Lower output switching frequencies would have higher amplitude (filtered less) while higher frequencies would increase timing errors and dissipated heat. Chosen frequency at about 500kHz, is likely a compromise. There is about 1% of this frequency residue clearly visible on the output but it is inaudible and also speakers won't play it. Cables won't radiate it either, being (for 3m) 1/200 of the wavelength. Capacitive coupling is always possible, but I'm not sure how spreading frequency would help. Speaker wires should be twisted, IMHO, since twisting reduces inductance (important) and increases capacitance (less important). Twisting wires makes them more immune to ambient electrical noise but also reduces radiated noise - including electrical field (capacitive coupling). I don't know of any spread spectrum class D, but who knows - it is still evolving.
Spread Spectrum Class D?
I was reading on switching power supplies from iFi and they claim they use spread spectrum outputs to minimize the ultrasonic noise produced. The idea is kind of simple. Instead of having a fixed switch frequency you spread the switching frequency around. This prevents a big spike of noise in the ultrasonic range and minimizes how steep the output filter has to be.
I'm curious if anyone has seen this idea used in high output Class D amps?
1 response Add your response