Do all Kundo electric clocks have parasitic oscillations?

Discussion in 'Electric Horology' started by praezis, Mar 29, 2015.

  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  1. praezis

    praezis Registered User

    Feb 11, 2008
    413
    15
    18
    Germany
    Country Flag:
    Region Flag:
    This question came up in the German DGC clock forum.
    link1 link2

    One poster showed his (less appropriate) measurements on his Kundo (type with transistor in the coil).
    Visible were wild oscillations on the transitions between 'transistor closed' and 'open'.

    kundo-1.gif
    Picture from an audio file, shows the wild oscillations only.

    The poster claimed that all Kundos will show this behavior, due to poor electronics at that time.
    My opinion is, his clock behaves faulty.

    What experience do other Kundo owners have?
    I am aware that some electronic knowledge and measurements are needed for answering my question.

    Regards,
    Frank
     
  2. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    63
    0
    Calif. USA
    I've seen this as well.
    From a design point, there are a number of ways to solve
    this type of things. One is to dampen it out. That
    causes more power loss. The other is to reduce the
    rate of response of the circuit. I've see both on other
    transistor driven clocks.
    Another is to add additional hysteresis.
    It looks bad but doesn't seem to effect the operation
    of the clock.
    Tinker Dwight
     
  3. eskmill

    eskmill Registered User
    NAWCC Fellow NAWCC Member

    Aug 24, 2000
    7,135
    27
    0
    Region Flag:
    I agree with the posters that the simple design of the transistor switched KundO battery clocks will show these "parasitic" high frequency oscillations during the motion of the pendulum depending on the velocity of the pendulum.

    The high frequency oscillations are most pronounced and continuous when the pendulum is stationary and without motion or vibration. The physical layout of the electronic components, variations in component characteristics and mechanical disturbances to the magneto-strictive effects of the solenoid and permanent magnet are all variables that contribute to the effect noted.

    It would be interesting at a "wizardly" level to compare the effect between the single and dual transistor drive circuits. Altering the circuit to suppress or eliminate the unwanted oscillations would likely cause pendulum motion to be severely reduced in my opinion.
     
  4. MartinM

    MartinM Registered User

    Jun 24, 2011
    2,961
    79
    48
    Male
    Medical Insurance Systems Analyst
    El Dorado, CA
    Country Flag:
    Region Flag:
    I may not be seeing the problem, correctly, here; But, on the larger scale of transitionong the impulse(s) to a magnetic field and having that field interact on the pendulum, the timing is such that it really has little effect on the physical system. Would adding a capacitor smooth the signal? Or would it just throw off the timing and not provide the impulse at the optimum time?
     
  5. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    63
    0
    Calif. USA
    A question of where would you put the capacitor.
    What you want is hysteresis. That is hard to do with
    a single transistor. The polarity of the signals are wrong
    if the coil comes off the collector.
    All you have is the transformer action and that is what causes
    the oscillation in the first place.
    You might be able to do it with the coil on the emitter side.
    Tinker Dwight
     
  6. eskmill

    eskmill Registered User
    NAWCC Fellow NAWCC Member

    Aug 24, 2000
    7,135
    27
    0
    Region Flag:
    MartinM asks if adding a capacitor smooth the signal?

    Yes. The unwanted oscillations are the result of coupling the sense and drive windings which share a common space on the solenoid spool. Some trials could be made on the two-transistor versions that have the printed circuit external to the solenoid.

    (earlier versions with the single transistor and damping resistor assembled with the coil spool are not ameniable to dis-assembly and re-assembly)

    I agree with Tinker Dwight about the spurious oscillation current is miniscule when compared to the sense or feedback current induced in the base circuit.
     
  7. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    63
    0
    Calif. USA
    Here is my thought.
    Put the drive coil on the emitter.
    Take the bottom end of the sense coil
    and have a resistor to the bottom of the drive coil.
    Add a capacitor to the junction of the emitter
    and drive coil. The other end of the capacitor to
    the junction of the sense coil and the resistor.
    The idea is that when the sense coil starts to turn
    on the transistor, you get positive feedback from
    both the transformer action of the coils and the
    turned on drive coil.
    As the magnet swings more, the capacitor discharges
    through both the base and the resistor.
    When the magnet is no longer pushing the sense enough
    to maintain forward bias on the drive coil, the polarity
    of current across the capacitor reverses, causing the
    transistor to hard turn off.
    Time constant of the capacitor and resistor should be just
    long enough to get past the oscillation time.
    Tinker Dwight
     
  8. praezis

    praezis Registered User

    Feb 11, 2008
    413
    15
    18
    Germany
    Country Flag:
    Region Flag:
    Thank you very much for all your comments!
    To be honest, I hoped that someone said, his clock does not show such oscillations. I really cannot imagine that Kundo did not work on that issue and supplied them like that.

    There is no hysteresis in the circuit, but it does have positive feedback:
    circuit

    Increasing collector current will increase the voltage in the sensor coil.
    Shouldn't then the duration of oscillations at least be shorter than these 60ms?

    Some time ago I tested the newer 2-transistor Kundo with circuitboard in the base. There was no wild oscillating, IMHO there is some hysteresis.

    Frank
     
  9. kdf

    kdf Registered User

    Aug 26, 2011
    238
    1
    18
    BScEE
    Beograd
    Country Flag:
    In the early days transistors were expencive, so producer wanted simplest working circuit, I guess. In this case some parasitic oscillations are inevitable, but if they do not affect movement operation and battery life, it shouldn't be a problem and I don't think that it is some fault.
     
  10. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    63
    0
    Calif. USA
    It might even make the clock more accurate than a
    more snappy turn on and off. A more snappy turn on
    would tend to be more related to the state of the
    battery.
    Tinker Dwight
     

Share This Page