Text

theses

With a growing world population placing ever increasing demands on agriculture
and an increasing scarcity of water suitable for irrigation, there is a need for efficient
irrigation techniques utilizing comprehensive knowledge of soil water content. While
commercial solutions are available that provide the required of information, none were
found that are low cost, low maintenance, and high durability. This thesis presents a sensor
and data backhaul design that addresses these three key parameters with measurable results.
The design uses a TPIP-K transmit-only wireless platform for remote reporting of the
measurement data. Since soil moisture measurements need only to be taken at regular but
not precise intervals and individual data points are not critical by themselves, no bidirectional
communications are necessary, leaving communication channels open to
measurement data only and allowing more sensors to be deployed for a given portion of
wireless spectrum. Additionally, since the most energy consuming component is the
wireless radio itself, the transmit-only design lends itself to long battery life because the
radio is only active during the brief times data needs to be reported. A long battery life
coupled with wireless communication means deployment is simple and unobtrusive to
other agricultural procedures. The design of the moisture sensor itself is based on a variable value capacitor whose dielectric material is the water in the soil, making it simple and
inexpensive to manufacture. The dielectric constants of air (εr = 1) and water (εr = 80) have
a large enough difference that the concentration of water per unit volume in the soil will
correspond to a specific capacitance value of the sensor. Making use of hardware features
specific to the MSP430 microprocessor on the TPIP-K, precise and repeatable
measurements can be made while maintaining low energy usage. Individual sensor
calibration for a particular soil is achieved with measurements taken at low and new battery
voltages, to account for battery aging, and approximately 25-30% and 100% field capacity,
the point at which any additional water is drained from the soil by gravity and made
unavailable to plants. Using bilinear interpolation based on the calibration data, the sensor
can than report the volumetric water content of the soil between the calibration points to
better than 3%, which is the range and detail of information most useful for agricultural
irrigation purposes. The resulting sensor solution meets the set design goals of developing
an inexpensive, low maintenance, and robust soil moisture sensor.

ETD_9335

M.S.

Includes bibliographical references

by Jakub Kolodziejski

doi:10.7282/t3-apcm-3a53

ETD graduate

The author owns the copyright to this work.