I have two questions on placement of freewheel diode in a H-Bridge circuit:
1) In reference to the general H-Bridge circuit below, wouldn't it be safer to place the freewheel diodes directly across the inductor? Wouldn't this solve the problem of high voltage spikes getting into the supply rail?
2) If it is ok to place the diodes (TVS) directly across the inductor, what would the orientation look like? Would the connection in the schematic to the left (with L2 and D6 through D9) works?
D1 to D4, D7 & D9 are TVS.
simulate this circuit – Schematic created using CircuitLab
Best Answer
A H-bridge with an inductive load and fly-back diodes in-built into the MOSFETs (or supplemented) relies on dumping the fly-back inductive energy into a capacitor across the supply and therefore the perceived spike becomes a relatively small ramp up in voltage.
Within one PWM cycle (for instance) that small voltage ramp is nulled out. What you see over several cycles is a small triangular ripple voltage superimposed on the power rails. What you get is energy recovery and an increase in efficiency due to storing and reusing that energy.
So, determine the leakage inductance of the load, estimate the peak current in the inductive load and calculate the power supply capacitance needed to limit the ripple to (maybe) 1Vp-p.
I'm all for using the H bridge as efficiently as possible.