2
\$\begingroup\$

I am trying to make 125 KHz oscillator based on CMOS inverter and feedback circuit consisting of Pi C-L-C section, as shown below. Since I need sine wave I will ommit the second inverter. schematic of the oscillator circuit

Before starting, I thought that this circuit will oscillate for every combination of L and Cs, but unfortunatelly it is not like that. So, my question is how to calculate L and C in order to make this work at 125 KHz?

The values shown in the figure are not relevant for 125 KHz frequency.

So, I decided to start from the begining and to solve somehow this circuit. In the first step I want to recognize the vital parts of the circuit, but I am not sure how to do that actually. Here we have:

  1. Amplifying element is inverter with negative transconductance.
  2. Feedback loop consists of an inductor.
  3. C2 is load capacitance
  4. For C1 i am not able to say what is it

Having this in mind, the type of feedback must be Shunt - Shunt, but the logic says it should be Series - Series.

My question is how to start with solving this circuit?

\$\endgroup\$
11
  • \$\begingroup\$ 125 kHz crystals are readily available, and will make design much easier. \$\endgroup\$ Commented Dec 1, 2014 at 15:07
  • \$\begingroup\$ Where do the shown component values come from? Do you know the working principle of the oscillator? \$\endgroup\$
    – LvW
    Commented Dec 1, 2014 at 15:12
  • \$\begingroup\$ That circuit will not work with 100nF on the output of a gate. \$\endgroup\$
    – Andy aka
    Commented Dec 1, 2014 at 15:43
  • \$\begingroup\$ It is not sufficient to have the -180deg at the desired frequency. The necessary loop gain (passive damping factor multiplied by the active gain) is deermined also by the L and C values. \$\endgroup\$
    – LvW
    Commented Dec 1, 2014 at 15:48
  • 1
    \$\begingroup\$ Yes it is solved. \$\endgroup\$
    – Nexy_sm
    Commented Mar 27, 2015 at 15:13

1 Answer 1

2
\$\begingroup\$

The design process is similar to a crystal oscillator, since the crystal will 'look' like an inductor in the steady state.

You treat the gate as a high-impedance input/high-impedance output amplifier- the output as a current source that is controlled by the input voltage, so it's characterized by the transconductance.

One requirement is that the transconductance multiplied by the product of the capacitor reactances must be > 1 at the desired frequency of oscillation to satisfy the Barkhausen criteria. Here is the typical transconductance of a 74LVU04 vs. supply voltage.

enter image description here

At 3V (just say) the transconductance is about 30 mA/V. So C < 0.22uF, and 0.1uF might be acceptable if you're using this particular chip at 3V or so, at 125kHz (or less).

You also want resonance at 125kHz, and the capacitors are effectively in series for that calculation, so 0.05uF, and with a 1mH inductor it should resonate at 22.5kHz.

\$\endgroup\$
3
  • \$\begingroup\$ Which parameter of the transistor will dictate max C and why? BTW, I will use like 10 V supply. \$\endgroup\$
    – Nexy_sm
    Commented Dec 1, 2014 at 16:12
  • \$\begingroup\$ As I said above, transconductance. \$\endgroup\$ Commented Dec 1, 2014 at 16:28
  • \$\begingroup\$ I see, but I don't understand why that comes from Barkhausen criteria? \$\endgroup\$
    – Nexy_sm
    Commented Dec 1, 2014 at 18:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.