Hi,
Do you agree that the attached 200W pushpull (11.4V input, 350V output, 100kHz) absolutely needs the Active SEPIC snubber with it?
Without this, the primary snubber losses would be way too high.
The SEPIC clamps the leakage energy to under 60V, and regenerates the leakage energy back to the input.
So its "lossless", and <60V rated FETs can be used for the pushpull, so low Rds(on)
Absolutely? For a 200W supply?
Hmmm…
How much power does the snubber lose, 10%? If so, does the added complexity and cost will offset some -but not all- of those losses? let’s say that you will recover 16 watts?
IMHO, this added complexity would make sense for a much larger supply.
The 0.5LI^2 energy in the small leakage inductance is high even though power is low...because vin is also low, and so iin is high, and the i^2 term of 0.5.L.I^2 is thus high...as such it is needed.
The SEPIC is dead simple, and being a sepic, it has no high side drive, just one fet.