Cannot Drive Perm Mag DC Motor with PNP

Discussion in 'General Electronics Chat' started by dtour, May 13, 2013.

  1. dtour

    Thread Starter New Member

    May 13, 2013
    3
    0
    Hello all,

    I'm attempting to switch power to a permanent magnet 12V DC motor that pulls 130mA at no load. The switch in use is a BD238 PNP BJT (Q1). The BJT is controlled with an 2N7000 NMOS transistor (M1). The BJT's emitter is connected to a 12V wall wort plugged into 120Vac. I've attached a circuit diagram to this post. It is not shown in the images, but I have added a freewheeling diode across Q1.

    Initial tests (img1):

    • Direct 12V application to motor works fine.
    • Complete circuit can switch a resistive load in place of the motor
    • Complete circuit will turn on motor briefly, but NMOS fails after about 2 seconds when motor is used for load in complete circuit
    Secondary tests (img2):

    • Isolate PNP circuit, can switch resistive load without incidence
    • Isolate PNP circuit, CANNOT turn on motor
      • It's like the PNP is not turning on all the way
      • tried to increase Veb which marginally increased the voltage across the motor (1.3V >> 1.6 V, should be ~12V)
    Questions:

    • Why does the NMOS fail? It seems that it is only creating a voltage divider circuit. Could it be related to the PNP Veb drop when it the BJT turns on, driving the voltage to 0.7V below Ve?
    • Why can't the PNP turn on the motor when isolated?

    Any thoughts are helpful; thanks.
     
  2. ScottWang

    Moderator

    Aug 23, 2012
    4,848
    767
    The Vgs for mosfet is too less, you have to raising the Vgs or using a npn bjt to replace the mosfet.
     
  3. dtour

    Thread Starter New Member

    May 13, 2013
    3
    0
    Scott,

    Thank you for your reply.

    Why does the Vgs of the NMOS matter? All the NMOS is doing is creating a voltage divider circuit, which it aptly does when the load is purely resistive.

    This also does not explain why the PNP is unable to operate the motor when isolated.
     
  4. ScottWang

    Moderator

    Aug 23, 2012
    4,848
    767
    For a logic level Vgs that when it's on(sat) >= 4.5V.
    For a normal level Vgs that when it's on(sat) >= 10V.
    You can check the datasheet to get more details.

    Decreasing the values of R5 to get more -Ib and to drive more -Ic.
     
  5. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    656
    To saturate a BJT, you need Ib≈Ic/10. This means that R5 needs to be about 1k. 20k is WAY too high. If you look closely at the PNP datasheet, you will probably see that Vce(sat) is specified when Ic/Ib=10. You are probably using the beta (Hfe) spec, if you looked at the datasheet. The problem with this is that beta is specified with Vce>>0V, like maybe 5V.
    Your MOSFET may be being destroyed by motor flyback voltage when you turn it off (although I can't see a path for that). Put a diode (1n400x) across the motor, anode to ground.
    Also, 3V is not enough to turn on a MOSFET unless it is a logic-level device, specified for Rds(on) when Vgs=3V.
    You can replace the MOSFET with an NPN, with a series base resistor of about 1.8k.
     
    Last edited: May 13, 2013
    dtour likes this.
  6. #12

    Expert

    Nov 30, 2010
    16,247
    6,744
    I would think even lower resistance considering 130ma at no load, the start current might be several times as much. I'd do the math for 1/2 to 1 amp.

    The 2N7000 can barely pass a tenth of an amp at 3.3 volts Vgs. R5 needs to be about 120 to 220 ohms. Then you need to measure the voltage on the 2N7000 to make sure it has very little voltage across it or it will get too hot at 312.5 Centigrade per watt. This might be what's killing your 2N7000. I think the 2N7000 is the weak link here. You can try adjusting R5 as high as it needs to be to protect the 2N7000 and see if the result will drive the PNP. If not, you need a different transistor.
     
    dtour and Ron H like this.
  7. Dodgydave

    Distinguished Member

    Jun 22, 2012
    4,964
    743
    Last edited: May 14, 2013
  8. dtour

    Thread Starter New Member

    May 13, 2013
    3
    0
    Hey thanks to everybody for your replies. I hadn't played around with discrete components on a bread board for a while and forgot about all the "fun" that follows.

    Per Ron's suggestion, I replaced the NMOS with an NPN that has Vebo = 3V and R5 = 1.5k. The circuit now functions properly and nothing heats up. Fortunately my friend just informed me of a little mom and pop electronics shop in town or else I would be out another couple of days for a Digi-key order and shipping to Hawaii. Thanks Ron.
     
Loading...