The purpose of this study was to test the output characteristics of a computer controlled fertigator. The specific objectives were: (1) to compare the measured electro-conductivity (EC) and Hydrogen ion concentration (pH) with the target EC and pH using software inputs that represented possible operating ranges of the fertigator; (2) to determine if the fertigator was capable of producing nutrient solutions that were chemically different, but had the same EC; (3) to determine a set of inputs and fertilizer salts for calibrating the fertigator.
The fertigator was programmed to deliver 80 L of water containing of various ion concentrations. The nutrient solutions were based on the standard concentrations of Hoagland's solution. A formula was used to calculate ECs of nutrient solutions and it was used to develop the test for nutrient solutions that were chemically different, but had the same EC.
Two formulas were developed to predict the pulse rate of the injectors based upon several parameters that needed to be know before the fertigator could be programmed properly. One formula predicted the pulse rate of the acid injector and the other formula predicted the pulse rate of the nutrient injectors. The Q-COM system could not pulse the injectors more 15 times per minute. The reason these two formulas were needed was that if the pulse rate was too high the system would not inject. A nonlinear regression analysis was used to predict the output EC based on the observed data collected. The nonlinear regression contained parameters modeled the time response and the final values of the EC It was the final value from the regression analysis that was used to compare the output characteristics of the nutrient solutions tested.
The results of this study showed that the fertigator was capable of creating nutrient solutions that were within 12% of the target value for five of the seven nutrient solutions tested. The standard deviation of the variability in the data was inversely related to the rate at which the injectors pulsed: the higher the pulse rate, the
lower the standard deviation. A regression analysis was conducted to determine the relation of standard deviation to pulse rate.
The fertigator was shown to make solutions that had the same EC but very different composition. Three out of four of the solutions with the same EC were within 10% of the target value; two of the four were within 5.3% of the target value. A calibration solution was designed to have the lowest variability in the data to allow for an accurate calibration. The calibration solution was a hybrid of the Mid EC solution and the High EC solution. The results of this study show that the calibration solution does have the lowest variability in the data.