Magnet decoupling

Discussion in 'The Projects Forum' started by digielex, May 22, 2007.

  1. digielex

    Thread Starter Member

    May 21, 2007
    16
    0
    Hi,

    I have a project in which a magnet rotates under the influence of a rotating magnetic field (provided by a motor). I need to detect the point(speed) at which the rotating magnet decouples with the field(assume this happens!) and raise an alarm when this happens.

    My idea is to use a RF signal and make it pass through the the area where the magnet rotates, and receive it at the other end. I expect to find some disturbance at the receiving end when decoupling happens.

    Do you think this will work, are there devices available to detect this kind of scenario??

    Thnx
     
  2. cumesoftware

    Senior Member

    Apr 27, 2007
    1,330
    10
    I can give a sugestion. If the magnet decouples from the field, there should be a current increase in the coil. You could try to experiment this principle.
    Then you could make an RM network (a sort of voltage divider). A comparator , that can be made with an LM324/LM358 op amp, would compare the voltage between the node between the resistor and motor and the voltage from another voltage divider (RR network used as reference). If the voltage on "+" terminal of the op-amp exceeds the one present on the "-" terminal, the op-amp will produce a voltage relative to ground.

    The principle: current increase at the motor coil = motor has less resistance = voltage decrease between the motor teminals.

    But this will only work with DC motors.
     
  3. recca02

    Senior Member

    Apr 2, 2007
    1,211
    0
    just an idea:
    we use magnetic pickup to monitor speed,
    so as it is already a magnet i think small circular coil(with low turn so as to negligibly load the magnet) might pick up change in rotation speed.(its a bar magnet ,right?).

    just curious why use a motor to make a rotating magnetic fiel. windings like in armature (or armature arrangement itself) can produce magnetic field when ac is applied to it.
    another thing i dont think a decoupling might happen due to speed change in
    magnetic field ,it might happen due to load on the rotating part but again i may be wrong.

    perhaps a little more understanding of the project might help me come up with an idea (not sure about rf though might just work ,dunno)

    cumesoftwares idea is a bit better.
     
  4. digielex

    Thread Starter Member

    May 21, 2007
    16
    0
    Sure, thnx.

    It's a DC Motor too. I'll try it out.
     
  5. digielex

    Thread Starter Member

    May 21, 2007
    16
    0
    I missed a point there, I have a strong magnet (about 1 Tesla) mounted on the motor. That's how the field comes about. Gosh! I can't believe I missed that, but I guess you got the point.

    Yes, you are right. To make it clearer, I am dealing with a magnetic stirrer that contains a stir bar, and a magnet below it provides the necessary field for stirring action. At some speeds (I managed to calculate those mathematically) the stir bar just decouples. As you can see, I do not have access to the stir bar for measurements (thats why I can't get current values etc). I must use some kind of wave to detect changes to fields.

    Hope this helps.
     
  6. cumesoftware

    Senior Member

    Apr 27, 2007
    1,330
    10
    One more thing. Are you refering to a chemical laboratory stirrer with those bar magnets attached magnetically to a magnet disk, being the last one driven by the motor?

    If that is the case, there will be no direct interaction between the motor coil and the magnet. I think in that case, the motor accelerates if the magnet decouples from it, so the current might decrease a little.
     
  7. digielex

    Thread Starter Member

    May 21, 2007
    16
    0
    Yes, it is a lab stirrer! You are absolutely right, I do expect that motor current will decrease due to decoupling.

    I am looking at modifying my motor controller to detect this abnormality and make some kind of speed adjustment.

    How does that sound?
     
  8. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    You might find that a couple of reflective opto couplers will work. When the signal frequency from the rotating magnet drops below that of the motor shaft, then that shows the decoupling. Use several shiny and dark strips to give more than one pulse per revolution.
     
  9. digielex

    Thread Starter Member

    May 21, 2007
    16
    0
    I looked at the optocouplers you mentioned, however, the operating temperature for the experiments for the stirrer are extremely high.

    Is there any other way this can be done?
     
  10. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    Use a source and sensor outside the beaker. Mouser carries light to voltage converters by TAOS. They respond to light from medium IR to UV. Shine a source in the plane of the stirring magnet so it blocks the light twice per revolution. Use a flag in the motor shaft to do the same. A blue LED would give a big signal.

    Mouser part #856-TSL12S @ $1.26 would probably work. Ambient light wouldn't be a problem with a little shielding.
     
  11. cumesoftware

    Senior Member

    Apr 27, 2007
    1,330
    10
    I think in that case it should really decrease (but I'm using only my intuition), since the magnet will no more cause a delay on motor speed. When the magnet decouples, it will cause acceleration in some half cycles and deceleration in the other half cycles, but I think that will correspond to a motor rotating without a magnet.

    Also, you could use some sort of magnetic pickup or detector (like a magnetic tape reading head). So, if the magnet stops its rotation, the magnetic field would stop oscillating and no more current would be induced. But you should do this in a way that the pickup will not interfere (atract) with the magnet. Try using an air core inductor (with most turns as possible) around the magnet and container (toroidal shape recommended).
     
  12. digielex

    Thread Starter Member

    May 21, 2007
    16
    0
    Thanx

    I will explore this option further..
     
Loading...