Doubt in MOSFET

Discussion in 'General Electronics Chat' started by sudharshan.panduranga, Jun 8, 2010.

  1. sudharshan.panduranga

    Thread Starter Active Member

    Apr 10, 2008
    61
    1
    Hi.. to all,
    I have some doubts in selection of MOSFETS using Datasheets

    1. For SMPS what are the parameters to be consider in the datasheet.
    I ll consider Vds, Id, Rise time and fall time. How to find whether it is suitable for switching frequency.

    2. How to calculate maximum temperature rise in the MOSFETs. What are the parameters related to temperature rise.

    Please give with one example.


    Regards,
    Sudharshana S. P
     
  2. Darren Holdstock

    Active Member

    Feb 10, 2009
    262
    11
    Sorry, no calcs, this will be a suck-it-and-see answer. The first thing you need to know is the switch current the MOSFET will be expected to handle. Usually the SMPS datasheet will give you a handy equation or graph here; expect the switch current to be several times larger than the SMPS output current. Add about 20-30% reliability margin, and there's your drain current spec. The MOSFET Vds spec is fairly trivial - again add about 20-30% margin. Check the SOAR graph on the MOSFET datasheet to check the viability at maximum current and maximum voltage (you can't subject a transistor to both of these stresses simultaneously).

    Now comes the tricky bit, but there's an easy way to do it. You're quite right Sudharshan to be concerned about the switching frequency and power consumption, and as with many things there are tradeoffs involved. First you need to choose your switching frequency, and what you choose here will depend upon your design priorities. A low frequency will give you better efficiency/lower heat generation, but will require bigger, more expensive inductors and capacitors. Higher frequencies will make for a smaller, cheaper power supply, but the efficiency will be lower, and it will run hotter. Now you've picked your frequency, it's time to choose the switch transistor. A high frequency SMPS will be more efficient with a higher Rds(on)/lower gate charge FET, while a lower frequency SMPS will be more efficient with a lower Rds(on)/higher gate charge device. It's a trade-off between switching losses driving a higher gate charge and static losses with a higher Rds(on). It's simplified a bit in that a low Rds(on) implies a high gate charge, so that narrows down the choice.

    Here's where you cheat: Ignore the dissipation calcs, they're complex and the chances are they'll give the wrong answer anyway. Buy a range of different FETs that meet the current, voltage and SOAR criteria, and audition them by using in the SMPS circuit and measuring the efficiency under the full output load. The one that is the most efficient wins! Take 3 temperature measurements while you're doing these auditions if you're curious - ambient, the SMPS IC, and the MOSFET. These temperature readings should be quite illuminating.
     
  3. sudharshan.panduranga

    Thread Starter Active Member

    Apr 10, 2008
    61
    1
    Thanks for replying


    Regards,
    Sudharshana
     
Loading...