Design and Improvement of an Automated Tool for Quality Control and Performance Assessment of PV Modules

Abstract

Photovoltaic (PV) systems are at the heart of the energy transition, providing an essential source of clean, renewable energy for applications such as solar pumping, which is essential for irrigation and rural water supply. However, their efficiency depends directly on the quality and performance of the modules, which are often affected by defects or unfavorable environmental conditions. This article presents the development of an innovative automated tool designed for advanced characterization of PV modules by analyzing key parameters such as voltage and current. The system integrates measurement sensors (voltage, current, temperature, etc.), an Arduino Mega board and an SD card, enabling real-time data collection, processing, and recording under various environmental conditions. The results of the experimental tests demonstrate a significant improvement in the PV panel selection process, ensuring optimized choices at the time of purchase and rigorous monitoring during operation. This innovation contributes to maximizing energy performance and extending panel longevity, reinforcing their role in the transition to a sustainable energy model.

1. Introduction

Photovoltaic (PV) systems play a central role in the global energy transition, offering a sustainable and renewable energy solution for electricity generation [1,2]. Thanks to their ability to convert solar energy directly into electricity, these systems are used in a wide variety of applications, ranging from large-scale power plants to stand-alone systems in remote areas. Their efficiency and reliability depend on the quality of the PV modules and their ability to operate optimally in a variety of environmental conditions [3,4]. Among the many applications of PV technologies, solar pumping systems have proved particularly advantageous, especially in rural areas for agricultural irrigation and water supply. These systems make it possible to reduce dependence on fossil fuels [5] while providing access to water in areas where connection to the electricity grid is limited or non-existent. However, to maximize their potential, it is crucial to ensure rigorous quality control of the PV modules used and to continuously optimize their energy yield. In solar pumping systems, as in other PV systems, the performance of PV modules fluctuates according to climatic conditions, their lifespan, and the condition of the panels themselves (solar cell degradation, manufacturing defects or soiling). It is therefore essential to implement a tool that automatically assesses the performance of photovoltaic modules, not only to ensure that these systems operate efficiently and sustainably [4], but also to guarantee the quality of the modules when they are purchased. Such a tool must be capable of measuring key parameters such as voltage and current in real time and assessing power, series and parallel resistance, form factor, conversion efficiency, etc., in order to detect any faults and adjust operations to maximize the energy efficiency of these panels, which still have a low conversion efficiency [6].

Measurement of the current–voltage (I-V) characteristic curve is usually carried out by connecting a variable load to the terminals of the photovoltaic panel, which forces the operating point to cover the entire I-V characteristic. To do this, different types of variable resistive loads, capacitive load and active load based on linear circuits (based on transistors) are used in the literature to obtain this curve.

In approaches based on variable resistive loads, the photovoltaic (PV) array is loaded by a matrix of configurable resistors. Each configuration corresponds to a specific overall resistance value, forcing the PV array to reach a distinct operating point. By adjusting the load configuration manually [7] or electronically [8], it is possible to collect operating points along the I-V curve [9]. This method is very simple to implement, straightforward and cost-effective, and produces notable results for characterizing low-power PV modules [10]. However, the number of possible configurations is limited, and the overall resistance values are fixed; moreover, for high power ratings, this method becomes cumbersome, bulky and expensive [11]. Despite this, the variable resistance approach is still widely used, as it allows fast, static measurements.

The electronic charging method generally uses power transistors, such as BJTs, MOSFETs or IGBTs [12]. When a MOSFET operates in its linear region (ohmic region), it acts as a variable resistor, where the resistance between drain and source (R_DS) varies as a function of the control voltage between gate and source (V_GS). Conversely, a BJT operates as a regulated current source, providing a controllable impedance by progressively modulating the input current ramp signal [13]. The study in [14] proposes the design of an I-V curve plotter using an electronic load approach, combined with a low-cost wireless sensor network (WSN), allowing automated data logging for distributed photovoltaic modules [11]. Using this method allows the I-V curve to be refined and allows scans in any direction, from open circuit to short circuit and vice versa. However, MOSFETs have a finite value of R_DS(on) in the saturation region, which leads to measurement errors near I_sc. Furthermore, MOSFET transistors cannot generate I_DS in the cut-off region, which leads to incorrect measurement near V_oc [11]. Transistors are only used to characterize PV modules with low power ratings, as they can only maintain high power for a few milliseconds as a load. For high power ratings, a heat sink and cooling fans are required for safe operation [10].

The capacitor charging approach relies on the capacitor’s ability to emulate a variable resistance. When a DC voltage from the photovoltaic module is applied to the capacitor, the capacitor becomes charged, resulting in a gradual decrease in current magnitude and an increase in voltage [13]. This method allows passive charging of the capacitor until the voltage exceeds that of the open circuit while recording the data points of the characteristic curve. It uses multiple high-voltage capacitors connected to active switches to meet different requirements for monitoring the current–voltage (I-V) curve as the load changes [10]. This approach is commonly used in commercial portable I-V curve tracers due to its good scalability properties as well as its low volume and specific weight. In addition, I_sc and V_oc measurements are made with minimal error and ripple. However, this technique requires an additional discharge circuit to safely dissipate the capacitor’s stored charge. Additionally, when characterizing high-power cells or modules, the size of the capacitor becomes large and bulky.

