Connect with us

motor surge current limiting or soft starts

Discussion in 'Electronic Design' started by Mark Harriss, Jan 2, 2007.

Scroll to continue with content
  1. Mark Harriss

    Mark Harriss Guest

    I have a problem integrating a microprocessor board with an existing
    TRIAC based motor speed controller. The speed controller outputs an
    AC waveform which is then bridge rectified and fed via a centre-off
    reversing switch to a 1/15th horsepower 180 VDC motor.

    The 2.5V logic micro board resets when the motor is switched on at full
    speed or is reversed at full speed. Various methods of shielding and
    tying inputs on the micro board to logic High or Low via resistors has
    given some improvements but does not entirely cure the crashes.

    Can anyone suggest a startup spike current limiting solution or some
    motor speed control circuit with a soft startup.

    Mark Harriss
  2. I had this on a PIC, starting/reversing a 3-phase
    10KW AC motor. As you have experienced, tying
    inputs to a logic level improved things but was
    not a complete cure. It did give a hint though
    that lower resistance input tie-downs could be
    the answer.

    So the desperation software sequence ran like this.

    1. Read all uP inputs.

    2. Set all inputs to outputs, at the same logic level.

    3. Do the switching of the motor reversing contactors.

    4. Pause for the nasties to settle, (200mS in my case).

    5. Put the inputs back as inputs.
  3. Mark Harriss

    Mark Harriss Guest

    Tony Williams wrote:>

    Thanks for the idea Tony, I could possibly link the manual
    speed controller to the logic board to do this by triggering
    a outputs only mode.

    My particular application is for a digital angle readout
    using a MAXQ2000 micro which can directly drive a 4 1/2 digit
    seven segment LCD, IOW a capacitive load which I now have 100K
    resistors terminating.

    A snubber across the motor leads does seem to improve things
    slightly and the only experiments with axial inductors on the
    motor leads made things much worse most likely due to magnetic
    leakage, if I can get some toroidals I'll try it again or
    even a low value power resistor to see if it helps.

    Short term this may work, for the medium term I'll remove the
    reversing switch from the product and long term I'll try and
    design a slow start controller possibly PWM.
  4. default

    default Guest

    Inrush limiting thermistors - the kind used for current limiting on
    capacitor input power supplies (TV sets and the like).

    Zero crossing solid state relays may also work for light inductive

    DC motors with brushes may also cause EMI that is worse while
    starting. A filter on the motor input will fix that.
  5. john jardine

    john jardine Guest

    For quick fix I'd be inclined to mod' the +/0/- switch, causing it to switch
    an extra relay with a power resistor across it.
    Something like a 330ohm 10W resistor in line would restrict peak current to
    1/2 amp until the relay contacts closed, putting full voltage and current
    back onto the motor. Standard relay operating time of 10-20ms is probably
    enough for that small a motor. Extra delay by adding a cap feeding the new
    relay operating coil.
    If the reset always occurred at switch time I'd have fitted big snubber caps
    across the changeover relay coils (but it doesn't).
  6. Joerg

    Joerg Guest

    Hard to say without seeing the setup plus schematics but a software fix
    to patch up the effect should IMHO be the last resort. Have you tried
    running the motor leads through a common mode choke? How is the ground
    between the two boards? If it's just a wire and not a solid plane that
    could also spell trouble.
  7. KB have these off the shelf, however rectification is the first step. They
    work well.
  8. The problem could be anything, voltage spikes, current spikes, magnetic
    field spikes, switch spark EMI, and probably more.

    I'd start with a software fix-- limiting the rate of change of voltage
    and current across the motor.
    I know it's real cool to be able to "instantaneously" change the
    voltage or current directinons, but you have to pay the price of such
    abrupt changes-- lots of spikes. That may help a lot, but won't do a
    thing for human-initiated switch flipping-- that will spike and spark
    like crazy.

    Next I'd try a big snubber across the motor leads-- say 3 ohms and 1uF
    for a start. And maybe a 220 volt varistor to clip the peaks. And
    similarly across the switch contacts.
  9. Mark Harriss

    Mark Harriss Guest

    I believe it's based on a very old KB unit no longer made, Homer.
  10. Mark Harriss

    Mark Harriss Guest

    I thought they would be ideal except for the 60 second cool down time
    which may be a problem.

