I am trying to simulate a part of a power supply, but after I added lead inductance values for the IGBT model, ltspice ends with time step too small error. Maybe @Bordodynov knows how to make it work?

 

Hi.
The coupling coefficient, which differs from unity, can greatly slow the calculation.

 

Thank you. What are the advantages of your version with coupling caps? I understand the cap on the primary should prevent flux walking and saturation of the core, but beacause the transformer will be driven with dead time between positive and negative pulse, where both ends are connected to ground I think that should be enough to reset the transformer.

 

See

 

Ok, but that is with single ended drive. I will be using a bi-polar drive analogous to the waveform in a full bridge converter, see the waveform of the source in the original simulatuion. Then I don´t think that the DC-restore caps provide any advantage.

 

Dead time

 

This is what I am simulating, the controller already provides dead time between the positive going and negative going pulses. http://www.utaes.g6.cz/Invertor_130/schema.png