A few random questions... (power mosfets & grounding)

Discussion in 'General Electronics Chat' started by sdrawkcab, May 14, 2007.

  1. sdrawkcab

    Thread Starter New Member

    May 14, 2007
    Okay, sort of new at this, hopefully I've worded everything so it makes sense...

    1. How feasible is it to drive a pretty hefty power mosfet using only a standard comparator? Turn on/off delay is absolutely not an issue here, as long as it isn't up in the millisecond range or something. Knowing the maximum output current of the comparator, the threshold voltage of the mosfet, and the input capacitance of the mosfet, is there an easy way to calculate the expected turn on/off times? Is the output current of a comparator internally limited, or will I have to externally limit it?

    2. I'm working on a battery-powered project that also uses a transformer. What kind of grounding considerations will I have to take on the secondary side? My case is non-conductive, so I'm assuming I'll have to put in some sort of large neutral conductor. How would I size this? Or am I completely on the wrong track here? Also, are there any grounding considerations for the primary (battery) side?

  2. John Luciani

    AAC Fanatic!

    Apr 3, 2007
    To calculate the turn-on time use the equation I = C dv/dt. You should also
    determine the average power dissipated during turn-on and verify that you
    do not exceed the maximum power rating. You need to look at the
    transient thermal response to determine the maximum power rating.

    I am not sure about what you are asking in question two or what you are
    trying to design. In general you want low impedance paths for power supply
    and ground. There are many trace width calculators on the www that will
    help you calculate resistance ---
    (e.g. http://www.geocities.com/CapeCanaveral/Lab/9643/TraceWidth.htm)

    If you have high frequency switching currents you will need to calculate
    impedance and use proper layout techniques to reduce radiated noise.

    (* jcl *)
  3. gootee

    Senior Member

    Apr 24, 2007
    What kind of comparator? I am not knowledgable-enough to know what is considered to be a "standard" comparator. The few that I AM a little-bit familiar with, such as LM311 and LM339, don't really have what I would call an "output current", at least not in the way I usually think of it.

    Someone will correct me if I'm wrong, I hope, but those comparators basically either short their output pin to their V- supply pin, or not, depending on the inputs' state. It's called an "open collector" type of output.

    To work as a basic comparator, they need a "pull-up" resistor, from their output pin to a voltage source; let's say it's the positive supply voltage.

    When the comparator's + input pin's voltage goes above the - input pin's voltage, the internal transistor switch (whose collector is the comparator's output) stops conducting current through the pull-up resistor (which, if current is defined as flowing from + to -, and the pull-up R is connected to the V+ supply, WAS flowing out of the V+ supply, then through the resistor, then into the comparator's output pin, and then out through its V- supply pin, while the output pin was at about 0v if single supply or about V- if dual supply), which then allows whatever else is connected to the comparator's output to suddenly find itself connected to the V+ supply rail, through the pull-up resistor.

    So, to answer one of your questions: Assuming that my rambling above was more-or-less correct, then the current that comes out of a basic comparator circuit's output (as described above), when the comparator's +in_v > -in_v, would be limited by the value of the pull-up resistor and the positive supply's voltage. And it would come straight from the V+ supply through the pull-up resistor, and not from the comparator itself.

    Good luck.

    - Tom Gootee


  4. John Luciani

    AAC Fanatic!

    Apr 3, 2007
    The value of the pull-up resistor you choose is limited by the comparator's ability to
    sink current. Too low a value of pull-up and the comparator will not change state.

    (* jcl *)
  5. sdrawkcab

    Thread Starter New Member

    May 14, 2007
    To follow up -

    For some reason, whenever I try to stick a pull-up resistor on the comparator output (I'm using LM 311s), the output just pegs high. However, using a pull-down resistor to ground works fine. (I'm working with a single supply, so ground is also my V-).

    The problem now is that when using a pull-down resistor, my output high from the comparator is much lower than V+. For example, at 12V+ supply my output high will be ~9.4V (diode drops?). The output low is a solid 0V. This works great for driving my NMOSes, but presents an obvious problem when I'm trying to drive a PMOS (they're always on).

    Any ideas why putting a pull-up of any size (below the Mohm range) results in the output pegging high?

    Thanks for the help everyone.
  6. beenthere

    Retired Moderator

    Apr 20, 2004
    I think you're going to have to post up your circuit before we can help further. I have driven FET's with LM311's before, with none of the problems you describe.