Improving Spectral Responsivity Measurements by Correcting for Filter Bandwidth

Spectral responsivity (SR) measurements are key in characterizing photovoltaic (PV) devices and calculating spectral mismatch correction factors (MMF). SR is measured using different wavelengths of monochromatic light. In practical measurement systems, this light has a finite bandwidth, meaning that it is only quasi-monochromatic. The data analysis, however, typically assumes perfectly monochromatic light. The accuracy of the calculated SR then depends not only on the actual bandwidth of the monochromatic light used during the measurement but also on the intensity distribution inside the bandwidth. In our SR measurement system, we use quasi-monochromatic light obtained from a solar simulator and a range of bandpass filters. First, we measured and analyzed the spectral content of light transmitted for each filter. Then, we developed an iterative correction scheme for the measured SR. This numerical approach assumes no previous knowledge about the SR of the device under test (DUT). Applying this scheme for bandwidth correction to a range of PV cells with varying SR curves, we found that the corrections for most wavelengths are of minor importance. However, for those wavelengths where the shape of the SR curve of the reference device and DUT differ notably, the corrections are significant up to 20% and higher. We investigated the change in the spectral MMF, which was found to be negligible in most cases but not always. Therefore, the bandwidth correction is mainly important not only for improving SR measurements of PV devices but also relevant for the determination of spectral mismatch factors.