Overall I'm working on a single-phase full bridge inverter, but before I implement the full bridge I wanted to start with a simpler circuit. Below I have a circuit diagram of the half-bridge circuit I'm using. And I'm trying to drive the MOSFETs using the IRS21084 half-bridge driver. I’ve looked at and followed the advice of some of the design and application sheets from Infineon on how to implement the circuitry of the gate driver, but I'm having a major issue. The issue is that I'm getting an output voltage of about 15 V at the high-side gate driver output (HO) and about 6 V from the low-side gate driver output (LO). If I'm not mistaken the correct output should be the same for both the high and low sides. This is troublesome because the low-side output voltage is not going to be able to turn on the low-side MOSFET and overall the half-bridge won't work.

Half-bridge circuit

The following is a brief explanation of how I have set up and configured the IRS21084 half-bridge driver.

  1. I’m providing a constant 15 V with the LM317 for the supply voltage (VCC)
  2. I’m using the teensy 3.5 microcontroller to provide a PWM input at 20 kHz for both the HIN and LIN pins.
  3. I’m using a 100 kΩ resistor for the dead time.
  4. For the bootstrap diode, I'm using the UF4007 from Vishay. It has a reverse recovery time of 75 ns and a voltage drop (Vf) of 1.7 V.
  5. For the bootstrap capacitor, I followed the DT04-04 application sheet from Infenion called ‘Using Monolithic High Voltage Gate Drivers’ and ended up with a capacitor value of 2.2 uF. I used a ceramic capacitor.
  6. For both the decoupling capacitor, and the bypass capacitor, I couldn't find any information on how to determine the values for them, so I used the same value capacitor from the bootstrap, for both decoupling and the bypass capacitor.


When testing the gate driver circuit does it need to be connected to the bridge circuit? Also, how do I determine the values of Vge-min and Vce-on for my circuit? For the bootstrap circuit do I need to include a bootstrap resister if so, how would determine the value for it? I used ceramic capacitors for all three capacitors, does that affect the circuit in any way or do I use different types of capacitors for the decoupling or bypass capacitors?

Lastly below are the provided datasheets of the components I'm using 91188.pdf (vishay.com) uf4001.pdf (vishay.com) IRS2108-IRS21084 (S)PbF (infineon.com) https://www.infineon.com/dgdl/Infineon-Using_Monolithic_Voltage_Gate_Drivers-AN-v01_00-EN.pdf?fileId=5546d462584d1d4a01585242c11947b1

  • \$\begingroup\$ Are you providing an IN-phase or and OUT-OF-phase input to Lin? Did you notice in the datasheet it's inverted? The line above Lin tells you it's inverted compared to the output for the lower MOSFET. \$\endgroup\$ Commented Apr 25, 2023 at 20:53
  • \$\begingroup\$ For both the HIN and LIN inputs of the gate driver I provided the same 50% PWM duty cycle. I am aware that, at the LIN pin my signal is going to get inverted, and that's actually (for testing purposes) what I want so that the half-bridge works as intended. At least from what I know, but I'm not exactly sure how that could be the problem of having two different output voltages or at least the low side. I was wondering if perhaps it could be that I didn't choose the correct bypass and decoupling value. I've looked at Infineons application sheets, but I couldn't find anything helpful. \$\endgroup\$
    – Norbert
    Commented Apr 26, 2023 at 21:44

1 Answer 1


The only thing I see that could be wrong is your PWM signal to LIN on the IR21084.
The datasheet specifies it as "out of phase" or inverted compared to the output to the lower MOSFET.
Here is the relevant input/output timing diagram:
enter image description here

I have added the green lines where it was hard to trace the signals vertically.
As you can see, the upper MOSFET is ON only when both HIN and LIN are high.
The lower MOSFET is ON only when both HIN and LIN are low.
This means that you should just tie them together and use only one PWM output from your MCU to this IC.
When you add the second IR21084 to complete the H-bridge, it will also have both inputs tied together, but will be run with a PWM signal of opposite phase or inverted from the other one.

  • \$\begingroup\$ First, I wanted to thank you for your reply. I did tie the PWM inputs together from the microcontroller and made it output a 50% duty cycle at 20 kHz to clearly see on output traces, to both the HIN and LIN inputs of the driver. I used the oscilloscope and saw that I'm getting the correct output traces, with the high side being a 50% duty cycle while the low side being the inverted 50% duty cycle. But the overall issue of having a low output voltage from the low side is still present. I also saw examples where people use pull-down resistors at the gate, but I don't think that relates to me. \$\endgroup\$
    – Norbert
    Commented Apr 26, 2023 at 21:34

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.