Hello,
Can someone provide a suggestion of what it would be a good topology and recommend any paper of book for a Class 1 isolated DC-DC with the basic requirements: Input=400Vdc, Output = 24Vdc, output power up to 500W, high efficiency like >85% at full load. I really would appreciate any input from someone about that. Thanks.

Input range will affect the choice.

The basic options to me would be:

If you need a reasonable range for holdup time, etc:

1) if you want to do it fast, two-switch forward is the least complicated, most rugged.

2) If you want more efficiency and lower noise, the phase shifted bridge

If the input and load range is not too big and you don't mind spending a lot more time on it

3) the half-bridge LLC

for all of the above you can raise efficiency and complication with synchronous rectifiers. For the first two, you need to figure out how to clamp the secondary ringing.

I'm sure you will get many more suggestions here. Everyone has their favorite circuit, often the one they have just built so they know exactly how to do it!

Many thanks for your comments Ray. Actually the input voltage will be well regulated at 400Vdc, as it will be provided by a PFC stage. The output load in reality will change a bit as its impedance will not change, but the output voltage will be adjustable from 15V up to 28V, in order to have some control of the total output power.

Do you know a good way to have this control on the output voltage without major complications? A digital control can be also considered.

Thanks,

Since your output voltage is not constant, the third option (LLC) will not be optimal. In your case phase shift will be a good option.

An alternative would be to use the 2-switch forward converter with a passive regenerative turn-off snubber of the type suggested in US patent 4675796 – “High Switching Frequency Converter Auxiliary Magnetic Winding and snubber circuit” – Gautherin et al.
Since the converter has negligible turn-on loss due to the transformer leakage inductance anyway, this should lead to a significant improvement in efficiency as a result.

I support Ray's suggestion. If you require just >85% efficiency then, two-switch forward should an easy option. Especially when one has already experience on it (in fact it applies to each converter).

Even with a PFC input, the input voltage is not well regulated - if you have any ride-through time requirement. Typical requirements are to maintain the output in regulation with either a one half cycle (10 ms) or full cycle (20 ms) outage of the ac input power. In either case one must trade off the range of regulation of the dc-dc stage with the size of the hold-up capacitor.

With a dc bus regulated at 380 Vdc it would be nice to be able to use half the energy stored in the hold-up capacitors during the hold-up time, but that requires operating down to 270 Vdc. That usually requires too much of a trade off in normal operation efficiency. More typically one designs to operate down to 330 Vdc which only uses 25% of the stored energy.

Efficiency spec of only 85%. You could get past 90% with a simple 2-switch forward or a full bridge. If you go with a phase-shift and/or synch-rect you should be able to get much better efficiency, but it sounds like you don't need to. Do you have a weight and size spec? Your size requirement, i.e. heat sink size is a key factor.

Designing the power transformer will be your main concern since you have to take into account the output voltage range (1:2 ratio) and the inherent 120Hz ripple voltage superimposed to the regulated 400Vdc.
One approach would be to design a high efficiency 400Vdc/36Vdc power stage followed by a buck converter (synchronous rectifier). That way, the power transformer would be optimized in size and efficiency while meeting the required galvanic isolation class.
The power stage topologies described above are all well suited for your project but the one that interests me the most is the Microchip one.

Two stage power conversion will result in lower efficiency compared to full bridge /2 switch forward converter with synch rect

As far as the overall efficiency is met, 85% for this project, it doesn’t really matter which topology is chosen. But, this project requires an adjustable output voltage from 15 V to 28V and that’s the main concern. My approach would be to use a high efficiency power stage where the duty cycle would be independent of the output voltage. Then, a high efficiency step-down converter would be used to adjust the duty cycle to provide the desired output voltage