How calculate maximum PWM frequency for MOSFET

[SolarTrap]

Hello all,

I would like to know what the maximum PWM frequency would be that I can drive this MOSFET:

http://www.vishay.com/docs/72439/sum110p06-07l.pdf

There are a lot of delay times given but even if I take the highest number (200ns) for 'turn-off time' this would result in 5 MHz. But this would mean that even on a 95% duty cycle the MOSFET would still need it's 200ns to go off... So the Hz number has to be much lower.
Here its seems that I would need some insight on how to calculate this. Let's say my PWM resolution is 256 just as a calculation example.

Thanks
Markus

 

[itsmoked]

Yep 200ns to go OFF and 30ns to go ON. 130ns.
f = 1/T = 1/130ns = 7.7MHz to go on/off as fast as possible.


Of course that's pretty useless since just about no energy would be transferred. You probably want your PWM period to be 10 to a 100 times larger than 130ns so the ON/OFF delays remain insignificant.

Call it 1us.
Your maximum ON period would be 255us and your shortest period would be 1us.

We don't know what you're using this for so even those periods might be idiotic.



Keith Cress
kcress -

http://www.flaminsystems.com

 

[SolarTrap]

It is used for a charge controller. The typical 500Hz generates some audible noise and 100kHz makes the mosfets really hot so I was looking for the best alternative.

 

[ScottyUK]

At high frequency you will need a fair bit of curent to drive the gate capacitance. That doesn't seem a big FET for operation at 100kHz so if it is getting really hot maybe take a look at the gate drive.

 

[iop95]

PWM frequency is mainly limited by losses in magnetics not by MOSFET performance.
There was made converters up to 70MHz, but efficiency is not so good to became an commercial product.
For a MOSFET in a power converter, maxim frequency is not a simple 1/(ton+toff); need to take care for gate charge remove time, that is a complex mechanism, very dependent of driver current and slew rate capability.
Also, to fast slew rate mean high di/dt that have implications in EMI and develop high voltahes on parastic inductance.

 

[itsmoked]

iop95 has covered it well for a charge controller.

Faster slew = more efficient.
Faster slew = more EMI
You need to find the sweet spot.
Use a good 'gate driver'.

Having only 3.3V will certainly cramp your gate driver selection. Probably down to one poorly data-sheeted Microchip part.

If you're doing high-side switching that often requires isolation. N-MOSFETs switch faster. And can used high-side once you're using a driver and maybe isolation.

Keith Cress
kcress -

http://www.flaminsystems.com