yes, as described here already, 10kw/m^3 is a good benchline for a maximum power density...and of course, you must note whether you are usinf flyback type or bridge type transformer.....as dont forget the flyback only magnetises the core in one direction, so you get a bit higher allowance with that, ...bearing in mind that the curves in these datasheets refer to bipolar stimulation of the core....and usually of course, they refer to stimulation by a sinusoid, and of course, thats often not what you have in an smps.
Treez, thank you for your post.
Actually that was my very first choise. The transformer design of the converter (Push-pull, 500 W, Vin=24 V, Vout= 200 V) lead me to a core (https://en.tdk.eu/inf/80/db/fer/e_65_32_27.pdf), which is not available neither in Mouser nor DIgikey nor Arrow, and it would be available in march. Instead, there are N27 cores... that's the background of my question.
respectfully, 10kW/m^3 is 10,000 watts per 100,000 cm^3, i..e. 0.1W per cm^3
these results are measured on a small toroid, in a large core the flux crowding in the corners leads to higher overall dissipation - hence 0.1W / cm^3 often needs heatsinking at this level...
Thanks, yes, we reverse eng'd a 3kw battery charger once, and the ferrite LLC transformer had a few gap pads pressed against it, and these gap pads were also against metal structures (including the metal case)...so heatsunk....it always makes me wonder though, with ferrites poor thermal conductivity, just how effective it is......obviously a darn site better than no heat sink, yes.