switching coil polarity

Discussion in 'General Electronics Chat' started by floomdoggle, Dec 24, 2010.

  1. floomdoggle

    Thread Starter Senior Member

    Sep 1, 2008
    217
    2
    Hi all,
    What I would like to do is change the polarity of a coil. I'd be using two opto interrupters to fire in either direction; north, off, south on a single coil. I'm looking to be pointed in the right direction. Thanks in advance.
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    If your response time needs are low, you might use a pair of SPDT relays.
    If your response time needs are fast, you'll need something like an H-bridge.

    You haven't said anything about your coils' current and voltage requirements, so that's all up in the air.
     
  3. floomdoggle

    Thread Starter Senior Member

    Sep 1, 2008
    217
    2
    12 volt, 1 amp. Will an H bridge be full power?
     
  4. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    There will always be some power loss, however slight.

    How fast do you want to reverse the polarity? More than a couple times a second? If so, you'll need an H-bridge, as relays won't be fast enough.
     
  5. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,771
    930
    Yes. except for the tiny bit of power dissipated in the transistor switches.

    Something else you might look into. If you wind your single coil. with a PAIR of wires. Ending up with one single mass with four ends. Ground both wires on one end and then drive the coils from the remaining two wires. Always send a positive voltage to one and always a negative to the other. The field reversal at the time when current through one coil stops will aid the other winding in the coil in building its field instead of opposing it.

    I can try to make a sketch of how to wind it, if the word picture doesn't develop
     
  6. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,004
    1,522
    Kermit2 - Is that whats known as a 'bifilar winding' ?
     
  7. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,771
    930
    Yes and no. It is bifilar in that it has two wires wound on together, but the idea of winding bifilar is to use it in anti-phase operation, - to cancel any magnetic field created. fold the wire in half but don't cut it. Wind it like a normal solenoid. When power is applied, one half of the winding makes a north pointing field and the other half makes a south pointing field and the fields cancel since current in a series circuit is the same. It is used to make low resistance power resistors with manganin(sp?) wire.
     
  8. jjohnson_vng

    New Member

    Apr 23, 2011
    7
    0
    Kermit I am not sure I follow you on how to wrap the coil.

    I was thinking you would start the wrap from opposite ends of the coil.

    In a bifilar you wrap your 2 wires at the same time (preferably twisted around each other). It would seem to me you would want to wrap one clockwise and one counter clockwise to get the correct polarity.

    Am I missing something?

    Also I was wondering would it be possible to use a 4 transistor circuit with a single wrap coil. 2 transistors to control each direction of flow.

    If X and Y are the wires at their ends then I would use a 2 transistor to put X to positive and Y to ground and the other 2 to put Y to positive and X to ground.

    I want to flip my coil 10K per minute.

    Thoughts would be much appreciated.
     
  9. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    Yes, I think you are missing something. You can use a double-wound coil using a bifilar approach, connecting the ends appropriately: one wire at each end is grounded, leaving two inputs giving opposite effects.

    This does mean that the full coil voltage will appear on two adjacent wires, which may be a disadvantage.

    The "H-bridge" referred to by SgtWookie is a four element driver - possibly the four transistor circuit you are contemplating is of this kind.

    Finally, 10k a minute? Do you mean 166.66... Hz? What is the coil inductance?
     
    Last edited: Apr 23, 2011
  10. jjohnson_vng

    New Member

    Apr 23, 2011
    7
    0
    Ok thats what I was thinking basically using a non twisted bifilar coil with both wires being the same gauge.

    Or

    A Single wire coil with an H-Bridge type circuit that would flip the polarity up to 10,000 per minute. I have not wrapped the coils yet so not sure the the ohms yet. I am still in the design phase of this project.

    What I am trying to accomplish is a high speed coil oscillator that will vibrate between 2 perm magnets. I would control the speed of the oscillation via a arduino or parallel port on my laptop.

    Wanting to use short bursts of energy to force the coil to move from one point to another point on a pendulum.

    Make sense?

    Thanks for the response...
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Well, 10,000 reversals per minute is 166.666.. reversals per second. You're going to have kind of a hard time getting the current to reverse directions quickly if your inductor is very large.
     
