Mosfet Drive

Discussion in 'The Projects Forum' started by Ashkar, Jul 18, 2011.

  1. Ashkar

    Thread Starter Member

    Feb 7, 2011
    71
    0
    I want to use my microcontroller pwm output to drive a power dc motor of 12vdc .I want to use irfz44n mosfet.just the circuit is to be able to increase or decrease the motor speed nothing else.So what configuration to implement.
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,156
    3,063
    Could be very simple, but you need to specify the current demand (including startup and stall current) of your motor.
     
  3. Ashkar

    Thread Starter Member

    Feb 7, 2011
    71
    0
    Continous 2.0 amp @ 12v
     
  4. wayneh

    Expert

    Sep 9, 2010
    12,156
    3,063
    OK, well that MOSFET should be plenty for that. At the simplest, you need to send a ~12v "digital" (full on at 12v or full off, 0v) signal directly to the gate of the MOSFET. That will open and close a path to ground (the FET's source pin) for the current coming through your motor connected to the drain pin.

    I should have asked about the output from your MCU, the frequency and voltage. If it's 5v, you need to either choose a logic level MOSFET or send that logic signal thru another device (comparator, transistor, schmitt trigger, etc.) to get it up to at least 10v. Otherwise the FET won't fully turn on and will generate heat. At 2A and with a heat sink on the FET, you might get away with a bit of that but it would be better to design it away.

    If your frequency is "high", say 10kHz or more, you'll need input from others here. Your FET is essentially a capacitor and the current required to charge and discharge the gate goes up with frequency. So a current limiting resistor might be needed to limit that current. Also, it's a good idea to put a pulldown resistor on the gate to discharge it if the MCU fails.
     
  5. Ashkar

    Thread Starter Member

    Feb 7, 2011
    71
    0
    So what would be a better and cheap option to drive those mosfets(irfz44n)
     
  6. AnnexGR

    New Member

    Dec 2, 2010
    7
    4
    Ashkar, this really depends on what the capabilities of your uC is. Could you provide us more details like:

    Switching Frequency?
    Voltage output of your uC? (3.3V, 5V, etc...)
    Source and sink current of your uC output?
     
  7. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    I believe this should do.
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    That'll turn off pretty quickly, but the 10k resistor in the gate charge path will cause a very slow turn-on time, and much power dissipation in the MOSFET.

    See the attached for a more robust driver made from discrete components.
    Not shown is a 0.1uF capacitor from the collector of Q1 to ground; without that, the turn-on time of the MOSFET will be slowed considerably.

    Note that this driver is inverting; when the input from V1 (simulated PIC I/O pin) is high, the MOSFET is turned off; when the input from V1 is low, the MOSFET is on. You should set that I/O pin high very early in the power-up sequence.
     
  9. Ashkar

    Thread Starter Member

    Feb 7, 2011
    71
    0
    @AnnexGR I am using a 5v o/p uC Its ATMEL ATMEGA 32 deviceit's about 20mA sink current.
     
    Last edited: Jul 21, 2011
  10. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Yes, my rise time on the gate will be pathetically slow. Thanks for correcting me. I like your circuit as it mimics TTL.
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    It would really be better to use a dedicated MOSFET gate driver IC. They are far faster and can source/sink much more current than the circuits presented thus far. That means the MOSFET can turn on and off much more quickly, lowering the power expended as heat in the MOSFET itself.
     
  12. newbie217

    Active Member

    Apr 12, 2009
    52
    0
    I want to understand this circuit in more detail. Could someone explain the purpose of D2 and L1, or what they are suppose to represent? Also, why is D1 needed?

    Thanks.
     
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    L1 represents the inductive load of the motor.
    D2 is a "flywheel" diode. When the MOSFET turns off suddenly, there needs to be a path for the current that is flowing through the motor; if there is no path for current, the voltage can become very high, very quickly. The high voltage would damage/destory the MOSFET, as it would exceed its' breakdown voltage.
     
  14. newbie217

    Active Member

    Apr 12, 2009
    52
    0
    SgtWookie,

    Thanks for the response. I'm trying to come up w/ a simple, but effective circuit to do fast load transients for resistive loads. I just started using LTSpice, and came up w/ a circuit somewhat similar to the one you posted. Basically, a push-pull output stage driving an NMOS.

    I don't have Q1 which also makes me curious on why its needed. At first glance it looks like a common emitter that inverts the input signal. Does it have something to do w/ increasing the input impedance? What's the best way to simulate input impedance using LTSpice?

    Here is an image of the circuit I came up with and SgtWookie's circuit moded for a resistive load. Just trying to learn and improve from the pros here.

    Thanks!
     
  15. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    In your circuit on the left, your input voltage from Vsource ranges from 0v to 5v.
    This means that the emitters of Q1 and Q2's output will be ~0v+0.63v to 5v-0.63v. or 0.63 to 4.37v. The 0.63v is OK, as the threshold voltage is higher. But, that MOSFET requires Vgs to be 10v in order to be considered fully ON. You are liable to see that MOSFET going up in smoke very quickly.

    Adding another transistor in a common emitter configuration to drive the bases results in a range of 0.63v to 11.37v, since the supply in 12v.
     
  16. oldtech33709

    New Member

    Sep 24, 2011
    26
    0
    An IRF3717 may make a good choice for the FET. It's a 20V, 20A part and at 10Vg the device is hard on.
     
  17. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Nice that the IRF3717 has Rds(on) specified for operation with Vgs=4.5v. Qg (total gate charge) isn't bad either at 22nC's.

    Only thing with it is that newcomers may have difficulties with the SO-8 package; it's pretty tiny, and one would have a hard time plugging it into a breadboard without adapters.

    I've been warming up to IR's IRLU series that are in the IPak (TO-251AA) package; sort of like a mini TO-220 package that you can actually plug into a breadboard.
    http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=IRLU024NPBF-ND
    The IRLU024Npbf is a 17A, 55v, Rds(on)=80m when Vgs=5v, Id=10A; Qg=15nC. Not a bad little MOSFET.

    These IRLU7807's are 43A, 30V, Rds(on)=14.5m @ Vgs=4.5v, Qg=7nC:
    http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=IRLU7807ZPBF-ND
     
Loading...