0
\$\begingroup\$

I have a total of 20 IRF3205 mosfets in a H bridge configuration for switching a transfromer to get pure sine wave output. (Five mosfets at each end of thr H bridge).

  • Datasheet says total gate charge of 150nC max for each mosfet. Switching frequency is 26khz max. Gate voltage is limited to 12V max because regulated 12V supply is fed to driving IC.
  • Driving IC is HIP4082 with peak current of 1.25A.
  • Gate resistors for each mosfet is around 110 ohm. Single pulldown resistor for each 5 mosfets(100k). 18V Zener diodes for protection.
  • Max load current is 200A. They are all mounted on seperate beefy heatsinks and I will be using heatsinks as drain terminals.

Questions

  • Will I be cooking those fets in transition state or will they just do okay?
  • Is there a better value of gate resistors and pulldown resistors for each mosfets?

Edit:

Question

*Should I even care about drive problems? Frequency is just 20-30khz.

\$\endgroup\$
16
  • 1
    \$\begingroup\$ When you say "switching a transformer to get a pure sine wave output" that raises some red flags for me. Alternatively, in another interpretation of your post, it sounds like you want to run the MOSFETs as a linear devices which is oxymoronic because you also say H-bridge. If you gave us a schematic and more info, then I would not need to do all this guessing. \$\endgroup\$
    – DKNguyen
    Commented Aug 9, 2019 at 18:37
  • \$\begingroup\$ @DKNguyen He's probably synthesizing a "pure sinewave" using pwm. \$\endgroup\$
    – Drew
    Commented Aug 9, 2019 at 18:45
  • \$\begingroup\$ @Drew Yes, spwm. \$\endgroup\$ Commented Aug 9, 2019 at 19:00
  • \$\begingroup\$ I just want to know if they will blow up at 200A (100A average for each group of 5 fets) because of insufficiant gate drive and consequently overheating. Thats why I am asking for correct gate resistor value for each fet. \$\endgroup\$ Commented Aug 9, 2019 at 19:05
  • \$\begingroup\$ Mosfets are for turning on and off, if the voltage is in between, then the mosfet will dissipate significant current as heat. If they are paralleled this might be acceptable, simulating in a SPICE package is probably the easiest thing to do to find out the thermal numbers \$\endgroup\$
    – Voltage Spike
    Commented Aug 9, 2019 at 19:11

0

Your Answer

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

Browse other questions tagged or ask your own question.