Designing a Colpitts Oscillator

A local oscillator for use in 40 meter amateur radio bands, as well as the 7.2-7.3 MHz shortwave broadcast band is designed and built. The oscillator uses two variable capacitors for coarse and fine frequency adjustment. A common-base BJT amplifer is used to provide amplification. Feedback is provided through a capacitive voltage divider, making it a Colpitts oscillator. Frequency range is chosen based on a 455 kHz intermediate frequency downmix.

Requirements

• Frequency range of 6.545 to 6.945 MHz
• Easily tunable to 1 kHz steps
• Should use 10 pF to 160 pF variable capacitor
• Should use T68-2 toroids
• Powered with 12 Volts

Design

Tank Circuit

Using an inductance of 1 μH and the minimum and maximum frequencies, the required capacitance can be found using the equation below. The minimum, median, and maximum frequencies and capacitances are shown in the table below.

$$ C = \frac{1}{4 π^2 f^2 L} $$

Capacitors

Essentially, a base capacitance of 525 pF is needed in addition to a variable capacitance of 0 to 66 pF.

$$ C = C_1 + C_2 + C_{ce} + (C_3 || C_4) $$ $$ C = C_1 + 8 \text{ pF} + 4 \text{ pF} + 270 \text{ pF} $$ $$ C_1 = 243 \text{ pF} $$

$$ C_1 \text{ is the parallel capacitance of the tank circuit}$$ $$ C_2 \text{ is the output capacitance of the oscillator and thus the input capacitance of Q1 } C_{bc}$$ $$ C_{ce} \text{ is the output capacitance of Q2 }$$ $$ (C_3 || C_4) \text{ is the capacitance of the feedback voltage divider created by C7 and C8 }$$

This base capacitance is provided by C5 and C6.

The variable capacitors must swing from 0 pF to 66 pF. The C1/C2 series provides a capacitance swing from 5 pF to 9.41 pF. and the C3/C4 series provides a swing from 9.09 pF to 61.54. Collectively, this is an capacitance adjustment range of 14.1 pF to 80.0 pF. This should give enough range to adjust within target frequency ranges. The minimum capacitance adds to the base capacitance, but that will be trimmed out with the capacitor or by adjusting the inductor windings.

Inductor

The 1 uH inductor was wound on a T68-2 toroid. The required number of windings was found using a calculator. In this case, 13 turns were used.

Building and Testing

The oscillator was built on a single-sided copper-clad FR4 board. I used a desktop CNC mill to cut "lands" into the board. The board is powered by clipping leads to the power and ground nodes. The board has a U.FL connector for LO output.

I terminated the output into a 50 Ohm resistor and used an 10x oscilloscope probe to see the waveform. The output is really ugly, but the main LO frequency is still the most prevalent. I am thinking that a bandpass filter will clean the signal up enough for the circuit to be used in a real receiver. While connected to the oscilloscope I was able to verify that the power output of the oscillator was not too powerful for my spectrum analyzer. The peak-to-peak voltage is roughly 400 mV, which corresponds to a power output of -4 dBm.

The circuit was connected to my TinySA signal analyzer and a picture of the output is shown below. The harmonic distortion is ridiculously high. In my defense, the second harmonic is about the same as a Baofeng.

What's Next?

I think this circuit could use some improvements. Other oscillators that I have built in the past were not nearly as distorted (some were even almost sinusoidal). I need to better analyze this circuit and determine what is causing the distortion. I suspect that it is poor biasing in the first stage.

However, my larger goal is to build a receiver. I do not want to spend too much time trying to build the perfect oscillator. I feel that this oscillator may be useable with a decent bandpass filter, so that is the next thing that I will build.