Accurate flash testing

of high-efficiency PV modules with QuickSun® 
solar simulators using Capacitance Compensation (CAC)

High-efficiency solar cells – such as passivated emitter and rear contact (PERC) cells, heterojunction technology (HJT) cells, and interdigitated back-contact (IBC) cells – are currently gaining in production volume and market share. However, the solar simulators used in many manufacturing lines cannot accurately measure the maximum power and other characteristics of these devices. Typically, maximum power is underestimated, causing manufacturers to sell their products with lower power, and thus lower price, than necessary.

The difficulty of accurately measuring these products is caused by the high capacitance – ability to store electric charge – of high-efficiency technologies. Due to well-passivated surfaces and high effective minority carrier lifetimes, these architectures store a considerable amount of electric charge within the devices. Consequently, charging and discharging of the device during flash testing distorts the measured I-V curve and maximum power when short flash pulses are applied.

As steady-state solar simulators and flash testers with extensively long flash pulses are unnecessarily expensive and cumbersome, Endeas has developed a solution that is applicable to flash testers with relatively short flash pulses. The Capacitance Compensation (CAC) method enables accurate I-V curve and power measurements of any high-efficiency photovoltaic (PV) product with QuickSun® solar simulators using a flash pulse of less than 40 milliseconds.

How the Capacitance Compensation method works

During flash testing, a current-voltage (I-V) curve is obtained by sweeping the output voltage from open-circuit to short-circuit conditions (or vice versa) while recording the output current induced by flash illumination.

The topmost figure shows uncorrected I-V curves corresponding to a high-efficiency solar cell with an open-circuit voltage of 720 mV. A linear sweep of the output voltage with a duration of 20 ms was applied in both the forward direction (from short to open circuit) and the reverse direction. The measurement artefacts due to capacitive charge storage in the solar cell are visible around the maximum power point (MP) as an underestimated output current during the forward sweep and an overestimated output current during the reverse sweep as compared to the simulated steady-state I-V curve. Note that many solar simulators typically determine the I-V characteristics based only on a single forward sweep of the voltage, causing the maximum power to be underestimated when high-efficiency products are measured.

The second figure shows the equivalent circuit of a typical c-Si PV cell, including the steady-state components and the capacitive components causing the measurement artefacts during short flash pulses. It is worth noticing the basic difference between these two component groups. The current going to equivalent circuit diodes and the shunt resistance depend solely on voltage and are therefore independent of the rate at which the voltage is swept during a flash pulse. In contrast, the current going to capacitances depends not only on the voltage but also on its rate of change, which makes the cell behave differently depending on the rate at which the voltage is swept during the flash pulse. The capacitive components are negligible in traditional PV cells, enabling the voltage to be swept at any rate, but this is no longer the case in recent high-efficiency PV technologies.

The Capacitance Compensation (CAC) method developed by Endeas measures the capacitance of the PV cell or module based on the current and voltage recorded during a normal flash pulse. By observing how the measured current behaves while a forward sweep charges and a successive reverse sweep discharges the device, the method accurately measures the device’s capacitance. Once the capacitance is known, the influence of capacitive components is eliminated and the steady-state I-V curve and the maximum power are constructed. A comprehensive explanation of the method is available in the whitepaper downloadable from this link.

Experimental proof of the accuracy of the Capacitance Compensation method

The accuracy of the CAC method in correcting experimental data has been evaluated using a state-of-the-art high-efficiency solar cell with an open-circuit voltage of 720 mV. In the experiment, forward- and reverse-swept I-V curves were measured and used to obtain the CAC-corrected current. A point-wise averaged I-V curve (denoted as simple average) was also calculated for comparison. The deviation of the efficiencies determined from these I-V curves from the steady-state efficiency are compared in the last figure with different sweep times.

The forward sweep marked in blue corresponds to a traditional flash measurement, which can be seen to include unacceptable error even with the longest experimental sweep times. Importantly, the efficiency determined via the CAC method stays within <0.1%rel error with a sweep time of 20 ms or longer. In contrast, the efficiency determined by the simple averaging of the I-V curves approaches the steady-state efficiency with increasing sweep time but introduces a significant error (>0.1 %rel) below 60 ms. The CAC method thus provides a considerable improvement in comparison to the simple point-wise averaging of the I-V curves.

Because the CAC method allows decreasing the sweep time below 20 ms, even in the case of the investigated state-of-the-art solar cell with a high open-circuit voltage of 720 mV, both the forward and reverse sweeps can easily be measured within a single flash pulse of only 40 ms. This is a tremendous technical advantage from the perspective of flash system durability and total testing time.

The above-presented cell results are fully transferrable to the module level. Hence, the CAC method allows accurate determination of steady-state I-V parameters even in state-of-the-art high-efficiency solar cells and modules whose open-circuit voltages exceed 700 mV/cell. The CAC method is included in all current QuickSun® solar simulator models from Endeas.