A Systematic Approach

EMC and ripple problems in power designs are easy to misdiagnose, because the symptoms appear at the system level while the cause usually lives in the layout. This article presents a systematic procedure for diagnosing converter and supply designs, using the Hawun converters and supplies as references, so that the cause is found by measurement rather than by guessing.

Step 1: Inspect the Switching Loop

The loop that carries the switched current is the primary source of both radiated emissions and ripple, because it carries a fast-changing current and it has parasitic inductance. Make the loop as small as possible: place the input capacitor close to the converter, keep the loop compact, and return the current directly beneath the trace so the enclosed area is minimized. A tight loop reduces EMI and ripple at the same time, and it is the single most effective layout rule.

Ground Return

The ground return must be continuous and direct beneath the switching path. A split or a long ground return increases the loop area and the emissions, and it also adds noise to the output. Use a solid ground plane and connect the converter's ground pad to it with multiple vias.

Step 2: Check the Output Capacitor and Filter

Output ripple depends on the output capacitor and its placement. Confirm the capacitor meets the ripple and the load-transient requirement, and place it close to the load rather than only at the converter. Where the ripple must be lower, add a second-stage filter or post-regulate the rail. Measure the ripple at the load, because that is where the circuit sees it.

Input Filter

The input filter limits the conducted emissions the converter produces. Confirm the filter is present, that its components match the datasheet, and that its layout is compact, because a filter with a poor layout does not attenuate the emissions it should.

Step 3: Verify the Measurement

EMC and ripple measurements are easy to get wrong. Use the correct probe and bandwidth, measure at the point of interest, and keep the setup consistent between measurements, because a moved cable or a wrong probe changes the result. For conducted emissions, measure with a LISN against the standard's limit rather than with an arbitrary setup.

Frequency and Harmonics

Conducted emissions appear at the switching frequency and its harmonics, so the spectrum shows where the problem is. A peak at the switching frequency points to the loop and the filter; a peak at a higher harmonic points to the switching edge, which can be slowed with the gate drive or the snubber.

Step 4: Confirm the Thermal Design

EMC fixes sometimes trade against thermal performance, so confirm the design still meets its thermal requirement after any change. A larger gate resistor or a snubber reduces emissions but raises loss, so the thermal design must be revalidated. BeiLuo's power lab can measure the emissions, the ripple and the temperature together, so the trade-off is understood before production.

When to Escalate

If the design still fails after the loop, the filter and the measurement are corrected, escalate with the measured data: the spectrum, the ripple at the load, and the temperatures. BeiLuo's FAE team can review the layout against the reference and characterize the design in our lab so the cause is identified before a board respin.