Sensing a Current Spike

Discussion in 'The Projects Forum' started by roro36, Sep 8, 2011.

  1. roro36

    Thread Starter Member

    Sep 7, 2010
    52
    0
    I have a project that requires the sensing, not measuring of a current spike that must not "connect", for lack of better word to the line. So I thought to use a simple current transformer and either have a latching system or interrupt a micro to display that a current spike was detected? Does this sound correct or am I missing something?
     
  2. #12

    Expert

    Nov 30, 2010
    16,346
    6,831
    It sounds good if the current is in a range that is above the noise levels in the current detecting circuit. Is a "spike" 100 amps or 100 nanoamps?

    Better questions get better answers.
     
  3. roro36

    Thread Starter Member

    Sep 7, 2010
    52
    0
    My bad, that would help. The current will be in the 20mA to 30mA range. Lasting for not more than a few miliseconds?
     
  4. #12

    Expert

    Nov 30, 2010
    16,346
    6,831
    I've never done this but, a hall effect sensor might be the way to do that.
     
  5. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Is the 20 to 30mA spike superimposed on another "steady-state" current? Or it just a case of the current spike being on or off?
     
  6. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    If it is a case of current being either on or off with no background value then you could feed the current carrying conductor through a powdered iron or ferrite ring of suitable dimensions. Wind an isolated secondary of say 100 turns to generate an emf at the current spike. Then use a fast comparator to detect the secondary emf. The only proviso is the unknown rate of change of the current spike. For such a simple idea to work would probably require a rate of change in current of the order of perhaps 1A/usec.
     
    Last edited: Sep 8, 2011
  7. wmodavis

    Well-Known Member

    Oct 23, 2010
    737
    150
    A key factor in monitoring transient events is to be sure the frequency response i.e. bandwidth of the transient monitoring portion of the circuit is appropriate to see the spike. My guess (because I dont have response data in front of me) is that a hall effect sensor may have a faster response time than a current transformer. At least use a high frequency current transformer if you choose to go that way.
     
Loading...