DCM Full Bridge SMPS goes into CCM if transformer leakage inductance is too much?

[treez]

Hello,
The attached (LTspice sim and pdf schematic and excel design doc) show a DCM Full Bridge SMPS with
Vin=390vdc
Vout = 400vdc
Pout = 100w
Fsw = 100khz
DCM

The actual exact leakage inductance value in the transformer plays a big part in the design calculation. This is because the secondary side transformer leakage inductance appears in series with the output inductor when it is ‘discharging’. Therefore, it is possible that having too much leakage inductance will take the converter into CCM and induce subharmonic oscillations. This is proven by noting that the attached simulation suffers subharmonic oscillation when transformer coupling is 0.995, but suffers no subharmonic oscillation at all when the transformer coupling is unity.

We need to find the exact amount of secondary side transformer leakage inductance so that we can deduct that from the output inductor value, so that we stay in DCM.

The coupling of k=0.995 gives a leakage seen from the primary of 156uH. What proportion of that leakage inductance is in the secondary side? Do you agree that its 134uH?

 

[Easy peasy]

it's normal to divide the leakage equally side to side, taking into account the turns ratio...

Our E55 and E65 transformers (2.5 & 3.5kW) 35kHz, 250VDC: 65V have about 1-2uH leakage on the pri side

This gives very good results under all conditions and only minimal snubbing needed to clean up the waveforms, helping with RFI...