Two-MOSFET Forward Converter: One MOSFET Gets hotter by 33% ???

Discussion in 'The Projects Forum' started by davefromnj, Oct 14, 2008.

  1. davefromnj

    Thread Starter New Member

    Oct 14, 2008
    4
    0
    deletemeplease
     
    Last edited: Nov 26, 2008
  2. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    A schematic would be helpful buddy.
     
  3. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    You have to use a p-channel mos for the high side mos because to fully turn on the high side mos you need to apply a voltage 10 volts (as to fully turn it on) above the bus voltage. If it is less than 10 volts above the bus voltage it works in the linear region and gets overheated. Also, this circuit when you turn both mos on will oscillate at a very high frequency (the highest the mos can turn on/off) if your gate voltage on the high side mos is less than Vbus+5 Volts (approximately). This is because when the mos are first switched on the voltage on the source of the high side mos will rise to almost Vbus, this will turn off the high side mos, then the voltage on the source of the high side mos will drop to almost zero, then the upper mos will turn on again and so on.
     
  4. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    It looks like you're beating up your top FET. The snubber diode D2 should go around only the transformer coil, not that and Q2. The gate is possibly not remaining 10 volts above the source.

    Your losses are going to be proportional to the transit times each device spends going from full conduction to full turn off. You can see what each FET is capable of from data sheets, but the drive circuitry - not shown - is what makes the FET's turn off and on.

    Question - why are you using two FET's in this circuit? What does Q2 contribute?
     
  5. nanovate

    Distinguished Member

    May 7, 2007
    665
    1
    The schematic is correct for his SMPS topology.

    Allows you to use lower voltage rated FETs and to reduce the amount of leakage inductance energy that has to be dissipated.

    If the OP can post his driver circuitry then we might be able to help more.

    The OP can try adding a Schottky diode across the upper FET. Its reverse voltage rating should be about 30% higher than the line voltage. You do not need 500V FETs if the input is indeed 240V -- for your topology you do not need 2X the voltage rating.
     
    Last edited: Oct 14, 2008
  6. thingmaker3

    Retired Moderator

    May 16, 2005
    5,072
    6
    How is the upper MOSFET driven? What is the Vgs for it? Are you sure it is being fully turned on?
     
  7. nanovate

    Distinguished Member

    May 7, 2007
    665
    1
    There might be current flowing through the body diode to the input cap when the FETs are turned off.
     
  8. nanovate

    Distinguished Member

    May 7, 2007
    665
    1
    Most power designers end up measuring this but a good first order equation is:

    Psw = ( V^2 * fsw * Id * C_rev_transfer_cap ) / Igate

    C_rev_transfer_cap may also be called Crss in the datasheet
     
Loading...