I have an SB200 linear and on AM it is equivalent to a DX100.
I increased the bias on it so to the point just beyond cutoff for AM so
that
it requires a little more to drive and it reaches negative 100% modulation
slightly before the driver does. But this seemed to help the modulation
linearity and efficiency just a little.
That's an old trick that I first read about in a 1930's RADIO article.
Overbias the AM linear to increase the negative peak modulation percentage.
It also can have a distortion advantage, especially if the modulator driver
stage uses grid modulation. Most grid modulated finals become nonlinear
near 100% negative modulation, or are simply incapable of cutting the
carrier off completely, so to maintain good modulation linearity, limit the
modulation to 90% or less in the negative direction, then overbias the
linear output stage until the negative peaks in the output reach near 100%.
This will retain modulationr linearaty on negative peaks and reduce overall
distortion levels.
