It’s showing me " doAnalyses: TRAN: Timestep too small; initial timepoint: trouble with node “probe_int_gndpwr_rload1_1”
run simulation(s) aborted "
First, your power supplies are incomplete. Fix that first, then come back.
Cheers.
ngSpice needs a node with the net name “0” (Zero), but KiCad is smart enough to translate the commonly used “GND” net to the Zero that ngSpice needs. Without that reference simulations don’t work properly. I don’t know whether the “GNDPWR” gets translated properly.
For the rest, your schematic is an huge mess. There is hardly anything in it and still it took me a few minutes to get an idea of what is going on. Looks like the two FET’s are in series and the FET drivers have their own floating power supplies.
Getting any sort of simulation working is finicky, especially with KiCad + ngSpice. It’s not as well integrated as with other simulators, and very silly mistakes can be hard to spot. Therefore I recommend to always start with a very simple simulation, make sure that works first, and then extend in small steps. For example: If you start with one FET connected to a PWM / pulse source and one power supply, then you can verify that the FET model works, and you can exclude problems with the FET model and it’s pinout. In another step, verify that the UCC535 model works at all.
in ngSpice “Timestep too small” is a catch all error that happens if ngSpice can’t converge for some other reason. If it can’t converge to a solution, then it lowers the timestep and tries again, and at some point it gives up with a “Timestep too small”.