Shoot through current inherent in half-bridge SMPS's

Hello,

Have you ever heard of the problem, with a half-bridge SMPS, that when the bottom FET is switched ON, the sudden
falling voltage at the source of the upper FET, means that unfortunately the upper FET gets turned on for a short time as well
and enormous shoot-through current occurs.

This is due to the upper FET's Gate-Source capacitance suddenly getting charged up above its ON threshold.....it gets charged up via the upper FETs own drain-gate capacitance.

Have you heard of this problem, and do you know how to mitigate it?

The problem is well known with badly designed gate driver circuits, that aren't able to pull-down the gate during turn-on of the opposite switch. High power IGBT designs have a negative gate voltage for the same reason.

A certain current pulse is however caused by charging Cds, you should watch the gate voltage to detect shoot-through.

I think using gate drive ICs such as IR2110 and L6385E will not cause this problem.

One more thing that I always do when using MOSFETs is that I always connect 1k resistors between the gate and source of the MOSFETs. I've seen many a times where my circuits have failed due to absence of that resistor. It may help discharge the gate-source capacitance, maybe.

Hope this helps.
Tahmid.

the 1K gate-source resistor is there for when your gate drive IC loses control voltage and you want to ensure your mosfet has a tendancy to turn OFF.

FvM is correct, best way to protect against parasitic turn ON is to apply a negative voltage to the gate. some chips have this ability, some do not. IGBTs require more negative than MOSFETs.. i usually do -10V to +15V on IGBT (with bi-direction TVS for protection set at 18.5V).

also, a good gate drive layout is important. you can have emitter copper pours under your gate drive IF and ONLY IF your circuit is well isolated from neighbouring gate drives. when you lift the board in the air you should see LIGHT GREEN colour all around your gate drive circuit indicating that there is absolutely no copper traces or pours sneaking through the region of the isolated gate drive circuit.

finally. .with a good de-saturation circuit you should be able to detect shoothrough and turn OFF IGBT within 5us.