Posted by Al Sekela on November 21, 2002 at 12:54:34:
In Reply to: Bi-Amplifying M1 posted by Gordon Meltzer on November 06, 2002 at 19:38:57:
Gordon,
I own Magnepans, not Sound Labs, and visit this site to keep up with Roy's misadventures with his A3s. However, I could not help noticing the confusion that has erupted in the thread below. Let me provide some general background that may help.
Your question is a good one. Bi-amplification in general can have benefits that are not obvious from first order analysis. The question here is whether a solid-state amp with low output impedance is better for the bass feed to Sound Lab speakers than a second tube amp; I think everyone would agree that tubes (and especially OTL amps) are better for the treble feed, since the speakers are likely capable of revealing more audio detail than any known electronic device can produce.
Cone woofer damping is one of the chronic difficult questions in audio. Amateurs and unprepared engineers find it hard to understand. The relevant math is Newton's law: a second-order ordinary differential equation that describes the motion of a mass connected to a spring and a damping force proportional to velocity. There are three different kinds of solution to this equation, depending on the magnitude of the damping force:
-Overdamped describes a situation where the system moves slowly back to equilibrium after a disturbance;
-Underdamped is where the system moves more quickly but overshoots and rings; and
-Critical damping is where the system moves back to equilibrium in the shortest time without ringing.
Please see the Dan Tomcik article posted on the ASOG site (URL below) for a non-mathematically-challenging discussion of how this applies to cone woofers.
This is relevant to cone woofers, not electrostatic panels. The damping force for a cone woofer is the regenerative braking force from current developed in the voice coil as the cone moves towards equilibrium. Amplifier output impedance controls the magnitude of the current and the degree of this braking force, and, as Tomcik shows, can be used to adjust the woofer damping to critical, which is optimum for bass power with minimum distortion. Lower amplifier output impedance allows more braking current and increases the damping of cone woofers.
Electrostatic speaker membranes are under mechanical tension and subject to "drum head" ringing (see the FAQ section of the Sound Lab web site for a good discussion of this). The damping is due to air resistance and the membrane is always underdamped. When installed in panels and charged up, the membranes respond to stator voltage, not current. Is there an electrical influence on the membrane damping comparable to what happens with cone woofers?
The charge on the membrane induces "image" charges in the two stators. As far as the membrane is concerned, the stators look like two electrical mirrors with images of the membrane (but with opposite sign of charge) in them. If the membrane is disturbed and the input signal is restored to zero, what will happen? The image charge in the closer stator attracts the membrane: it acts to increase the membrane displacement. A lower amplifier output impedance makes it easier for the image charges in the stators to respond to membrane motion (since current must flow in and out of the stators). However, this force is proportional to membrane distance from equilibrium, not velocity, so it looks like a net reduction in membrane tension. The damping force is unaffected but the resonant frequency is lowered by some amount which may be insignificant.
Now, all successful electrostatic speakers, and Sound Labs in particular, are designed to work with this drum head ringing property of the membranes. To know whether increased output impedance, such as with OTL amps, helps improve the bass fidelity, you have to know what kind of amplifier the designer used to "voice" the speaker. I don't have this information, but a Sound Lab owner may, or you may want to contact Dr. West again.
With regard to even power response over the frequency range, again, it is not obvious whether the impedance dip at the crossover frequency responds better to low or high amplifier output impedance. An amplifier which pumps more current into the system at the impedance dip may cause a rise in the sound pressure depending on what the system does with that current. However, with respect to rising impedance at lower frequencies, a higher amplifier output impedance is better for delivering more constant power as the frequency is reduced.
With regard to heat dissipation in OTL amps, the relevant comparison is with solid state amps with some hope of reasonable transparency: IME these all run with such high bias current that they are too hot to touch in operation and dissipate comparable heat loads into your listening room.
Please let me know if you have more questions.
-Al