enter image description here

I am trying to design a H-bridge to drive 24V motors. No issue on the low side but I would like to use the FOD3184 Opto/Gate driver on the high side of the H-bridge to drive a P MOSFET and I am somewhat worried about the P-channel MOSFET being pulled to GND by the gate driver.

2 Questions for the community:

Question 1: On the high side I would like to used a 12V Zener diode to protect the P MOSFET and limit the voltage drop to -12V. I am also thinking about adding a resistor to limit the current pulled through the diode to GND via the gate driver. I am not sure where to place the resistor (which value?) or even if this is the right approach. I already have the gate driver and the MOSFET so i would like to make use of them instead of buying new parts.

Question 2: I also read threads on adding a capacitor across the MOSFET to control the slew rate but the thread all talked about this in the context of driving capacitive loads. In my case the load will be inductive in nature (Motors). Should i worry about this here?

I am using the Fairchild FQP47P06 and FQP50N06L on each side of the H-bridge with the FOD3184 to drive them. Supply voltage is 24V DC. PWM frequency 10-15 KHz. the motors are small... RC car type of things. I have 2 steeper and 2 BLDC running at less than 1Amp current.

Proposed circuit change based on the answer from Rioraxe:

-12V Negative power supply: (is the polarity of the capacitors correct?). In this circuit the LM7912 makes me think of a normal voltage regulator connected upside down Power to GND Pin and GND to the Input pin... i'll have to test this on a breadboard.

Let me know.

enter image description here

  • \$\begingroup\$ Please add a circuit diagram, with placeholders for the locations you are not sure about. There is a schematic tool in the editor. \$\endgroup\$ Sep 6, 2016 at 7:34
  • \$\begingroup\$ Thanks. included a screenshot from Upverter of the circuit as is for now. The Zener would come across the P MOSFET parallel to the pull-up resistor. The resistor to channel the current to ground from the diode is where i am less sure. \$\endgroup\$
    – user122764
    Sep 6, 2016 at 10:38
  • \$\begingroup\$ U22 will not work. In should be conected to pwr and gnd to gnd. I build up my schematics in LTspice (free) and simulate it to get confidence about the voltage levels and the transient behaviour. \$\endgroup\$
    – schwdk
    Dec 22, 2018 at 6:49
  • \$\begingroup\$ With that isolated driver IC you should be able to make a bootstrap configuration that allows using a N-FET as top FET. That would eliminate your worries about driving the gate as Vgs would be either 0V or the bootstrap capacitor voltage you choose. \$\endgroup\$ Jan 26, 2019 at 11:33

1 Answer 1


The simple zener across gate-source would waste a lot of power. You would need to add a current limit resistor from driver to gate and you want to size the resistor for significant driver current, that significant driver current would persist through the zener and the resistor when driven to ground by the gate driver.

I would create an intermediate supply for the Vss of FOD3184s. The average current requirement would not be high, perhaps less than 10mA per driver-MOSFET. If a 7912 or similar is handy, that may be what I would use. Connect the regulator common/reference to PWR. Put a few uF of decoupling cap next to each optocoupler-driver.

I don't see how adding capacitors across the MOSFET outputs would be beneficial.

  • \$\begingroup\$ Thanks Rioraxe. I'll include ~30 OHMs resistor on the Gate line between the MOSFET and the gate driver. I saw several schematics with such resistors. The capacitor and the slew rate question ... I should have been more precise. The capacitor would be between Gate/Drain on the P MOSFETs for example and between Gate & Source on the N MOSFETs if needed i guess. The application note i was referring to is AND9093/D - Using MOSFETs in Load Switch Applications but this always refers to capacitive loads. \$\endgroup\$
    – user122764
    Sep 7, 2016 at 9:51
  • \$\begingroup\$ I am not 100% familiar with negative voltage sources. I looked at the datasheet of the LM7912. I edited my initial question. Would you be able to take a look at the proposed circuit to generate the -12V? (especially the polarity of the electrolytic capacitors) \$\endgroup\$
    – user122764
    Sep 7, 2016 at 15:12
  • \$\begingroup\$ AND9093/D deals with power supply regulator applications of charging a capacitor. In your application of PWM to inductive load, you want the switching on and off to be hard and fast to minimize switching loss. The inductive load would limit the current ramp by itself. The 30Ohms probably work fine for the size of MOSFET and switching frequency you have chosen. But I would start with 10Ohms for faster switching to get further incremental reduction in switching loss. \$\endgroup\$
    – rioraxe
    Sep 7, 2016 at 19:01
  • \$\begingroup\$ The 7912 connections look right. Just nitpicking -- for many of us who have done this a long time, we are very accustomed to more positive voltage on top. You can get that if you redraw the 7912 symbol to have GND/COM on top. You may even see the checking goes easier in this case. \$\endgroup\$
    – rioraxe
    Sep 7, 2016 at 19:08
  • \$\begingroup\$ hum... i'll need to return my brand new LM7912. The FOD3184 needs a Vdd-Vss voltage of at least 15V. at 12V i am right in the transition phase of the transistor. Should have been more careful. Thanks for the tip though. this was helpful. \$\endgroup\$
    – user122764
    Sep 9, 2016 at 23:18

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.