Limitations of the Clapp-Gouriet Oscillator

Limitations of the Clapp-Gouriet Oscillator

Soil Moisture Sensor Project

The Limitations of the Clapp-Gouriet Oscillator

Drunk with success, I decided to test the limitations of the Clapp-Gouriet Oscillator. For a good reason of course. It may “sort of” work on the desk when the cat is not trying to push it off the table. But it will need to survive in the ground while being baked by the heat of the sun. So it’s time to see where it will stop working.

If you don’t want to read lots of boring writing, the quick version is this: It didn’t work.

Using this design

The first thing I wanted to test was whether or not I could use a smaller piece of circuit board as a sensor. This was to serve two purposes. Firstly, a smaller value capacitor would mean that the oscillator would run faster. Secondly, a smaller value capacitor would reduce the range over which the capacitance would vary when its environment changes from air to water. If you start with a capacitor that has a value of 10pF in air, it will have a capacitance of 800pF in water. But if you start with a 50pF capacitor (air), its capacitance will be 4000pF when it is in water. My thinking is that a smaller range will be easier to measure than a larger range.

The frequency increased when I used smaller sensors and decreased when I put the sensors in water. This meant that the overall idea worked well, and that it was time to build a proper prototype.

Testing on circuit board:

Testing the circuit on circuit board instead of breadboard revealed problems that the breadboard had masked. Firstly, the oscillator refused to work with the smaller sensors. It was not a case of the oscillations being very slow. There were no oscillations at all. It refused to work unless I placed a capacitor in parallel with the sensor board, increasing the capacitance. Secondly, there was very little difference in the output when the sensor was in air compared to the sensor being in water. To make matters even worse, the whole thing would stop working when I touched the sensor. As far as I can tell, I’m 70% water. The sensor circuit was so sensitive that it could detect my presence from approximately 1m away.

I initially thought that instead of building a moisture sensor I could patent a security sensor that detects people from 1m away. When the dollar signs fell from my eyes I remembered seeing this kind of thing somewhere… perhaps I had accidentally built one half of a Theremin.

Perhaps the circuit board has parasitic capacitance…

My first thought was that there could be some weird parasitic capacitance in the circuit board that I had designed. I had designed the board to have two layers. The top layer was where the components would be, the the bottom layer would act as the “ground” or “negative” side of the circuit. Was it possible that the two layers, being just over a millimetre apart, were acting as a capacitor?

In order to test this, I did two things. The first was to build another circuit with all components connected only to each other. The second was to disconnect all the components on the circuit board from the ground layer and connect then to a “ground” wire.

Components connected directly together
Components connected directly together

Neither of these options had any effect. The circuit still stopped working when I touched the sensor boards, and there was little difference between the sensor in air and the sensor in water.

What could this mean?

It is possible that the parasitic capacitance on the breadboard had actually kept the circuit working. And I had removed this parasitic capacitance when I moved the circuit from the breadboard to the circuit board.

The internet shed some light on my predicament. As it turns out, the Clapp-Gouriet oscillator is limited in how far you can vary the frequency before it becomes unstable. “Unstable” is probably an electricla engineer’s euphemism for “Stops working, warranty voided, kills someone”

Have a read of the answer by Russel to this question, particularly the last part that refers to the range of frequencies that can be generated. The Clapp oscillator can have a frequency range of 1.8, which probably means that if the base frequency is 100MHz, the highest frequency could be 180MHz. The problem is that the capacitance of the sensor board can vary from 10pF to 800pF (or more if the board is larger). Such a large change in capacitance will have a large effect on the LC resonant tank of the oscillator.

Brendan’s thoughts on designing this project (art vs science):

Dark Arts (*)———-—————————————————————– Science

escher stairs
Good luck

Go back to the main page.

Leave a Reply

Your email address will not be published. Required fields are marked *