It seems the power dissipation capability of each MOSFET is being exceeded. A suggestion would be to parallel enough of them to allow each to work within its safe operating area.
To dissipate a kilowatt, even for just 50ms, it will probably require several of them in parallel. Some datasheets I've seen for MOSFETs characterize their surge power dissipation rating for just 1ms as the longest increment. At 50ms, the surge rating probably won't be much higher than the steady-state figure.
I'm wondering why the gate resistor value is a very large 10k ohms. Gate input resistors usually have a much lower value than that, no higher than in the 10s of ohms range. High current MOSFETs usually have gate resistors of less than 10 ohms.
MOSFET
Re: MOSFET
A series gate input resistor can often be as high as you want since the gate current is nearly 0. Where it counts is in switching time as this R and the gate Capacitance set a maximum rate that it can switch. A higher R also tends to protect the gate from ESD damage and electrical overstress. A high power FET would have a large gate (capacitance) so maybe that's why the low resistor in those designs
Re: MOSFET
In this case a high-valued gate resistor probably wouldn't hurt since there shouldn't be quick-changing wave forms present on the MOSFET drain. However, if there happened to be rather high voltage transients present on the drain, too little external impedance between the gate and the source could cause the maximum gate-source voltage rating to be exceeded because of the internal capacitance between the gate and the drain. A task of the gate resistor in such cases is to remove charge from the gate.
Re: MOSFET - paralleling MOSFETs
I got reminded of my experience with paralleling MOSFETs. I had been reading beforehand that when doing so, separate gate resistors should be used on each one. However, using separate resistors like that caused the MOSFETs to keep failing in my switching power supply project.
The solution turned out to be to tie all the like MOSFET pins directly together and so use just one gate resistor for all of them. (All the gates were tied directly to the other gates, drains to drains, and sources to sources.)
What I think could have been happening is that with separate gate resistors, the voltage on one of the gates was being taken too far negative (below -20v) because of the drain-gate capacitance. I recall just one MOSFET in each parallel set failing at a time. But loads of MOSFETs got destroyed before it occurred to me to go against popular thought when paralleling MOSFETs.
Ever since then, I have been using just one gate resistor for each set of paralleled MOSFETs in various projects, and the MOSFETs haven't ever failed that way again.
The solution turned out to be to tie all the like MOSFET pins directly together and so use just one gate resistor for all of them. (All the gates were tied directly to the other gates, drains to drains, and sources to sources.)
What I think could have been happening is that with separate gate resistors, the voltage on one of the gates was being taken too far negative (below -20v) because of the drain-gate capacitance. I recall just one MOSFET in each parallel set failing at a time. But loads of MOSFETs got destroyed before it occurred to me to go against popular thought when paralleling MOSFETs.
Ever since then, I have been using just one gate resistor for each set of paralleled MOSFETs in various projects, and the MOSFETs haven't ever failed that way again.
Who is online
Users browsing this forum: No registered users and 0 guests