One of the LTC7890 phases failing in a 2-phase synchronous configuration

Hello,

I am having issues with the LTC7890 GaN-FET controller configured in a two-phase setup and operating as a current source to drive a diode load. The dev board with EPC FETs was reconfigured for such an application and was successfully implemented into the prototype design. Please see the attached schematic to understand the working principle and connections.

The issue is that with the new prototype board, two of GanSystems FETs are used for the high side, and the SW1 node is being damaged easily.

Tests Conducted:

1. Tested the board with a single FET on the high side under various load conditions. Everything worked as expected.
2. Tested the board with two FETs on the high side but with a DC setpoint voltage fed straight into the ITH1 pin. Everything worked under various load conditions and duty cycles.
3. With feedback connected and two pairs of FETs on the high side, the LTC fails. Typically, the bootstrap voltage supply for the HS FET drops to 0V for one phase, while the other phase functions normally. Sometimes EXTVCC, connected to a 6V supply from a lab bench PSU, shorts out. Disabling and reenabling the IC via the EN pin results in SW2 functioning while SW1 does not. In these cases, the LS FET of SW1 remains open, shorting SW1 to ground.

Initially, I suspected Cboot, so I increased it from 100n per phase to 2.2u and later added a 4.7-ohm resistor to remedy transients. However, this did not help.

Switching frequency does not affect the failure. With a DC setpoint voltage fed into the ITH1 pin, the frequency was swept up to 2MHz, and the device continued to function. With feedback connected and fsw set to 500KHz, it continued to fail.

Attached are the following for reference:

1. Schematic Diagram: Highlighting the areas where issues are suspected.

Any insights or suggestions on what might be causing this issue would be greatly appreciated.