motor control with PWM

Discussion in 'General Electronics Chat' started by mik3, Aug 13, 2008.

  1. mik3

    Thread Starter Senior Member

    Feb 4, 2008
    4,846
    63
    Hi guys,

    I want to control the speed of a 12v DC series wound motor with PWM. I would like some suggestions on which transistor to use. The motor draws about 35 Amps at 12 volts (3300 rpm) and 100 amps at 10 volts (990 rpm). I want to chop the 12 Volts DC from a big lead acid battery to make an average voltage of 10 Volts. I would like to use a transistor which can handle a continuous current of 200 amps (safety factor of 2 because the maximum current at 10 volts is 100 amps) but it can handle large surge current until the motor speeds up. Can you suggest any?
    What is more, if someone has a good circuit to control dc motors i would like to share it to take some ideas.
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I believe I'd use several IRF3703, IRF1404 or IRF3205 N-ch power MOSFETs in parallel, well heat-sinked of course. Although they're individually rated for 130A to as high as 210A, the package size gives a practical limit of around 65A. You undoubtedly know already about the ease of paralleling MOSFETs due to their positive temperature coefficient; if one gets warm, the others take over.

    The challenge then would be to ensure the gates get charged/discharged as quickly as possible. You might do that using some NPN and PNP transistors capable of several amps of current wired as emitter followers.

    Then you'll have the tradeoff of keeping the PWM frequency low to minimize the transition time of the MOSFETs, or 20kHz+ to get it out of audible range. Running a motor with that kind of power might get a tad noisy if you're running at a kHz or so.

    You'll need some rather heavy duty diodes across the motor to take care of the reverse EMF, too - with some small caps (perhaps 330pF to 1nF) across them to allow them time to conduct. Big diodes = slow diodes - but you wouldn't want the caps to be very large.
     
  3. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,671
    899
    The last high-power motor controller I made was about 8 years ago. I used 5-parallel IRF1010E's or IRF3205's top and bottom in different versions (i.e., 10 total for the controller). The half bridge controller was an LT1158. PWM with soft start was generated by a TPIC2101. The motor was a 6V ford starter running at 12V, so it drew plenty of current, especially when stalling the motor briefly. System has worked well enough, so it hasn't needed any fixes
    But, if I were to start a design again, I would use more modern components and would seriously consider an industrial IGBT or MOSFET from e-bay, rather than so many paralleled TO-220 packages. John

    Edit: Just saw Sgt Wookie's post. I run mine at about 20KHz to limit the noise.
     
  4. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Here's a pretty new ST Microelectronics STRIPFET:
    http://www.st.com/stonline/products/literature/ds/13301/stv300nh02l.pdf
    Id=280A, Vds=24V, Rds(on)=0.8mOhms. Unreal.

    Since your voltage is so low and your app is basically on/off PWM switching, might be a good idea to stick with power MOSFETs instead of going to IGBT's. If you were over a couple hundred volts and/or were doing something linear, than IGBT's would be the way to go.
     
  5. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,671
    899
    @SgtWookie,

    I agree completely, particularly with a 0.8 mOhm RDS(on). The focus of my earlier suggestion was to get the big industrial packages with screw terminals at a reasonable price. It is really the package limitations of the TO-220 that are a pain.


    That ST mosfet is definitely something I want to try. It certainly qualifies as a more modern component, and I like that package. Thanks for the link.

    John

    Edit: I just compared it to the IRF1010E. Gate charge 120 nC vs. 220 for the 1010E. Price only $5.10 at DigiKey for single quantities. Wow.
     
    Last edited: Aug 13, 2008
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    John,
    I was stunned when I'd first found the IRF3703; 210A @2.8mOhms with a total gate charge of 209. Reading thru the datasheet again, package limitation is 75A; the 210A is theoretical if the leads wouldn't get melted off the TO220 case. ;) Downrating it to 65A isn't a bad idea.

    But that new STV300NH02L is just unreal. 57.4% of the gate charge, 28.7% of the Rds(ON), and look at that tiny footprint! 10mm x 15mm x 4mm would be more than it needs. Not even a special board layout! You'd need some mighty fat (and thick!) traces on the board to carry that much current, though - and I'd want to make sure that the gate driver was mighty beefy; capable of a couple amps. Nothing would blow that IC off the board faster than a slowly changing gate voltage.
     
  7. girish patel

    New Member

    Jan 3, 2012
    1
    0
    Hi guys
    I want to control dc shunt motor operating voltag 24 v and wattage is 2400. I want to use three mosfet IRF3205 parellal but i dont know how can i design mosfet driver myself and i want to use NE556 as a oscilator. please give me some idea
     
Loading...