Tube vs solid state power amp volumes and speaker load

[Malcolm Irving]

I think I’m beginning to get it! How about this:

If we can take the ideal-transformer part out of the picture by referring all the quantities to the primary side, we have the following circuit loop:
Current source which can get switched off (i.e. approximate model of an output pentode), in series with total leakage inductance referred to primary, in series with speaker impedance referred to primary.
If the current source tries to cut off, we get a big voltage spike across the leakage inductance, because the inductance is doing its best to keep the current through it (and hence through the whole loop) constant.
Constant current through the speaker does not represent any spike, as far as the speaker is concerned (even if it is inductive as well as resistive).

(Speakers are capacitive for part of their frequency response, but let’s leave that for another day.)

EDIT: Bit of a fallacy here. If the tube does cut off then there can be no current in that loop, despite how hard the inductance is trying to keep it flowing by producing a big voltage spike. I think I need some capacitance in there somewhere to absorb the current and make it work? Or maybe the magnetising inductance can help out by providing a parallel path?

[Mike Sulzer]

I think I’m beginning to get it! How about this:

If we can take the ideal-transformer part out of the picture by referring all the quantities to the primary side, we have the following circuit loop:
Current source which can get switched off (i.e. approximate model of an output pentode), in series with total leakage inductance referred to primary, in series with speaker impedance referred to primary.
If the current source tries to cut off, we get a big voltage spike across the leakage inductance, because the inductance is doing its best to keep the current through it (and hence through the whole loop) constant.
Constant current through the speaker does not represent any spike, as far as the speaker is concerned (even if it is inductive as well as resistive).

(Speakers are capacitive for part of their frequency response, but let’s leave that for another day.)

EDIT: Bit of a fallacy here. If the tube does cut off then there can be no current in that loop, despite how hard the inductance is trying to keep it flowing by producing a big voltage spike. I think I need some capacitance in there somewhere to absorb the current and make it work? Or maybe the magnetising inductance can help out by providing a parallel path?

Resolving the fallacy: I do not think it is necessary to introduce any capacitance in specific places. It will do to recognize that the tube cannot shut off instantly, for whatever reasons, but rather causes the current to decrease rapidly, causing what we have been discussing.