Spatial and Spectral Corrections for Integrating Sphere Photometry and Radiometry

The energy efficiency of lighting is expressed in terms of luminous efficacy. It is the ratio of the visible light emitted by the source to the power consumed in the process. The total amount of useful light produced by a source is described as luminous flux, which is often measured using integrating sphere photometers. These instruments consist of a hollow sphere and a detector whose output signal is proportional to the luminous flux emitted by the device under test. Non-ideal characteristics of photometers and integrating spheres induce measurement uncertainty on the luminous flux measured – and thus the resulting energy efficiency.

With solid-state lighting superseding the conventional incandescent and energy-saving lamp technologies, the associated measurement techniques need to be revised as well. As integrating sphere photometers are generally calibrated using incandescent light sources, the measurement uncertainty is increased when determining the luminous efficacy of solid-state lighting products such as LEDs. This uncertainty can be reduced by employing correction factors that take into account the imperfections of the measurement system.

In this dissertation, a method based on a fisheye-lens camera was developed to reduce measurement uncertainty due to spatial non-uniformities of integrating spheres. In order to calculate the spatial correction factor, the relative angular intensity distribution of the lamp under test is required. Traditionally, obtaining such a distribution has involved time-consuming and resource intensive goniophotometric measurements. With the fisheye camera method, the distribution can be measured in seconds using a fisheye camera installed into a port of the integrating sphere.

To reduce the measurement uncertainty due to differences in the spectra of the calibration source and the device under test, a new LED-based reference spectrum was developed for calibrating photometers. The reference spectrum is based on one of the eight LED illuminants that were developed in the same study to be employed in colorimetry. For the LED products and photometers tested, the new reference spectrum reduced the average spectral mismatch errors by a factor of two, when compared with an incandescent calibration source.