spectrum of PWM

[Haible Juergen]

Does anybody have a formula for the spectrum of a PWM
signal ?

(I'm aware there are spectra published for certain fixed
pulse widths, but that's not what I'm after.)

So, is there a closed expression for the PWM spectrum
with Triangle carrier and Sine modulator, with modulator
fequencies up to carrier frequency ?

To make things easier, let's assume initial PW = 50%
and modulator amplitude < carrier amplitude.


JH.