"Reg Edwards" wrote in message
...
The harder the modulated tube is driven into class-C conditions and
saturation, the more linear does the plate modulation become.

Operation of the tube becomes independent of curvature in the tube's
characteristic. The plate current operating angle is small.
It behaves more like an on/off switch.

The more non-linear it is, the smaller the operating angle, the more
linear is the modulation.
======================================

PLATE MODULATION.
It's really all very simple.

Imagine a class-C triode amplifier with very small operating angle and
running nearly into saturation.

The plate load is a tuned tank circuit having a high impedance at
resonance. Or it can be a Pi-tank circuit. Makes no difference!

With the high impedance load, conditions are such that whatever is the
DC plate voltage, the RF plate volts swing down to a very low
plate-to-cathode voltage.

Ideally it should be zero volts. But in practice it cannot fall below
the positive, peak, instantaneous, RF grid volts. This corresponds to
the instant of peak plate current.

The RF voltage across the tank is then very nearly EQUAL to the DC
plate voltage regardless of the tubes characteristic curves. Curvature
doesn't matter. It is obscured by the small operating angle. The tube
is conducting only for a small fraction of the time.

Modulate the DC plate voltage at an audio frequency. With 100 percent
modulation the DC plate voltage swings between a very low voltage and
twice the DC supply volts. And so do the RF volts across the tank.
The job is done. You have an almost perfect linearly modulated
amplifier.

It is necessary only to ensure grid drive is just sufficient to drive
the tube into saturation when the DC plate volts is twice the DC
supply volts. It will then remain saturated at all lower voltages.

With a triode, saturation occurs when the RF plate voltage swings down
to not much more than the peak RF grid volts.

With a beam tetrode, saturation occurs when the RF plate voltage
swings down only to something less than the DC screen-grid volts and
100 percent linear modulation cannot be achieved. But 100 percent
modulation is always undesirable because of the risk of
over-modulation.

With a bipolar transistor, modulation can be even more linear because,
with the high impedance tuned tank, the device saturates or 'bottoms'
at very nearly zero RF collector volts. About 0.7 volts.
----
Reg, G4FGQ.