MOSFET is not turning off accurately at high switching frequencies (above 6 kHz)

[Rehma Anjum]

I am trying to run the MOSFET IRF3710 as low side Mosfet. As I increase the frequency above 5khz, the falling edge of the waveform starts getting slanting (as shown in the attached video file). After 22kHz it enters into Continuous Conduction Mode. I want to run the MOSFET on at least 30kHz and as I have to run the MOSFETs in H-Bridges for Dual Active Bridge Converter, so I can not work on lower frequencies.
I am using a 3.2k ohm resistive load. I have tried using multiple MOSFETs to understand the issue such as IRF540, IRF3205, 20N60C, IRFP460, F22N60E and IRF840 but the same happens with each one of them at high frequencies.
I also gave a direct PWM signal at the gate (without gate driver), but the same issue persists.
I have attached the circuit design in the attachment. Kindly help me figuring out the problem. Thanks

 

[betwixt]

I'm surprised you got as far as 5KHz!
The problem is the effective capacitance you drive between the gate and source (and drain to some extent) through a 1K resistance. The data sheet quotes about 3nF so the gate voltage has to charge and discharge that capacitance through the resistor. You need to increase the current available to drive the gate. The TLP250 is rated to supply 0.5A so you can drop the 1K considerably, I suggest 27 Ohms but you might be able to go lower.

Brian.

 

[danadakk]

Sim with 1K in gate :

Sim with 27 ohms :

I ignored output snubber as output Z is dominated by its 410 pF source to drain C.

10 ohms in gate did not seem to make much difference in turnoff, even though peak gate current
reached .6A Effective Toff constrained by 3.2K dis-charge path due to Miller ???? Just tried it with
200 ohm drain load, much faster, Toff.


Regards, Dana.

--- Updated Apr 19, 2022 ---

Here is the 200 ohm drain load sim (with 50 ohms in gate to lower drive requirements):

Regards, Dana.

--- Updated Apr 19, 2022 ---

Here is the 200 ohm drain load sim (with 50 ohms in gate to lower drive requirements):

Regards, Dana.

 

[Rehma Anjum]

Sim with 1K in gate :

View attachment 175562

Sim with 27 ohms :

View attachment 175563


I ignored output snubber as output Z is dominated by its 410 pF source to drain C.

10 ohms in gate did not seem to make much difference in turnoff, even though peak gate current
reached .6A Effective Toff constrained by 3.2K dis-charge path due to Miller ???? Just tried it with
200 ohm drain load, much faster, Toff.


Regards, Dana.

Thank you for explaining. I will replace gate resistance with 10 ohms or 27 ohms. Also, I would test it by reducing the load to 200 ohms.
But, I did not understand the part, where you mentioned 'Toff constrained by 3.2k discharge path due to Miller'. It would be great help if you could elaborate it a bit more. Thanks.

 

[danadakk]

Miller C in a MOSFET :

Miller effect - Wikipedia

en.wikipedia.org

Basically when a device in active region its G x its parasitic C, an effect from
gain in a circuit.

https://www.infineon.com/dgdl/Infineon-Use_Gate_Charge_to_Design_the_Gate_Drive_Circuit_for_Power_MOSFETs_and_IGBTs-AN-v01_00-EN.pdf?fileId=5546d46267354aa001673ba630970081

MOSFET switching :

https://www.vishay.com/docs/68214/turnonprocess.pdf

https://www.vishay.com/docs/73217/an608a.pdf

Regards, Dana.