    The particular TRIAC based speed controller always switches some time
    after zero crossing which would prevent this solution.
    The motor is about two feet away, it always seems to crash at start up
    or reversing while the motor is spinning which seems to indicate a
    current related magnetic surge.

    I've just realized about 2:00 am this morning that there's a mains
    side inductor that is axial/ solenoid in construction and is aimed at
    the micro board, I'll replace it with a toroidal and see how it affects

    I tried a motor side filter based on these same inductors which made
    things worse.
  11. Mark Harriss

    Mark Harriss Guest

    Joerg wrote:>

    Hi Joerg, I've haven't tried tried a common mode filter yet, a snubber
    did help though. The logic board was placed in a wire earthed PCB copper
    foil box with no success and then inside a 12" length of steel 3"x3"
    pipe with no effect either which did surprise me, after that I stacked
    about 6 Kg (12lbs) of transformer steel around the micro board to
    no effect.

    There's no doubt the speed controller board is an ancient and very
    noisy design, basically a rectified output DIAC/TRIAC dimmer circuit.

    There is
    a solenoid type inductor on the active mains wire which would be
    magnetically leaky that I'll substitute with a toroidal inductor to try
    and reduce magnetic leakage. I have the mains and motor wiring tightly
    twisted in an attempt to cancel the magnetic fields of the internal
    wiring but this does not affect the current in the copper tracks on the
    speed controller board.

    I have a tightly confined box to place the micro and speed board in
    so they are close to each other and overlap slightly, unfortunately the
    client did not listen to any advice about using a cast aluminium box
    with separate compartments or about making it bigger to have some

    Mark Harriss
  12. Mark Harriss

    Mark Harriss Guest

    The current design has two entirely independent circuits in the same
    box. A future design would have a micro reading a pot and outputting PWM
    to a big FET.
    I'm discussing with the client an interim fix of removing the
    reversing switch as it's really not at all necessary and earlier models
    did fine for 40 years without one, it's more of a "sales feature", this
    would eliminate the problem altogether as they then turn it on and vary
    speed with a switched pot, slowly starting from a halt.

    That may help a lot, but won't do a
    That's exactly what the problem is caused by: ignoring the instructions
    and reversing at speed or setting full speed and turning on via the
    reversing switch.
  13. default

    default Guest

    The cool down is much less than 1 minute for the size you'd use - more
    like five/ten seconds. If it is the reversing condition that requires
    a quick cool down, two could be used one for each direction - soft
    start is desirable and if not thermistors could also be incorporated
    in the triac control - but that would take more understanding and

    I had this robot - small lab Cartesian robot on a large table. All
    the gizmos on the table would reset the computer when switched on - we
    solved the problem by installing one commercial "brick wall" filter on
    the supply to the ancillary machines - just a super filter using many
    stages of differential and common mode LC filters.

    One of the homoginizers would still occasionally do a reset when it
    switched - on the far end of the table with long leads passing by some
    of the robot's feedback leads. It would happen once in 200 times and
    wasn't considered severe enough to bother with. Watchdog timer would
    fix things and we'd only lose an hour of time when it happened - late
    at night when no one was around.

    The homoginizer also had another fluke that was more vexing - the
    still liquid in the vessel would fly out when the motor started -
    ramping it up fixed the problem. I went to a radio supply place in
    town and bought ten inrush thermistors - six in series made the motor
    start slower and fixed the liquid splash problem and also fixed the
    rare computer reset problem.
    It would be a less than ideal solution and might not fix your problem
    - but one never knows. Do one?
    It is foolish, as a rule, with most motors except steppers and some
    synchronous motors to reverse direction on a motor that is already
    turning. That appears as a short to the supply - and draws heavy
    current for a time. In automation applications one almost always lets
    the motor coast to a stop before reversing (or electronic braking is
    used to quickly stop it) - big motors complain by popping fuses if you
    reverse them without a stop first.

    Slow Syn made a synchronous motor that turned 600 rpm at 60 cycles -
    that particular motor could reverse direction all day long without
    hurting it - but it needed a variable frequency drive to change the
    speed or a variable sheave pulley type speed adjuster.
    You need an LC network to cut noise - a single inductor is not usually
    the answer especially with motor brushes and phase controllers

    Most of the crap (EMI/RFI) that hits processors is common mode - so a
    common mode choke with some low inductance capacitors on both sides to
    ground and across the line are called for.

