Hi side MOSFET

Not 40A, but this safely handles 10A using a BUZ11, just search digikey for higher amps.

It depends on the swithcing frequency as to how much gate charge current is neded to switch it fast enough to not dissipate power. You could make an isolated DC supply and optocouple to the input a 1A or better driver to drive a large high side N-FET.

A more reasonable solution is to put the N-fet on the low side with the source referenced to ground. Then you can use the many low side drivers and drive it as hard as you desire.

The series resistor close to the gate is to prevent undershoot and overshoot (shic could get as serious as oscillations in some cases) gaused by the gate capacitance, stray inductance and high Di/dt diring the charging edge. It usually only takes 10 ohms or so to damp it. Yes it will slightly slow down you switching but this can be a good thing as a compromise between the radiated EMI and the MOSFET power dissipation. Fast edges and wires are good RF tramsmitters.

Most n-channel high-side MOSFET drivers include a provision for a flying-capacitor gate-driver supply voltage, using the PWM signal itself for the charge-pump action. You have to add the capacitor, and sometimes a diode as well. Some include a low-voltage disable.

low to drive a P75NF75 mosfet.

The IR2117 is meant for high-voltage offline use. There are lots of good choices for low-voltage high-side drivers, and most have much higher current capability. Check out TI, MicroChip, LTC, etc.

ST's 75nf75 is a modestly-serious MOSFET, and 400mA gate-drive current would be a bit on the weak side, but 6A will likely be more than you need. Anyway, there are drivers to 12A and beyond.

There are a few that have internal self-oscillating charge pumps, so they can be used at DC, but they're rarely used for PWM.

I would recommend you use an IX6R11S3 -- high side/low side gate driver with 6Amps peak output current. You can get info off the IXYS web site --

Sam