High efficiency is one of the most significant factors in power electronics and the most challenging parameter to obtain. Wideband switches support achieving this objective, but they add to the system’s expenses. That is where LLC resonant converters come in; this sort of converter offers a number of benefits over others, including good electrical isolation, high power density, low EMI, and high efficiency.
Furthermore, LLC resonant converters may be employed in a variety of applications, including consumer electronics and renewable energy applications such as PV, wind, hydro, and geothermal, among others. This article compares the Si and SiC MOSFETs modeled in a 3KW half-bridge LLC converter with a large input voltage range. The results reveal that SiC MOSFETs are more efficient than Si MOSFETs for high-frequency power applications. Despite this, Si MOSFETs are still popular in low-voltage and low-power applications due to their low cost.
Working of Half-Bridge LLC Resonant Converter
The circuit of a half-bridge LLC resonant converter is shown in the diagram above. At the frequency fs, two MOSFETs, Q1 and Q2, are turned on and off 180o out of phase. Each MOSFET has the same on time, which is slightly less than 50% of the switching period Ts = 1/fs. A small dead-time Td is also added between the turn-off of one MOSFET and the turn-on of the other. This dead-time is critical for the converter’s performance since it ensures that Q1 and Q2 do not cross-conduct and permits soft-switching (zero-voltage switching ZVS at turn-on) for both MOSFETs.
Even though the resonant tank has three reactive elements (Cr, Ls, and Lp), it has two resonance frequencies. One is associated with secondary winding conduction: only Ls is active, whereas Lp vanishes due to a constant voltage across it reflected back from the secondary side; its value is:
The other resonant frequency is associated with the secondary winding(s) being open. As Ls and Lp are effectively in series, the tank circuit switches from LLC to LC:
Frequency fR1 has a higher frequency than fR2. The distance between fR1 and fR2 is determined by the Lp to Ls ratio: the greater the ratio, the greater the distance between the two frequencies, and vice versa.
Power losses in Half Bridge LLC Resonant Converter
On the primary side, the conduction power losses are located in the power MOSFETs, the resonant capacitor Cr, and the transformer. Additional losses are found in the secondary rectifiers as well as the output capacitors.
Power MOSFETs have significant switching losses. The magnitude of the switched current with a large load is a trade-off between the need for both MOSFETs to have zero voltage switching (ZVS) and their turn-off losses. The larger the switching current, the greater the margin for ZVS.
Half-Bridge LLC Resonant Converters with Si and SiC MOSFETs
Technology advancements and the usage of wide bandgap (WBG) power devices such as Silicon (Si) and Silicon Carbide (SiC) have offered a replacement for traditional approaches. A more robust system with higher dependability can be attained because of its superior switching speed, less switching loss, higher efficiency, and higher power density. The key advantages of high-frequency operation include a smaller transformer and EMI filter, as well as an integrated resonant inductor into the transformer, which minimizes converter size even further. Due to the reduced cross-talk afforded by ZVS, MOSFETs can operate reliably even in the absence of a negative bias drive voltage, cutting the cost of the driving circuit.
To examine the performance of silicon carbide (SiC) MOSFET (SCT3080KL) and silicon (Si) MOSFET (STW45NM50), a 3 KW half-bridge LLC resonant converter was modeled. In one model, the SiC MOSFET was employed, whereas, in another, the Si MOSFET was used. The converter was designed and controlled using frequency modulation technology, which controls the output voltage by changing the frequency. The following are the calculations for conduction losses, switching losses, and total losses:
Calculation of conduction losses across a MOSFET
Pcond = Irms2 ∗ Rd ∗ Duty cycle
Calculation of switching losses across a MOSFET:
Psw = Vin ∗ I peak ∗ toff ∗ fsw
Total losses in all the MOSFETs will be
Ptotal = 2 ∗ (Pcond + Psw)
The graph above portrays loss comparisons of SiC and SiC at full and half load. It implies that switching and conduction losses in SiC MOSFETs are lower than in Si MOSFETs. Silicon carbide (SiC) MOSFETs are more efficient than silicon (Si) MOSFETs.
The graph above depicts the switching and conduction losses of Silicon and Silicon Carbide MOSFETs at various input voltages. It demonstrates that when input voltage increases, switching losses increase. Conduction loss is greatest at low voltage. The switching losses of Silicon Carbide SiC MOSFETs are substantially lower than those of Silicon Si MOSFETs, according to the results.
Outcomes of Simulation
The efficiency analysis of silicon and silicon (Si) carbide (SiC) switches for different loads is shown in Figure (a). The load efficiency of the silicon MOSFET converter is substantially lower than the load efficiency of the SiC MOSFET converter at a load of 25%, according to the analysis. The efficiency analysis for changes in input voltage is shown in Figure (b) where efficiency is greatest at nominal voltage, and declines as voltage climbs above nominal voltage.
Wide bandgap silicon (Si) and silicon carbide (SiC) devices have provided numerous options for high efficiency, high power density, and power conservation in a variety of applications. The simulations, together with collected waveforms, show that Silicon Carbide MOSFET outperforms Silicon MOSFET in a Half-bridge LLC resonant converter. When compared to Silicon (Si) MOSFETs, Silicon Carbide (SiC) MOSFETs perform better in higher frequency power applications. However, because of their low cost, Si MOSFETs are still used in low voltage and low power applications.
- Citation: Farooq, H.; Khalid, H.A.; Ali, W.; Shahid, I. A Comparative Analysis of Half-Bridge LLC Resonant Converters Using Si and SiC MOSFETs. Eng. Proc. 2021, 12, 43. https://doi.org/10.3390/engproc2021012043