I want to build a soil moisture measurement probe based on this idea:

enter image description here

An oscillator feeds a transmission line burried in a soil via a termination resistor R1, XOR gate acts as a phase shift detector - it will be high untill a reflection comes back from the transmission line. More water in soil - higher the dielectric constant, longer it takes for the reflection to come back, higher the voltage after R2C1 filter. I'm going to implement the transmission line as a trace on a PCB, i'm thinking about frequency of 80MHz and transmission line length of ~40cm.

My question is will an oscillator like CB3-3C-80M0000 be able to drive the transmission line or do I need some buffering? If so, what buffer would be suitable for speeds like these?

  • 1
    \$\begingroup\$ To be honest, I don't know. But I assume you know about the oscillator's load capacitance. In which case, I'm not sure how you'd match the transmission line (especially when the purpose is to change the characteristics) and sensor capacitance to that of the oscillator. My best advice would be to buffer it for safety. But then maybe a neat alternative sensing trick would be to measure the oscillator frequency based on water level. of course I have no idea what else affects the oscillator. \$\endgroup\$
    – lm317
    Jan 25, 2013 at 1:49
  • \$\begingroup\$ 40cm at 80MHz isn't really a transmission line; it's not long compared to the wavelength. If you want a sensor that size, you'd be better off just sensing the total capacitance at a lower (even audio) frequency. \$\endgroup\$ Mar 24, 2018 at 0:49
  • \$\begingroup\$ The "dielectric constant" (aka relative permittivity) isn't constant at all; it is very frequency dependent in soils. Switching from 80MHz to audio frequencies would definitely change permittivity measurements. At lower frequencies soil salinity/conductivity will also affect the permittivity and cause inaccuracies in water content estimations. \$\endgroup\$
    – remcycles
    Nov 15, 2019 at 19:57

1 Answer 1


datasheet says output load (CL) up to 30 pF up to 80 MHz, then up to 15 pF. That seems to me a very small capacitance easily achieved with those long traces (40 cm) and a high epsilon-r medium. If you are already using an x-or gate, why not using another one to buffer it for free?

  • \$\begingroup\$ yes, I've put a buffer and don't regret about that :) \$\endgroup\$
    – miceuz
    Feb 8, 2013 at 7:36

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.