Component advice for an H-Bridge

Discussion in 'The Projects Forum' started by Dream-, Aug 23, 2011.

  1. Dream-

    Thread Starter Active Member

    Nov 26, 2010
    69
    0
    I am currently designing an H-Bridge that needs to operate on the range of 160-180 VDC with a current load in the neighborhood of 15A. The load is non-inductive and it will be driven by a microcontroller PWM and appropriate gate drivers.

    Should I seek MOSFET or IGBT transistors for this application? Is it practical to increase the current output by using transistors in parallel?
     
  2. Dream-

    Thread Starter Active Member

    Nov 26, 2010
    69
    0
    It seems that doing parallelization with the MOSFETs is the natural way to go.

    So if I have a 180V 15A load, my power is 2700W. If I use 4 transistors per block, I would need a minimum of 675W of power dissipation on each one. I am assuming that my total power dissipation will be less than the sum of the 4. So I am thinking of getting 700W rated MOSFETS.

    Here's two that could work. One is much more expensive but the package might be better suited for this project, because I am thinking that regular PCB traces might not be good enough for this much power. (Then again if they sell them in through hole packages, they must work)

    http://datasheet.octopart.com/FDH44N50-Fairchild-datasheet-50972.pdf
    http://datasheet.octopart.com/IXFN80N50-IXYS-datasheet-546085.pdf
     
  3. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Power dissipated in the (switching) MOSFET will be voltage drop on RDSon x current + switching losses. (Not multiplying Busbar voltage with current)
    Switching losses will depend on switching frequency and gate drivers/trace inductances etc.

    You should be good with 1 of these transistors on each side (4 in total).

    There are also several good options from International Rectifier with higher current and less voltage (irf.com).

    Provide a good heatsink for the MOSFETs and make sure you have a stable DC / good bypass capacitors (low ESR) and short traces to them and between the MOSFETs.
     
    Dream- likes this.
  4. Dream-

    Thread Starter Active Member

    Nov 26, 2010
    69
    0
    Of course! I don't know what I was thinking of when computing the power dissipation that way. If all voltage was being dissipated there would be no output, just a hell lot of heat, hehe.

    Thank you for the correction. Now I realize that my life is much easier since my requirements are easily met with many MOSFET models without the need for parallelization.

    With an Rds=0.27Ohm, and a 15A current, the voltage drop is V=4V, and the power dissipated due to the MOSFET resistance P=60W. Not counting switching losses, all seems good.
     
Loading...