In some notes on transmitters I have found the statement that "inductive coupling between transmitter amplifier stages reduces the generation of harmonics", with no further explanation. Can anyone please explain why this is so, or convincingly refute it if it isn't?
Absent any other context, I have a couple guesses.
A ferrite core in a transformer has losses that increase with frequency, so consequently the higher frequency energy in the harmonics is dissipated as heat in the transformer. Provided of course that the core is not saturated to the point where its non-linear properties introduce more harmonics than are removed.
Additionally, any real inductor, of any core material, or a transformer winding, has a parasitic capacitance due to the proximity of the windings. The inductance and this capacitance define a self-resonant frequency for the inductor, and above this frequency the inductor presents a capacitive reactance instead of an inductive one. If the transformer is designed such that the harmonics are above the self-resonant frequency, those harmonics will then see a shunt capacitance to ground, effectively low-pass filtering them.