ERROR ESTIMATION IN INDUSTRIAL SOLAR CELL SORTING (Part I)

摘要

Industrial sorting of photovoltaic solar cells is mainly based on the measurement of a unique I-V point, understandard conditions, at a known voltage close to the one of the maximum power point. The current applicableregulation deals more with solar simulators (what they can achieve), and with measurement for a complete I-Vcharacteristic (I-V curve), than with classification itself. The purpose of the present work is to make a study ofseveral factors influencing the measurement used for industrial photovoltaic solar cells sorting.Factors to be under study are mainly spatial uniformity of the irradiance over the cell under test, temporalstability and spectral match of the lighting source, cell temperature, bias voltage use in sorting, and measurementratio. This work try to establish which of all the factors seems to be more critical for a right sorting. Although all thisstudies are integrated in an ambitious project named SORTERR (acronym for SortSorting Err ing Errors), in present paper weintroduce results for two aspects only: Spatial non-uniformity, and time instability.A measurement protocol has been established for each factor under study, based in a simplified model of solarcell. Experimental measurements have been drawn using both a Class A xenon solar simulator and a halogen dichroiclamps based one, so comparisons between different lighting sources could be drawn (both are "steady-state" solarsimulators). Temperature conditions have been imposed by means of a thermostatic chuck, so cells temperature couldbe adequately controlled. Measurement and record instruments have been remained as a constant for both lightingsources.The main conclusion extracted from experiments done in this Part-I is that applicable regulation to solarsimulators is relatively lax in some aspect, namely temporal stability of the lighting source, what have beendemonstrated to be critical.