Benchtop EMC Testing: A Two-Inductor Fix

Benchtop EMC testing gets a lot easier once you stop guessing and start separating what kind of noise you’re actually fighting. In this highlight from our recent webinar, Sebastian (Baltic Lab) takes a dirt-cheap LM2596-style buck converter that fails conducted emissions, tries the “obvious” snap-on ferrite clamp, and then shows—on the screen—why it barely moves the needle.

Benchtop EMC Testing: Why the Ferrite Didn’t Save It

The ferrite clamp does change the trace a little, but only in a couple of narrow spots. That’s your clue: clamp-on ferrites mainly suppress common-mode current on the cable, not the differential-mode current that’s sloshing back and forth between + and – on the input. If the dominant offender is differential-mode, a random ferrite “because EMC” is basically a superstition with a bill of materials line item.

Common-Mode vs. Differential-Mode: Find the Real Culprit

To make this visible, Sebastian uses two LISNs and a mode-separation accessory (a “LISN Mate”-style combiner) so the spectrum analyzer can show common-mode and differential-mode components independently. Once you can see both plots side by side, the earlier mystery makes sense: the ferrite only nudged the common-mode bumps, while the big failure is clearly differential-mode.

Benchtop EMC Testing Fix: Two Inductors, Done

The fix is refreshingly boring (which is a compliment). Two series inductors—470 µH each—knock the differential-mode noise down to (pre-)compliance levels in the CISPR 25-style setup (Class 3 limits in the demo). For convenience they’re placed right before the LISNs during the measurement, but the real design lesson is placement: in a redesign, you’d put that impedance at the converter input and treat it as part of the input filter network. Benchtop EMC testing doesn’t have to mean “buy a magic box”; it often means “identify the mode, then choose the part that actually targets it.” For a deeper look at the kind of LISN add-on used here, see the LISN Mate accessory.