Connect with us

High current FET application

Discussion in 'Electronic Design' started by Dude Man, Dec 11, 2003.

Scroll to continue with content
  1. Dude Man

    Dude Man Guest

    I need to have a very large current capcity while varying the voltage. I
    have an op-amp set up with the right gain and such, but instead of buying
    one FET to carry all the current (20amps or so) I plan to put a few in
    parallel. I understand that they won't share current evenly 100%. Is there
    a rule of thumb as to how much more current one could flow over another?

    Thanks for any help.
     
  2. Ian Stirling

    Ian Stirling Guest

    If hard on, then it's only Rds variations you have to worry about.
    If not, you have to worry about Vgs thresholds causing drastically varying
    current.

    The first case, a few tens-hundreds of milliohms will make the currents
    come out similar.
    You may not even need any ballast.
    In the second case, it can vary significantly.
    Good design practice is to assume that the threshold may vary as much as
    the specifications allow, and to work out the current spread in
    that case.
    One FET may well be the cheaper way.
    What sort of vooltage is across the FET?
     

  3. Seeing the circuit you have in mind might help. But, I *imagine* what
    you're doing is building some sort of power supply, using the FET in its
    linear range as a series regulator?

    Anyway, see other active threads about MOSFETs: in the linear range (that
    is, in between "all the way off" and "all the way on"), many of them have
    negative Vth temperature coefficient. That is, as they get hotter, they
    turn on more, meaning they take more of the current, meaning they get hotter
    still. Really the only way around that problem is to use series resistors.
    App notes abound; look at recent threads on paralleling transistors to get
    more current out of an LM317 (search in this ng for "LM350" to find one such
    thread).
     
  4. Dude Man

    Dude Man Guest

    Thanks Ian, Walter,

    I understand the issues you are talking about. I checked the catalog
    and I think I will take the advice and just buy the right part in the first
    place. It's not nearly as expensive as I was thinking. That way I don't
    have to get in to ballast and tungsten or anything. It just means I have to
    wait longer....


    : I need to have a very large current capcity while varying the voltage. I
    : have an op-amp set up with the right gain and such, but instead of buying
    : one FET to carry all the current (20amps or so) I plan to put a few in
    : parallel. I understand that they won't share current evenly 100%. Is
    there
    : a rule of thumb as to how much more current one could flow over another?
    :
    : Thanks for any help.
    :
    :
     
  5. Dude Man wrote...
    Be careful, the datasheet-stated power dissipation and current-handling
    capabilities are for 25C case temp, which implies a perfect contact with
    no insulator to a perfect heat sink whose surface never goes above 25C.
    And under these perfect conditions the junction temp goes right up to
    the absolute maximum rating, which you want to stay well below!

    Why not tell us more about what you're doing and include some numbers?
    And the part you've picked?

    Thanks,
    - Win

    whill_at_picovolt-dot-com
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-