Power-dependent small-signal model for fluorescent lamps based on a double-pole double-zero transfer function

In prior literature, the small-signal incremental impedance of fluorescent lamps has been characterized using a single-pole single-zero transfer function. This transfer function provides a good estimation of the lamp dynamics at rated power. However, when this model is used at low power levels, the accuracy is severely reduced; especially at high perturbation frequencies (5-20 kHz). In present work, a new double-zero double-pole transfer function is used to model the lamp dynamics with great accuracy even at low power levels. The best results have been obtained using a transfer function with two real zeros and a double complex pole. Using a new characterization procedure, the small signal model of a 26W compact fluorescent lamp has been obtained. All the terms of the transfer function (static gain, poles and zeros) have been calculated at different power levels. Different polynomial expressions have been used to approximate these terms as power dependent functions. The proposed characterization procedure also allows obtaining the equivalent resistance of the electrodes. This feature has been used to measure the effect of the extra heating imposed by the IEC60901 during dimmed operation.