DC Dc converter efficiency

[skn96]

Hi All
I'm trying to write down some requirement for power supplies between 1 to 2 kW range. I do not know how much is the reasonable efficiency numbers that people achieve these days for PFC , LLC resonant and phase shifted full bridge DC-DC converters between 1 to 2 kW range at full load? Is 94% for 1.2kW PFC and 95% for 1.2kW half bridge LLC and 96% for 2kW phase shifted full beige too optimistic?

 

[Warpspeed]

Its difficult, because efficiency costs dollars.

Going from 95% to 97.5% means halving the overall losses for only a 2.5% improvement.
So you very rapidly slam into the very solid brick wall of diminishing returns.

Management and sales departments get really excited about efficiency.
But when you cost it all out they are then often far from amused.

 

[Velkarn]

efficiency depends on output voltage.
what is the value of output voltage?

 

[mtwieg]

None of those numbers are infeasible (for peak efficiency, not average efficiency), but they require you doing an excellent job of selecting the power components, and having the tools to estimate the losses of each component in the actual prototype (hope you have a nice thermal camera).

Going from 95% to 97.5% means halving the overall losses for only a 2.5% improvement.
So you very rapidly slam into the very solid brick wall of diminishing returns.

It's a matter of perspective. That 2.5% improvement means the converter generates half as much heat for the same power level. If your application is temperature constrained, then that means you could push twice as much power through it. Not a trivial return at all. Hence why many SMPS like laptop chargers (which always have very little cooling available) go out of the way to get very high efficiency.

 

[skn96]

My output voltage for both LLC and phase shifted full bridge is 170V and input 400V. For PFC the input is universal 85Vrms to 265Vrms and output 400V.
I understand that I can have very good efficiency numbers by spending money and sacrificing size on magnetic and semiconductors but we are not designing this product for best efficiency value in Market. These are audio amplifiers and efficiency not the most important spec. I just wanna know what is reasonable range of efficiency that I can achieve with reasonable traditinal design? For example: 92% on PFC for 1/3 of power and 93% on LLC and full bridge for 1/3 of power is reasonable or too aggressive?

 

[Velkarn]

i designed FSFB converter Vin=100-200V; Vout=100V; Pout=600W; no sync rectification; with peak eff=95% but it was hard optimized with eff goal.
for PFC stage you can refer to some refdesigns:
https://www.fairchildsemi.com/reference-designs/RD-355.pdf
**broken link removed**

so i think "92% on PFC for 1/3 of power and 93% on LLC and full bridge for 1/3 of power" is ok, you can write even 93-94% for PFC stage

PS. "1/3 of power..." you mean for 30% of max load?

 

[skn96]

Thanks for your response. Yes by 1/3 of power I mean 1/3 of maximum load.