Richard Fry wrote:
"Please explain the reason why the bandwidth characteristics needed to
generate a single transistion from one DC level to another is different
than when repeating that same transition any number of times before or
after."

No answer from Reg Edwards has appeared on my screen and it`s 6 hours
later in England than in the central U.S.A., so Reg may be sleeping.

The transitions are likely quite similar from one repetition to the next
but more is involved. It`s true that the shorter the length of the
pulse, the higher the efficiency of the harmonic generator. It is also
true that there is an optimum pulse length for maximum output at the
harmonic frequency. Pulse duration as well as on-off transition times
are involved. Output of a harmonic generator is a multiple of the pulse
frequency.

Terman has harmonic generator information starting on page 473 of his
1955 edition. This is slanted toward Hollow-State devices which were
still dominant at the time.

Best regards, Richard Harrison, KB5WZI