2
\$\begingroup\$

I'm trying to make a buck converter that uses a P-MOS for high-side switching using an STM32's PWM signal.

I can't figure out why it would be useful to drive the MOSFET using a push-pull MOSFET gate driver because the two MOSFETs that are going to be used in a push-pull configuration will also need to have their gates discharged/charged, and the charging current is still able to harm your microcontroller GPIO pins. Am I missing something here?

enter image description here

\$\endgroup\$
4
  • 1
    \$\begingroup\$ The best approach is to use a dedicated gate driver ic for example tc4420cpa is a strong driver and quite cheap to.but if want to use transistors you could use some logic level mosfets.the gate voltage and capacitance of logic level fets are very low compared to normal power mosfet. \$\endgroup\$ Aug 16, 2021 at 11:57
  • 4
    \$\begingroup\$ Consider how much current is needed to quickly charge/discharge the big NMOS on the right, then it becomes obvious that Sp and Sn are needed. \$\endgroup\$ Aug 16, 2021 at 11:59
  • 1
    \$\begingroup\$ You can use much smaller MOSFETs on the left. \$\endgroup\$
    – user253751
    Aug 16, 2021 at 12:22
  • \$\begingroup\$ The driver MOSFETs amplify the gate current to the power MOSFET so it switches quickly and is not in linear (high dissipation) mode for very long; if it sits in linear mode, the magic smoke usually escapes the casing. Other design factors to consider: what happens when the microcontroller crashes, when its PWM output is not yet initialized and floats, when Vcc is out of tolerance? \$\endgroup\$ Mar 15, 2022 at 23:35

2 Answers 2

8
\$\begingroup\$

Driving MOSFETs are usually picked with much lesser gate capacitance, compared to power MOSFET. Lower gate capacitance in real world means much less Drain-Source current. So that is why in fact the power MOSFET will have higher Id and Qg and due to this you have to use some small driving MOSFETs with lower Id/Qg.

While they are "small", they still can handle much more current compared to MCU I/O thus can charge/discharge gate of the power MOSFET much faster.

PS: Driving DC-DC PWM signal from MCU are usually considered a bad practice. Code flaws can cause a disaster and adds a lot of additional points of errors. Dedicated PWM/controller IC, configurable from the MCU if required, would provide less error prone process.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Your final paragraph is phrased as on-chip vs off-chip, while it sounds as if you are actually contrasting software vs hardware PWM. In fact hardware PWM is a very common feature of MCUs, so the "dedicated PWM controller IC" provides many disadvantages (increased latency, decreased throughput, need for inter-chip connection, complexity of controlling the inter-chip port, difficulty of synchronizing with other MCU activity etc) and for the majority of application no advantages vs the integrated DMA-and-interrupt-capable hardware PWM peripheral at all. \$\endgroup\$
    – Ben Voigt
    Aug 16, 2021 at 22:14
  • \$\begingroup\$ @BenVoigt OP was talking about buck converter (and I've mentioned DC-DC application). So he also needs a feedback loop and with STM32 that would definitely involve parts of software calculations & control. While it's certainly possible there are some drawbacks and care should be taken. A bit more explanation could be found here: electronics.stackexchange.com/questions/292172/… \$\endgroup\$
    – NStorm
    Aug 17, 2021 at 5:15
3
\$\begingroup\$

am I missing something here?

If you get the data sheet of the logic and pre-drivers you'll find this: -

enter image description here

\$\endgroup\$

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.