Plan A (Reprise)

Plan A (Reprise)

Plan A (Reprise): Returning to Variable Frequency Oscillators

I returned to variable frequency oscillators after moving away from using low pass filters.

There were two problems I wanted to overcome to make these oscillators work. Firstly, last time I used them, they seemed to stop working when I put water or my hand near the sensor. Secondly, I have no accurate way of measuring the frequency generated by the oscillator.

Problem 1: Stability

On reflection, it should not be hard to build an oscillator that works at almost a couple of hundred MHz. The reason for this is so simple, I was almost embarrassed to realise that I had missed it when I was looking at this a few months ago.

People have been building hobby FM transmitters for years. I made one when I was a kid. They just work. Using components that you can get from Jaycar.

A common FM transmitter that keeps coming up is the Colpitts Oscillator. I originally rejected the Colpitts Oscillator because it didn’t use a variable capacitor. But when I was looking at various ways of building an FM transmitter, I saw some designs with a variable capacitor used as a trimmer capacitor. What a difference one word can make.

This was easy to rig up on breadboard, and it worked straight away. Perhaps I was overthinking the previous design attempts.

Problem 2: Measurement

I had been given a great idea on how to measure the activity of the oscillator – use a radio receiver. The oscillator will emit some radio frequency radiation that I can detect using a radio receiver that can scan for and show these emissions. Easy.

If you want the slightly more technical details: It is a thing called an ‘SDR” or Software Defined Radio. Instead of turning a dial to tune into the required station, you can use your computer to do the tuning for you. Even better, you can use the computer to show you how far the frequency of an oscillator shifts if you happen to adjust the variable, or trimmer, capacitor.


I won’t go into all of the nitty gritty details here because this is about my attempts to build a soil moisture sensor. There are lots of sites that can help you set up software defined radios if you want to use them.

The important parts are on the left two-thirds of the screen. The white line about a quarter of the way from the top is what the receiver detected across this narrow range of frequencies. Spikes in the white line indicate a radio signal that is stronger than the background noise (also known as static).

The colourful part in the bottom to thirds is called a ‘waterfall’. It is a way of showing you what the receiver detected in the past few seconds. The top of the ‘waterfall’ is updated a few times per second and the old information is pushed down the screen until it falls off the …. waterfall. Think of it as a graph where frequency is on the horizontal axis and time is on the vertical axis. The colours indicate the strength of the signal. Blue is weak, yellow is intermediate, and red is strong. Where there is a peak in the white line, there will be a red streak in the waterfall.

Now for the fun part.

The frequency of the oscillator is inversely proportional to the values of the inductor and capacitors. If any one of them increases, the frequency decreases. As I moved towards and then touched the moisture sensor, the red line showing the strong signal moved to left. This meant that the frequency generated by the oscillator decreased.

In short, it worked. But I noticed something interesting.

Before I touched the sensor, there is a thin red line on the waterfall indicating that there is a strong, but stable frequency being generated. But when I touch and hold the sensor, the frequency seems to jump around very quickly.

I don’t know what this means.

During my first attempts to use an oscillator to detect the presence of moisture, the oscillators ‘died’ when the sensor was submerged in water. Could be possible that the oscillator didn’t actually die, but the signal became too erratic to detect?

Main menu