1

I have a buck converter (Half bridge) as shown in the schematic below. (48V output). V_D goes to an external inductor. Buck Converter half bridge

I have measured the inductor current for different loads (Nominal(upto 4A), Overcurrent(5 to 8A) and short circuit case(above 8A). For the Nominal case, the inductor current waveform has triangular waveform as expected.

For overcurrent and short circuit case, I get non-triangular waveform as seen in figure. Inductor current for overcurrent case (yellow) The period is 3 times and there is non-linear slope. The current waveform when I load the output from no load to overcurrent is shown in figure. Inductor current (yellow)

Why do I get this incorrect waveform?

Note: When I increase the gate driver supply voltage from 12 to 14V, I see correct triangular wave for the overcurrent case. But I cannot change the voltage to 14V in the final circuit because of technical reasons. I tried reducing the high side turn on gate resistance to 15 ohms, but I still get non-triangular inductor current waveform

I want a solution that does not involve voltage change.

Summary

  1. I see non-triangular inductor current when the load at output of the buck converter is high (Overcurrent/short circuit current)
  2. When I increase the gate resistance the problem is solved upto certain extent (for overcurrent case). The problem still persists for the short circuit case.
  3. When I increase the gate supply voltage to 14V the problem is solved even for short circuit case.
  4. Instead of increasing the gate supplied voltage, I increased the bootstrap capacitance. There was no change in the waveform received.
  5. The problem seems to be temperature dependent. When the temperature increases the problem of non-triangular waveform occurs even for lower loads like 8 ohms (Still in over-current case)

In conclusion I want answers to

  1. Why do I get non-triangular waveform and how do I fix it without changing the gate driver supply voltage?
  2. What can be changed in the gate driver circuit?
Ishani Engineeŕ
  • 49
  • 7
  • I don't see an inductor in your schematic. Please show the complete schematic including drivers. – Andy aka Feb 22 '23 at 10:35
  • You gate driver scematic is too small to read. What is producing the driver inputs? – Bruce Abbott Feb 22 '23 at 11:23
  • 2
    I see triangular ramps, operating in partial continuous-conduction-mode. – rdtsc Feb 22 '23 at 12:41
  • 2
    Please show the logic input signals as well. – Tim Williams Feb 22 '23 at 13:57
  • Now when I increase the high side turn off gate resistance, and the low side gate resistance, the problem goes away. I wonder that may be the gate driver was overloaded and may be there parasitics in the gate drive loop causes the non triangular current at higher loads. How doess the gate parasitic inductnace influence the switching loss, such that I get non-triangular inductor current waveform ?? – Ishani Engineeŕ Mar 06 '23 at 09:37
  • 2
    That looks kind of like subharmonic oscillation? – Hearth Apr 05 '23 at 13:23
  • Why do you think it's a problem to begin with? It may be a problem, or may be acceptable depending on application. – Kuba hasn't forgotten Monica Apr 05 '23 at 15:05
  • Its a problem since after few tests the converter gets damaged, especially the gate-driver – Ishani Engineeŕ Apr 06 '23 at 06:54
  • In my opinion this question needs some more editing. What relationship is there between the simulation schematic and the actual circuit? I see a 500uH inductor in one picture and 3x266uH inductors in another. You also need to show oscillographs of the gate drive waveforms. Tweaking the gate resistances without measuring isn't a good way to troubleshoot - are you at risk of shoot-through? Overload behaviour generally constrains the duty cycle but should not cause cycle-to-cycle oscillation unless your control loop is fighting with your OCP. – Adam Lawrence Apr 17 '23 at 16:16

0 Answers0