Hi everyone,
Now a more conceptual point.
I have implemented frequency modulation using the continuous bidirectional ramp. Checking the result with a spectrum analyzer, I see that the bandwidth is a bit higher than it should be, I guess something acceptable, due to the fact that it's not an analog modulation (I have used 1000 memory positions, trying to make it as smooth as possible). My implementation seems to be OK, as the distance between sidebands is exactly the modulating frequency.
What is bugging me is the fact that the Bessel nulls are not where they should be. In other words, with a modulation index of 2.405 or 5.520 (and so on), the carrier doesn't vanish - as a matter of fact, the index is not even close.
Any thoughts?
Thanks