Viewing 2 reply threads
  • Author
    Posts
    • #10834
      LectronFan
      Moderator

        Hi all,

        Remember last project where we described a sawtooth oscillator ?
        In the last experiment, we noticed that the slope of the sawtooth wasn’t lineair at all.
        This is a natural phenomenon when we charge a capacitor.

        We have explained this earlier, look here : Capacitor tester

        So, to overcome this, we need to charge the capacitor with a constant current.

        A method to use is a Widlar constant current source.

        We will use 2 Si PNP transistors T3 & T4.
        T4 has its Base and Collector tied together. So the current flowing through the Base and Emitter will be constant.

        This current is mirrored to T3 (Bases are tied together), so T3 will charge the capacitor with a constant current.
        The rate of charging current is determined by R5 (100 K).

        As Michael mentioned, this circuit is build with the modern Silicium transistors and has a negative ground.
        The legacy transistors with their build in resistors can’t be used.

        Here’s a diagram and setup.
        When you look at the Oscilloscope, you’ll notice a perfect lineair slope.

        Many greetings

        Sawtooth Widlar

        Widlar setup

        • This topic was modified 3 years, 3 months ago by LectronFan.
      • #10836
        Michael
        Keymaster

          Hi Frank,

          Thank you for your latest learning lesson.

          Why is the linear slope waveform so important if the oscillator is not used for precision purposes like the music instrument amplifier application?

          Hope that you have a great weekend!

          Best,
          Michael

        • #10973
          LectronFan
          Moderator

            Hi Michael,

            The linearity can be quite important when you want to achieve a pure sawtooth. Also when you want to use it as a measuring tool for OpAmps.

            In the next experiment, I will propose an simpler circuit with the same results.

            Many greetings

             

        Viewing 2 reply threads
        • You must be logged in to reply to this topic.