This is Brian from Rythmik. I have to say after one hears good bass, it is hard to go back without it.
John himself has gone through several subs and he is pretty happy with bass from directservo. Here I would like to give a sensible explanation of why it can sound better than other subs.
What servo achieves is something that would take a lot of physical asset (such as larger magnet or lighter cone mass) to do. For instance, very long time ago, there was an article in "speaker builder" magazine trying to answer why with the same motor, a midrange driver would sound better than a similar midbass driver from the same manufacturer (at that time it is a common practice of doing that from European makers). The reason is simple, the midrange driver has the characteristic that the midbass driver has to take away: 1) low moving mass (sorry midbass driver's cone needs to be heavy in order to make fo value, also the voice coil needs to be longer for the required excursion), 2) low inductance (sorry midbass drivers needs a lot more turns of voice coil to make the BL value), and 3) lower Qts value (with the heavier cone of midbass driver, the Qts cannot be low). Qts is actually an indicator about how well the cone movement can be controlled. Anyway, in subwoofer, not only all of these rules are violated, they are doubled, or even tripled. Eventually all of these broil down to cone control. First, high moving mass takes larger magnet to control (keep in mind, all analog signal is continuous, if the cone is in the wrong position because the distortion at time x, how long does it take for the cone to come back to correct position at the next moment?, or the system should just inherit that error forever? And if it does recover, what is the recovery time. Servo has way faster recovery time). Second, high inductance has a similar effect as high moving mass. What servo does is it tries to gain the control back so that the cone movement follows what the amplifier asks, and no others. Why would the cone not listen to the amplifier? For instance, the reflection from enclosure within. Sound is generated from both sides of the cone, and we shouldn't assume the energy from the back of the cone just magically disappears. That energy actually pushs the cone and causes the second, third.... transmission. The purpose of the servo is to hold the cone at bay with respect to these reflections, and at the same time follow the signal from the amplifier to move the cone with the exact speed that it wants. As soon as the sensor sense some cone movement in the wrong direction, the feedback will cause the amplifier to output a certain voltage to counteract the external force. How much voltage is will output depends on the servo loop design.
When people discuss servo, they cannot avoid talking about Velodyne and other accelerometer based servo approaches. We uses sensing coil. The main difference is the accelerometer achieves high cone mass in equivalent circuit, and our sensing coil approach achieves higher damping. This is best explained with a push test on the cone when the system is powered up and the driver is in free air. Actually you cannot do that with Velodyne because the sub will destroy itself if it is not properly sealed. When you push our cone, one would get a sensation of oil pot. That is very similar to those oil pot sitting next to a tone arm. On the other hand, accelerometer based sub will give you the sense of a heavy brick. We all know which is better. When the reflection hits our cone, our cone moves less than nonservo cone, and during this process, it dissipate the energy, and reflection back a smaller energy, etc.
I think this is a 3000 ft view of what servo does. BTW, our servo uses only two resistors and 1 cap between the sensor output and the amplifier feedback summing point. Even nonservo amplifier needs to use these many components. So our approach does not have any catch. It has no active component on the feedback path of sensor. Some people call it passive servo, and I prefer to call it direct coupling servo. There is no servo hiss sound, there is very little hum noise (no hum noise if the system grounding is good), there is almost zero coloration.
Brian Ding
Rythmik Audio
