DC/DC Boost converter with 95% eff.? GaN MOSFETs

[Velkarn]

i designed dc/dc 20-30v to 24v, 100W, 90%eff. converter with active clamp forward topology (hard switched and not optimized for high eff.).
soft switching may be archived in active clamp forward by adding inductor in series with primary.

We build 24 / 24V isolating converters at 400W at 96.5% efficiency

what topology did you use?

 

[Easy peasy]

push pull with lossless snubbers which port the turn off energy to the output...

 

[BradtheRad]

I saw this circuit described as a sepic converter. (It may be a sepic, it may be something else.) Watching it in simulation...

Snubber energy is created at the primary side. It is steered (via diodes and capacitors) to the secondary side. It is added to the output. Hence this may be a lossless snubber as mentioned in post #22.



To obtain 150W, notice the size of current spikes going through various components. There is no doubt a more efficient topology.

 

[schmitt trigger]

No SEPIC....SEPIC requires two inductors or a single coupled inductor.

Would you mind sharing the web page where you found this snubber? Thanks

 

[Warpspeed]

That is a pretty unusual circuit Brad.
That transformer will be huge (for the power level).
3mH 30A primary and 9:1 step up.
A lot of copper in that.

And the switching device still sees the full peak dc output voltage.
Only advantage I can see over a plain vanilla boost converter will be the much lower peak rectifier currents.
Or maybe I am missing something...

My choice for this would be a diagonal half bridge flyback.
Lossless snubbing, all excess energy returned to the source.
Lowest voltage mosfets required, very low rds on possible even with two devices in series.
Flyback transformer provides voltage step up with resulting reduced peak secondary current.

If you rally want to get fancy, two of those in parallel running interleaved would reduce the input and output ripple currents.

 

[BradtheRad]

No SEPIC....SEPIC requires two inductors or a single coupled inductor.

Would you mind sharing the web page where you found this snubber? Thanks

Sorry I did not make a note as to where I found this particular design. It's one of many switched coil schematics I've played with.

This one was called a sepic and it resembled schematics I've seen called 'sepic'. I forget if it had one coupled inductor, or two separate inductors. I had to rearrange it completely just to make sense of it. (In fact it might be a flyback.)

I was experimenting with it in the past year and a half. Finally I realized it made efficient use of snubber energy. The primary snubber current needs a path to ground, so perhaps that is a characteristic of a sepic. From what I can see, there are a variety of sepic designs.