Full bridge inverter using UC22510 Not working LTspice

Hi everyone,

I am currently working on the design of a Wireless Power Transfer (WPT) system for an EV Charger. I am at the phase of simulating and validating the primary-side inverter driver stage before integrating the full resonant tank and secondary side.

I am using the UCC21520 isolated gate driver in LTspice, but I am facing persistent simulation stability issues that are blocking me from moving to the next phase of the project.

The Setup:

• Application: EV Wireless Charger (Primary Inverter Stage).
• Driver: UCC21520 (Isolated Dual-Channel Gate Driver).
• Simulator: LTspice.
• Current Phase: Standalone validation of the Gate Driver circuit.

The Problem: I am unable to get a consistent output at the gate pins. The simulation behavior is erratic:

1. D3 & D7 Dependence: When the circuit is complete, I get no output. If I remove diodes D3 and D7, I get output only on the positive cycle (Low-side works, but High-side fails/floats).
2. Intermittent Success: On one specific occasion, the simulation ran perfectly without any circuit changes. I have not been able to reproduce this successful run since.

My Analysis: Since this is a high-frequency EV application, I suspect the issue might be related to the bootstrap circuitry (D3/D7) interacting with the driver model, or a simulator convergence error (given the "random" success).

My Questions:

1. Are there known convergence issues with the UCC21520 model in LTspice for high-frequency WPT applications?
2. Could the "random" success indicate an Initial Operating Point issue?
3. Do D3/D7 require specific parameters (e.g., ultra-fast recovery vs. standard) to satisfy this specific driver model?

I have attached the schematic of the driver stage below.

Any insights would be appreciated to help me get past this driver validation phase.

Thanks,