EMC Question of the Week: December 12, 2022

Conducted emissions testing of a 12-volt device reveals that the noise on the power wire (V_BATT+) is 12 dB over the limit. The noise on the power return wire (V_BATT-) is 12 dB below the limit. This product would benefit from a pi-filter

1. on the V_BATT+ side
2. on the V_BATT- side
3. on both sides (same filter)
4. employing a CM choke

Answer

The best answer is “a.” The V_BATT+ side is the one that requires filtering.

This answer may seem obvious, but sometimes engineers overthink it. The differential-mode and common-mode components of the measured noise are,

$\begin{array}{c}{V}_{DM}=\frac{V_{BATT+}-V_{BATT-}}{2}\approx \frac{V_{BATT+}}{2}\\ V_{CM}=\frac{V_{BATT+}+V_{BATT-}}{2}\approx \frac{V_{BATT+}}{2}\end{array}$

There is a differential-mode and a common-mode component. Both are 6 dB over the limit.

It's tempting to conclude that both differential-mode and common-mode filtering are required. However, it's far more likely that the device-under-test is simply unbalanced. The V_BATT- side is likely well-connected to the circuit board ground and/or the system chassis ground at the measurement frequencies. There are no significant sources driving V_BATT- relative to the ground at the measurement point. The noise source drives V_BATT+ relative to circuit ground. Simply filtering the V_BATT+ side to the circuit board ground is the best option.

Of course, it would be possible to meet the conducted emissions requirement with a much more complicated filter. A common-mode choke would reduce the noise on one side and increase the noise on the other side. Both sides could then be filtered to the chassis ground. And while this might work, the filter would be much bigger and more expensive than necessary.

Have a comment or question regarding this solution? We'd like to hear from you. Email us at This email address is being protected from spambots. You need JavaScript enabled to view it..