It is clear from the above work that one of the remaining major challenges is the development of an automated solution for capturing I-V curves with increased accuracy while minimizing the cost and complexity of the system and maximizing its durability and adaptability to various applications.

This article proposes the development and enhancement of an automated tool for assessing the quality of the energy efficiency of photovoltaic systems using an approach based on the automated characterization of photovoltaic modules. The proposed tool rapidly measures and analyzes the electrical parameters of PV modules at various load levels to identify optimal performance points and ensure maximum utilization of the panel’s energy potential.

The document is structured as follows: Section 2 details the electronic design of the proposed system. Section 3 is devoted to the development of the system control algorithm and the approach used for parameter extraction. Section 4 describes the experimental implementation of the device testing. Section 5 analyzes and discusses the experimental results. The conclusion of this work, together with the outlook, is presented in Section 6.

2. Electronic Design of the Proposed System

Figure 1 depicted the wiring diagram for the proposed device. It works as follows: after powering up, the panel to be characterized is connected to the device. Pressing the TEST button starts the characterization operation, which lasts about 10 s. The Arduino Mega first sends a control signal to the relay, which closes its contact, then generates a variable duty cycle PWM signal applied to the base of the 2N2222A bipolar transistor. The latter acts as a power interface for the IRLZ34NPbF MOSFET, used as a variable load. The shunt resistor is used to measure the current delivered by the panel, while the 150 kΩ and 12 kΩ resistors connected as a bridge divider are used to estimate the panel voltage. The LM35DZ temperature sensor measures the panel’s operating temperature. Additionally, 4.7 V Zener diodes protect the Arduino’s inputs from overvoltage, and the MBR10100 diode protects the device from reverse polarity. Once the characterization is complete, the Arduino sends a control signal to the relay, which this time opens its contact. The extracted parameters are then displayed on the 20 × 4 LCD screen. To save the data to the SD card, it is necessary to press the SAVE button. The input characteristics of the proposed system are listed in

Table 1. Input characteristics of the proposed system.

Measurement Ranges	Values
Voltage (V)	50
Current (A)	15
Temperature (°C)	0–150

.

The choice of the Arduino Mega as the control platform for our proposed system was primarily motivated by its affordability, ease of programming, broad community support and compatibility with a variety of sensors and actuators. These features make it an attractive option for prototyping and experimental validation. However, it is important to recognize that the low cost of the Arduino Mega and other components comes with some drawbacks related to power consumption and durability, which are discussed below:

• Power consumption

The Arduino Mega is based on the ATmega2560 microcontroller (manufactured by Microchip, a company based in the United States of America), which is not specifically optimized for low-power applications. Its typical consumption varies from around 50 to 70 mA in operation, which can be considered high compared with that of other microcontrollers designed for energy-saving systems.

• Durability

As a general-purpose development board, the Arduino Mega is not designed to operate in industrial or harsh environments. Factors such as temperature variations, humidity, dust and mechanical stress can adversely affect its performance and longevity. In addition, its use of standard rather than industrial quality components may reduce its reliability over a prolonged period of use.

Despite these limitations, the Arduino Mega remains a judicious choice for the current stage of this research because of its simplicity and its compatibility with the proposed control strategy.

2.1. The Photovoltaic Generator

In this work, the photovoltaic generator using the single diode model was modeled and simulated. This model is not only simple to implement but also accurate [15]. Furthermore, studies [16,17] conclude in that this model is the most appropriate for monocrystalline cells. The electrical circuit of this model is shown in Figure 2.

By applying Kirchhoff’s law to the circuit in Figure 2, the expression in Equation (1) is obtained, which represents the current delivered by the PV cell.

$$I=I_{ph}-I_o\left[exp\left({\displaystyle \frac{V+R_{s}I}{n{N}_{sc}{V}_t}}\right)-1\right]-{\displaystyle \frac{V+R_{s}I}{R_{sh}}}$$

(1)

With

$$V_t={\displaystyle \frac{kT}{q}}$$

(2)

where V is the voltage of the PV panel, V_t is the thermal voltage, I_ph is the generated photocurrent, q is the electron charge, I_o is the diode reverse saturation current, n is the diode ideality factor, k is the Boltzmann constant, T is the surface cell temperature and N_sc is total number of series cells, R_sh and R_s are the parallel and series resistances, respectively.

The photovoltaic (PV) panel used in this article is the solar photovoltaic Helios module. Its data sheet is presented in

Table 2. Electrical parameters of the solar photovoltaic Helios module.

