**A valid question.**

I have always been curious, and ever so slightly uneasy, about the label applied below 200Hz on the impedance curve of the Quad 63 that reads “level dependent”. I have not satisfied myself as to why this non-linearity exists to a significant enough degree that Walker felt it necessary to call it out. Here are some possibilities, thrown out without a lot of analysis or research on my part:

1.) Mark, with your “turn off the bias” comment, you seem to be implying that constant charge assumption may no longer be applicable, and that diaphragm charge is varying with signal. Maybe I’m reading too much into what you said. But perhaps the 10M series HT resistor and neon bulb plus the diaphragm resistance is not enough resistance to meet the 2fRC>1 criterion at these bass frequencies. This seems plausible. Still the Quad’s bass seems very clean to me (at least up to a limit), compared to the many significant non-linearities of moving coil bass cones.

2.) The input impedance due to primary inductance of the step-up transformers drops gradually below 200Hz. The secondary is lightly loaded and the input current is a non-linear function of voltage due to B-H non-linearities. Perhaps if driven by a true voltage source amp, much of this factor of impedance non-linearity is not transferred to the air.

3.) The intended operating range of ESLs is where the radiation resistance exceeds diaphragm stiffness reactance (due to stretch or springiness). Think of a series RL circuit. As frequency decreases, stiffness reactance starts to become comparable to and then exceeds the radiation resistance. The stiffness reactance of a stretched plastic film is bound to be non-linear, and the linear radiation resistance gives way to this non-linearity at low frequencies. The effect would show up in the impedance curves, as well as in the airborne response.

4.) Mark, you said “I THINK the variation with power that occurs at low frequencies is because the amount of power coupled to the air is a larger (smaller?) fraction of the total power dissipated.” I don’t see how radiation loading by itself is a non-linear function, but perhaps you were thinking of these other factors.

So, I don’t know which of these is at “fault”. It could be that more than one of these factors is at play at once.

Any further insights from anyone?