Getting lucky

Yesterday's post was filled with excitement for the show and setup day. Then reality strikes. I tend to forget, year after year, what a pain in the keester it actually is. What awful rooms we're given in hotels, never once designed to house a stereo system, of the multiple trips to the hardware store, the last minute fretting over this and that. That in the end it rarely turns out as good as your highest expectations. But magically, somehow, it seems to work out. The system sounds good, Sprout Lounge looks amazing. Come by and meet Ted Smith, Bascom King, Arnie Nudell and Scott McGowan if you're at RMAF. And, speaking of Bascom King, the designer of our new power amplifier and one of the few experts on servo woofer systems in the world, he spent some time setting the record straight on servos. Consider this a free lesson for anyone interested in learning about making better sounding subwoofers. This was very generous of Bascom.

"I feel compelled to put out the truth about acceleration feedback! In a closed box, or a ported box for that matter, the near field acoustic pressure (what a microphone would measure) of the woofer cone is related directly to the acceleration of that cone. I have attached a little sketch of the relationship of acceleration, velocity, and displacement as how they vary with frequency. Then the second diagram shows the acceleration with frequency of a closed box woofer with a resonance of about 60 Hz with high frequency compensation necessary for loop stability with several values of feedback applied. (Click on the diagram to enlarge it).

Lots of things can be learned by the diagrams. For instance, with no feedback or the first diagram case, below resonance, the displacement becomes flat and constant. Another thing that is discoverable, that for constant acceleration, the displacement must rise at 12 dB/octave over the whole range. Putting the acceleration feedback in, causes that displacement to rise at 12 dB/octave below resonance down to where the loop runs out of gain as shown in the bottom figure.

So bottom line - if ones wants to flatten frequency response, use the accelerometer as the sensor. If one wants to use velocity, differentiate it (multiply it by a + dB/octave function, ie. a high pass first order filter with a cutoff way above the band of interest - makes for HF noise in the circuit) to make it look like acceleration. If one wants to use a position sensor, double differentiate it (making it more noise prone yet).

I have done all of these things at one time or another and the accelerometer is the best way to do it. Of course there are ways to use the other sensors and not fully turn their output in to acceleration and then equalize the response ahead of the feedback loop part."

Thanks Bascom. While I am certain this is more technical than most of my readers want, it's rare you get a real lesson from the master himself.

And, for those subwoofer manufacturers wishing to add servo control to their woofers, you just got lucky getting this info. While all seemingly technical, the entire control circuit is very low cost, consisting of a couple of op amps and a piezo sensor. My hope is that more manufacturers and designers will apply this technology; simple enough to implement. But oh my, what a wonderful improvement the world of home music reproduction gets, the more models of subs that use this technology, first introduced in 1968.