How does this Marshall Hi/Lo power switch work?

[Gregg]

Concerning the OT primary Marshall are using their "regular" type OTs so nothing special about that.
In one of my 100W amps I wired only two of the tubes to be switchable in Pen/Tri mode and there's a clear audible difference in tone at low and higher levels.

[Zozobra]

I guess other factors in triode connected screen operation is that there is no screen sag compression and the PI has to swing harder for equivalent(ish) volumes/output so there is more PI dirt in the tone.

[Gingertube]

Thanks for the explanation Ian. That triode "darker tone" - some of my customers love it. Some, not so much.

I'm sure the hi fi amps that use triode have OT's made to match impedance. Whether there's some treble pre-emphasis (or feedback compensation) to make up for the Miller effect treble loss, that's optional and I'm sure what makes the difference one brand of amp to another. Here's another place for the smart custom amp builder to use his/her brain & ears.

For HiFi there are usually several things going on.
1st - they usually have beefier phase splitters or dedicated drivers after the phase splitter to better drive that increased Miller Capacitance at the output tube grids. Thats why HiFi guys like Concertina (Cathodyne) Splitters - they have very low output impedance.
2nd - The output impedance of the output tubes themselves is much lower in triode mode. That means better drive of the primary inductance of the OT (so the bass roll off [fromed by teh outtube output impedance and the primary inductance] will be lower) AND better drive of both the leakage inductance (primary to secondary) and interwinding capacitance (promary) of the OT so the TWO High frequency roll offs will both be moved higher. So Triode Mode gives better Low and High frequency response from the OT.

Cheers
Ian

[Mike Sulzer]

I think dissipation in the tubes is about the same in the two modes. Compared to the pentode mode, the triode mode can draw less peak current, but the voltage drop across the tube is higher when this peak current occurs and so V*I would be similar. I think you would have to measure or model to really know.

It is what happens when the screen is not kept at a high enough voltage to keep accelerating the electrons to the plate.

[Pedro Vecino]

Dissipation is the same in that amplifier. The main problem of the triode mode is the dependence of the symmetry on the iddle currents of the tubes. It is much more susceptible to generating hum. In fact, a difference of 4 or 5 mA. (negligible in penthode mode) presents a hum in triode mode more than obvious.

[Malcolm Irving]

At the usual operating point, the plate current in a pentode is not very sensitive to plate voltage (in other words the plate characteristics are almost flat). A hum voltage on the supply to the plate will therefore not create much hum current in the speaker.
The plate current in the pentode is more sensitive to screen voltage, but that is invariably controlled by a further power supply smoothing stage
In triode mode, the plate current is sensitive to plate voltage and hum voltage on the plate supply will create greater hum current in the speaker.