Toshiba's use of aux driver for correction

[JRT]

I am posting this because I thought some here might find the underlying idea perhaps a little thought provoking after cutting through the amusing translated marketing BS. For example, consider the use of secondary transducer(s) to cancel sound transmitted in the enclosure (structure borne acoustics) and related modal resonances excited by same. ...or interfering with sound in the enclosed air, maybe reducing delayed reflections passing back through the cone, maybe interfering with spectra that would otherwise excite poorly damped resonances that are otherwise difficult to correct. ...etc.

Here is the article, below (link to source):

New Speaker System Clarifies Sound by Canceling Its Own Noise

17 March 2009

Toshiba Corp developed a new technology to accurately reproduce sound by using an auxiliary speaker.

The company developed the "accurate sound reproduction technology" in collaboration with Shiro Ise, an associate professor at the Department of Urban and Environmental Engineering, Kyoto University Graduate School of Engineering, and Actimo Ltd.

The technology uses sound from an auxiliary speaker to eliminate noise (sound other than the original sound) generated from the main speaker. The noise results from insufficient rigidity of a chassis and supporting parts. And the new technology cancels it out with opposite-phase sound waves generated from the auxiliary speaker attached to the main speaker, enabling to exactly reproduce the original sound, according to Toshiba.

In general, about 40% of the sound from a speaker is sound other than the original sound, and the percentage increases in the case of an ABS chassis, a thin chassis that is normally used in small and thin AV equipment, according to the company.

This time, Toshiba confirmed that such non-original sound can be eliminated by more than 90% in an experiment using equipment incorporating this technology. The company plans to conduct research and development for introducing this technology in its products.

Toshiba and its partners will announce this technology at the 2009 Spring Meeting of the Acoustical Society of Japan, which takes place from March 17, 2009, in Tokyo.

Yukiko Kanoh, Nikkei Electronics

[brianpowers27]

Sounds good. Think noise cancelling headphones. I wonder where they obtain the signal for the noise. Is this a tiny mic placed near the speaker?

[cole1]

Take this pill for headaches. Side effects include vomiting, diahrea, fatigue and heartburn.

[Pete Schumacher ®]

Take this pill for headaches. Side effects include vomiting, diahrea, fatigue and heartburn.

[Ryan_M]

From their little write up it sounds like this is an attempt the reduce the demons of "thin wall ABS" speaker boxes. Who here builds their cabs out of thin ABS? I suppose there's a profit margin benefit for a mass producer to implement an additional speaker and associated amp vs. the cost and shipping weight of heavy thick walled cabs.

[romanbednarek]

How do they get that number of 40% of the sound doesn't come from the driver directly? That may have been true for the first speaker cabinets that I built out of poplar and plywood without any bracing but it certainly isn't true for anything I've built since.

It just seems to me that this theory of improvement is like taking the first straw house that the three little pigs built and coating it with glue versus just building a brick house from the start.

Now what we would be interested in is a way to use this technology in a solid speaker enclosure, not to remove wall resonances but internal resonances but I find this difficult to fathom because the location of the "cancellation" driver would be critical and it seems like you would need more than one for it to be effective. Who knows what unwanted side-effects might arise as well.

[JRT]

Sounds good. Think noise cancelling headphones. I wonder where they obtain the signal for the noise. Is this a tiny mic placed near the speaker?

They might be using ANC technology, similar to that used in headphones. Toshiba has thrown money into using that technology to quiet HVAC systems.

But they may be simply using panel exciters inside the box. The modal response is predictable, so they shouldn't need a microphone in a feedback loop, and the associated acoustic delay might create more problems than it solves.

[Pete Schumacher ®]

They might be using ANC technology, similar to that used in headphones. Toshiba has thrown money into using that technology to quiet HVAC systems.

But they may be simply using panel exciters inside the box. The modal response is predictable, so they shouldn't need a microphone in a feedback loop, and the associated acoustic delay might create more problems than it solves.

accelerometers on the panels right next to the transducers could be used to provide negative feedback, allowing excitation signals to keep the accelerometer outputs to near zero.

[johnnail]

From their little write up it sounds like this is an attempt the reduce the demons of "thin wall ABS" speaker boxes. Who here builds their cabs out of thin ABS? I suppose there's a profit margin benefit for a mass producer to implement an additional speaker and associated amp vs. the cost and shipping weight of heavy thick walled cabs.

Well, if you DID have thin-walled ABS housings, couldn't you quiet them down by simply lining them with some sort of bitumenous damping material on the inside? Like some of that asphalt/loaded vinyl stuff that PE used to carry but doesn't anymore? It was 1/8" thick, but added a lot of mass and did a really good job of deadening a panel. Just as good as the Dynamat stuff, but at a 10cents on the dollar cost. I did that to a couple pairs of speakers that I got from Best Buy on sale a few years ago...though they had aluminum housings....the Celestion AVP302 series on glass baffles.....had to do some serious redesigning of the crossover to get the balance just right...they were way, way forward, but once I got the crossover "adjusted" and applied some of the PE damping to the walls, they sounded really sweet.

John

[JRT]

accelerometers on the panels right next to the transducers could be used to provide negative feedback, allowing excitation signals to keep the accelerometer outputs to near zero.

Would the feedback loop be fast enough? Similar motional feedback techniques have been used in some servo subs.

(link to PE's $15 panel exciter, p/n 300-375)

[Pete Schumacher ®]

Would the feedback loop be fast enough? Similar motional feedback techniques have been used in some servo subs.

(link to PE's $15 panel exciter, p/n 300-375)

Sure. If the tactile transducer has a suitably wide bandwidth, it could be used to oppose motion in the panel created by the drivers. And since most resonances are limited to the midrange at best, you'd only need to dampen up to maybe 1KHz to take care of most of the junk.