12 windings mutual inductors

PLECS
1

Hello everyone,

I am having trouble understanding how to correctly model multiple mutually coupled inductors in PLECS.

My system consists of 12 windings in total. I am studying a simple power supply with a single load, modeled as one inductor and a R series, but I would like to include the interaction between this main winding and the other 11 windings in order to observe the effect of the voltages and current induced by them on the main one.

I already have the complete 12x12 mutual inductance matrix, which I entered into the PLECS Mutual Inductors block. To study the effect of the other windings on the first one, I closed each of the remaining 11 circuits using a controlled current source, forcing the exact current in each winding. Esch imposed current is derived from the following equation:

V_Li = R_i * i_i + L_i * di_i/dt + sum(M_ij * di_j/dt), with j ≠ i

However, this approach becomes very difficult to manage in PLECS, mainly because in the mutual inductor block all windings are tightly interconnected and the model becomes hard to read and control.

For this reason, I was considering an alternative approach: splitting the problem into multiple two-winding subsystems. Specifically, I would like to create several 2x2 mutual inductance blocks that model the coupling between winding 1 and each of the other windings (1–2, 1–3, 1–4, …, 1–12). Then, I would connect the primary winding (winding 1) of each block in series, in order to reproduce the overall effect of the other 11 windings on the main one, but putting in series i think i will sum correctly the mutual voltage terms but i sum also the main inductor voltage for 11 times.. am i wrong?

In your opinion, is this approach conceptually correct, is possible to do or you can please suggest me a way to do it? thanks in advice

2

Dear @Gianluca_97

From your description, I would also suggest that the Mutual Inductor block is the best fit for your application. Since you have already tried this approach, I would be interested to know what exactly you find cumbersome about using this block.

I can understand that the connectivity may feel inconvenient, as all 12 windings are arranged sequentially, but from a modeling perspective this is still the most straightforward and physically consistent solution. To improve readability, you might consider defining the inductance matrix in the Initialization and then simply referencing it in the Mutual Inductor block. This often makes the schematic much cleaner.

You can also place the Mutual Inductor block inside a Subsystem and arrange the external connections in a way that better suits your needs.

Another possible option would be to move entirely to the signal domain and model the system using a State Space component. Whether this approach is actually simpler or more convenient depends on your specific goals and is open to discussion.

If you are able to attach your model, it would likely make it easier to understand your setup and provide more targeted feedback.