The Basic Structure and Principles of poe Network Transformers

Due to the parasitic capacitance of non ideal transformers, some electromagnetic interference signals will be coupled to the primary end of the transformer through parasitic capacitance. Therefore, in order to further prevent electromagnetic interference from propagating between the primary and secondary coils, a single T-piece network transformer can be used, and then K pieces can be connected in series to the primary or secondary side of the transformer. This is a T-piece+K-piece network transformer.

Adding K components to the network transformer is because K components do not attenuate useful data voltage signals, but can attenuate EMI.

The balanced data voltage signal uploaded by UTP is applied to Pin 6 and Pin 4 of the secondary coil of the network transformer. The hollow arrow indicates the direction of signal transmission, and the solid arrow indicates the direction of signal current. According to the direction of current, when the signal current passes through the upper and lower coils of the K-piece, the magnitude of the current is equal and the direction is opposite, that is, one flows towards the same end and the other flows out of the same end. The magnetic flux changes inside the magnetic ring of the K-piece cancel each other out, and the magnetic flux changes are zero, which means that the inductance of the upper and lower coils of the K-piece is zero. Therefore, the K-piece has no attenuation effect on the useful voltage signal.

The above figure illustrates the attenuation effect of K components on EMI. The EMI from UTP is applied between Pin 6 and Pin 4, and the solid arrow indicates the direction of EMI current. When EMI current passes through the upper and lower coils of K-piece, its current direction is the same, that is, both currents flow towards the same end. The magnetic flux changes caused by them inside the magnetic ring of K-piece are superimposed on each other, and the inductance presented by K-piece increases linearly with the frequency f.

The inductive reactance is expressed as ZL=2 π * f * L. Among them, L is the inductance of K pieces. At present, the inductance of K components in network transformers produced by various manufacturers is about 10 μ m.

The K-piece is connected in series in the EMI circuit, because a considerable portion of the EMI from UTP needs to drop at both ends of the K-piece before reaching the secondary coil of the T-piece, so the common mode choke can play a role in blocking EMI. The blocking effect of K-piece on the mid to high frequency components of EMI is better, because the higher the frequency f, the larger the ZL, and the greater the EMI at both ends of K-piece.

For current type PHY, common mode inductance cannot be placed on the PHY side. The reason is that the current IS flows from the center tap to one or the other winding of the common mode choke coil according to the switch state, and the current is provided by the bias voltage VCC. The PHY chip controls the flow of current. When the instantaneous current of a normal data signal passes through one or two coils, the changes in magnetic flux inside the K-piece magnetic ring are superimposed and not zero, which means that the K-piece will generate high impedance to the normal signal, thereby affecting the transmission of the normal signal.

As shown in the figure, it is a typical circuit of current type PHY. There are no K components here, but not all cases require the use of network transformers with K components. The decision to use them is based on EMI test results, cost, and other factors. When laying out this circuit, a 0.1 μ F capacitor should be placed near the center tap of the network transformer, while a 49.9 Ω terminal resistor should be placed near the port of the PHY chip.

The structure of this type of network transformer is not particularly commonly used, but it may be used in certain situations, which will be explained together here. The structural diagram of the network transformer composed of T pieces that can be three wire looped through K pieces is shown in Figure 8. This type of network transformer consists of K pieces, where the two ends of the stimulation coil and the three leads of the middle tap are parallel on the magnetic ring of the K piece, with the same number of turns wound.

As mentioned above, for current type PHY chips, placing K pieces on the PHY side will result in high impedance for normal signals. If K pieces are placed on the UTP side, it will affect the impedance matching of the center tap. Therefore, in order to solve the above problems, a network transformer consisting of T-piece and three wire loop K-piece has emerged. This type of network transformer is suitable for scenarios where K-piece is placed on the PHY chip side. As can be seen from the picture:The instantaneous current of a normal data signal flows out from the terminal, passes through the middle coil, and then enters the PHY chip through the upper and lower coils. In the figure, the total current flowing out through the middle coil is equal in magnitude and opposite in direction, and equal in magnitude and opposite in direction. The total current change in the K pieces of the three wire loop is always zero, that is, the total magnetic flux change is also zero. Therefore, data signals can be transmitted unobstructed in the K pieces of the three wire loop.