SLVAFJ9 March 2023 TPSF12C1 , TPSF12C1-Q1 , TPSF12C3 , TPSF12C3-Q1
Since magnetic components for both CS and VI are large-sized and likely custom parts (offsetting the size reduction enabled by AEF), it’s a good idea to select an AEF topology that precludes the use of additional magnetic components. The VSCI implementation leverages capacitors in combination with low-voltage active circuits for sensing and injection, and thus achieves a smaller size [5].
Figure 7-1 shows a simplified single-phase schematic to illustrate the basic principle of the selected FB-VSCI circuit in a CM filter setup. As mentioned previously, the main idea with this AEF topology is to use an injection capacitor with a value similar to the Y-capacitance in an equivalent passive filter in order to reduce the values of the CM chokes, which are the largest components in high-power filters.
The Thevenin-equivalent CM noise source consists of voltage source vS in series with source impedance ZS, which is considered capacitive. The mains impedance ZGRID is typically inductive. CM chokes, designated as LCM1 and LCM2, also function as decoupling elements to achieve high source and load impedances, as required for a high-attenuation FB-VSCI design (see row "d" in Table 6-1).
With Y-rated sense and injection capacitors connected to the AC lines, the purpose of the circuit is to reduce the total filter volume, yet maintain low values of the low-frequency earth leakage current using an active circuit that shapes the frequency response of the injection capacitor – effectively increasing its value for high frequencies. In turn, this amplified injection capacitance over the frequency range of interest is the key to lower CM choke inductances relative to values of a passive filter with equivalent attenuation.
The circuit advantages are: