Part Number: PMP20637
I watched with great interest comparison video between Phase Shifted Full Bridge & FB-LLC --- done a while ago by someone at TI
Let us recast some conclusions in light of superb performance of LLC by many new controllers.
#1 the comparison does not mention any output inductor. This is a significant drawback- of any topology which has to handle very low current to high. In a world of GaN devices, the size of any DC inductor will kill the compactness of any product. Not to mention huge loss & then damping one has to do. For small ac ripple filtering, a small inductor can be easily accommodated.
#2 It says that LLC topologies have penalizing output rms & ripple and require 9 times more output capacitor. In my model( obtained from mathworks public repository of models in Simulink), & in many of TI's LLC design guides, this does not appear to be true. Especially for high power case of Interleaving & with SR, without the need for DC inductor, LLC appears to be very competitive.
#3There are key failure modes in PS-FB topologies which is a serious drawback. Also not mentioned.
#4 ZVS is very hard to extend and hard switching is a nuissance.
One thing is very well discussed is the inability of LLC to be synchronized and variable frequency control. I do not know about audio applications, this has not been a problem in my experience in very critical application we handled over decades.
SO unless sync is a requirement, I would proceed with our plans to implement all GaN HB LLC using UCD3138.
Any contrary reviews will be highly appreciated.
-robin
