holder (Axel Ridtahler) has a model but he's not sharing it per agreement with Modal Akustik who he licensed to market a Ripol sub he collaborated on (they don't want any competition).

What is available are comparison measurements reported on several websites. Here's one: https://www.lautsprechershop.de/hifi/aka_sub_sonder_en.htm

There are also comparison measurements presented in Ridtahler's patent here: https://patents.google.com/patent/DE19830947C2/de

The patent is in German but I was able to copy/paste excerpts into Google Translate until I found the section reporting measurements of a woofer with an FS of 25Hz, resonating at 40Hz in a 150 liter sealed box, versus the same woofer resonating at 17.5Hz in Ridtahler's enclosure.

Absent a model, I'll give you my uneducated country boy take on it:

- On the backside, the woofers are not constrained by a trapped air volume, as they would be in a sealed box where pressurization would push the resonance upward. Rather, the rear chambers are large and open on one end; allowing very little pressure buildup.

- On the front side we have two opposing woofers in close proximity firing directly into each other (push/push), in a much smaller chamber that's open on one end.

I can see how each woofer would impact the other, and how the air pressure and density within the small chamber must be above baseline atmospheric--- and to the extent that it is, would impede the woofer from resonating as it would in free air, and thereby force its resonance downward as reported in the patent, and as supported elsewhere by comparative physical measurements.

There consensus among the Ripol gurus one the DIY Audio Forum ("Rudolf", "Calvin" and others) is that reducing the area of the front chamber opening forces the resonance downward-- until a point of diminishing returns is reached where the efficiency loss becomes unacceptable.

The general guidelines dictate that the area of the front chamber opening should be 1/3 to 1/4 the woofers' combined s/d (piston area), and as follows:

For woofers with X-max >10mm, target 1/3 area
For woofers with X-max <10mm, target 1/4 area

 

[TLS Guy]

I don't blame you for being skeptical. It's a bit hard to get one's head around at first, but if you think about if for a while it starts to makes sense.

Unfortunately I can't provide a model. I heard some years ago that Martin J. King (a well known math wizard / transmission-line speaker modeler) was working one but I don't know what became of that, or whether anyone has created a usable model. I'm betting one exists.

I also heard that the patent holder (Axel Ridtahler) has a model but he's not sharing it because he's marketing a sub licensed to Modal Akustik.

What is available are comparison measurements reported on several websites. Here's one: https://www.lautsprechershop.de/hifi/aka_sub_sonder_en.htm

There are also comparison measurements presented in Ridtahler's patent here: https://patents.google.com/patent/DE19830947C2/de

The patent is in German but I was able to copy/paste excerpts into Google Translate until I found the section reporting measurements of a woofer with an FS of 25Hz, resonating at 40Hz in a 150 liter sealed box, versus the same woofer resonating at 17.5Hz in Ridtahler's enclosure.

Absent a model, I'll give you my uneducated country boy take on it:

- On the backside, the woofers are not constrained by a trapped air volume as in a sealed box which would otherwise allow pressurization that would push their resonance upward. Rather, the rear chambers are fairly large and open on the back so very little pressure buildup could be expected.

- On the front side we have two opposing woofers in close proximity firing directly into each other (push/push), in a much smaller chamber that's open on one end.

I can easily imagine how under those conditions the impact each woofer has on the other, and that the air pressure and density within the front chamber must be above baseline atmospheric, and to the extent that it is, presents a resistance that would prevent the woofer from resonating as it would in free air, but would force its resonance downward, as reported in the patent, and as supported elsewhere by comparison measurements.

There consensus among the Ripol gurus one the DIY Audio Forum (Rudolf & Calvin, specifically) is that reducing area of the front chamber opening forces the resonance ever further downward until a point of diminishing returns is reached where the efficiency loss becomes unacceptable.

The general guidelines are that the area of the front chamber opening can be 1/3 to 1/4 the woofers' combined s/d (piston area).

For woofers with X-max >10mm, target 1/3 area
For woofers with X-max <10mm, target 1/4 area

The impedance peak has lowered, but that is not the same or equivalent to Fs. The FR suggests that the Fs has NOT been lowered.

A speaker is a spring. And two identical springs will have the same resonance unless you change the weight on the end of the spring or the Young's modulus of the spring. I see nothing about that design that would change either. The rear of he cones are open to free air, and the fronts widely open to the air one side. The will give it output on the rear of the cones with a null at the back and front. There will be output on the open side from the front of the cones, but that output will be out of phase with the rear of the cones, and therefore there will be cancellation from the out of phase rear part of the cone allowing a meagre output from that side port. So you will have a figure of 8 output from the rear of the cones, and an out of phase output from the opening in the middle.

