Category: Hardware
Product Number: LT8306
I have an open support ticket (CS-592333-Q3Z9D1) but have not received a response yet-- posting here as well in case the team is more responsive.
I have a flyback design using the LT8306 that works great in steady-state. Pulling the chip into shutdown while the flyback is operating at full-load, however, routinely causes the mosfet to fail short-circuit due to what appears to be an overcurrent event. I've attached an LTspice simulation that is accurate to the test condition that produces this failure, as well an image of the actual design schematic. I've also captured the failure event on the oscilloscope twice:
First failure captured on oscilloscope (no modifications to circuit yet). Yellow is enable voltage. Green is drain voltage. Blue is gate voltage. After shutdown, gate voltage remains high for over 100us, resuling in an overcurrent event that fails the mosfet short circuit. Our working theory: the enable voltage isn't pulled under 200mV due to weak GPIO, and perhaps there is trapped charge on the mosfet gate when the gate drive goes high impedance after shutdown.
Second failure capture. Changes to circuit include implement an open-drain shutdown fet as seen in LT8306 datasheet figure 27, and add a 1kohm resistor between mosfet gate and source for extra bleed path. We were running an endurance test toggling the enable every 3 seconds. On the 50th iteration the circuit failed and the attached waveform captured. Inexplicably, the gate is driven high for a long duration of time ~20us after shutdown. Yellow is the enable (now active low), blue is the gate voltage.
