Problems with motor drive circuit.

Discussion in 'General Electronics Chat' started by rasputin666, Mar 18, 2009.

  1. rasputin666

    Thread Starter New Member

    Aug 22, 2007
    8
    0
    I am trying to drive a 36V 200W electric motor for an electric bicycle, the original controller having bit the dust.

    This is the circuit I have built:
    [​IMG]

    All tests fine, as far as I can tell, with only a multimeter to work with. Using a 100 ohm test load, output ramps nicely from 0 to V-batt as I adjust the throttle. However, when I connect the motor I get essentially nothing.

    I hear a very faint squeal (which I presume is the PWM signal in the windings of the motor) and the wheel kicks back and forth slightly.

    The orriginal controller was a DKY3616, for which I can find no info on the web.

    Can anyone offer me any suggestions as to what I am doing wrong, and what I could do to fix it?
     
  2. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Are the MOSFETS getting hot with the motor in the circuit?

    Have you checked the motor runs OK from the fixed 36V supply - no switching and just directly connected?

    What type of freewheeling diode are you using?

    The 100 ohm test load is quite modest when compared to the 200W 36V motor operating at rated value - nominally 6.5 ohms.

    What is the motor stalled/starting current likely to be?
     
  3. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Also check the diode is OK ...:)
     
  4. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    I mean the big one in the high power path ....:rolleyes:
     
  5. rasputin666

    Thread Starter New Member

    Aug 22, 2007
    8
    0
    MOSFETS hot? No. Tho the first one did explode when I forgot the Back EMF diode. Along with the resistor and both the BCs. Whoops.

    Motor? Check. Ripped the bike out of my hands when I tried that stunt.

    freewheeling diode? The big bugger? Can't recall the part no, but it's dimensions are roughly 10mm dia x 10 mm long with hefty leads. Multimeter says it's OK.

    100 ohm test is all i had. And it did what it was supposed to do confirming good pseudo voltagte which varied properly with the throttle settings.

    Starting current? couldn't say, but the DC resistance of the windings is about 4 ohms.

    I don't know if the motor is brushed, or brushless, It has 2 leads and works fine with a direct feed.
     
  6. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Motor in. Then. You should be able to test the switching between 0 and 100% speed by applying a suitable voltage (0 or +5V?) at the 680Ω opto-isol input to signal ground. Unfortunately the motor will get a pretty hefty kick if the signal is applied for too long! Can you get the PICAX to send a low repetition rate (say once every 5 seconds) short duration pulse (say 1 second at high). You could try this manually with a battery - just need to be nifty with the on/off sequence!

    If the motor doesn't try to spin up with the enable high then there's something else wrong - maybe insufficient drive voltage, although there seems to be plenty available from the 9-12V source.

    Is the back EMF diode a fast receovery type? Presumably it worked with the previous driver - was that at 10kHz? :)
     
  7. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Hi Gundam001,

    I'm guessing the PWM is required to adjust the motor power - subject to the riders' requirements at the time. Higher PWM duty gives more power.

    A 47Ω resistor in the source ? For what purpose? There wouldn't be much voltage left for the motor.

    :)
     
  8. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    There is no 47 ohm resistor.
    Each Mosfet has its source connected to ground.

    Each 4.7 ohm resistor is in series with the gate of each Mosfet to keep them from oscillating at a very high frequency.
     
  9. hgmjr

    Moderator

    Jan 28, 2005
    9,030
    214
    It sounds like your opto-isolator may not be able to switch effectively at 10KHz.

    hgmjr
     
  10. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Hi hgmjr,
    Except that everything "worked" OK with a 100Ω replacing the motor as the load.

    Hi Audioguru,
    I stumped on this - I thought, for some now inexplicable reason, I was responding to a follow-up post by one Gundam001 (?) which I now can't see or find ....?? Old age I guess. Among several curious suggestions in this "phantom" post there was one that a 47Ω resistor could be inserted in the source. I was puzzled by the suggestion - that's all ...

    I'm still interested in the freewheeling diode - if it's just a power diode (??) the reverse recovery time may be too slow for 10kHz operation. Since there's been no follow-up from rasputin666 we may never know the outcome.
     
  11. ifixit

    Distinguished Member

    Nov 20, 2008
    638
    108
    You need to determine the inductance of the motor winding. Generally, the higher the inductance, the lower the PWM frequency will need to be. This allows current (torque) time to build up to get the motor moving.

    A good starting point is; the high time of a 50% PWM signal should equal 1 time constant of the motor L. Tweak from there.

    Shotgun approch... try 100Hz.

    Good Luck
    ifixit
     
  12. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Reduce the PWM frequency by a lot.

    Try 50Hz-100Hz, it may hum at low speeds, but it will work. avoid frequencies that your ears are most sensitive to, roughly over 500Hz and under 10kHz.

    Is there enough mass/flywheel involved that a 10-20mS pulse would be mostly absorbed from perspective of the rider?
     
  13. rasputin666

    Thread Starter New Member

    Aug 22, 2007
    8
    0
    My appologies, real life events have kept me from doing anything on this. I haven't even had a chance to try feeding a 5v signal to the opto yet. (Heading out now to try this.)

    As for adjusting the PWM frequency downwards, there's a problem, the PICAXE has a minimum PWM frequency of 3900 Hz. Unless I can find the POKE address which lets you overide this.
     
  14. rasputin666

    Thread Starter New Member

    Aug 22, 2007
    8
    0
    And that said John is that.

    Direct driving the input resistor of the opto did indeed start the motor, with one slight problem, it didn't stop. And about 20s later all the magic blue smoke came out of the MOSFETs.

    Since I've already spent more time on this project than I can afford, I guess the next step is to buy a replacement controler module and be done with it.

    Thankyou very much for your help people.
     
  15. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Sorry to hear that.

    Hope it all works out with the plans for the new controller.

    :(
     
Loading...