My CMOS always destroyed..

Discussion in 'General Electronics Chat' started by mahmoud shendy, Aug 31, 2008.

  1. mahmoud shendy

    Thread Starter Member

    Dec 23, 2007
    21
    0
    Dear all, nowadays I develop a design that controls speed of 220Vac universal motor, the control basically depends on rectifying the mains voltage using a full wave bridge, then applying the rectified voltage to the motor and switch it using IRF740 "n-channel CMOS" in a PWM fashion to control the speed, despite of connecting a Free Wheeling Diode across the MOTOR, the CMOS get destroyed many times!! How to overcome this problem?
    Also, how to smooth current of the motor as I use 7KHz chopping frequency and the mains of 50Hz..

    Thanx a lot ...SHENDY
     
  2. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Please post circuit diagram.
    Are you sure the motor can work with 220V DC?
     
  3. mahmoud shendy

    Thread Starter Member

    Dec 23, 2007
    21
    0
    here is the schematic!!
     
  4. blocco a spirale

    AAC Fanatic!

    Jun 18, 2008
    1,440
    368
    Does the MOSFET get hot ?
     
  5. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    please answer all these questions
    1)Are you sure the motor can work with 220V DC?
    2)What is your PWM frequency?
    3)What is the value of R4?
    4)What type is the diode?
     
  6. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Add a couple of low omage resistors to the emitters of the BJT's, you are going to generate current surges through the transistors, but it will reduce it some.

    Don't forget you will have a PEAK voltage of 308V pulsating DC on the MOSFET (which I assume is what you're talking about). Even a small cap (around 10uF) will reduce this somewhat, though it may not be practical in this application.

    The speed of your swamping diode may be an issue, spikes could be getting through just because of thier speed. I'm not sure it applies, but you could try a Schottky diode to see if it makes a difference. You might also try a small cap around your inductive load to see if it makes a difference.
     
  7. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I threw together a somewhat modified version of your schematic.

    Included is a cap across the swamping/flywheel/anti-EMF diode Bill was referring to. Large rectifier diodes tend to have slow switching times; during the time it takes to start conducting, the reverse EMF can build to extremely high voltages. The first time that happens, your MOSFET will be destroyed.

    Also, you're attempting to drive an emitter follower arrangement from TTL levels. Output from a standard TTL device may only be as high as 2.8v. With your emitter follower arrangement, this puts the gate of the MOSFET at somewhere around 2.1v-2.2v; or in the threshold (linear) range. This means the MOSFET will be dissipating a heck of a lot of power, probably more than the motor itself.

    I added a level conversion using a 2N2222 transistor. This inverts the PWM signal, but allows driving the gate of the MOSFET at 10v.

    Many of the values and components I used are simply guesses, and were what was available in the stock LTSpice library. D5 won't carry enough current for your needs, and the cap may need to be larger; perhaps as much as 1nF (1,000pF)

    No clue what your motor's resistance might be. You have not provided power information.
     
  8. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Hi Sgt. Wookie,
    Sorry for nit-picking but the voltage driving the original emitter-followers is +5.0V from a Cmos micro-controller, not TTL.
    Also the gate voltage of your Mosfet is about 9.3V, not 10V.
     
  9. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    OK, I didn't see the part where our OP mentioned CMOS uC; just that it was a uC with TTL level outputs. Still, even if his output was nearly 5v, it would drop down another 0.6v-0.7v out of the common emitter driver, giving them perhaps 4.3v-4.4v on the gate. Unless it was a logic level MOSFET, it would be operating in the linear region.

    Agreed. Using an emitter follower like that, the voltage range on the gate will be from about 0.7v to 9.3v. That's certainly far better than 0.7v to 4.3v. Having Vgs ranging from 0v to 10v would certainly be preferable.
     
Loading...