The whole thing is extremely inefficient and wasteful of power. It sounds to me like one of many crazy contraptions that German Audio has specialized in for years.

 

[jazzman53]

I'm not saying you're wrong-- just that a lot of knowledgeable people disagree with you. And aside from your point about poor efficiency (valid but no worse than other dipoles I've heard), I'm not aware of any published research that supports your conclusion.

I'm no expert myself, so I can't be the arbiter. I rely on research and guidelines from sources I respect and pass it on if it seems helpful.

For me the bottom line is what I hear when I crank up some tunes, and I have no complaints. I would love to invite you over for a listen!

 

[lovinthehd]

I'm not saying you're wrong-- just that a lot of knowledgeable people disagree with you. And aside from your point about poor efficiency (valid but no worse than other dipoles I've heard), I'm not aware of any published research that supports your conclusion.

I'm no expert myself, so I rely on research and guidelines from sources I respect and pass it on if it seems helpful.

For me the bottom line is what I hear when I crank up some tunes, and I have no complaints. I would love to invite you over for a listen!

You'll rarely meet someone more impressed by his own opinions on audio and speakers than TLS Not sure why your method is particularly an advantage otoh.

 

[jazzman53]

You'll rarely meet someone more impressed by his own opinions on audio and speakers than TLS Not sure why your method is particularly an advantage otoh.

I wish I could claim it but it's not "my" method That honor belongs to the inventor.

In fact, I must confess that I've always struggled with math, and my knowledge of electronics is only rudimentary at best (I can recap an amp but I couldn't fix one if it doesn't play).

The Ripol's are pretty amazing though. I can't believe how LOW they go. It doesn't seem logical that a dipole (with it's accompanying phase cancellation), could go so low.

TLS's point about efficiency is correct though. My rule of thumb for dipoles (in the lower octaves) is that you need about 4X the piston area to get equivalent volume. Their sound is their redemption.

The Ripol's are the only subs I've found that blend seamlessly with my dipole main speakers, which are hybrid wire-stator electrostats with an OB mid bass woofer on a modified H-baffle.

BTW; I will be posting a thread on my homebuilt ESL's at some point. I've been perfecting the design for 15 years and I consider it my masterpiece.

 

[TLS Guy]

I wish I could claim it but it's not "my" method That honor belongs to the inventor.

In fact, I must confess that I've always struggled with math, and my knowledge of electronics is only rudimentary at best (I can recap an amp but I couldn't fix one if it doesn't play).

The Ripol's are pretty amazing though. I can't believe how LOW they go. It doesn't seem logical that a dipole (with it's accompanying phase cancellation), could go so low.

TLS's point about efficiency is correct though. My rule of thumb for dipoles (in the lower octaves) is that you need about 4X the piston area to get equivalent volume. Their sound is their redemption.

The Ripol's are the only subs I've found that blend seamlessly with my dipole main speakers, which are hybrid wire-stator electrostats with an OB mid bass woofer on a modified H-baffle.

BTW; I will be posting a thread on my homebuilt ESL's at some point. I've been perfecting the design for 15 years and I consider it my masterpiece.

I can see why it might well blend well with electrostatic speakers. As the sort of polar pattern I describe would indeed be an advantage in that situation.

However, I do not see why the design should lower Fs of the drivers. That would be easy for you to put the the test with a DATS woofer tester. All you would have to do, is measure one driver and then two in that enclosure of yours. That would put the theory to the test and decide the issue definitively as to whether Fs is lowered. I have never seen a claim for any design that claims lowering of driver Fs, and see no reason I can think of why that one would.

So if I had built that unusual enclosure, I would be doing a thorough investigation of the physics of the situation.

 

[jazzman53]

I can see why it might well blend well with electrostatic speakers. As the sort of polar pattern I describe would indeed be an advantage in that situation.

However, I do not see why the design should lower Fs of the drivers. That would be easy for you to put the the test with a DATS woofer tester. All you would have to do, is measure one driver and then two in that enclosure of yours. That would put the theory to the test and decide the issue definitively as to whether Fs is lowered. I have never seen a claim for any design that claims lowering of driver Fs, and see no reason I can think of why that one would.

So if I had built that unusual enclosure, I would be doing a thorough investigation of the physics of the situation.

Hello again TLS Guy,
I don't consider myself having the smarts or testing gear to resolve this question.

I have been pondering your assertion that the impedance peak doesn't necessarily correlate to the woofer's resonance. It does in free air, according to the Eminence speaker website, but I don't know whether that necessarily holds true in an enclosure. Yet I still can't yet imagine a cause for an electrical impedance peak other than the woofer's resonance peak.

