LM5021: Power supply is failing in conducted emission test CISPR22 Class B

Hi Libin,

There are a few methods to tackling an EMI issue. You’ll want to identify your main sources of conducted EMI: Differential Mode (DM) and Common Mode (CM) and determine if one is dominant. Differential mode noise is conducted on the power line and the GND line in opposite directions. DM noise behavior is defined by paths with high di/dt, or paths where the current is switched on and off. Common mode noise is conducted on all lines in the same direction, and can be identified through nodes with dv/dt. A common source of CM noise is a transformer.

Typically, EMI measurements are taken on L1 and L2 using LISN (Line Impedance Stabilization Network) to provide a defined source impedance. The parameters of this testing must be closely controlled (receiver properties, bandwidth, frequency range, etc.) and are defined in standards such as the CISPR 22 or EN55022. The EMI source can then be broken down into its DM and CM equivalent circuits, where the appropriate combination of inductance and capacitance can be used to provide attenuation at a given frequency range.

Please refer to the following resources for additional information on how you can solve EMI issues:

1. An Introduction to Power EMI and Safety https://training.ti.com/introduction-power-emi-and-safety
2. Power Supply Design and Layout Optimization: Minimizing EMI and Instability https://training.ti.com/power-supply-design-and-layout-optimization-minimizing-emi-and-instability?cu=1134585
3. Flyback Transformer Design Considerations for Efficiency and EMI https://www.ti.com/seclit/ml/slup338/slup338.pdf 

Best Regards,

Emily