At the end of Russell McMahon's answer to this question (High-stability oscillators (non-crystal)) there is this reference:

"Shows that Gouriet-Clapp, Seiler, and Vackar oscillators have equivalent frequency stability given equal resonator Q. They differ only in how much the amplitude of oscillation changes when they are tuned. The three circuits are useful over frequency ranges of 1.2, 1.8, and 2.5 to one, respectively." (J. K. Clapp, "Frequency Stable LC Oscillators," Proc. IRE, Aug 1954.)

Does this mean that if I design a Gouriet-Clapp oscillator to have a frequency of 100MHz, I can change the tuning capacitor to give me a frequency range of 100 to 120MHz?

Background: I want to take advantage of the dielectric properties of soil and use it as a tuning capacitor in a variable frequency oscillator running at >80MHz (I need frequencies >80MHz in order to minimise the effect of minerals in the soil). The only catch is that the dielectric constant of the soil will vary between 4 and 80 depending upon how wet the soil is, which means that the value of the soil capacitor will vary from 10pF (calculated) to 400pF.

  • \$\begingroup\$ Way, way over the top - if your capacitance changes that much you don't really need a high stability oscillator; you just need an oscillator like a conventional colpitts. \$\endgroup\$
    – Andy aka
    Sep 26, 2017 at 10:21
  • \$\begingroup\$ There are probably more issues in the way how your sensor works then int he oscillator electronics. I would build the oscillator and try it with one of the old 500pF trimmer caps and with you sensor and see how it behaves with both. \$\endgroup\$ Sep 26, 2017 at 14:08
  • \$\begingroup\$ @Andyaka How do you make sure that such a large change in capacitance doesn't upset the feedback to the LC tank circuit (because it will change either C1 or C2)? Could I place the soil sensor in parallel with either one of the tank circuit capacitors? \$\endgroup\$ Sep 26, 2017 at 21:30
  • \$\begingroup\$ Or could I place the sensor in parallel with both C1 and C2? \$\endgroup\$ Sep 26, 2017 at 22:47

2 Answers 2


The maximum frequency is ultimately limited by the transistor you use. Or to be more precisely its frequency dependent amplification. When increasing the frequency the amplification will drop. At some point it will get so low that the transistor cannot compensate the losses anymore and thus not sustain the oscillation.

Another limit is the transistor having a frequency dependent phase shift. It is possible that the phase shift get so large that the Barkhausen condition cannot fulfilled anymore and thus oscillation becomes impossible. But this is actually a limit that can be worked around, by adjusting the impedance matching to compensate for the phase shift.

Todays high frequency transistor allow to build oscillators well in the GHz range. So, unless you select a slow transistor, this will not be the limiting factor in your circuit.


Consider this topology


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ Is there a name for this topology? \$\endgroup\$ Sep 27, 2017 at 23:53

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.