Guru
Joined: 02/02/2017
Location: AustraliaPosts: 1471
| Posted: 01:08am 15 Feb 2026 |
I agree with you re. long term low power running with some shoot through is bad
for reliability. The failure mechanism is a puzzle for me.
I agree failure results when gate oxide layer degrades enough.
These FETS are robust in the other areas though so some ST will heat
the device and reduce conversion efficiency but leave the oxide layer undamaged?
I took above suggestions and tried 1R2 and 2R2 gate resistors (in place of 10R)
and on the 2R2 side, I tried 270pF and 470pF caps, connecting Gate to Source with short leads.
more DSO captures again..
first is 2R2 with no caps
low side switches on (Yellow is low side gate voltage)
high side switches on (dare blue is high side gate voltage)
and this is where the trouble is located.
and then there are the 1R2 gate resistors. in both cases switching speeds seem a bit faster.
and
there is a lot of high freq activity on the gate voltages, indicating
wild oscillation of the state of the half bridge switching process.
I now put 270pf ceramic cap across the gate and source pins of both
low and high side. maybe less oscillations?
and
and now replaced with 470pF caps
and
not a lot of improvement, sadly.