Changing the type of oscillator

Changing the type of oscillator

Soil Moisture Sensor Project

Changing the type of oscillator

It is 2017, time to stop trying to build my own oscillator. It wasn’t too hard to build an oscillator that worked at a fixed frequency. But when hours becomes days of trying to get that frequency to change reliably a little thought begins to gnaw away: “It’s 2017 – there’s got to be a better way.” It’s time to start changing the type of oscillator I am using.

My aim was to build a simple and reliable, high frequency sine wave oscillator whose output frequency would change depending upon the capacitance of the soil. Two things I hadn’t counted on when I started this project have held me back.

Specialised knowledge

Firstly, there is a lot of specialised knowledge that goes into high frequency electronics. For example:

  • a Clapp-Gouriet oscillator needs you to select a transistor with a “transition frequency” that is high enough so that the “gain” of the transistor “provides enough feedback that is near enough to the correct phase to satisfy the Barkaushen criterion”,
  • this feedback is provided into a “capacitive voltage divider” that is part of a LC resonant tank,
  • the values of the inductor and capacitors will affect the final frequency according to the equation 1/(2*pi*√(L*C)),
  • the ratio between the values of the capacitors in the voltage divider will affect the feedback,
  • and I have no idea whether the inductor should be bigger than the capacitors or the other way around,
  • and perhaps a lot more.

Designing or adapting high frequency electronics appears to be a less suited to “Dick Smith’s Funway into Electronics”, and more suited to the final year of a Bachelor of Engineering. This means that oscillators are EITHER reliable or simple when you move into high frequencies. Not both (reliable means that they just work, simple means that I can build it using stuff I have in my electronics box.)

Large changes in capacitance cause large changes in frequency

Secondly, the capacitance of the soil changes by a very large amount when water is added to the soil. On paper it didn’t seem to be that much; the dielectric constant of soil ranges from 3-5 up to somewhere near 80. This means that the capactiance of the soil will be approximately 20 times greater when wet compared to dry. Now that’s a problem. To get an LC oscillator working at high frequencies, you need small capacitors and inductors. If one capacitor suddenly becomes at least 20 times larger, the frequency will change by a large amount.

How does the frequency change if you increase the capacitance?


If you have a LC tank circuit made of a 10nH inductor and 2 x 47pF capacitors, it’s resonant frequency will be 164MHz. Once you add a soil capacitor:
10pF soil capacitor = 156MHz
200pF soil capacitor = 92MHz
400pF soil capacitor = 71MHz

A large change in capacitance creates a very large range of frequencies for a single oscillator to cover. Recall from a previous post that oscillators can only cover a limited range before they become unstable.

It’s 2017. It’s time to change the type of oscillator that I am using. But more on that another day.

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

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

Go back to the main page.

Leave a Reply

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