Chokes don't make sense to me...

[Hardtailed]

I'm having a problem understanding why we use chokes in tube amps.

Basically, the job of a power supply (beside filtering) is to provide a constant voltage source independant of the current demand. At least, that's what an ideal power supply would do, and in the case of tube amps, we do expect it to sag a little on high current demands.

But a choke, if I understand correctly, actually tries to regular current, not voltage. Therefore it will generate a few volts if current goes down, and cut a few if current goes up, therefore resisting the current variation process that is tube amplification.

I guess I must be wrong somewhere, or maybe it's specifically those shortcomings that make them desirable in tube amps?

[Steve Conner]

"Choke" is just the old-fashioned word for inductor.

Inductors perform the opposite function to capacitors. A capacitor stops DC current from passing, but allows AC through. An inductor allows DC to pass straight through, but "chokes off" AC current.

By combining a choke with capacitors, we can make a filter that is very good at letting the wanted B+ voltage through while getting rid of unwanted 120Hz buzz. The capacitors dump the 120Hz ripple to ground, while the choke blocks it from going through.

The two different kinds of filtering reinforce each other, and the result is that a CLC filter will have less ripple and less sag than a CRC with the same size of capacitors. On the downside, chokes are heavy and expensive.

If you overdo it, using a very large inductance and small capacitors, the transient response will be poor. Load transients will kick the filter into resonance, making the B+ voltage bounce around, and I guess that could be what you called "resisting the process of tube amplification" or whatever. With the values normally used in guitar amps, this doesn't seem to happen. However, really old amps from the days of choke-input filters and 8uF paper caps may have suffered from it.

[Enzo]

If there is a momentary peak in the demand from the power supply - current demand - the voltage will tend to drop. The capacitors resist this drop by delivering their stored charge. The inductor resists this drop by trying to boost the voltage back where it was. It does this from the energy in its winding. They are fighting the same battle.

[jrfrond]

By creating a "Pi" filter with three reactive components, you create a steeper rolloff of the hum components. Of course, the feel is different too, as it creates a stiffer supply. However, if you go back to the early days of guitar amp design, feel was not a design parameter.