A Text About EMI, PCB Design and Switching Frequency

A Text About EMI, PCB Design and Switching Frequency

 

The article guides

 

Since the development of the power module, engineers have focused on how to make the module more compact and lightweight. In fact, everyone knows that the power density of the product can be improved by increasing the switching frequency. But why hasn't the module's volume changed so far? What limits the increase in switching frequency?

 

Switching power supply products are dominated by the market, and increasingly require small, lightweight, high efficiency, low radiation, low cost and other characteristics to meet a variety of electronic terminal equipment, in order to meet the current portable electronic terminal equipment, the switching power supply must be small, The light weight features, therefore, increasing the operating frequency of the switching power supply has become a problem that designers are paying more and more attention to. However, what are the factors that limit the frequency increase of the switching power supply? In fact, it mainly includes three aspects, switching tube, transformer and EMI and PCB design.

 

  1. The switch tube and switching frequency

Switching tube is the core device of switching power supply module. Its switching speed and switching loss directly affect the limit of switching frequency. The following is an analysis of everyone.

 

(1) Switching speed

The loss of the MOS transistor is composed of switching loss and driving loss, as shown in Fig. 1: the turn-on delay time(on), the rise time, the turn-off delay time(off), and the fall time.

Figure 1 Diagram of MOS tube switch

 

As shown in Figure 2: FDD8880 switching time characteristics table.

 

Figure 2 FDD8880 switching time characteristics table

 

For this MOS tube, its limit switching frequency is: fs = 1 / (td (on) + tr + td (off) + tf) Hz = 1 / (8 ns + 91 ns + 38 ns + 32 ns) = 5.9 MHz, in practice In the design, since the control switch duty cycle realizes voltage regulation, the conduction and cut-off of the switch tube cannot be completed instantaneously, that is, the actual limit switching frequency of the switch is much smaller than 5.9 MHz, so the switching speed of the switch tube itself limits the switching frequency.

 

(2). Switching loss

The corresponding waveform diagram when the switch is turned on is shown in Fig. 3(A). The corresponding waveform diagram when the switch is turned off is shown in Fig. 3(B). It can be seen that the VDS voltage of the switching transistor and the flow through the switch tube are turned on and off each time the switch tube is turned on and off. The current ID has overlapping time (yellow shaded position in the figure), resulting in loss P1, then the total loss PS=P1 *fs in the operating state of the switching frequency fs, that is, the number of times the switch is turned on and off when the switching frequency is increased The more the loss, the greater the loss, as shown in Figure 3 below.

 

Figure 3 Diagram of switching tube loss

 

  1. Transformer iron loss and switching frequency

The iron loss of the transformer is mainly caused by the eddy current loss of the transformer, as shown in Figure 4.

 

 

 

When a high-frequency current is applied to the coil, a varying magnetic field is generated in the conductor and outside the conductor perpendicular to the current direction (1→2→3 and 4→5→6 in the figure). According to the law of electromagnetic induction, a changing magnetic field generates an induced electromotive force inside the conductor, and this electromotive force generates eddy currents in the entire length direction (L side and N side) of the conductor (a→b→c→a and d→e→f→d). Then, the main current and the eddy current are strengthened on the surface of the conductor, and the current tends to the surface. Then, the effective AC cross-sectional area of the wire is reduced, resulting in an increase in the AC resistance (eddy current loss coefficient) of the conductor and an increase in loss.

 

Figure 4 Diagram of transformer eddy current

 

As shown in Fig. 5, the transformer iron loss is proportional to the kf power of the switching frequency, and is related to the magnetic temperature limitation. Therefore, as the switching frequency increases, the high-frequency current flows in the coil to generate a severe high-frequency effect. Thereby reducing the conversion efficiency of the transformer, resulting in an increase in the temperature of the transformer, thereby limiting the switching frequency.

Figure 5 Diagram of transformer iron loss and switching frequency

 

  1. EMI and PCB design and switching frequency

 

Assuming that the power device loss described above is solved, the real high frequency also needs to solve a series of engineering problems, because at high frequencies, the inductor is not the inductor we are familiar with, the capacitor is not the capacitor we know, all the parasitic parameters. Corresponding parasitic effects will occur, which will seriously affect the performance of the power supply, such as the parasitic capacitance of the transformer's primary side, the leakage inductance of the transformer, the parasitic inductance and parasitic capacitance between the PCB wiring, which will cause a series of voltage and current waveform oscillations and EMI problems. The voltage stress of the switch tube is also a test.