  12. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    Yes, 167Hz easily qualifies as an audio frequency. You are going to need a low mass coil if the coil itself is the moving element.

    In this case, I would forget about double windings so as to keep the mass down. A single low-mass winding along the lines of a loudspeaker voice coil would seem more appropriate.
     
  13. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,771
    930
    I get the feeling the OP needs a large mass, if he is to see any change in the pedulum swing period the mass must be a significant portion of the mass of the pendulum.
     
  14. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    I would think the force exerted by the coil divided by the total mass would be the figure of merit here.

    It might be interesting to think about how long said "pendulum" is to be. Since it is to be excited at 167Hz, we might assume that this would be its natural frequency. In this case, its period is 6ms. The period of a pendulum is given by T = 2π√(L/g),

    so we get T^2 = 4π^2 L/g , or L= g(T^2)/(4π^2) = 9.81*36E-6/39.48 = 8.9μm Can that be right?
     
  15. jjohnson_vng

    New Member

    Apr 23, 2011
    7
    0
    As for the pendulum I am hoping oscillate it about 20mm.

    As for the force to be honest I am not sure have to go back to my college physics books to review force equations LOL.

    I made single 300 turn 5mm air core 10mm W and 20mm Diameter using 23 AWG.

    The issue I had with this configuration was heat build up. To get a strong enough magnetic field I had to pump 12VDC at 3.7amps into the coil.

    Unfortunately this caused the coil to super heat and the result was not good.

    The idea of using and Audio frequency to control the timing is great.

    Think I need to go back to coil wrapping school :(.

    The key to getting the force up would be to balance the mass of the coil to the maximum amount that the magnet fields could oscillate at the desired frequency.

    Thoughts...
     
  16. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    I would think that a spring system might help, so that the system could be tuned to resonate at 166.66...Hz. That would let you use less drive power. Clearly a gravity pendulum won't cut the mustard.

    This begs the question though - are you only interested in 166Hz? If you want a broad frequency response, spring tuning won't be so much use and you will indeed require a lot of grunt to drive it.

    All of this would be easier to understand if you could explain what you want to achieve as an end product.
     
  17. jjohnson_vng

    New Member

    Apr 23, 2011
    7
    0
    No 166hz is not a specific target.

    I am going want to be able to adjust the frequency.

    As for the purpose I am simply going to try and drive a push rod from the end of the pendulum. The push rod will be connected to a small flywheel to turn the oscillation into a rotational force.


    So the plan was to pivot a coil between to neo magnets at a variety of rates.

    Going to use a Linux box to control a H-Bridge circuit to turn on and off power to the coils and reverse the polarity of the coils to increase the force.

    My biggest issue atm is that I keep frying my Pot. Using a 12V battery array to a Pot to the Coil to ground and as soon as I open up the Pot to allow any current flow it just goes puff. So I decided to replace the Pot with a set of resistors, but at 330ohms it burns and at 1k ohms I get no magnetic field out of the coil.

    If I take the resistors out of the circuit and simply run the current to the coil and use a DC power supply that limits the amps to 4 the coils super heat and melt.

    I know I am just missing something in my circuit, which is driving me crazy. Maybe I need 2 resistors. 1 between Bat and Coil and one between Bat and Ground.

    Thoughts
     
  18. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    I suspect that you may have a more fundamental problem. With an air core coil moving between some magnets (even quite strong ones), the flux density is not likely to be very big.

    Commercial motors make extensive use of magnetic materials so that a high flux density can be achieved. This allows substantial counter-emfs to develop at moderate motor speeds.

    Unless your motor is able to generate a useful counter-emf due to motion through the field, the coil current will be limited only by its resistance and its self-inductance, and with an air core the latter will not be large.
     
  19. jjohnson_vng

    New Member

    Apr 23, 2011
    7
    0
    So I obviously need to switch to Soft Iron Core Coils.
     
  20. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,771
    930
    Then we get to a pesky little problem with frequency. With increased frequency the magnetic domains in the iron cannot change direction fast enough to follow the AC signal. The electromagnet will not have time to become fully magnetized in either direction before the electrical signal has changed polarity again. At 166Hz this won't be a huge problem, but it will be noticeable
     
Loading...