This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM5066: Operation Issue

Part Number: LM5066

Hey Team,

I have a customer using the LM5066 and they are experiencing a ~70ms drop in the output as the input switches from 10V to 28V.

We have tried adding an RC filter to reduce the inrush current and have also explored adjustments to the TIMER pin, but worry that this may cause some issues. Even when bringing the TIMER pin to a minimum capacitance we see about 8ms of voltage drop. 

Any additional advice is greatly appreciated. This is urgent and has become a bottleneck to development.

Best,
Dajon McGill

  • Hi Dajon,

    A Vin step cause large inrush current to flow to charge the output cap. This creates large differential voltage across Rshunt which triggers circuit breaker event and restart. This is the expected behavior of any hot-swap protection controller.

    To confirm, please share Vin, Vout, GATE, FET current waveforms in a single scope shot. 

    If it confirms, we need to slow down the Vin step slew-rate by higher Cin or passive L filters. 

    Best Regards, Rakesh

  • Vin_10_to_28

     

    VIN_28_to_10

    I_IN_28_to_10

    I_IN_10_to_28


     

    Timer_28_to_10


     

    Timer_10_to_28


     

    I couldn’t record the timer and current simultaneously due to the current setup, but I got screenshots of the same exact transitions (from 10V in to 28V in and the other way around), one with the timer, and one with current input.

    • In all screenshots, the green curve is the gate signal
    • In the screenshots starting with “I_in” and “timer”, the yellow curve is Vout
    • In the shots starting with “VIN”, the yellow curve is Vin
    • Lastly, in the ones starting with “timer”, the blue curve is the timer signal
    • In the ones starting with “I_IN”, the purple line is the voltage across the shunt
  • Thanks.

    Kindly capture actual current and confirm whether the device is hitting circuit breaker event during 10V to 28V transition. If it hits circuit breaker event, we have to either slow the ramp-rate of 10V to 28V or blank the information through large enough filter across the Rshunt.

    Best Regards, Rakesh

  • I haven't gotten the chance to get a good grab of the current going through the channel, but I do have a follow up question: Regardless of the current seen, if the LM5066 hate the gate fault mask asserted (D7h), is there any chance for this behavior to still appear?

    Thanks,

    Constantinos

  • Hi Constantinos,

    If GATE_MASK register bit is set HIGH, it will avoid turning-OFF the GATE but there is associated risk as it blanks the GATE for real fault also.

    Best Regards, Rakesh