SNVSCB7A november 2022 – april 2023 TPSF12C1-Q1
PRODUCTION DATA
Figure 9-2 shows a schematic diagram of a 7.4-kW high-density OBC with conventional two-stage passive EMI filter. The CM chokes and Y-capacitors provide CM filtering, whereas the leakage inductance of the CM chokes and the X-capacitors provide DM filtering. Similar to TI reference designs TIDM-02013 and PMP22650, the circuit uses a two-phase totem-pole (TTPL) power-factor correction (PFC) front-end followed by a full-bridge CLLLC topology with active synchronous rectification.
The TTPL PFC stage runs at a fixed switching frequency of 100 kHz. The CLLLC isolated DC/DC stage runs at a variable frequency from 200 kHz to 800 kHz (500-kHz nominal) and provides galvanic isolation in addition to battery voltage and current regulation. Even though the use of LMG3522-Q1 GaN switches enables an open-frame power density of 3.8 kW/L, the conventional passive EMI filter occupies over 20% of the total solution size.
Note that the DC/DC stage in particular increases the CM EMI signature based on the high dv/dt of the GaN power switches, the transformer interwinding capacitance as well as the various switch-node parasitic capacitances to chassis ground.
This application example replaces the two Y-capacitors, designated as CY3 and CY4 in Figure 9-2, with a single-phase AEF circuit using the TPSF12C1-Q1. See Figure 9-3. The AEF circuit provides capacitive multiplication, which reduces the inductance value and thus the size, weight, and cost of the CM chokes, now designated as LCM1-AEF and LCM2-AEF. The total capacitance of the sense and inject capacitors is kept less than or equal to that of the replaced Y-capacitors, which results in the total line-frequency leakage current remaining effectively unchanged or reduced.