Hello guys,

I am wondering if a Full Bridge LLC or a Phase Shifted Full Bridge is a better choice for a 400V to 100V (up to 1500W) converter with a variable load.

Both are suitable for high-power applications. Additionally they have a similar component count.
So the main difference is the Method of operation as well as the stress in the components.

Full Bridge LLC
- limited output voltage range (due to gain/frequency)
- operates between fmin-fmax with a variable frequency
- the variable frequency increases the complexity of the magnetic component design
- highest efficiency at resonance
- "always" ZVS even at no load condition
- synchronous rectification is more complicated (due to variable frequency)

Phase Shifted Full Bridge
- wide range of output voltage
- operates at a fixed frequency, decreases EMI
- simpler component design
- good efficiency
- ZVS only at heavy load condition
- synchronous rectification with fixed frequency

It seems like the only main drawback of the Phase Shifted Full Bridge is that it does not support ZVS at low/no load condition. But still the majority of on-board chargers use a Full Bridge LLC.

What are you thoughts about benefits & disadvantages of the two converters?

 

I guess due to the high power of on-board chargers they use LLC, due to the fact that this topology has in total less EMI. Thanks to the sharp current slope/edges high frequency content is introduced in the PSFB topology, which creates higher EMI. The design of EMI filter is easier for the PSFB because of fixed switching frequency. But when considering that the converter is drawing the power from an intermediate DC voltage bus the EMI doesn't get so problematic.

The efficiency of the LLC seems to be equal or even worse than the one of the PSFB when it's working over a wide load range. This efficiency can be increased by lowering the DC bus voltage, which is created by the PFC converter and lowers the switching frequency of the LLC. But that approach is not always an option.

The power dissipation is at a narrow load range higher in the PSFB. In a narrow load range the resonance tank can be designed that it works optimal in the needed range.

In both topologies the main switching losses in the primary switch are the conduction losses (depending on Rds_on). Unless the converter is running at very high frequency, which would make turn-off losses dominant. The turn-on losses can be ignored due to the possibility of ZVS. Additionally the turn-off losses can be neglected, when a GaN transistor is being used in the design (no reverse recovery).
The PSFB experiences in those primary switches higher RMS current (due to the sharp current wave form) and therefor higher conduction losses.

It should be noted, that EMI and power dissipation is not the same and does not necessarily correlate with each other.