"Michael Black" wrote in message
news:alpine.LNX.2.02.1502151330260.12310@darkstar. example.org...
Since it was Armstrong's method, and he was interested in wide deviation,
the amplitude would have been removed by the multipliers to get it up to
the desired signal frequency.

Michael

Thanks for this historical information I didn' knew
And in those times the varicap diode was a dream so it was very clever.
It seems to me that adjusting levels of the carrier and sidebands
allows quite wide phase shifts
So may be frequency multiplication was not imperative.
A C class PA is quite good at removing AM
But this is only by figuring things as rotating vectors as I am too poor a
mathematician to be affirmative. :-)
Laurent
..