Half Bridge DC-DC Converter

[Gadzooks]

half bridge dc-dc converter

I've built a half bridge DC-DC converter but have a problem. I used a NCP5106A mosfet driver; Vd = 30 V; SUP40N10 NMOS from Vishay; fs = 40 kHz.
The switching signals on the gates of the mosfets are correct. However, when a mosfet is switched off the output point between the 2 mosfets rings a little bit (for 360 ns) and then the mosfet that just turned off switches back on for around 1 us (by itself the gate signal is low).

Why???

 

[Prototyp_V1.0]

half bridge dc/dc converter

If you have atached a long (?) cable, that could cause ring because a cable will behave as many capacitors in parallell and inductors in series.
A motor will also behave that way.

One cause for this is probably that the nmos won't open so much that you want it to. That means that you get higher output impedance than you want.

 

[itaig]

I would say it is inductance on the source of the MOSFETs. The driver IC needs to be connected directly to the source pad not to a track carrying the load current.

It is a very common problem when driving MOSFETs or IGBTs. Due to track inductance the source can go slightly negative thus increasing Vgs thus turning the device back on.

A common fix is to put a cap directly across the gate and source right at the FET. This slows the FET down a bit decreasing the di/dt and thus decreasing the effect of the inductance. Down side is it increases switching losses.

Another point is do you have any snubbing?

 

[dick_freebird]

The Cds of the low side FET might be overcoming the
gate drive strength on the ring overshoots; too high
an edge rate will increase the ring amplitude, too much
output capacitance at the switching node will increase
the ring energy and duration.

Weak decoupling that is drawn down by the output is
also a possibility, and any ground inductance at the
driver (or relative to the low side switch source) can
also make the driver bounce relative to the prime
powertrain ground-point.