1

I built a half-bridge smps based on IR2110S, it is working but the oscillograms seems weird and mosfets are a little hot.

LOW side mosfet Vds.

Two oscillograms, mosfets midpoint as oscilloscope ground, probe on rail and probe on gnd. Lower oscillogram is inverted, to show real picture. It doesn't look for shoot-through? Spikes on picture looks like an unexpected opening.

No load, transformer primary winding oscillogram. Schematics

What can be the problem?

Erch
  • 41
  • 2

2 Answers2

0

I have built his type of power supply. Place a diode from the AC end of 10uH to ground will reduce the diode losses.
If 10uH is large, during dead time the AC end of the inductor will be one (or two) diode drops below ground. This “shorts out” the transformer. There will be only a small voltage across the transformer. Looking at the wave form there are really only three voltages. “0V”, almost ½ supply and 31V. When any MOSFET is on you will have the supply across the transformer. When no MOSFET is on there should be very little voltage across T1.
That “unexpected opening” is caused by leakage inductance. (mosfets are off during this time)

ronsimpson
  • 866
  • 1
  • 7
-1

The no-load primary winding voltage droop hints that there's too low inductance for the used voltage (=about 155 volts) and pulse length. The current grows during a single pulse higher than mosfets can output with the existing Vgs drive. That drive is unknown

No induction law calculations are possible without knowing the transformer spec, gate drive oscillograph and timing details. Such things are asked already by others in comments.

snakeoil_2023
  • 1
  • 1