0
\$\begingroup\$

I need a high side switch. I thought about using a p-channel MOSFET. The source will be connected to around 50V-60V. How can I drive this MOSFET? I can't find a driver which can switch between 60V and 45V in order to put the MOSFET on and off. They all switch between ground and Vcc, which is generally below 25V. How is switching a high-side p-MOSFET at 60V done? My control signal is at 3.3V.

EDIT: I want to avoid a second power supply. The microcontroller used for the signal generates a signal for a low side n-MOSFET as well.

EDIT 2: The goal is to drive a chopper for a motor, so the switching frequency I was thinking about is around a few kHz.

\$\endgroup\$
3
\$\begingroup\$

What you're talking about is generally called a level shifter. Assuming you don't want high-frequency switching, you can use a circuit like

schematic

simulate this circuit – Schematic created using CircuitLab

A few tips. The NPN must be rated for more than 50 volts. You can play with R3 and R4, but the total resistance should give a current at 50 volts of about 10 times the NPN base current or less. Larger values will dissipate less power in the resistors, but will also interact with the FET gate capacitance to slow switching and increase the power dissipated by the FET. Adjusting the ratio of R3 and R4 controls the gate drive voltage, so don't get carried away. A 1 to 3 ratio at 50 volts gives 12.5 volts and 15 volts at 60, which should be just fine.

\$\endgroup\$
  • \$\begingroup\$ What do you mean with 'no high frequency' \$\endgroup\$ – Brent Thierens Nov 17 '18 at 16:13
  • \$\begingroup\$ @BrentThierens - That actually depends on all sorts of things, including the resistor values of R3 and R4, the gate capacitance of the FET, the power-handling capacity of the FET and how well the FET is heat-sunk. Oh yes, and how sharply you want the load voltage to rise and fall. With none of these specified, it's impossible to give a hard number, but usually something like 1-10 Hz or so is a reasonable cutoff. Faster than that and a more complex circuit is probably a good idea. \$\endgroup\$ – WhatRoughBeast Nov 17 '18 at 16:17
  • \$\begingroup\$ That is too slow, added it in the question \$\endgroup\$ – Brent Thierens Nov 17 '18 at 16:20
  • 1
    \$\begingroup\$ @BrentThierens a few hundred kilohertz is not too fast for this configuration (btw: you can find two FETS in the same package designed for this). Even a few megahertz if you are careful with the design and add some speed-up elements. By “slow” means that ns switching times might be out of the question, and this is faster than optocouplers. \$\endgroup\$ – Edgar Brown Nov 17 '18 at 16:28
0
\$\begingroup\$

Use just a normal isolated gate driver and isolated power supply. Connect supply's output to PMOS source. Voila.

\$\endgroup\$
  • \$\begingroup\$ I edited the question, I hope it is more clear. Since the microcontroller also drives another NMOS, the 3.3 can't be on another level since then I have problems with the other driver \$\endgroup\$ – Brent Thierens Nov 17 '18 at 16:11
  • \$\begingroup\$ Yes, i know. Just please find isolated gate driver and isolated 15V power supply, you will understand. If you wish, i can sketch, but later. \$\endgroup\$ – Gregory Kornblum Nov 17 '18 at 16:13
  • \$\begingroup\$ A sketch would come in handy :) \$\endgroup\$ – Brent Thierens Nov 17 '18 at 16:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.