Hi all,

I have designed an active-clamped fly-back converter, currently operating at 30V input voltage and generating a high voltage. The active clamp seems to work well at clamping the voltage on the main switch when it turns OFF, however when the main switch turns ON and the clamp turns OFF, there are some significant oscillations in the drain-source voltage of the fly-back FET which also is causing problems with the FET current.

The main issue is that the oscillations are creating EMI issues which are carrying over to my feedback voltage and operational amplifiers that will attempt to control the converter when I close the loop.

I have seen multiple documents that address clamping the voltage as the main fly-back FET turns OFF, but not particularly how to damp the oscillations when the FET turns ON. I can only assume that this is the parasitic COSS of the switch. This is a GaN switch so has very fast rise and fall times - particularly fall times, which seems to be quite a bit more rapid than the rise time, probably due to the clamp capacitance.

Can this just be fixed in the same way as a buck-converter in synchronous configuration, with a resistor-capacitor damping network from the switch node to ground? This would then mean the voltage is clamped at turn OFF with the active-clamp capacitor circuit, but also damping the oscillations at turn ON when the FET parasitic capacitance rings with some inductance in the circuit. In the attached image the bottom FET voltage is shown in blue and shown in purple is the coupled noise through the feedback pin - I took this measurement by attaching the positive terminal of the probe to the ground pin and the ground of the probe to see how much of the signal was real and how much was just noise coupling. Green is the output voltage which also has some noise but most of that seems to be pickup from the 1000x attenuation probe. The oscillation frequency in the feedback stage where the amplifiers are seems to line up nicely with the oscillation frequency of the FET voltage/current at turn on.

Any thoughts?

Best rgds, SiC

 

I don't see the problem. Much of the noise is from how the scope is connected. Some of the ringing is from the power supply capacitors self resonant frequency.
We don't dampen a closed switch (mosfet) because placing a RC across a 0.04 ohm MOSFET that is on does nothing.
You could slow down the Gate by increasing the resistance.
No schematic? No layout pictures?