Thanks for the additional datapoints, Rodman. I don’t think the pronounced pop at the moment the battery was connected (or when the cap was discharged into the driver) is surprising. For a theoretically ideal capacitor i = C(dv/dt), of course, ("i" = current; "C" = capacitance; "dv/dt" = change in voltage per unit time), so completing the circuit by touching wires or leads together would result in a nearly instantaneous change in the voltage applied to the cap, resulting in a large (dv/dt). And discharging the cap into the driver would have similar consequences, as the voltage across the charged cap would abruptly be forced toward zero by the low resistance of the paralleled driver.
But in the event of an amp failure resulting in a large DC output, how likely is it that the DC voltage will be ramped up in a comparably abrupt manner? I don’t know the answer to that.
And what I don’t understand in your latest results is how the C cell could hold a 200 watt woofer at full excursion. Even if the DC resistance of the driver is as low as 2 ohms, the battery would be putting not much more than 1 watt into it. And 200 watts into any reasonable woofer resistance corresponds to vastly more than 1.5 volts, of course.
Also, it seems relevant that the capacitors used in many tube-based preamps and other tube-based components to couple the signal at the plates of various interstage and output tubes to the grids of a subsequent stage or to the input of another component, such as a power amp, are in many cases used to block DC on those plates of well over 100 volts (at least if a cathode follower is not used, in the case of output stages). Yet that never seems to be a problem. And in many cases those caps have values that are not much different than the 0.1 uF cap Rodman experimented with. And I seem to recall that some McIntosh designs, at least, have used fairly large electrolytics for their output coupling capacitors, having values not all that much lower than the 33 uF he also experimented with.
Finally, although it’s more of an academic point than one having practical significance, if a large DC voltage is suddenly applied to a capacitor, and a large current briefly flows corresponding to C(dv/dt), the nearly instantaneous change in voltage means that spectral components are present at non-zero frequencies, at and near that instant. Which by definition means that the voltage is not DC, at and near that instant.
Regards,
-- Al
But in the event of an amp failure resulting in a large DC output, how likely is it that the DC voltage will be ramped up in a comparably abrupt manner? I don’t know the answer to that.
And what I don’t understand in your latest results is how the C cell could hold a 200 watt woofer at full excursion. Even if the DC resistance of the driver is as low as 2 ohms, the battery would be putting not much more than 1 watt into it. And 200 watts into any reasonable woofer resistance corresponds to vastly more than 1.5 volts, of course.
Also, it seems relevant that the capacitors used in many tube-based preamps and other tube-based components to couple the signal at the plates of various interstage and output tubes to the grids of a subsequent stage or to the input of another component, such as a power amp, are in many cases used to block DC on those plates of well over 100 volts (at least if a cathode follower is not used, in the case of output stages). Yet that never seems to be a problem. And in many cases those caps have values that are not much different than the 0.1 uF cap Rodman experimented with. And I seem to recall that some McIntosh designs, at least, have used fairly large electrolytics for their output coupling capacitors, having values not all that much lower than the 33 uF he also experimented with.
Finally, although it’s more of an academic point than one having practical significance, if a large DC voltage is suddenly applied to a capacitor, and a large current briefly flows corresponding to C(dv/dt), the nearly instantaneous change in voltage means that spectral components are present at non-zero frequencies, at and near that instant. Which by definition means that the voltage is not DC, at and near that instant.
Regards,
-- Al