Shop OBEX P1 Docs P2 Docs Learn Events
Using FETS on DC motors — Parallax Forums

Using FETS on DC motors

ArchiverArchiver Posts: 46,084
edited 2004-07-02 14:33 in General Discussion
Can you suggest a "high speed recovery diode" to be used with a MOSFET
driving a motor (and source)?


>No where in the previous messages did I see anyone
>mention the need for a high speed recovery diode across
>the DC motor.

>Without one, the DC motor will develop one high voltage
>spike which will fry the MOSFET when trying to turn off
>the motor.

>The diode needs to be rated at least 30% of the nominal
>current rating of the motor and have low nanosecond
>recovery time.

>The diode clamps the EMF generated on turn off and
>keeps current flowing in the motor.



[noparse][[/noparse]Non-text portions of this message have been removed]

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2004-07-01 20:38
    FYI,

    No where in the previous messages did I see anyone
    mention the need for a high speed recovery diode across
    the DC motor.

    Without one, the DC motor will develop one high voltage
    spike which will fry the MOSFET when trying to turn off
    the motor.

    The diode needs to be rated at least 30% of the nominal
    current rating of the motor and have low nanosecond
    recovery time.

    The diode clamps the EMF generated on turn off and
    keeps current flowing in the motor.

    Sincerely,
    Ron

    '********************* Previous Messages
    *********************

    Message: 4
    Date: Wed, 30 Jun 2004 21:39:07 -0000
    From: "Dave Mucha" <davemucha@j...>
    Subject: Re: MOSFET keeps burning out

    --- In basicstamps@yahoogroups.com, laurasdog@w...
    wrote:
    > At 08:02 AM 6/30/2004, nathanchronister wrote:
    >
    >
    > >Hi. I'm using the microcontroller to control a DC
    motor. I have a
    PWM
    > >output hooked up to an N-channel MOSFET. (I'm not
    using an H-
    > >bridge because I don't need reverse.) Even though
    the MOSFET is
    rated
    > >25 amps and should easily handle the ~10 amp motor
    current, it
    keeps
    > >burning out. After this happens the motor won't shut
    off.
    >
    >
    > I'm not real familiar with using the stamp PWM
    feature, but I'd be
    curious
    > what frequency it switches on and off at...
    >
    > MOSFET's have a very high DC input impedance on the
    gate. so it
    takes
    > almost no gate current to turn on with a *DC signal*
    but there can
    also be
    > a very large gate capacitance, which means at high
    frequencies it
    can take
    > a LOT of gate current to charge and discharge that
    capacitance.
    > The data sheet for your device shows an input
    capacitance of 1100
    pF which
    > is pretty large.
    >
    > The stamp is probably unable to source enough current
    to fully turn
    the
    > MOSFET on and off at the PWM frequency....
    >
    > Steve


    The is very true. the gate has a large capatacance
    component that
    will not be satisfied with the logic signal from the
    Stamp.

    Check out MOSFET DRIVERS. these tiny chips handle 4
    amps at your
    voltage to drive the FET.

    The problem stems from how the FET works. As the
    voltage on the gate
    slowly (in FET speed) begins to allow the FET to pass
    current, the
    slow voltage does not saturate the FET, but begins to
    activate parts
    of it. The result is your circuit attemps to pass the
    full current
    on a tiny portion of the FET and burns out sections.
    This cascades
    across the rest of the FET and your chip smokes.

    Dave
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-01 22:10
    Ron,
    Good point!... inherently there is a diode formation
    from the substrate of the MOSFET anyway, but this is
    usually considered a "weak" diode and should not be
    relied upon for absorbing EMF spikes in any application.

    BTW) The diode could be placed across the S/D of the
    MOSFET and/or the terminals of the motor.

