PCB Technical - Analysis of Hidden Features of PCB Passive Components in EMI/EMC Design

Traditionally, EMC has been regarded as "black magic". In fact, EMC can be understood by mathematical formulas. However, even if there are mathematical analysis methods available, those mathematical equations are still too complicated for actual EMC circuit design. Fortunately, in most practical work, engineers do not need to fully understand those complex mathematical formulas and the theoretical basis existing in EMC specifications. As long as simple mathematical models are used, they can understand how to meet EMC requirements. .

This article uses simple mathematical formulas and electromagnetic theory to illustrate the hidden behaviors and characteristics of passive components on printed circuit boards (PCBs). These are the requirements that engineers have to design beforehand when they want to make their electronic products pass EMC standards. Must have basic knowledge.

wire and PCB trace

The seemingly inconspicuous components such as wires, traces, fixing frames, etc. often become the best transmitters of radio frequency energy (that is, the source of EMI). Every component has an inductance, which includes the bond wires of the silicon chip, and the pins of resistors, capacitors, and inductors. Each wire or trace contains hidden parasitic capacitance and inductance. These parasitic components will affect the impedance of the wire and are very sensitive to frequency. According to the value of LC (which determines the self-resonance frequency) and the length of the PCB trace, self-resonance (self-resonance) can be generated between a component and the PCB trace, thus forming an efficient radiating antenna.

At low frequencies, the wire generally only has the characteristics of resistance. But at high frequencies, the wire has the characteristics of inductance. Because it becomes high frequency, it will cause the change of impedance, and then change the EMC design between the wire or PCB trace and the ground. At this time, the ground plane and ground grid must be used.

The main difference between wires and PCB traces is that the wires are round and the traces are rectangular. The impedance of a wire or trace includes resistance R and inductive reactance XL = 2πfL. At high frequencies, this impedance is defined as Z = R + j XL j2πfL, and there is no capacitive reactance Xc = 1/2πfC. When the frequency is higher than 100 kHz, the inductance is greater than the resistance. At this time, the wire or trace is no longer a low-resistance connecting wire, but an inductance. Generally speaking, wires or traces that work above audio frequency should be regarded as inductance, and can no longer be regarded as resistance, and can be a radio frequency antenna.

The length of most antennas is equal to 1/4 or 1/2 wavelength (λ) of a certain frequency. Therefore, in the EMC specification, wires or traces are not allowed to work below λ/20 of a certain frequency, because this will suddenly turn it into a high-performance antenna. Inductance and capacitance will cause the resonance of the circuit, this phenomenon will not be recorded in their specifications.

For example: Suppose there is a 10 cm trace, R = 57 mΩ, 8 nH/cm, so the total inductance value is 80 nH. At 100 kHz, an inductance of 50 mΩ can be obtained. When the frequency exceeds 100 kHz, this trace will become an inductance, and its resistance value can be ignored. Therefore, this 10 cm trace will form an efficient radiating antenna when the frequency exceeds 150 MHz. Because at 150 MHz, its wavelength λ = 2 meters, so λ/20 = 10 cm = the length of the trace; if the frequency is greater than 150 MHz, its wavelength λ will be smaller, and its 1/4λ or 1/2λ value will be It is close to the length of the trace (10 cm), so a perfect antenna is gradually formed.

resistance

Resistor is the most common component found on PCB. The material of the resistor (carbon synthesis, carbon film, mica, winding type... etc.) limits the effect of frequency response and the effect of EMC. Wire-wound resistors are not suitable for high-frequency applications because there is too much inductance in the wires. Although carbon film resistors contain inductance, they are sometimes suitable for high-frequency applications because the inductance of its pins is not large.

What people often overlook is the package size and parasitic capacitance of the resistor. Parasitic capacitance exists between the two terminals of the resistor. They can damage the normal circuit characteristics at extremely high frequencies, especially when the frequency reaches GHz. However, for most application circuits, the parasitic capacitance between the resistor pins is not more important than the pin inductance.

When the resistance is subjected to overvoltage stress (overvoltage stress) test, you must pay attention to the change of resistance. If an "electrostatic discharge (ESD)" phenomenon occurs on the resistor, something interesting will happen. If the resistor is a surface mount component, the resistor is likely to be penetrated by the arc. If the resistor has pins, ESD will find the high resistance (and high inductance) path of this resistor and avoid entering the circuit protected by this resistor. In fact, the real protector is the inductance and capacitance characteristics hidden by this resistor.

capacitance

Capacitors are generally used in the power bus to provide decouple, bypass, and maintain a fixed DC voltage and current (bulk) functions. A truly pure capacitor will maintain its capacitance value until it reaches the self-resonant frequency. Beyond this self-resonance frequency, the capacitance characteristics will become like an inductance. This can be explained by the formula: Xc=1/2πfC, Xc is capacitive reactance (unit is Ω). For example: a 10μf electrolytic capacitor, at 10 kHz, the capacitive reactance is 1.6Ω; at 100 MHz, it drops to 160μΩ. Therefore, at 100 MHz, there is a short circuit effect, which is ideal for EMC. However, the electrical parameters of electrolytic capacitors: equivalent series inductance (ESL) and equivalent series resistance (ESR), will limit this capacitor to only work at frequencies below 1 MHz.

The use of capacitors is also related to the pin inductance and volume structure. These factors determine the number and size of parasitic inductances. Parasitic inductance exists between the welding wires of the capacitor. They cause the capacitor to behave like an inductance when it exceeds the self-resonance frequency. Therefore, the capacitor loses its original function.

inductance

Inductance is used to control EMI in the PCB. For an inductor, its inductive reactance is proportional to the frequency. This can be explained by the formula: XL = 2πfL, XL is the inductive reactance (unit is Ω). For example: an ideal 10 mH inductor, at 10 kHz, the inductance is 628Ω; at 100 MHz, it increases to 6.2 MΩ. Therefore, at 100 MHz, this inductance can be regarded as an open circuit. At 100 MHz, if a signal passes through this inductance, the quality of the signal will decrease (this is observed from the time domain). Like the capacitor, the electrical parameters of this inductor (parasitic capacitance between the coils) limit this inductor to only work at frequencies below 1 MHz.

The question is, if inductance cannot be used at high frequencies, what should be used? The answer is, "ferrite bead" should be used. The iron powder material is iron-magnesium or iron-nickel alloy, these materials have high permeability (permeability), under high frequency and high impedance, the capacitance value between the coils in the inductor will be the smallest. Iron powder beads are usually only suitable for high-frequency circuits, because at low frequencies, they basically retain the complete characteristics of inductance (including resistance and resistance components), so they will cause slight losses on the line. At high frequencies, it basically only has a resistance component (jωL), and the resistance component will increase as the frequency rises, as shown in Figure 1. In fact, iron powder beads are high-frequency attenuators for RF energy.

In fact, iron powder beads can be regarded as a resistor and an inductor in parallel. At low frequencies, the resistor is "short-circuited" by the inductor, and current flows to the inductor; at high frequencies, the high inductance of the inductor forces the current to flow to the resistor.

In essence, iron powder beads are a "dissipative device" that converts high-frequency energy into heat. Therefore, in terms of performance, it can only be explained as a resistance, not an inductance.

Figure: Characteristics of iron powder materials

Transformer

Transformers usually exist in power supplies. In addition, they can be used to insulate data signals, I/O connections, and power supply interfaces. Depending on the type and application of the transformer, there may be a shield between the primary and secondary coils. The shield is connected to a grounded reference source to prevent capacitive coupling between the two sets of coils.