Power amplifiers in general have a DC offset at their output, which may be anything from under a millivolt for circuits with well-designed servo systems to half a volt or so in some DC-coupled amplifiers. In an ideal world we don't want any DC component in the voltage across the speaker terminals - partly because it's a waste of power, partly because it leads to unnnecessary heating of the crossover inductors and speaker coils (which may or may not be significant), and partly because it takes the speaker magnetic system away from its most linear operating region. Here I show how the significance of at least the third of these can be assessed.
Using the standard theory, the displacement x of the cone for a given applied DC voltage V is given by
where BL is the BL product in N A-1, Cms is the compliance in m N-1, and Re is the total DC resistance in ohms (voice coil plus crossover).
Now we need a criterion to decide how large an offset we will tolerate. A parameter that is almost always suppled by driver manufacturers is the maximum linear excursion xmax. If we insist that the displacement x under zero signal is less than some fraction a of xmax, then the maximum permitted DC voltage is
If we can put up with 10% of xmax, the T-S parameters for the two HDA drivers I use in my three-way speakers give Vmax=240mV for the HM210Z0, and 800mV for the HM100Z0 (note that the units given in Audax' main spec sheets are incorrect!). Of course one would prefer smaller displacements than this, and generally we look for an offset of less than 100 mV.
Alex Megann, December 2000
Back to DIY page.