At 10/24/2007 10:08, you wrote: >At 10:10 AM 10/23/2007, you wrote: > > >No, it's not linear. First off, it doesn't satisfy the superposition > >principle since it will produce harmonic when fed by a pure sinusoid carrier > >(hopefully we can agree to that without added discussion), so right off the > >bat it's nonlinear. Furthermore, it's also non-reciprocal; what enters port > >1 exits port 2 with little loss, but the converse is not true. > ><---I though it was linear unless it becomes saturated. Circulators >are ferrite devices and as long as the ferrite core isn't saturated, >it remains linear.
True. I once measured the output of a 220 MHz circulator being pumped with ~80 watts with a spectrum analyzer & couldn't see any 2nd harmonic output from the circulator. I think the circulator itself was rated for 350 watts but not sure; of course the reject load had a much lower rating. I'm not saying that circulators don't need harmonic filtering, just that there's no guarantee that one will in fact act in a nonlinear manner, particularly if driven below saturation. You have to measure it to be sure, otherwise be safe & use a LPF or bandpass cavity. Bob NO6B
