Frequency Response of an Agricultural Fence and the Implications for Data Transmission
McMullan, J. (2014). Frequency Response of an Agricultural Fence and the Implications for Data Transmission (Thesis, Master of Philosophy (MPhil)). University of Waikato, Hamilton, New Zealand. Retrieved from https://hdl.handle.net/10289/8832
Permanent Research Commons link: https://hdl.handle.net/10289/8832
The electric fence has been used as a data transmission medium in Gallagher Products for a number of years. This has allowed the energizer to be remotely controlled and, with the current generation, to monitor the performance of the fence remotely. Very little investigation has been conducted into determining the optimum frequency bands to transmit in to give optimum performance. We propose that the fence spectrum be split into three frequency bands. A Fence Pulse Guard Band which extends from DC to 10 kHz, a Low Frequency Channel which extends from 10 kHz to 250 kHz, and a High Frequency Channel which extends from 250 kHz to 10 MHz. The fence frequency response is dependent on the length of the fence and is dominated by transmission line effects and radiative losses. For the test fence, the spectrum up to 250 kHz is flat without any frequency selective fading. Above 250 kHz, the spectrum is very unstable and the frequency selective fading can exceed 15 dB. Operating in this region requires an advanced system to utilise the available bandwidth. The impedance of the human operator in the system is best characterised as a fractional capacitor in series with a resistor. Higher frequencies are attenuated less up to 10 MHz after which the impedance is dominated by the resistance. The impedance of an insulating joint is best characterised as a capacitor in series with a resistor. Higher frequencies are attenuated less and are the preferred method for reducing the effect of insulating joints. The Low Frequency Channel is suitable for less robust systems which cannot tolerate frequency selective attenuation. The High Frequency Channel is suitable for robust systems which prioritise performance. We present a number of possible solutions for improving the efficiency of the modulation and error correction strategies. Solution 3 utilising Phase Shift Keying (PSK) with eight symbols and Trellis Code Modulation (TCM) is recommended as the first solution to be implemented and evaluated. A forward error correction strategy as outlined in Solution 1 is also recommended for implementation first. This research suggests that the electric fence system could be significantly improved in performance and reliability using the methods mentioned above but at some cost.
University of Waikato
- Higher Degree Theses