    >FYI,
    >
    >No where in the previous messages did I see anyone
    >mention the need for a high speed recovery diode across
    >the DC motor.
    >
    >Without one, the DC motor will develop one high voltage
    >spike which will fry the MOSFET when trying to turn off
    >the motor.
    >
    >The diode needs to be rated at least 30% of the nominal
    >current rating of the motor and have low nanosecond
    >recovery time.
    >
    >The diode clamps the EMF generated on turn off and
    >keeps current flowing in the motor.
    >
    >Sincerely,
    >Ron
    >
    >'********************* Previous Messages
    >*********************
    >
    >Message: 4
    > Date: Wed, 30 Jun 2004 21:39:07 -0000
    > From: "Dave Mucha" <davemucha@j...>
    >Subject: Re: MOSFET keeps burning out
    >
    >--- In basicstamps@yahoogroups.com, laurasdog@w...
    >wrote:
    > > At 08:02 AM 6/30/2004, nathanchronister wrote:
    > >
    > >
    > > >Hi. I'm using the microcontroller to control a DC
    >motor. I have a
    >PWM
    > > >output hooked up to an N-channel MOSFET. (I'm not
    >using an H-
    > > >bridge because I don't need reverse.) Even though
    >the MOSFET is
    >rated
    > > >25 amps and should easily handle the ~10 amp motor
    >current, it
    >keeps
    > > >burning out. After this happens the motor won't shut
    >off.
    > >
    > >
    > > I'm not real familiar with using the stamp PWM
    >feature, but I'd be
    >curious
    > > what frequency it switches on and off at...
    > >
    > > MOSFET's have a very high DC input impedance on the
    >gate. so it
    >takes
    > > almost no gate current to turn on with a *DC signal*
    >but there can
    >also be
    > > a very large gate capacitance, which means at high
    >frequencies it
    >can take
    > > a LOT of gate current to charge and discharge that
    >capacitance.
    > > The data sheet for your device shows an input
    >capacitance of 1100
    >pF which
    > > is pretty large.
    > >
    > > The stamp is probably unable to source enough current
    >to fully turn
    >the
    > > MOSFET on and off at the PWM frequency....
    > >
    > > Steve
    >
    >
    >The is very true. the gate has a large capatacance
    >component that
    >will not be satisfied with the logic signal from the
    >Stamp.
    >
    >Check out MOSFET DRIVERS. these tiny chips handle 4
    >amps at your
    >voltage to drive the FET.
    >
    >The problem stems from how the FET works. As the
    >voltage on the gate
    >slowly (in FET speed) begins to allow the FET to pass
    >current, the
    >slow voltage does not saturate the FET, but begins to
    >activate parts
    >of it. The result is your circuit attemps to pass the
    >full current
    >on a tiny portion of the FET and burns out sections.
    >This cascades
    >across the rest of the FET and your chip smokes.
    >
    >Dave
    >
    >
    >
    >
    >
    >
    >
    >
    >
    >
    >
    >To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    >from the same email address that you subscribed. Text in the Subject and
    >Body of the message will be ignored.
    >
    >Yahoo! Groups Links
    >
    >
    >
    >

    Beau Schwabe Mask Designer II National Semiconductor Corporation
    500 Pinnacle Court, Suite 525
    Mail Stop GA1
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-02 14:33
    Use a good quality fast diode, we use EGP10G or FE6D diodes. Otherwise the
    diode in the fet, being faster than the one in parallel, will take the brunt
    of the energy, and nothing will be gained<G>. Just like the FYI person
    explained. Standard 60 Hz diodes, 1N4007, for example, are not fast enough
    to ensure that they will work.



    Original Message
    From: Beau Schwabe [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=w3uZE7-4S68bMTvT9zDqH1_B2UyBucQXA-fAlngIr9N2GDxsGuYW72EDmIo-bF_Nk48KeChPt9tcaWwDSZdeOWc]bschwabe@a...[/url
    Sent: Thursday, July 01, 2004 5:10 PM
    To: basicstamps@yahoogroups.com
    Subject: Re: [noparse][[/noparse]basicstamps] Using FETS on DC motors


    Ron,
    Good point!... inherently there is a diode formation
    from the substrate of the MOSFET anyway, but this is
    usually considered a "weak" diode and should not be
    relied upon for absorbing EMF spikes in any application.

    BTW) The diode could be placed across the S/D of the
    MOSFET and/or the terminals of the motor.

    >FYI,
    >
    >No where in the previous messages did I see anyone
    >mention the need for a high speed recovery diode across
    >the DC motor.
Sign In or Register to comment.