Detector using Capacitance

Discussion in 'The Projects Forum' started by emilj726, Apr 1, 2014.

  1. emilj726

    Thread Starter Active Member

    Oct 1, 2010
    52
    2
    Hello,

    I am trying to come up with a trigger circuit that will tell me when a pair of stripping blades touch the conductor of a insulated wire. Can anyone suggest something?
     
  2. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    Is this an automatic wire stripper?
    Max.
     
  3. emilj726

    Thread Starter Active Member

    Oct 1, 2010
    52
    2
    Yes it is.
     
  4. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    I have only serviced Eubanks units, and usually if the machine is set up OK there is no need if the correct die is used etc?
    What make of machine?
    Max.
     
  5. emilj726

    Thread Starter Active Member

    Oct 1, 2010
    52
    2
    I understand what you are saying but I want to be able to detect when the stripping blades hits the conductor that way I can tell the controller to back off a little that way there is no cut on the wire.
     
  6. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    I think you have set yourself a challenge, presumably you can change the strip blades on the fly?
    I don't quite see doing it by capacitive method?
    It would make be easier if the core conductor on the drum was accessible but you have a rotating drum to contend with.
    Max.
     
  7. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,515
    1,246
    Yeah, but I think there would be a large enough change in capacitance or inductance that some kind of AC-excited bridge circuit would catch it. Fox and hound wire tracers work in the telecom industry without a return connection.

    One problem to solve would be getting a connection to the blade that does not include the metal mass of the entire die assembly. Through a rotary coupler. In the available space. OK, three problems...

    ak
     
  8. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    The part I would like to see is changing the strip depth on the fly?
    With a production machine you cannot stop every time the strip is wrong if this is a problem, it sounds like a mechanical deficiency?
    Normally the blade and dies are all at the machine ground potential.
    Max.
     
  9. emilj726

    Thread Starter Active Member

    Oct 1, 2010
    52
    2
    The blade holders are specifically coated so that they are isolated from the rest of the machine. I have a pcb that I have designed but I am having a hard time detecting because the wire diameter is so small and the blade is thin so there is not a significant enough of a capacitance change for the circuit to react.
     
  10. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,515
    1,246
    If the Core Sense terminal is grounded through a resistance, a capacitance, or both, it looks like the osc freq is lowered. If that is the entire sensor circuit, then all of that other digital goop seems like a lot of work for a hi-low freq discriminator. Clearly I'm missing something.

    ak
     
  11. emilj726

    Thread Starter Active Member

    Oct 1, 2010
    52
    2
    The digital counter is used as a indicator that the oscillator is working. It will count to a certain value and I use that as a Sencore PCB is installed to the machine. Than low signal goes to our main controller letting it know that this pcb is installed in the machine. All that other "digital goop" that you mention is needed for that Flip-Flop to turn on or off when needed. What I am asking if there is something I can add to c5 so that it detects better or easier.
     
  12. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,515
    1,246
    Probably not with the circuit you have. With only a single oscillator and a simple hi-low freq discriminator, there's not much else you can do to extract more information from the Core Sense input. Making C5 larger would mean it would take less of a disturbance to shift the freq, but it also would lower the noise immunity of the system.

    What you have now is an absolute detector. If the freq moves from here to there, you get an output. I would think this needs adjustment whenever dies or wire sizes change, and maybe not catch very small changes relative to the main carrier freq.

    A better (and more complicated) system would use ratios of frequencies rather than absolute values for frequencies. Convert the oscillator output to a voltage level that varies in proportion to the freq, like an FM demod. This could then be lowpass filtered to establish a reference level for the main detector that would vary with long-term (multi-second) drifts in the sensor environment. Sort of like an AGC or automatic level detector, the threshold would track the main signal as it wanders around, and detect only short-term disturbances.

    The long-term reference level is one input to an analog comparator. The other input is a dc level that represents the instantaneous (or short-term) frequency. The output indicates that the input is below the recently-established norm. This should give you greater sensitivity to small changes over a larger range of operating conditions.

    ak

    ps. I built this system in the early 80's to detect the beginning and end of spoken syllables. It worked without adjustment for all voices, young/old, deep/high, soft/loud, and took one quad opamp.
    .
     
  13. emilj726

    Thread Starter Active Member

    Oct 1, 2010
    52
    2
    Any circuit to reference from?
     
  14. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,124
    266
    Hmmm? Would the stripping blades not intermittently contact the conductors even when set at the perfect correct depth?

    Would it not have a lot of false triggers, leading to a too large setting of the blades?

    Just guessing here...
     
  15. emilj726

    Thread Starter Active Member

    Oct 1, 2010
    52
    2
    Analogkid what do you mean by ratios of frequencies?
    I originally thought about going the PLL route but I came up with this circuit to remove complexity.
     
Loading...