Scan-Speak releasing new extreme drivers with Eliptical voice-coils

[Flint]

What started as a hearty joke about innovating in loudspeaker transducer design has resulted in a pair of drivers, a tweeter and a mid-woofer, which have elliptical voice-coils and shockingly low distortion as a result.

http://www.scan-speak.dk/?page_id=638

As I am primed to start designing my next set of speakers, I have to wait for these to ship and hear them in person. The specs are impressive, including higher sensitivity since less dampening is required to reduce resonances.

Yummy!!!

 

[Flint]

 

[Dentman]

At a glance you would think this design would yield a few issues that they must have worked hard to overcome. I guess you feel good about the results.

 

[MakeMineVinyl]

I don't see how an elliptical shape to the voice coil could in and of itself reduce resonances. Maybe I'm missing something?

 

[Flint]

I don't see how an elliptical shape to the voice coil could in and of itself reduce resonances. Maybe I'm missing something?

According to Scan-Speak and some well-respected transducer design consultants who are commenting on them, the elliptical voice coil creates significantly more resonances along the speaker diaphragm due to the varying distance from the voice coil former to the edge (cone) or center (dome). However, the amplitude of those resonances are so vastly reduced that they reach the point of being audibly and measurably insignificant. As such the diaphragm requires less mass (the common method for reducing diaphragm resonance distortions is more mass from dampening or increasing the stiffness, and so on). Less mass in the diaphragm means a lighter mass to motor and effectively higher sensitivity.

Third party lab testing and real-world listening tests are not yet taking place, so this is all claims from the designers and those close to the product developments.

Scan-Speak used their proven and successful motor design for the magnetic structure, and applied many of their other trademark design principles. On story I heard was that they made the test midrange with a standard pressed paper cone and got amazing results. They then decided to apply their top of the line sliced cone approach to reduce distortion (at a cost to increasing the mass) and the difference in distortion were negligible, but the sensitivity was much lower. In the end they felt launching a new "flagship" without their previous flagship tech was the approach they took. They unintentionally solved an issue which they had been famous for solving with their complex sliced-cone approach. Pretty cool.

What I am most interested in seeing are 3D polar plots of their dispersion patterns, not just vertical and horizontal.

 

[MakeMineVinyl]

...the amplitude of those resonances are so vastly reduced that they reach the point of being audibly and measurably insignificant

Hmmm, I guess we'll see. Generally in cases like this with resonant masses, you can go for either one or a few large resonant peaks or many smaller ones. The catch though is that although the total resonant energy may be less, it can sound worse. It is kind of like having a large amount of 2nd harmonic amplifier distortion verses having much smaller amounts of very much higher orders of distortion. In the distortion case, the large, single 2nd harmonic is much less offensive to the ear than the smaller level high order distortion components. Amplifier manufacturers originally said that their new solid state designs with "low" but high order distortion was "insignificant". That gave rise to a whole generation of terrible sounding amps in the 70s.

Time will tell.

 

[Flint]

Hmmm, I guess we'll see. Generally in cases like this with resonant masses, you can go for either one or a few large resonant peaks or many smaller ones. The catch though is that although the total resonant energy may be less, it can sound worse. It is kind of like having a large amount of 2nd harmonic amplifier distortion verses having much smaller amounts of very much higher orders of distortion. In the distortion case, the large, single 2nd harmonic is much less offensive to the ear than the smaller level high order distortion components. Amplifier manufacturers originally said that their new solid state designs with "low" but high order distortion was "insignificant". That gave rise to a whole generation of terrible sounding amps in the 70s.

Time will tell.

I don't disagree with your concerns. I would like to hear them and see 3rd party testing and listening reviews. But I love advances in design concepts, and this is one I am watching close. Just like I am keen to get quality time with a few Beryllium dome tweeters.

 

[MakeMineVinyl]

I'm reminded of the "ear shaped" speaker driver that Yamaha came up with in the late 60s:

http://edsgoodstuff.com/edscart/index.php?main_page=product_info&products_id=1087

I never heard one, but the rationale I gather was that since it was shaped like an "ear" it would sound better. Didn't change the course of the industry, obviously.

 

[Botch]

Maybe if we made them dog-ear shaped, we could hear to 28.000 Hz?

 

[Flint]

I remember when time alignment became the popular issue in speaker design and companies like Technics and Sony made flat diaphragm speakers so the time arrival from each speaker was "perfect". Others designed their speakers so the tweeters or tweeters and midrange drives were offset so that when the speakers were aimed forward the acoustic center of each driver was aligned for the listener between the two speakers. While this is an important issue, it seems it is not as often addressed as the most important, and only Phase Technology (as far as I know) are still making flat front diaphragm is flat.

 

[MakeMineVinyl]

Time alignment was a well known issue as far back as the 1930s in movie theater speakers. It was solved in speakers like the Shearer horn by using a short bass horn in front of the woofer, thus placing the woofer and HF compression driver in the same physical plane. This is used in speakers like my Altec A7-s.

Horn speakers like the Klipschorn had massive time offsets between the bass horn and midrange/tweeter horns which can now only be solved by digital delay.

 

[Flint]

I didn't mean to say time alignment didn't matter, clearly it does. I was just commenting on the fun innovations released during a brief period when it was the only major issue being discussed in the market, and how many of those innovations never took off.