1
\$\begingroup\$

In a high voltage application (30-40V) am looking to drive a high side MOSFET which is sometimes required to be some for considerable amount of time (hours or days). [Edit] It is shown in the block diagram below.

High-side MOSFET

If I've understood correctly a bootstrap driver for a NMOS won't work in this case, right? Are there any other ways for driving high side NMOS which would require at times to be permanently on?

What about PMOS driver? When switching off it needs to connect to the supply voltage but for turning off its gate needs to be supplied around 10V below the supply voltage since connecting to GND will exceed the Vgs limit. Any idea how to do this? Any commercially available IC? Thanks a lot!

Edit: The circuit shown below works for switching a high side PMOS, albeit with some power loss and not at high frequency. If anyone knows any way to achieve higher frequency (100kHz), please let me know.

\$\endgroup\$
1
  • \$\begingroup\$ A schematic would help visualize the problem far better than a word-picture does. \$\endgroup\$ Commented Nov 20, 2013 at 7:06

3 Answers 3

1
\$\begingroup\$

Using Andy's and some other people's suggestion here is the solution that I've come up with

Part of a schematic dealing with PMOS switching

The pro of this approach is the simplicity, but the cons are that the driving speed is limited and the power consumed when the PMOS is ON.

Edit: More clarification: OutConn- can be any source voltage (Vcc). PGate2 is the connection to the PMOS gate to switch it. After actually testing this circuit, i can say that for a power MOSFET with gate capacitance of about 3000pF the switching frequency that can be achieved can be maximum of about 1kHz. If anyone knows how to achieve higher frequency, please let me know.

\$\endgroup\$
1
\$\begingroup\$

You could use a separate supply for the high-side driver rather than depending on the output switching to provide a bootstrap. Here, for example, is an IRF idea to get you started. An IGBT is shown, but it will work with an N-channel MOSFET.

enter image description here

Your P-channel idea can be made to work too.. but you need to split the power and signal paths. Create a supply that is -10V relative to the 30V rail, and put a nice fat reservoir capacitor on it. It has to have enough current capability to handle the gate switching at the maximum frequency.. a 3-terminal negative regulator might be better than a zener and resistor. Then use a gate driver chip connected to the +30 and (+30-10) rails. To level shift the signal into the gate driver, something like this:-

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
0
\$\begingroup\$

You can drive the P channel's gate via a resistor (say 10k ohm) and have a (12V) zener diode across the gate and source thus protecting the P channel FET's gate from over voltage. Also put a 100k ohm in parallel with the zener diode

If you don't mind consuming 3 or 4mA to turn on the FET then this should work just fine.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks for the answer Andy. I've updated the question with the block diagram, do you think your solution will work in this case? \$\endgroup\$
    – EarthLord
    Commented Nov 20, 2013 at 10:49
  • \$\begingroup\$ I've used the same on a prototype buck converter and it worked fine but in the end I went for an N channel FET with a high side driver to gain efficiency. \$\endgroup\$
    – Andy aka
    Commented Nov 20, 2013 at 11:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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