What simulation type to use in LTSPice to see a graph of varying OUTPUTS and not inputs

[MikeMl]

Even if you figure out how to model a varying resistance in LTSpice, you have lots of other issues...

The "resistance" of the inter-digitated probes on the PCB you showed in the picture is likely to be very high (>10megΩ), unless the PB is embedded in moist soil, where it might exhibit < 1megΩ. If it is embedded in soil, then the "sensor" will not last long if you put more than about 1Vdc between the probes due to electrolysis.

The correct way of measuring the "resistance" between the probes is to use an AC measurement, where there is no DC current component between the probes (basic electrochemistry)...

 

[SuzanneOC]

Hi Mike,

The sensors are not for soil measurement. Their application is to be embedded in a fabric. As the fabric gets wetter the output is recorded and converted into %age moisture reading. The expected lifetime is 14 days max and the 3v supply is from a coin cell battery.

I need to model this process.

 

[MikeMl]

Hi Mike,

The sensors are not for soil measurement. Their application is to be embedded in a fabric. As the fabric gets wetter the output is recorded and converted into %age moisture reading. The expected lifetime is 14 days max and the 3v supply is from a coin cell battery.

I need to model this process.

Have you done any measurements yet? I would expect that the resistance between the probes caused by loose contact with moist fabric will have an effective resistance of multi-MegΩ, or higher. The current at 10MegΩ will be I=E/R = 3/1e7 = 300nA, which is obviously not within the measurement range of the circuit you posted earlier.

 

[alec_t]

Pure water is a very poor conductor, so your measurements may reflect the concentration of dissolved salts/contaminants more than the % moisture.

 

[MikeMl]

...so your measurements may reflect the concentration of dissolved salts/contaminants more than the % moisture.

Sweat? Put the sensor in the arm pit?

 

[SuzanneOC]

And yet another way to simulate a variable resistor that doesn't require a pot.

It uses the Spice directive .step function to vary the resistance from 10 ohms to 1k ohms in 10 ohm steps (look in the Help file for .step).
(Note that R in the resistor definition must be enclosed with curly brackets).

The DC operating point simulation command then displays the output voltage versus the change in resistance.

Another advantage of this technique is that it directly plots the output voltage versus the resistance change.

View attachment 99077

Hi crutschow, this circuit works very well for my application. I see my error in not powering the opamp.
Many thanks