Mergence – what is it and why to use it?
The use of an inductor in series with a transformer is commonly needed for the operation of many power converters such as DAB, LLC, CLLC or ZVS PSFB.
Three approaches can be followed with the magnetic components:
- Use two separate components.
- Use a single transformer with poor coupling and high leakage inductance.
- Use a mergence where the inductor is placed on top of the transformer.
The mergence benefits from a higher power density than two separate components and depending on the application lower power loss than a poor coupled transformer. Why? Because option 2 generally requires the use of non-interleaved windings or even separated windings where the AC proximity losses could be significantly high.
By having a third core half with a center leg gap the series inductance can be easily controlled while optimizing at the same time the transformer winding structure and therefore minimizing the AC proximity losses. Moreover, the magnetic flux in the common core leg of the inductor and the transformer is generally cancelled and therefore the overall core losses of the magnetics can be also reduced.
Does it always make sense to use it? No, the mergence approach would not be always the best solution and each power converter needs to be analyzed carefully. Imagine a power converter where the transformer requires a big core such as an E71 and the series inductor can be done with a smaller core such as an E32. There it makes no sense to merge both components since 3 E71 halves would lead to a bigger and less efficient solution.
Is there any coupling between the inductor and the transformer in the mergence structure? Of course, sharing a common leg there is certain coupling. However, since the inductor requires a gap to store the energy its reluctance path would be much lower than that of the transformer and its flux will tend to follow within its core and the shared leg rather than through the transformer.
What are the major drawbacks? The coil former needed in the third half core is normally not standard. This means that unless you have a production volume relatively high the cost of opening a mold for a new plastic part might not be justified.
Is there any other benefit? Yes! By having the same current on the series inductor and transformer the same wire can be used. This allows the possibility to wind both coils at the same time and avoid the need of an external connection which means less termination losses and more efficiency.
Can this approach be extended to other inductors different than the one in series with the transformer? Yes, it can be done but special care needs to be taken with the fluxes in the shared leg. Imagine for instance the output inductor of the PSFB which has double the frequency of the transformer. During one half switching cycle the inductor and the transformer fluxes would be cancelled but also added. Such addition could lead to transformer saturation and failure.
Is there any power or frequency limitation? No as far as the flux cancellation and core and winding losses are properly computed.
To conclude mergence are not a panacea but could provide a really good performance and power density in many applications.