Hi,
I have problem with mosfet failure in my device with LTC4380IDD-4.
Below is my scheme, inspired by Figure 10. 12V Surge Stopper with Fault Timer Disabled from the datasheet.
The transistor I use is IAUT300N10S5N015ATMA1, current sense resistor is 1mohm.
The load are 13 BLDCs, dc bus capacitance is around 2mF in MLCC, with dc bias derating around 300uF + local 3x10uF MLCC and 100uF electrolytic capacitor(EEH-AZA1H101B). Power supply is li-on battery, 10S(30-42V) with ESR 300mohm.
Clamping Zener on the driver VDD is far above the power supply voltage, and I don't use current limit/monitoring features - device works as regular high side static switch.
We tested up to 700W of continuous power(around 20A) and it works fine, but during high transient loads(almost all actuators with full torque, probably current around 30A - we didn't have possibility to measure as it happens only during accidents).
We have BMS that turns off the battery if the voltage drop is below 30V for a longer time, so in case of such transient we have 9V drop. We didn't notice to BMS cut-off during transients that destroy the mosfet, so it was really short.
I'm not able to reproduce the failure in the safe way, so I wonder if you have suggestion what can cause, that after the high power peak there is mosfet breakdown?
The symptomes are shorted gate, drain and source - in two cases of the failure it was the same behaviour. It tested with the IPT007N06N mosfet as well(750uohm rdson) and it happened also.
I wonder how to protect the mosfet to avoid failures. My initial guess is gate breakdown, so I placed 15V Zener diode from source to gate and I'm doing tests(for now I didn't destroyed is, but I didn't pushed it too hard). I also thought, that the reason is maybe that gate voltage falls, so it works in the actively, and big transient exceed Avalanche energy, single pulse. For the better from tested transistor it is 1100 mJ(ID=150A,RGS=25Ω). Do you have recommendation what can I do to improve reliability during high power transients? It's not for sure thermal failure - we monitor transistor temperature and we didn't noticed overheating.
Thank you in advance
