FET and Power problems

Discussion in 'The Projects Forum' started by themotorman, Jun 13, 2009.

  1. themotorman

    Thread Starter Member

    Jun 13, 2009
    First I'd go with a PWM system , However the mosfet failuer mode might be simple. The DC brushed motor puts out a lot of high voltage transients and they increase with the power and current. Sinc eyou have no easy way to stop them you can abosorb them either by using transient suppressors across the motor or sometimes a capacitor and resistor combo. Both of thes eideas will reduce the sparking.. and HV problems.
    For a check try isolating the motor with a small inductor on each side of the motor as close to the brushes as you can and put a capacitor 10 nF seems about right and also put transorbs ( voltage limiting energy absorbers) across the brushes.
    If this works and your mosfets are safe then the cooling issue will come up .. if you insist on a linear operation then you will have to use enough mosftes and heat sinks to handle the power.
    Nice ideas to play with.
  2. themotorman

    Thread Starter Member

    Jun 13, 2009
    I have similar problems with DC motors driven by PWM . I think the failure is due to a gate failure although I have tried diodes across the gate-source and also a transorb ( non-linear semicoinductor) and a zener. I find that i have very high spikes with severe ringing at each switching time even though I have ultrafast diodes acorss the motor , ferrite and cap suppression for sparking of brushes. The probelm is one of high currents, low resistances and high voltages.. a difficult mix!
    I do not think it is the brushes as I can get a failure with the rotor locked no sparking then. . In th e ideal world I could keep all wires very short but in an EV it is hard to keep battery leads, and or motor leads so short that they do not present a high impedance to the very high instantaneous currents. Putting a large cap across the line seems a possible idea but the physical size makes this impractical. I have tried slowing down the MOSFET drive but this will increase the switching losses, and gets rid of the advantages of a MOSFET, and still had failures.
    I kknow it can be done as there are several commercially available controllers that claim to be bullet proof.
    I would like some ideas to try as I am building up a supply of blown MOSFETS... BTW if the gate is shorted to the drain after failure then the failure is usually due to gate oxide breakdown. This can occur in nanoseconds so difficult to see happening .
    any one with a bullet proof circuit for 50 amps 200 volts system??
  3. beenthere

    Retired Moderator

    Apr 20, 2004
    Welcome to AAC.

    I am not sure how well related your two posts are, but, since they both hijack this thread - http://forum.allaboutcircuits.com/showthread.php?t=16801&page=4 - I though I would keep them together.

    Possibly you could post up a schematic so we can comment on the hardware?
    Perhaps you see how difficult it becomes to reply to questions when they are posted in another's thread?
  4. DC_Kid

    AAC Fanatic!

    Feb 25, 2008
    a few suggestions.

    1. use avalanched rated fets
    2. use a zener to ground to help protect the fet
    3. what henries is the motor?
    4. ringing on the drain after fet stops can be squashed with a ceramic and resistor to ground (snubber)
    5. what fets were you using?

    50A @ 200V doesnt sound like much.