I would like to implement a boost PFC stage for an alternator that has a large leakage inductance (~40mH). When looking at the schematic, we can see that the boost inductance
Lb and the alternator inductance
La are connected in series (through the diodes
D2, depending on the half-cycle), and that the large value of
La means that it cannot be ignored (and that
Lb is useless and can be removed).
In theory we need to measure the instantaneous alternator voltage
Va, in order to shape the alternator current
Ia waveform the same way (scaled with the error on output voltage
Vo). This is clearly impossible as
La are intrinsically the same component, so
Va+ is not available.
At first I was planning to measure the current and rectified voltage
Vr (as should be done in the absence of leakage inductance
La), and to compute what
Va should look like, knowing the value of
La. But then I realized that as soon as I will start switching
M1, I will alter
Vr (it is the purpose), so I can't use them anymore.
I have a few ideas in mind about how to workaround this, but they all seem quite impractical (like measuring the derivative of the current
M1 is closed). In addition I was thinking that this problem should be classical, for instance in wind turbines, but I did not manage to find information about this on the web.
Is this a known problem, or is
La usually insignificant and ignored? Does a robust solution exist to implement a proper PFC in this situation?