This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. IEEE JOURNAL OF PHOTOVOLTAICS 1 Impact of Nonuniform Illumination and Probe Bar Shading on Solar Cell I–V Measurement Samuel Raj, Jian Wei Ho, Johnson Wong, and Armin G. Aberle Abstract—This paper examines the effect of nonuniform and probe-bar-shaded solar illumination in the I–V characterization of large area Si wafer solar cells. The illumination conditions were experimentally implemented by masking the solar cell during mea- surement. The results are examined in detail with simulations in Griddler, a finite element analysis software developed at the Solar Energy Research Institute of Singapore. The light I–V character- istics, particularly the fill factor, depend on the specific irradiance profiles relative to the current-collecting busbars due to effects on the effective series resistance R s . The effect is slight for small light nonuniformities (IEC 60904 Class A nonuniformity classification or <2%) but for light nonuniformities up to 20%, the standard deviation in the fill factor can reach about 0.19% absolute. Probe bar shading directly leads to an underestimation of the measured fill factor due to longer current paths and larger voltage drops. Index Terms—Current–voltage (I–V), Griddler, nonuniform illumination, probe bar shading, solar cell. I. INTRODUCTION T HE definition of standard testing conditions for solar cell current–voltage (I–V) measurements, precise definitions of the solar spectrum and classification of solar simulators [1], are consistent with the general motivation to render solar cell I–V testing a highly standardized and reproducible measurement between test laboratories. Most efforts in raising measurement accuracy, establishing traceability, and standardization focus on the prediction of device short-circuit current (I sc ) under the target spectrum [2]. Meanwhile, the cell open-circuit voltage (V oc ) is usually not sensitive to the probing configuration, and is accurate to within 0.5% uncertainty so long as the device tem- perature can be maintained to ±1°C. On the other hand, there is considerably less promotion of best practices toward obtaining I–V curves that yield accurate maximum power points (MPP) or fill factors (FF), although there are numerous factors related to light distribution and probing configuration that can impact Manuscript received March 20, 2017; accepted May 29, 2017. This work was supported by the Solar Energy Research Institute of Singapore (SERIS). SERIS is supported in part by the National University of Singapore and in part by the Singapore’s National Research Foundation through the Singapore Economic Development Board. (Corresponding author: Samuel Raj.) S. Raj, J. W. Ho, and J. Wong are with the Solar Energy Research Institute of Singapore, National University of Singapore, 119077, Singapore (e-mail: samuelraj@nus.edu.sg; jw.ho@nus.edu.sg; johnson.wong@nus.edu.sg). A. G. Aberle is with the Solar Energy Research Institute of Singapore and the Department of Electrical and Computer Engineering, National University of Singapore, 119077, Singapore (e-mail: armin.aberle@nus.edu.sg). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JPHOTOV.2017.2726558 these measured values. Implicitly, the ideal test condition of a so- lar cell prescribes close proximity of source and sense probes, identical voltages at current extraction points of like polarity, and perfectly uniform distribution of light over the cell plane. The impacts of deviations from these conditions have only been cursorily reported [3], while detailed work on the effects of nonuniform light illumination have mainly focused on concen- trator solar cells [4], [5]. The goal of this work is to investigate the typically encountered spatial variations in illumination in the testing of large area Si wafer solar cells, and examine whether routine setups for I–V testing suffice in achieving accurate FF measurements. There are namely two often encountered sources of light nonuniformity in the test cell plane: the first being nonuniformity in the incident light field itself, and the second being shading by the probe bars. Solar simulators that employ xenon short arc lamp, parabolic mirror, fly’s eye integrator, and collimating lens can typically meet the 2% spatial uniformity requirements for class A categorization, although bulb misalignments, degrada- tion in reflecting mirrors, nonuniformity developed in filtering elements, etc., can erode the spatial uniformity over time. Sys- tems that employ more than one lamp, or an array of light emit- ting diodes (LEDs) which involve multiple illumination sources that need to be integrated and homogenized over the test plane, present further challenges to achieving the required light uni- formity. Shading by probe bars is usually the result of finite probe bar thickness, misalignment between the probe bars and the cell busbars, bent probe bars made of printed circuit board or other flexible materials, and the casting of shadow due to large incident beam divergence. In the last scenario, if the beam divergence is 6°, a 5 cm high probe bar contacting the outer busbars of a three bus bar silicon wafer solar cell with 156 mm length would cast a shadow 1.45 mm wide. Probe bar shading can be avoided during the determination of I sc , by making use of Kelvin probes which only contact the busbars near the wafer edges for current extraction, but the same current extraction near wafer edges cannot be used for the determination of the I–V curve MPP, as the cell busbars are typically not designed to conduct large currents. This paper presents both experimental data on the variation of I–V parameters under different light distributions that might be encountered in realistic I–V testing conditions, as well as sup- plementary simulations using a detailed finite-element model [6], [7] of the solar cell plane under a large number of dif- ferent possible light distributions. The latter simulations are useful toward establishing the statistical distribution in the I–V 2156-3381 © 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications standards/publications/rights/index.html for more information.