When a Common Mode Filter Increases Conducted Emissions

In around 90% of CE (Conducted Emission) issues on the mains input, the solution is simple: just add a common-mode (CM) filter.Most of the time, I use Würth Elektronik filters — they’re reliable and predictable.

However, every once in a while, something strange happens.Instead of reducing the conducted emissions, the filter actually makes them worse — sometimes by several dB between 1 MHz and 30 MHz.Let’s see why this happens and how to fix it.

The Typical Setup

In a standard CE test, we have the following configuration:

• The EUT (Equipment Under Test) sitting on the table.

• A CM choke or filter inserted between the LISN and the EUT.

• A power cable, typically around 1.5 m long, connecting the filter to the LISN.

What’s Really Happening

When we add a common-mode choke, we increase the common-mode impedance seen at the input. If the EUT has a strong magnetic field in the range 150 kHz - 30 MHz. If this magnetic field couples into the power cord, it can induce a voltage that adds directly to the measured noise at the LISN.So instead of reducing emissions, the filter seems to amplify them.

This phenomenon is magnetic coupling, and the induced voltage is proportional to the time-rate of change of magnetic flux through the loop area of the power cable:

where:

If the magnetic field B from the CM choke passes through the loop area formed by Live–Neutral–Earth conductors, we get an unwanted induced voltage.The larger the loop area A, the larger the induced voltage.

The “Twisted Cable” Test

To confirm this hypothesis, I usually perform a very simple test:Replace the normal power cord with a twisted power cord.

If the CE emissions go down, our theory is confirmed — we are dealing with magnetic coupling.

Why does twisting help?Because every twist reverses the polarity of the induced voltage. Over the length of the cable, the induced voltages cancel each other:

Vtotal≈N⋅(Etwist−Enext twist)≈0V

In more intuitive terms, twisting reduces the effective loop area and therefore the magnetic flux linkage:

As the twist pitch increases (more twists per meter), the effective loop area decreases, and the induced voltage becomes negligible.

Practical Fixes

Once we’ve proved that magnetic coupling is the cause, we can look for practical solutions. Here are a few effective ones:

1. Use a twisted or shielded mains cable.The simplest and most effective fix. The twist breaks the magnetic coupling.

2. Shorten the unshielded section between the CM filter and the LISN.Keep leads as short as possible, ideally under 10 cm.

3. Add a grounded metal plate between the filter and the cable.This provides partial magnetic shielding.

4. Place the CM choke on the LISN side instead of the EUT side (if acceptable).

Conclusion

A CM filter is not always a guaranteed fix for CE problems.When emissions go up after adding a choke, don’t panic — it’s often a magnetic coupling issue.Verify the theory using a twisted cable, and once confirmed, you can apply one of the solutions above.

Sometimes, solving EMC problems isn’t just about choosing the right component — it’s about understanding the physics around it.