Electrostatic Shielding for Switching Power Adapters

One of the most challenging specifications in the design of switching power adapters is reducing the common-mode conducted RFI (Radio Frequency Interference) current to an acceptable level. This conducted noise is mainly caused by parasitic static electricity and electromagnetic coupling between power-switching components and the ground plane. The ground plane may consist of the chassis, cabinet, or ground wire, depending on the type of electronic equipment.

 

Designers of switching power adapters should thoroughly review the entire layout, identify areas prone to such problems, and implement appropriate shielding measures during the design phase. Correcting improper RFI design in later stages is often difficult.

 

In most applications, electrostatic shielding is necessary wherever high-frequency, high-voltage switching waveforms may capacitively couple with the ground plane or secondary output. This is particularly important where switching power transistors and rectifier diodes are mounted on heat sinks that contact the main chassis. Additionally, magnetic fields and capacitive coupling may introduce noise in components or lines carrying large switching pulse currents. Potential problem areas include the output rectifier, output capacitor mounted on the chassis, and capacitive coupling between the primary, secondary, and core of the main switching transformer, as well as other drive or control transformers.

 

When components are mounted on heat sinks thermally connected to the chassis, unwanted capacitive coupling can be mitigated by placing an electrostatic shield between the interfering component and the heat sink. This shield, typically made of copper, must be insulated from both the heat sink and the component (e.g., transistor or diode). It blocks capacitively coupled AC currents, which are then directed to a convenient reference point in the input circuit. For primary components, this reference point is typically the common negative terminal of the DC power supply line, near the switching device. For secondary components, the reference point is usually the common terminal where current flows back to the transformer's secondary side.

 

The primary switching power transistor generates high-voltage, high-frequency switching pulse waveforms. Without adequate shielding between the transistor case and the chassis, significant noise currents can couple through the capacitance between them. A copper shield placed in the circuit injects any substantial current into the heat sink via capacitance. The heat sink, in turn, maintains a relatively small high-frequency AC voltage concerning the chassis or ground plane. Designers should identify similar problem areas and apply shielding where necessary.

 

To prevent RF currents from flowing between primary and secondary windings or between the primary and grounded safety shield, main switching transformers typically include an electrostatic RFI shield on at least the primary winding. In some cases, an additional safety shield may be required between the primary and secondary windings. Electrostatic RFI shields differ from safety shields in their construction, location, and connection. Safety standards require the safety shield to connect to the ground plane or chassis, while the RFI shield is usually connected to the input or output circuit. EMI shields and terminal blocks, made of thin copper sheets, carry only small currents. However, for safety reasons, the safety shield must withstand at least three times the rated current of the power fuse.

 

In offline switching power transformers, the RFI shield is placed close to the primary and secondary windings, while the safety shield is located between the RFI shields. If no secondary RFI shield is needed, the safety shield is positioned between the primary RFI shield and any output windings. To ensure proper isolation, the primary RFI shield is often DC-isolated from the input power line via a series capacitor, typically rated at 0.01 μF.

 

The secondary RFI shield is used only when maximum noise suppression is required or when the output voltage is high. This shield connects to the common terminal of the output line. Transformer shielding should be applied sparingly, as it increases component height and winding dimensions, leading to higher leakage inductance and performance degradation.

 

 

High-frequency shield loop currents can be significant during switching transients. To prevent coupling to the secondary side through the transformer's normal operation, the shield connection point should be at its center, not its edges. This arrangement ensures that the capacitively coupled shield loop currents flow in opposite directions on each half of the shield, eliminating inductive coupling effects. Additionally, the ends of the shield must be insulated from each other to avoid forming a closed loop.

 

For high-voltage outputs, the RFI shield can be installed between the output rectifier diodes and their heat sinks. For low secondary voltages, such as 12V or lower, secondary transformer RFI shields and rectifier shields are generally unnecessary. In such cases, placing the output filter choke in the circuit can isolate the diode heat sink from RF voltage, eliminating the need for shielding. If the diode and transistor heat sinks are entirely isolated from the chassis (e.g., when mounted on a PCB), electrostatic shielding is often unnecessary.

 

Ferrite flyback transformers and high-frequency inductors often have significant air gaps in the magnetic path to control inductance or prevent saturation. These air gaps can store considerable energy, radiating electromagnetic fields (EMI) unless adequately shielded. This radiation can interfere with the switching power adapter or nearby equipment and may exceed radiated EMI standards.

 

EMI radiation from air gaps is greatest when the outer core is gapped or when gaps are evenly distributed between poles. Concentrating the air gap in the middle pole can reduce radiation by 6 dB or more. Further reduction is possible with a fully enclosed pot core that concentrates the gap in the middle pole, although pot cores are seldom used in offline applications due to creepage distance requirements at higher voltages.

 

For cores with gaps around perimeter poles, a copper shield surrounding the transformer can significantly attenuate radiation. This shield should form a closed loop around the transformer, centered on the air gap, and be about 30% of the winding bobbin's width. To maximize efficiency, the copper thickness should be at least 0.01 inches.

 

While shielding is effective, it introduces eddy current losses, reducing overall efficiency. For peripheral air gaps, shield losses can reach 1% of the device's rated output power. Middle pole gaps, by contrast, cause minimal shield losses but still reduce efficiency due to increased winding losses. Shielding should therefore be used only when necessary. In many cases, enclosing the power supply or device in a metal casing suffices to meet EMI standards. However, in video display terminal devices, transformer shielding is often required to prevent electromagnetic interference with the CRT electron beam.

 

The additional heat generated in the copper shield can be dissipated via a heat sink or redirected to the chassis to maintain operational stability.