I re-read Ridtahler's patent again and it definitely states that "resonance" was measured. The Google translation doesn't mention "impedance", but also doesn't specify how the resonance was measured, so I can't rule out an inference from an impedance peak.

After reading your post, I've also been pondering whether a box resonance could affect a measured electrical impedance in the driver (separately from the woofer resonance). So far I'm unable to mentally connect two such events, or isolate a cause other than the resonance itself.

I may take a while to post again because I'll be busy for the next few days packing and traveling to CarverFest (a two week hi-fi drunk-fest in the NC Mountains with a bunch of Carver gear owners). While I'm there I'll ponder some more and pick some brains and hope for enlightenment. Perhaps the Bobfather (Bob Carver) will weigh in (at the risk of further confusion on my part).

Charlie

 

[TLS Guy]

Hello again TLS Guy,
I don't consider myself having the smarts or testing gear to resolve this question.

I have been pondering your assertion that the impedance peak doesn't necessarily correlate to the woofer's resonance. It does in free air, according to the Eminence speaker website, but I don't know whether that necessarily holds true in an enclosure. Yet I still can't yet imagine a cause for an electrical impedance peak other than the woofer's resonance peak.

I re-read Ridtahler's patent again and it definitely states that "resonance" was measured. The Google translation doesn't mention "impedance", but also doesn't specify how the resonance was measured, so I can't rule out an inference from an impedance peak.

After reading your post, I've also been pondering whether a box resonance could affect a measured electrical impedance in the driver (separately from the woofer resonance). So far I'm unable to mentally connect two such events, or isolate a cause other than the resonance itself.

I may take a while to post again because I'll be busy for the next few days packing and traveling to CarverFest (a two week hi-fi drunk-fest in the NC Mountains with a bunch of Carver gear owners). While I'm there I'll ponder some more and pick some brains and hope for enlightenment. Perhaps the Bobfather (Bob Carver) will weigh in (at the risk of further confusion on my part).

Charlie

The only scientific way to resolve this, is to test the hypothesis, and that means measurement. Otherwise it is just all hot air. The DATS units are not expensive and easy to use. So measurement of Fs of two drivers in and out of the enclosure is what is required. This is especially important these days, as people claim all sorts of untrue nonsense. At this time I am high skeptical that arrangement lowers Fs of the drivers. The rest of it I can understand. However benefits at those low frequencies of that polar response is mute due to frequencies rapidly becoming omnidirectional.

 

[jazzman53]

Taking a short break from packing for my trip.

I continue to ponder the relationship between the woofer's measured impedance peak and it's resonance. Eminence states on their website that they determine the woofer's resonance by the impedance peak, and my understanding is that this is the industry standard method.

I'm thinking that the impedance would have to include the mechanical resistance of the air and suspension, and also the back EMF generated by the voice coil cutting thru the magnet's flux. The back EMF specifically would have to peak with the resonance peak, as both reflect the cone's motion.

Absent convincing evidence to the contrary, the only thing that makes any sense to me at this point, is that the mechanical resistance and back EMF correlate to the resonance, and these are generating the impedance peak, so the resonance and impedance peaks must correlate exactly.

 

[TLS Guy]

Taking a short break from packing for my trip.

I continue to ponder the relationship between the woofer's measured impedance peak and it's resonance. Eminence states on their website that they determine the woofer's resonance by the impedance peak, and my understanding is that this is the industry standard method.

I'm thinking that the impedance would have to include the mechanical resistance of the air and suspension, and also the back EMF generated by the voice coil cutting thru the magnet's flux. The back EMF specifically would have to peak with the resonance peak, as both reflect the cone's motion.

Absent convincing evidence to the contrary, the only thing that makes any sense to me at this point, is that the mechanical resistance and back EMF correlate to the resonance, and these are generating the impedance peak, so the resonance and impedance peaks must correlate exactly.

With all those openings, I cant see that the loading is significantly different from free air. It seems to me that most likely the physics is open air rather than tuned. There is nothing about that design that would make it a tuned design. Adding drivers does not change the Fs ever. In fact in any tuned design the only T/S parameter that changes is VAS, and the driver power, and radiating area. They are both doubled obviously and resistance and inductance are halved assuming a parallel connection.

 

[jazzman53]

It seems to me the predominant component of the impedance peak would be back EMF generated by movement of the voice coil. If that is so, it would be consistent with a resonance peak.

If the resonance did not generate the impedance peak, the we have to explain what did.