RF oscilator 33MHz to RF detector through pcb coplanar capacitor Reflectometry sensor

[vitoa]

Hi, i'm developing some sensor based on RF principles (Reflectometry Sensor)
First all have a (0,7mm*10mm)water sealed test pcb with two 2mm*10cm cooper strips with 1mm distance between them ( coplanar capacitor ), changing its value with surround moisture (variable coplanar capacitor )

This pcb dual striped probe is to measure moisture inside tree trunk, so the ideia is to create a RF field sorrounding this plates to measure some volume inside tree trunk.


In order to build a simple,low power oscilator and rf detector circuit I used two IC for this porpouse
One is a precise 33MHz oscilator outputing 3000mV pp wave between ground and Vout
To save time on a peak detector circuit with skottkey diode I found some simple RF power detector IC LTC5507.
It will provide an output voltage proporcional to rf in, according datasheet -18dbm(0.08Vpp) to 14dbm(3.17Vpp) input wil have almost linear Vout from 200mV to 2000mV, perfect for microntroler adc.

One strip is connected to 33MHz@3Vpp oscilator, the other strip connects to rf detector ic.
This ideia is to do one strip work as antena emitter and the other strip anntena receptor, the higher the moisture inside tree trunk the higher refletance of signal get back to detector strip, increasing output voltage on rf detetor.

I need some help to know if this ideia is correct for my porpouses.
Need also to calculate strips
width dimensions to have it optimized to 33MHz with a lenght of 10cm to go in tree trunk .
I will consider worst humidity case same as probe soaked into water, this case reflectance is maximum and best humidity case free air, almost no reflectance so detector should ouput minimum.

Any help is welcome,
Best regards,Vito

- - - Updated - - -

Reflectrometry TDR

 

[chuckey]

Have you tested the conductance of the "water" found in tree trunks? I was under the impression that it is sap, and contains sugars and other complex chemicals, that feed the branches from the roots. If it has a high conductance (low resistivity) then a resonant tuned circuit will have its Q altered, which would give a much larger output then your lossy capacitor method.
Frank

 

[vitoa]

High condutance means also more nutrient concentration. If same water applied but one tap water and other tap water with nutrients , relative condutance will show more nutrient concentration on plant for same quantity water aplied.
But now only need to know if this principle works.

For example the probe soaked into water will give full output, because water is acting like shield and waves reflected back to rf power detector strip.
A kind of wave guide, if air all waves go away and almost no reflectance, meaning low voltage output from power detector circuit.

The ouputed signal by oscillator will be 33MHz @ 3Vpp
The received signal by rf detector must be higher than -34dBm ( 12.617 mV )

Are 33MHz enought to use this reflectance principle? Would the signal transmission between one strip and other enought?

Does this principle works as I describe? Is 33MHz @ 3Vpp output enought to do this porpouse?

 

[D.A.(Tony)Stewart]

The medium is basically a dielectric or an insulator. It will have capacitance, and a dissipation factor with leakage. There will be differences between concentric layers and longitudinally as well as some branches more than others.

Water has such a high dielectric constant (60) compared to dry wood due to moisture content differences. The dominate effect will be capacitive coupling of the moisture. If you measure water with various types from deionized to tap water or rain water, you will find a huge variation in leakage resistance or in other words the quality of the insulator to hold a charge will improve as minerals are removed and deionized. Contaminants may or may not affect leakage currents. I doubt there is any significant RF property unless there is an acid, base or salt. If you compare microwaving a cup of water for 1 minute with these elements , you can tell by the temperature what affect each has on absorption. Salt will have the biggest effect, then minerals. For example pure beeswax is a good insulator and it is also pretty transparent to microwave. ( from my own personal tests of wax rings in 1kW oven)

Without some basis of material science properties and precise control of gap, surface cleanliness from sap etc, I would expect more false positives with this method.

I think it would be far more informative to measure from branch to root or branch to local ground using a BODE plot between trunk and roots or ground and measure the tan delta or the dissipation factor. They do this now for power transformers with cellulose and oil insulation.

 

[vitoa]

Hi, the conductivity of water is afted with mineral or nutrients added. So if nutrients add the conductivity will be higher
but this sensor is not to get an absolute moisture value but a relative value.

Need to finish prototype to get some conclusions

Here is the draw of the sensor attached to tree with adhesive for example

 

[D.A.(Tony)Stewart]

So are you planning to measure quadrature impedance of s11 and s12?
or standing wave ratio?

Try an RF Network Analyzer first and scan over a wide range of f from 0.1 to 100MHz. ( relative to 50 Ohms)

A Reflectometer uses a pulse to measures large conductive impedance changes such as opens and shorts. or is used to measure small characteristic impedance mismatch with a known matched source source and load impedance, which is not in your plans using 33MHz.

This is more a change of dielectric properties at various depths or treating the medium as homogeneous at a fixed frequency. Why 33MHz?


Is this a longitudinal study over time? or a lateral study comparing trees of similar types?

THe trunk diameter between the plates will affect the relative capacitance.

 

[volker@muehlhaus]

In order to build a simple,low power oscilator and rf detector circuit I used two IC for this porpouse
One is a precise 33MHz oscilator outputing 3000mV pp wave between ground and Vout
To save time on a peak detector circuit with skottkey diode I found some simple RF power detector IC LTC5507.
It will provide an output voltage proporcional to rf in, according datasheet -18dbm(0.08Vpp) to 14dbm(3.17Vpp) input wil have almost linear Vout from 200mV to 2000mV, perfect for microntroler adc.

One strip is connected to 33MHz@3Vpp oscilator, the other strip connects to rf detector ic.
This ideia is to do one strip work as antena emitter and the other strip anntena receptor, the higher the moisture inside tree trunk the higher refletance of signal get back to detector strip, increasing output voltage on rf detetor.

This is not a good measurement concept for this task. Simple, but not accurate and not reliable.

Also, it doesn't do what you think it does. . At 33MHz (=9m wavelength in air) it doesn't use the reflection/transmission principle that you have drawn in your sketch, because your sensor/test device is all small compared to the wavelength and there is no relevant "wave propagation" behaviour. No reflectometry measurement here. Your design is simply measuring the impedance of a lossy capacitor, with the tree as the capacitor's dielectric.

Your receive signal would depend on both capacitance and loss between the electrodes. From the measurement you can't tell what has changed, humidity (water) or conductivity of the water (what you call nutrient concentration). Not a useful design.

If you want to measure only humidity (i.e. effective permittivity) it would be much better to create a sensor where only capacitance is measured. For example, the capacitance of your sensor could be part of an oscillator, and you evaluate oscillation frequency. That enables you to determine sensor capacitance (effective permittivity).