My buck converter output is going to the input!

Discussion in 'General Electronics Chat' started by assassin___0, Apr 22, 2013.

  1. assassin___0

    Thread Starter New Member

    May 10, 2011
    So I designed a generic buck converter (with microcontroller feedback) for battery charging application. I also added voltage and current sensors. My concern is that when I don't apply any input to my buck converter and then attach the 12 V lead acid battery to the output, the battery voltage can be measured at the buck converter input even when the MOSFET is off!

    It's like the output is going to the input. However, when I apply like 20 volts to the input and run the algorithms such as constant voltage and constant current, the circuit behaves as it should, meaning I can supply the battery with constant current or constant voltage.

    Even though this is working... my concern is that the output voltage that is being transferred to the input voltage (when no input is applied) may damage some ICs, specifically the output current sensing IC since that is the only one giving me funky values. So is this normal or is something wrong with my design?

    I can supply the circuit schematic but even though it's not a big circuit, it's still spread out in like 5 pages. Let me know what other info you guys might need.
  2. crutschow


    Mar 14, 2008
    If you have a P-MOSFET for a switch, the intrinsic body diode will conduct in the reverse direction when the MOSFET is off, thus the output voltage will show at the input as you have found. You need to post a schematic if you want an opinion as to whether that can damage any of the circuit.

    One way to prevent this is to use to P-MOSFETS back-to-back (sources tied together) but driven by the same gate signal. That combo will not conduct in either direction when OFF, but will when ON. The disadvantages are that you have the ON resistance of two transistors in series and you now have to drive the capacitance of two gates.