Overview

Sooner or later, every power supply designer finds out the hard way that if anything has the potential to cause a return to the drawing board at the very last moment, it is either a thermal issue, a safety related issue, or a stubborn EMI (electromagnetic interference) problem. Of these, the first does get resolved relatively easily more copper, better heatsinking, and hopefully more air. The safety issues also melt away, with a little prescience during the design phase and later by some heat shrink tubing, tie-wraps, liberally applied hot-melt glue, RTV (room temperature vulcanizing i.e. silicone glue), and so on. But we discover that EMI turns out to be a veritable "balloon" if we try to "push" in the emissions spectrum on one side, it "bulges" out at the other. We manage to achieve compliance with regulatory conducted emission limits, only to find it has been at the expense of the radiated limits, or vice versa. And sadly, our trusted little bag of tricks (and that dusty Fair-Rite kit of beads) may also sometimes mysteriously let us down. It's then we realize rather acutely if we can't comply, we can't sell!

The stumbling block to a successful understanding of this very vital but misunderstood area of power conversion is that some of the terminology and descriptions used by signal integrity engineers to describe EMI have been bandied about a little too freely in power. Sure there are...