Below is a response that Brian Ding from Rythmik audio posted on this thread
http://www.audiocircle.com/index.php?topic=83870.msg811956;topicseen#newI didn't want to pollute that thread anymore then it already has so I figured I would repost Brian's response so we can continue here. Danny, if this is a problem, please delete the thread. I didn't think it would be a problem since you sell Rythmik products and were part of the discussion in the other thread.
BTW-Thanks for the reply Brian. I learned a bunch!
I highlighted what really sparked my interest...
============================================================================================================
I am not trying to sell servo here. But this reply is more towards the questions that some members has asked. If I don't answer here. They would continue to carry those misconception and spread everywhere else. So just bear with me for a moment.
First and formost, I respect Don Keele. His work on motor structure analysis is classic. I still have his paper in my paper collection. But two things are very imporant, first, just because someone is working in the field of acoustics does not mean he is authority on everything in that field. It is like doctors. You go to different doctors for different problems. That is why we have expert making wrong remarks like what Dr. Bose once said something to the effect that "in order to reproduce the sound of an instrument faithfully, the size of the speaker needs to be of similar dimension size. Or something like the citation Bob Carver used to get his patent approved and made the point that his compact sub with 2kW amp solves a problem that an audio textbook (I forgot which one) said cannot be done. For those ever apply for a patent know that solving a problem that "experts" deemed unsolvable is a strong ground for granting a patent. The patent was later disputed and revoked beause the author of textbook was wrong. I have been asked many times if my background is acoustics or electrical/electronics. Servo happens to be a field that requires knowledge from both fields. Second, it needs the right person with the right skils to make servo works. SEAS had been working several years on servo. One time I was introduced to their staff engineer in winter CES and he asked me if mine works because they have problem getting theirs t work. The trick is my patent uses a combination of sensing feedback and current feedback. It is the current feedback that solves many previous encountered problems.
Now answers and clarification:
1) "The servo can limit cone movement". A lot of customers during their first call told me servo could prevent bottoming. That is incorrect. I was surprised that they make comments like this. Amplifier does not have un-limited power. Once the amp is into clipping, it behaves the same as nonservo subs, that is, the feedback has no effect at all because there is no room for the amp to make further correction. So in servo subs from Paradigm and Velodyne, it is the limiter that prevents the amp from getting into clipping, and hence bottoming. But any nonservo amp can also add a limiter to prevent bottoming. Anyone can buy a 10 year old used Velodyne servo and bypass its limiter and see how the servo design can prevent bottoming. The answer is servo does not prevent the driver from bottoming.
2) "but it can't correct for things like low motor strength suspension limiting". Very true. That is why we use drivers with good motor size and uses deep basket to support the excursion that the driver would not bottom out during moderate to heavy amp clipping. Very interesting that you touched two subjects: motor strength and suspension. First motor strength. When I worked with NCA in north California (in early 90's), Glen told me a story that this customer designed a driver with huge motor structure, but later on found the high BL value (a trait from over-designed motor) has made it unsuitable for sealed sub application. Guess how the company solved the problem? Widen the gap. Instead of reducing the motor size, they chose to defeat the purpose of having a larger motor by widening the magnetic gap. In short, larger motor is often related to wide magnetic gap. In the era of Don Keele, that would be an absolutely no no. Now second subject: suspension. Spider and surround both have distortion and because it is the only part that actually change its shape during cone movement, they contributes a lot of dynamic distortion that measured like noise. Most peopl don't understand the difference between noise and distortion, not to mention dumb measurement equipments that require engineers to interpret the results. In Klippel test, it brings in another dimension time (which I will also touch later).
3)"A servo cannot correct non-linear distortions, like the type you'd see with poor Klippel BL measurements." But the servo do reduce the distortion. yet you do bring up a good point. Here is the missing link. We actually have two BL measuement curves: one for driver coil and one for sensing coil. They don't look the same at all. Our sensing coil is far more linear than driver coil because it is ultra overhang design. I will post one sensing coil curve when I return to office on Monday. But in short, it is the technique that does not affect the efficiency of driver coil. All techniques linearizing BL curve achieve the linearity by sacrificing the efficiency. If one looks at a normal BL curve from Klippel, it resembles a upside down U. The middle peak representing the max BL value location. What a linearization voice coil technique do is it will bring down the value of the peak BL value so that it can align with lower Bl values at say 10mm or 20mm excursion point. Our servo technique does not make that type of compromise. I would not elaborate further as I am just trying to make me point here in defense of the critics of servo design.
4) Sound improvement from servo vs nonservo. I personally compare sound from servo and nonservo over and over again. I only have one customer who bought a nonservo kit and servo kit at the same time and his comment was the nonservo is very good, but the servo is the sweet deal.
He was so excited and asked me when I can extend extend the technology to midrange drivers. I wish I can have 48 hours a day to finish that project. My personal experience is the servo improves the spatial clue of the sound stage. On the other hand, the nonservo version of our drivers tend to have more noise that correlated only occasionally to the source. Listener cannot really connect that type of noise/distortion to the original source. so coherence is the key. with servo, instruments no longer warble or with ill-defined image. When I switch from 28hz to 14hz extension, the sound stage becomes deeper, wider, and taller. The sound is so coherent that it conveys the emotion of the performers in totality, giving me goosebumps. Now back to the coherence in BL plots from Klippel. For those who are familiar with Klipple system, the BL curve was a "curve fit" result from a test window. If one move the window in the viewer program, it gives a different BL curve. In short, when we are into finer and finer granularity of analysis, everything is more or less related t statistical behavior. Let us look at fast Fourier transformation (FFT), it is a result of averaging over a window. However, our hearing don't do averaging the way those measurement instruments do. That is why there seems to be a disconnect from measurement results to actual listening experience. We cannot control our hearing like how we manipulate testing results. not to mention limitation such as how we measure the depth of sound stage with instrument. We are still exploring the direction not formally investigated and yet we have people constantly quote findings from 40 years. The latter completely undermines any new findings. In short, servo is the technique to improve coherence in bass.
Lastly, sorry for the long post. It is simply to hard to me to read tthe critics without answring them, in particular when I am in a hotel in Las Vegas that my wife is waiting for me to get out of Hotel room.Plase just thnk of as a point of record.