    Long leads to the motor are an antenna that can radiate EMI - the
    filter should be as close to the motor as possible - and another as
    close to the controller as possible if the leads are long between
    motor and controller - triacs tend to be noisy (AM radio near it will
    tell you if it radiates)

    Don't ignore snubbers. Snubbers are just a cap and resistor that go
    across switches (motor brushes) to snub the spikes. As close to the
    source of noise as possible - killing noise at the source is 10 X more
    effective than trying to eliminate it at the computer.

    When the level of voltage changes suddenly (like a brush leaving one
    motor commutator segment and picking up the next on DC motors) the
    capacitor absorbs the spike (which is usually very fast - MHZ range).

    A resistor is used in series with the cap, to lower the Q of the
    circuit so you don't inadvertently make a tuned circuit with some
    inherent inductance in the wiring. Resistor & cap is placed across
    the noise source close to the source. Point one microfarad, and 100
    ohms is a good place to start. Voltage rating of the non-polar cap
    should be two or three times the supply voltage (potted AC rated caps
    are most desirable)

    You may also need a snubber across the relay or switch that switches
    current to the speed controller and/or reversing switch and motor.
    Snubbers are cheap and often worth their weight in gold.

    Don't overlook a power supply glitch too. Some supplies allow the
    spikes through to the processor - scope out the digital supply when
    the motor is changing direction.
  14. Joerg

    Joerg Guest

    Well, as a start I'd try common mode filtering. At least for the motor.
    For the rest of the circuit it's hard to tell without schematics and
    some detailed photos.
  15. Jamie

    Jamie Guest

    those drive boards are not designed for proper isolation on the common
    lines for your control logic.
    You're getting an in imbalance on the common line which is
    causing a momentary short long enough to drive power on your uC section.
    You need input isolation along with isolation on your uC board.
    use a xformer supply to operate your uC board and put isolation coupler
    on the board that will give you a floating common that you can use for
    the drive signal.
    You can also get your self one of those ready made units for that job.
    If you have one of those small low current DC clamp on's , you can
    test the common line from the drive to your uC board and i am sure
    you'll see a current spike there.
  16. Mark Harriss

    Mark Harriss Guest

    Hi Jamie,
    The speed and logic boards are not connected in any way except
    the common mains supply, one of my early experiments was to run the
    logic off a battery supply and it was still affected by the sudden
    current draw of the motor turning on and off.
  17. Mark Harriss

    Mark Harriss Guest

    Thankyou to everyone for the technical advice offered about my
    particular motor problem. At this point I'll remove the reversing
    switch which is not really necessary and work on PWM speed control
    for later models with a soft start circuit.

    So far I have tried:

    1 Magnetic shielding of the computer using steel (no change)

    2 Electrostatic shielding of the computer using an earthed copper
    box (no change)

    3 Filtering of the computer power supply, including
    using battery power. (no change).

    4 Terminating all tracks on the computer board with resistors
    (about 60% improvement).

    5 Twisting the AC and motor wires to cancel magnetic fields
    (no change).

    6 Capacitor across motor leads (no change) but speed
    controller did not work properly.

    7 Various snubber circuits (slight improvement).

    8 Inductors to motor leads to filter current surge (made
    things worse, possibly due to magnetic leakage).

    9 Common mode choke across motor leads with and without
    snubber, with and without varistor (slight improvement).
  18. Switching a motor from full forward to full reverse
    used to be known as plugging afair, and is a bit
    brutal, both for the motor and for the contactor.
    A quick soft start possibility is to have a series
    resistor that limits the stall (short circuit)
    current to about 2x or 3x full load current. Say
    to about 1A for your 180Vdc 1/15th hp motor. Use
    the now redundant reversing contactor to short
    out the resistor after a small delay.

    Last minute thought: Did you ever try switching
    the reversing contactor without the motor connected?
    Just in case the contactor was the cause of the
  19. Mark Harriss

    Mark Harriss Guest

    Hi Tony,
    It'll be even worse in the completed units as it has a steel
    flywheel to store energy.
    The reversing contactor is a DPDT center off switch
    I did try it without a motor connected but it had no
    effect. It would be nice to have a soft start integrated
    with the TRIAC circuit but I haven't worked out how to
    do this.

    I'll experiment with a micro reading a POT so that
    the midrange area is off and either side of that is
    forward or reverse. The $2.50 saved on the reversing
    switch should pay for an 8 pin Zilog micro to handle
  20. Holy crap. That energy has to go somewhere. Can you switch in series
    resistance to absorb it?

Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day