IRF540 mosfet thermal runaway?

Discussion in 'General Electronics Chat' started by hobbyist, Aug 29, 2008.

  1. hobbyist

    Thread Starter Distinguished Member

    Aug 10, 2008
    764
    56
    I remember reading that a Mosfet is opposite of Bipolar transistors,
    in that resistance goes up as it heats up, so no thermal runaway.

    I have a 15v. supply, with a 5 ohm power resistor, connected to the positive supply rail, a 6volt 3amp dc drill motor then a mosfet heatsinked connected to the neg. rail.

    Vgs=4v. Vds=4v. Ids=1.6A. No fan blowing on mosfet.

    Vds=6.5v. @ Ids=1.2A with fan blowing on Mosfet.

    I'm working on a experimental circuit thats why I need the low Vgs,
    and high Vds. to run the motor at slower than normal speeds.

    But from this observation it seems like the internal apparent resistance goes up when mosfet is cooler.

    And resistance drops when it heats up.

    I'm just curious why it works this way when it supposed to have neg. resistance.
     
  2. theamber

    Active Member

    Jun 13, 2008
    318
    0
    Correct.
    I think you are using an improper Mosfet for your application or you are biasing it wrong.
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    You're running the MOSFET near the threshold voltage, which isn't a good thing.

    Power MOSFETS should be used either fully ON or fully OFF. You should consider using PWM drive for your gate (Pulse Width Modulation). With PWM, you can control the ratio of ON time to OFF time, which is vastly more efficient than using a MOSFET in a linear mode.

    If your Vgs must be limited due to logic level considerations, then I suggest using something like an IRLZ44 instead, which is designed to be used with logic levels on the gate.

    Just a simple fan blowing on a non-heatsinked MOSFET that is being operated near threshold voltage is nowhere near enough cooling.

    If you took those readings with Vgs being the same both times, in order, and simply switching the fan on, then it's likely that the MOSFET was getting a LOT hotter between the times you took the readings.
    4V x 1.6A = 6.4 Watts, Rds = 2.5 Ohms
    6.5V x 1.2A = 7.8 Watts, Rds = 5.416 Ohms
     
    Last edited: Aug 29, 2008
  4. hobbyist

    Thread Starter Distinguished Member

    Aug 10, 2008
    764
    56
    Thanks guys I appreciate the quick reply's in helping me with my questions.

    I will try to explain in full detail my project I'm trying to accomplish, and all my procedures. In this same thread.

    Bye for now.

    thanks
     
  5. hobbyist

    Thread Starter Distinguished Member

    Aug 10, 2008
    764
    56
    ADMINISTRATOR
    If this is to much writing for a thread than please let me know and feel free to delete it, I don't want to cause any problems, this is a very helpful forum.



    Thanks Guys
    I appreciate the time you take to help me in my questions.
    I'll try to give a major background on my project.
    First of all my hobby and fascination with electronics is trying to build from scratch commercial circuit functions using all discrete components. This is just the enjoyment I have in circuit design.
    So all my projects are mainly just for the challenge and accomplishment of learning and designing from scratch. I have learned so much about electronic circuits as a hobby doing it this way.
    The project I'm working on now is a PWM for a small cordless drill motor.
    I have read on the innernet all I can at this time of what PWM is and how it works. So I got the enthusiasm to try to design one of my own.
    The first one I designed was made for a small continuous servo motor,
    with torque controle. It works real well.
    Now here is how I'm designing my new one, and where I'm at this moment with it.
    I tested and got all the data for Voltages to run the motor at high speed and low. I gathered all the Vgs data as well as Vds. after I aquired all data necessary, I started with my Gate driver.
    simply a emitter follower. built and test and adjusted values.
    then I started to work on the I call it "wave converter circuit", which is a two transistor schmitt trigger circuit. built tested and adjust values until that worked with the rest of the circuit.
    Now I began designing a CE. class A power amp, with a irf610 mosfet connected to the emitter resistor, from ground.
    It is this part that will increase the torque to themotor.
    Then I worked on a feedback regulator to couple into the mosfet.
    Here is how my first small motor drive works using this circuit design.

    I adjust my signal generator to a low frequency about 40Hz. with proper input V. The signal is applied to the class A amp. after amplification it is sent to the wave converter which converts it to square wave output, which is sent to gate driver and through mosfet driver to motor. There is a 5ohm sense resistor in series with the motor.
    As the motor is stalled the decrease in voltage across the motor is sensed by a darlington pair of emitter follower trans. which is biased just below the motor voltage under normal running conditions.
    As the motor stalls the decrease in motor voltage forward biases these transistors, thereby sending that sense current to the base terminal of the regulator feedback trans. causing an increase in output of the regulator, that feeds into the gate of the irf610 mosfet which in turn drops the emitter voltage lower at the power amp. causing more input voltage to be amplified and sent to the schmitt trigger which in turn allows more pulse width thereby increasing motor voltage and under stalled load, increased torque. It works good for small milliamp motor.
    So when I tried to use the same scheme for this larger amp motor, everything needs to be redesigned. Concept should work but extra circuitry needed.
    First flaw, with this motor when it is stalled I only get 200 mv. change across it. With the voltages and currents this circuit needs this is not large enough voltage change, to cause same results as other circuit with smaller motor.
    So I am sidetracking and breaking down the whole design and trying to see if I can sense a millivolt signal change and amplify it.
    After about 10 diffrent approaches I finally hit on a dc amp. Through this I have learned a lot about transistor amplifiers I never knew before. A Blessing in disguise.
    I now have a 4 stage CE amp. that will drop 3 volts for a 220 mv. drop.
    AV appr. 14.
    I decided on this project to keep a notebook off all my work, calculations data, draw schematics and waveforms etc... this is fun to do. Because I can look back through my notes and see what I did to get these values.
    So Now I want to move to the next step and redesign this amp with fewer transistors using both npn and pnp and once I get that working properly I will start working on a voltage regulator to increase the gate driver when the motor dorps this voltage.
    So this is where I'm at right now with this project. I would like to share more as this project goes on.
    But I don't know if it's against any rules to have a long thread like this.
    I'll try to keep it short next time on this same thread.
    Ps. ADMINISTRATOR
    If this is to much writing for a thread than please let me know and feel free to delete it, I don't want to cause any problems, this is a very helpful forum.
    Thanks again

     
  6. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    The post is not too long, and if you can share schematics with us, we can probably offer suggestions and possible corrections.

    Simple-sounding projects often become lots more complex.
     
  7. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Your post isn't too long. However, it would help readability a great deal if you would use paragraphs, and avoid excessively long sentences.

    5 Ohms is far too much resistance for a current sense resistor in this project; even a 0.5 Ohm resistor would be mighty large, dissipating 750mW, and would give 1.5v across it at full load. A 0.1 Ohm resistor would be more like it.

    Try a 40% PWM duty cycle, upwards of 1kHz in frequency.
     
  8. hobbyist

    Thread Starter Distinguished Member

    Aug 10, 2008
    764
    56
    Quote {5 Ohms is far too much resistance for a current sense resistor in this project; even a 0.5 Ohm resistor would be mighty large, dissipating 750mW, and would give 1.5v across it at full load. A 0.1 Ohm resistor would be more like it.}

    Yeh, your right Thats why I have to use such a large supply voltage,
    Now that I'm getting a good handle on how to amplify this small change in voltage, I could lower the sense resistor as well as the supply.

    Thanks for the tip..

    I'll post some schematics later as well..

    Thanks guys.
     
Loading...