5
\$\begingroup\$

this is a series resonant half bridge, used with a loosely coupled coils (so there is significant leakage inductance) a decoupling capacitor for the half bridge is used (not shown in schema) (maybe it's not well tuned.. the half bridge circuit used is a tiny board with the capacitor) circuit

When we observe the ouput of the half bridge ( voltage and current ) we have this : at resonant first experiment near resonance the resonance is at 50kHz for the second one resonance is at 50 kHz

the magnetic coupling isn't the same for the two experimental tests. i need leads on the cause of the voltage overshoot. stray inductance in the mosfets? i don't thing so for this voltage leve (10V input) i am suspecting the recovery of the free wheel diodes?

thank you

\$\endgroup\$
4
  • 1
    \$\begingroup\$ First you should check if your measurement setup is fine, like not using the gnd antenna of your scope. If you suspect something going on at the diode, then do some measurements. Also without knowing any of the components, voltages and currents involved, we can only guess. \$\endgroup\$
    – PlasmaHH
    Nov 10, 2015 at 10:14
  • \$\begingroup\$ I see nothing unexpected in the pictures - what is bothering you about the pictures? \$\endgroup\$
    – Andy aka
    Nov 10, 2015 at 10:17
  • \$\begingroup\$ @Andyaka i need some leads on the cause of the voltage overshoot when there is signinfant current at the commutation. \$\endgroup\$ Nov 10, 2015 at 10:19
  • \$\begingroup\$ Search for snubber circuits. \$\endgroup\$ Nov 10, 2015 at 11:24

2 Answers 2

2
\$\begingroup\$

i need some leads on the cause of the voltage overshoot when there is signinfant current at the commutation

With perfect MOSFETs, switching from high to low with zero switching time, there will be no theoretical overshoot. Reality is a different matter.

You have current in the inductor and there is a brief period in time where there is a gap in the switching between one MOSFET and the other. That means the current in the inductor will create a spike and the only thing dampening that spike is the output capacitance of the MOSFETs. This will reduce the effect of the spike and turn it into a high frequency decaying sinewave.

\$\endgroup\$
1
\$\begingroup\$

One more thing that I see here is that to control the MOSFETs gates you're using a simple inverter gate. It obviously has a non 0 propagation delay. Therefore when 'b' (the upper MOSFET gate) transitions from low to high 'a' will still remain high for the propagation delay of the inverter. For this brief moment you're basically creating a current spike on your voltage source. It is better when 'b' transitions from high to low though as the delay works to your advantage.

You could add a circuitry to control the gates that would follow this pattern: OFF/OFF-OFF/ON-OFF/OFF-ON/OFF-OFF/OFF instead of ON/OFF-OFF/ON.

\$\endgroup\$

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.