1
\$\begingroup\$

The high side MOSFET becomes extremely hot when I try to run the half bridge, and the output does not seem right.

The PWM signals to Hin and Lin come from an Arduino - 47.5% duty cycle, dead time of 132ns, switching at 3.14kHz.

What could have been the problem?

  1. Is my deadtime too small? I tried increasing it to 40usec. and did not see a change.
  2. Is it the placement of ground right? Should it be moved from P1 to P2? If so, what should the polarities of C3 and C4 be?
  3. Should C1 and C2 be electrolytic or ceramic? (http://tahmidmc.blogspot.com/2013/01/using-high-low-side-driver-ir2110-with.html mentions that ceramic capacitors should go parallel to electrolytic capacitors).
  4. I am using 220uF, 250V capacitors for C3 and C4, could they be an issue as well?

In the picture below, the yellow wave form corresponds to the gate input of the high side switch, the blue waveform corresponds to the gate input of the low side switch and the purple waveform corresponds to the output across the output resistor Rload-1kohm.

enter image description here enter image description here

\$\endgroup\$
19
  • 1
    \$\begingroup\$ C1 should be ceramic. Electrolytic is too slow (too much inductance). You can use a scope to see how long your high-side MOSFET is taking to turn on. Your scope image is too small to really be useful but the waveforms look at at least from the low detail that I can discern. \$\endgroup\$
    – DKNguyen
    Oct 3 '20 at 16:42
  • \$\begingroup\$ Thank you for your comments! I shall do that, I have edited the scope screenshot. \$\endgroup\$
    – Ruhi
    Oct 3 '20 at 16:47
  • \$\begingroup\$ I almost want to say change your diode to a faster one like a 1N4448, but your waveforms don't show any refresh issues. But you should do that anyways since the 1N4004 is meant to rectify 60Hz, not refresh at kHz. Your grounds are correct. Do not change. The IRF2113 is a unipolar device and P2 is not the point of lowest potential in the system. There is some weirdness in how your LO side gate drive signal seems to try and follow the high drive gate-drive signal. \$\endgroup\$
    – DKNguyen
    Oct 3 '20 at 16:51
  • \$\begingroup\$ Great, thanks so much! I shall changing the diode and give an update! \$\endgroup\$
    – Ruhi
    Oct 3 '20 at 16:56
  • 1
    \$\begingroup\$ There is some weirdness in how your LO side gate drive signal seems to try and follow the high drive gate-drive signal. Find out what is causing that. What is the voltage on the LO side gate drive when the HI-side gate drive is high? \$\endgroup\$
    – DKNguyen
    Oct 3 '20 at 16:56
1
\$\begingroup\$

switching at 3.14kHz

You can't expect the bootstrapping circuit to work at such a low frequency with such low value bootstrap capacitors. In other words, you upper MOSFET gate drive signal runs out of steam and turns the upper MOSFET into a regular source follower dropping several volts across it.

\$\endgroup\$
9
  • 1
    \$\begingroup\$ I wondered about that but there's no evidence in the scope trace of any droop. \$\endgroup\$ Oct 3 '20 at 18:12
  • 1
    \$\begingroup\$ OP should provide a differential probe of the bootstrap cap voltage relative to the HI side gate drive signal. That would confirm. But to me it almost looks like the bootstrap is not there at all and the HI side is being directly driven by a GND referenced 15V, rather than running out of steam. \$\endgroup\$
    – DKNguyen
    Oct 3 '20 at 18:16
  • \$\begingroup\$ @KevinWhite I may be wrong but if you look at the yellow voltage waveform (high side gate switch) there is only about 15 volts p-p and I'm unsure where the DC level is pitched so, I guess what I'm saying is that I cannot tell from the yellow signal whether it's good or not. It might be but there's certainly no evidence that the boot circuit will work with 3 kHz. \$\endgroup\$
    – Andy aka
    Oct 3 '20 at 18:17
  • \$\begingroup\$ Thank you for your inputs, I shall attach the differential probe waveform for the bootstrap capacitor voltage soon! \$\endgroup\$
    – Ruhi
    Oct 3 '20 at 18:33
  • 1
    \$\begingroup\$ Dear all, thank you so much for your feedback. I changed the frequency from 3kHz to 25 kHz, my MOSFET did not heat up. So, I think it was the bootstrap capacitors. To confirm the same, I replaced the bootstrap with an independent DC source clamping the high side MOSFET and in this case too my MOSFET did not heat up. Thank you! \$\endgroup\$
    – Ruhi
    Oct 4 '20 at 13:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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