The solar cell IV testing system is used to characterize the electrical performance of photovoltaic devices, and the accuracy of its test results directly affects component selection, process optimization, and power plant design. However, various internal and external factors may lead to testing deviations. The following analysis will be conducted from four dimensions: environmental conditions, equipment performance, sample status, and measurement methods.
1、 Environmental condition control
1. Light stability
The standard test requires simulating the solar spectrum (AM1.5G), but actual light sources (such as xenon lamps and halogen lamps) have irradiance fluctuations. A deviation of light intensity within ± 2% can cause significant changes in short-circuit current, and a real-time radiometer feedback control system is required. In addition, insufficient spectral matching (abnormal UV/IR ratio) can lead to distortion in quantum efficiency measurement.
2. Temperature regulation
For every 1 ℃ increase in battery junction temperature, the open circuit voltage decreases by approximately 0.4%. When the temperature control module is not used, long-term testing can cause the battery to self heat up, forming a negative feedback loop. It is recommended to use liquid cooled or air-cooled heat sinks to stabilize the battery temperature at 25 ℃± 1 ℃.
3. Environmental cleanliness
Dust particles adhering to the surface of the battery will produce a shadow effect, reducing the effective light receiving area; Moisture infiltration may cause electrode corrosion. The test should be completed in a clean room and a dust cover should be installed if necessary.
2、 Equipment hardware performance
1. Uniformity of light source
Non uniform light will generate transverse current inside the battery, causing a virtual low fill factor (FF). High quality homogenizers can achieve illumination uniformity of>95% and require regular verification of spatial radiation distribution.
2. Electronic load accuracy
The dynamic load matching capability determines the sampling density of the IV curve. Low quality loads may experience oscillations near the open circuit voltage region, resulting in the loss of critical inflection point data. It is recommended to use a precision source table with a resolution of 1mV/μ A.
3. Wiring resistance and poor contact
Insufficient probe pressure or metal oxidation can increase the series resistance (Rs), resulting in a lower measured FF. Metallographic whisker welding combined with spring top pins can minimize contact impedance to the greatest extent possible.
3、 Sample preparation and status
1. Integrity of surface passivation layer
If the silicon wafer cutting damage is not fully hydrogenated and repaired, the carrier recombination rate will increase sharply, leading to a deterioration of the filling factor. The surface cleaning process before PECVD deposition of anti reflective film is crucial.
2. Gate line design and shading loss
The penetration of silver paste through PN junction during sintering will cause leakage channels, and the width of the main gate line and the spacing between the fine gates need to balance conductivity and shading area. Screen printing alignment deviation>5 μ m significantly affects current collection efficiency.
3. Hidden cracks and micro defects
Invisible lattice dislocations can easily form leakage current under reverse bias, and EL image detection can identify such hidden dangers in advance. The IV curve of the damaged battery shows abnormal bending.
4、 Measurement parameter settings
1. Voltage scanning rate
Fast scanning (<10ms/point) cannot capture the process of capacitor charging and discharging, while slow scanning (>1s/point) may cause hysteresis due to capacitor discharge. The optimal speed needs to be dynamically adjusted according to the battery capacitance value.
2. Bias voltage range
Scanning only up to 80% of the open circuit voltage will miss detailed information about the high resistance area. The complete curve should cover the range from -5V to Voc+5V to observe the reverse saturation characteristics.
3. Data smoothing algorithm
Excessive filtering will smooth out local maxima, and using a reasonable moving average method (window size ≤ 5 data points) can achieve a balance between denoising and detail preservation.