I wonder if it is possible to drive a logic level mosfet directly from a micro controller.
The IRLZ44N has a Vgs(TH) of 2v, while the µC supplies almost 5v at 25mA or so.
Will that be enough to drive the mosfet into saturation fast enough for a 20KHz+ pwm ?
What gate charge do I need to calculate the current needed to drive the gate ?
I have used the IRL520N TTL MOSFET which connects directly to a micro. I used the PIC18F452 no problem When I moved to the 4520 it didn't work, Microchip explained that the new chips don't latch properly as there is a build up of capacitance and the gate stays high..
Must it be a mosfet? I dropped them and I now use Darlington with the same current capability..The trouble here is the larger the device the slower it goes.. are you pulling much amps from the mosfet?
You can use Qg (total gate charge) to estimate turn on time. It is the turn on time in nanoseconds that 1 amp of gate current will provide. In the case of your transistor it is 48. So 48/.025 = 1920 ns. add the rise time etc and you have a little over 2 us, not to good for a 20 Khz PWM.
PS. They never seem to be as fast in real life as the spec.
You can use Qg (total gate charge) to estimate turn on time. It is the turn on time in nanoseconds that 1 amp of gate current will provide. In the case of your transistor it is 48. So 48/.025 = 1920 ns. add the rise time etc and you have a little over 2 us, not to good for a 20 Khz PWM.
It somewhat depends upon the current and voltage you are switching with the PWM. A 2µs rise and fall time is about 8% of the 20kHz period. For a rough calculation of the average power dissipated by the transistor from the switching time, multiply 8% by the MOSFET supply voltage times ½ the peak current (0.08 x V x I/2).