MOSFET going to continuous conduction

Hi Ravi, 

The 60V, 2A MOSFET seems ok for this application. But just ensure that the MOSFET is able to drive that much current for the gate voltage you apply(the Vgs/Id characteristics in the datasheet).

Also, I think it's better to use a gate driver circuit for the MOSFET(like the below one with your chosen MOSFET part). Btw, I am not sure how well this method is to measure the dynamic char of flyback converter.

Fundamentals of MOSFET and IGBT Gate Driver Circuits (Replaces SLUP169) (Rev. A) (ti.com)

Hi,

Continuous conduction means the MOSFET is damaged, one or more parameters was exceeded. To troubleshoot, you'd need a 'scope probe to see the gate voltage and the drain with respect to the source. It could be as simple as needing a 50 ohm resistor wired between the gate and source (otherwise the gate sees a signal which is different from what you have configured on the function generator, but you won't know that without looking at it with a 'scope). Also it will be noticeable at high frequencies that the function generator impedance might not charge/discharge the MOSFET gate fast enough. Also, long wires will act as inductors, causing voltage spikes when the MOSFET switches off. There are techniques to eliminate that, you could try a ferrite beads.. but really you're flying blind without looking at it. Before the MOSFET blows, start at a lower voltage or frequency and record the signals from your 'scope, and actually see what is causing it.

A FET in this configuration would likely only die from heat. If the device didn't get hot to the touch it means it would have been a very fast internal heating due to high inrush current.

If it did get hot, then maybe the drive signal isn't turning the FET on completely. What gate voltage is being used?

Is it possible to measure the transient current during turn on of the load?

One other possibility is if the load is inductive there could be a high voltage spike on turn off. This could be mitigated using a snubber diode across the load. 

hello shabaz,

i was observing the current through the MOSFET and voltage output of the flyback converter. there is a current spike of 8 times the set load current. unfortunately i was not able to save those plots as i was not sure what was happening in the moment. i will repeat the test with a plain R load and shorter leads and try to check on the parameters properly. i will try to replace the function generator with a proper gate driver and check again.

One other possibility is that the FET is working properly, but the drive signal just never goes to the right voltage to turn it off.

hello dougw,

i was observing a peaking current while turn on nearly 8 times the set point(500mA). i observed that initially heating was occuring so chose a fet with higher power rating. i am using a function generator with 10V amplitude.

mosfet seems shorted as i checked with turning off the load, power supply too and again turning them on

It should be very tough to kill the big Toshiba FET, although it may need a heatsink if it is getting too hot to touch. Is it possible it is just connected incorrectly (drain and source swapped) and the body diode is conducting?

One final thing to try - put a resistor from the gate to ground to ensure no residual voltage is left at the gate.

Connection was fine, as I checked at lower frequency switching is happening.