High current spikes during on and off of the MOSFET switch(PFC circuit)

[dann11]

Hi! I am designing a PFC circuit(in boost configuration, inductor is CCM, switching frequency is 100kHz), and it has a high current spike during on and off of the MOSFET switch. I used a bulb as its load. I am not sure if these high spikes are results of the MOSFET parasitic noise or a common mode noise injected at the input of the PFC circuit ? And how should I work out on these noise? should I add a snubber circuit? or should I adjust the capacitance value of my input filter? or the inductance value of the EMI filter? I have attached the test result for your reference.View attachment F0074TEK.BMP

 

[barry]

Schematic?

 

[dick_freebird]

1-2A peak gate current spikes are normal for power
MOSFET drivers (sometimes more, for fat FETs).
The Qgg displacement charge has to go somewhere,
out the drain, source or both. 'Scope photo is not
informative, about what's what.

Seems like if you're designing it you ought to be
able to tell where the current is from and to, and
determine whether it's anything more than Cgg*dV/dt.

 

[treez]

what you show there looks like "pickup" in the scope leads...but high current spikes can come through the fet drain to source if you use the wrong diode....and it has too long reverse recovery....you need a sic diode or one of the powerint.com qspeed diodes.....and need to up the gate resistor till you get lss spiking through the fet...because also the diode capacitance charges through the fet when the fet turns on....and the cds dischrges through the fet when the fet turns on.

 

[dann11]

1-2A peak gate current spikes are normal for power
MOSFET drivers (sometimes more, for fat FETs).
The Qgg displacement charge has to go somewhere,
out the drain, source or both. 'Scope photo is not
informative, about what's what.

Seems like if you're designing it you ought to be
able to tell where the current is from and to, and
determine whether it's anything more than Cgg*dV/dt.

With the picture, the first channel(yellow line) is the gate voltage of the MOSFET, while 2nd Channel is the output voltage(Blue line), the 3rd channel is the output current(pink? line), and the 4th Channel is the Isense voltage..

The input current is from a low power supply which is needed to be boost up to around 400Vdc, 800W. If current spike are normal to power MOSFET, do you have any suggestion on how can I at least eliminate these spikes? because these spikes are still present at my output voltages..

- - - Updated - - -

but high current spikes can come through the fet drain to source if you use the wrong diode....and it has too long reverse recovery....you need a sic diode or one of the powerint.com qspeed diodes.....and need to up the gate resistor till you get lss spiking through the fet...because also the diode capacitance charges through the fet when the fet turns on....and the cds dischrges through the fet when the fet turns on.

As I have checked the datasheet of the diode, its trr is 45ns, which is quite longer than the fall time speed of the MOSFET which is 30ns.. I did tried to increase the gate resistor of my MOSFET, from 4.3 ohms to 20 ohms but the I still measured the same current spikes..

Just another question relevant to the MOSFET switching, does the diode at the gate controls the turning off of the MOSFET or is it the resistor?

 

[treez]

i presume you have a diode with anode to gate in your gate drive circuit...this helps you turn off quicker, since discharge of the gate source capacitance then bypasses the series gate resistor.
Have you looked at the qspeed powerint.com diodes for PFC?....THESE HAVE A NICE RESPONSE FOR PFC (sorry about capital lettering)

 

[dann11]

i presume you have a diode with anode to gate in your gate drive circuit...this helps you turn off quicker, since discharge of the gate source capacitance then bypasses the series gate resistor.
Have you looked at the qspeed powerint.com diodes for PFC?....THESE HAVE A NICE RESPONSE FOR PFC (sorry about capital lettering)

I haven't checked it yet.. Yes, I did place a diode parallel to the gate resistor. I am still trying to tune the gate turn on time of the MOSFET to lower down the spike value. Do you it is also best to put snubber in my circuit?