Post Go back to editing

LTC4446 Noise Ripple

Category: Hardware
Product Number: LTC4446

Dear Supporters,

I'm working with the LTC4446, the input signals are generated from uP and the outputs drive four mosfets in full bridge configuration.

The switching frequency is fixed at 1MHz and the selected mosfet are the IRFP140N.

During the EMC test we had a problem with the radiated emission and the problem I found is the ripple on the gate charge and discharge.

The first picture show the scheme and the second picture is the signal on the BG pin of U4.

How I can reduce this ripple?

I tried also with 330nF bootstrap capacitor but the noise is the same.

Parents
  • The ringing may have more to do with your power stage's hot loop (input capacitors + top mosfet + bottom mosfet inductance) than gate ripple. When the top mosfet turns on, it will recover the bottom mosfet's body diode (if there was current flowing) and/or charge/discharge the bottom/top mosfet's output capacitance. This causes ringing. This can sometimes be helped by slowing the top mosfet's turn on (resistor in series with BOOST pin). The other possibility is that dv/dt and Crss is causing the bottom mosfet to turn on. First thing is to observe the bridge leg (bottom drain) and gate voltage at same time. Short gate traces can keep the bottom mosfet off. Note that VCC and BOOST-TS absolute maximum voltage is 14V. Stay below this voltage.

Reply
  • The ringing may have more to do with your power stage's hot loop (input capacitors + top mosfet + bottom mosfet inductance) than gate ripple. When the top mosfet turns on, it will recover the bottom mosfet's body diode (if there was current flowing) and/or charge/discharge the bottom/top mosfet's output capacitance. This causes ringing. This can sometimes be helped by slowing the top mosfet's turn on (resistor in series with BOOST pin). The other possibility is that dv/dt and Crss is causing the bottom mosfet to turn on. First thing is to observe the bridge leg (bottom drain) and gate voltage at same time. Short gate traces can keep the bottom mosfet off. Note that VCC and BOOST-TS absolute maximum voltage is 14V. Stay below this voltage.

Children
  • This is the H-Bridge scheme, the load is the primary side of transformer

    As you suggest this is the gate voltage and the drain voltage

    Add the resistor in serie at boost pin not made effect.

    I saw a little bit improvement replacing the MBSR260 with ES1D diode.

    I don't undestrand because the rising and falling time of gate voltage is around 40 ns, I would have expected 10ns with 1.4nF of gate capacitance