Parameters	Values	Symbols
Maximum power (W)	60	Pmpp
Temperature coefficient of Isc (A/K)	0.04	ksc
Maximum current (A)	3.75	Impp
Short-circuit current (A)	4.01	Isc
Maximum voltage (V)	17.3	Vmpp
Series cells	36	Nsc
Temperature coefficient of Voc (V/K)	−0.35	koc
Open-circuit voltage (V)	21.6	Voc
Parallel cell	1	Np

and gives specific parameters under standard test conditions (STC: T = 298.15 K and G = 1000 W/m²). However, some essential data for PV panel implementation, such as diode saturation current, photocurrent, ideality factor, and series and parallel resistances, are not included. To determine these parameters, a simple tool provided by MathWorks, called “PV Array” and available in Matlab R2020a, was used [18] to automatically generate the missing values from the data sheet. Subsequently, a photovoltaic panel model was designed and simulated using module surface temperature (°C), irradiance (W/m²), series resistance (Ω) and parallel resistance (Ω) as input parameters. For the purposes of this simulation, the values R_s = 0.01 Ω and R_sh = 1000 Ω were set. By varying the meteorological parameters, such as irradiance and PV module temperature, the characteristic curves of the photovoltaic panel were obtained and are presented in Figure 3.

The characteristic curves of the photovoltaic panel presented in Figure 3 show that, like any other semiconductor component, they are affected by variations in temperature and are also sensitive to fluctuations in irradiance [18,19,20]. It can be seen that the voltage V_oc decreased with increasing temperature. This was due to the temperature dependence of the band gap of the semiconductor material, which reduces the potential energy of the charge carriers. In addition, the current I_sc depends mainly on solar irradiation: higher irradiation generates more photons, thus increasing the number of charge carriers.

2.2. Electronic Load

When selecting MOSFETs for the electronic load, it is essential to take into account specific criteria to ensure the safety and accuracy of measurements. The drain-source voltage (V_DS) of the transistors must be at least 15% higher than the open-circuit voltage of the photovoltaic array, while the drain-source current (I_DS) must exceed the short-circuit current of the array by 10%. In addition, this choice is influenced by other important criteria, such as component availability, cost and minimizing the PCB footprint.

Table 3. Characteristics of the IRLZ34NPbF reference transistor.

Measurement Ranges	Values
VDS (V)	55
IDS (A)	30
RDS(ON) (Ω)	0.035

shows the characteristics of the MOSFET_S used, V_DS, I_DS, R_DS(ON) and the operating temperature range.

The MOSFET functions as an electronic switch. When the MOSFET is driven by a PWM signal, the apparent resistance of the MOSFET is an average of the resistance between two states (ON and OFF), as a function of the duty cycle or duty cycle of the PWM signal. If R_DS(ON) is the resistance when it is switched on and R_DS(OFF) is the resistance when it is switched off (which is practically infinite), the apparent resistance R_load perceived by the solar panel is a function of the duty cycle of the PWM signal applied between its gate and its source. This apparent resistance, in view of the characteristic curves of the MOSFET, can be approximated as the relationship in Equation (3).

$$R_{load}={\displaystyle \frac{R_{DS\left(ON\right)}}{D}}$$

(3)

–

If D = 0, this means that the MOSFET is switched off for the whole period and does not conduct any current, so R_load tends towards infinity.

–

If D = 1, this means that the MOSFET is always on and the apparent resistance is simply R_load = R_DS(ON).

In this way, the apparent resistance perceived by the load is reduced as the duty cycle D is increased, and vice versa.

2.3. Measurement Sensors

The current is measured using a 0.33 Ω shunt resistor instead of an ACS712 current sensor due to availability constraints. For voltage measurement, a simple resistor-based divider circuit is used to measure the photovoltaic array’s voltage. This voltage divider reduces the generator output voltage to below 5 V using a combination of low-tolerance resistors. Temperature measurement is achieved using an analog sensor with an accuracy of less than ±1 °C. This sensor can measure temperatures ranging from −55 to +150 °C with a sensitivity of 10 mV/°C.

3. System Control Algorithm Development and Parameter Extraction

3.1. System Control Algorithm

The control algorithm developed for this device is shown in Figure 4. This algorithm controls the duty cycle of a PWM signal applied to the electronic load by notionally varying the load across the photovoltaic array. It then measures the key electrical parameters (voltage, current) and analyzes the data collected to extract the essential internal characteristics of the array. It runs in a loop, automatically varying the duty cycle and recording the corresponding measurements to identify and extract the points of optimum performance.

The organization chart consists of several key stages. First, during system initialization, the output pins for the PWM signal, current and voltage sensors, LCD display and SD card are defined. The duty cycle values are initialized, and tables are created to store voltage, current, and power values. Additionally, the PWM frequency is configured to match the characteristics of the solar pumping system and PV modules. Next, in the duty cycle variation stage, the algorithm decreases the PWM signal’s duty cycle in 0.5% steps from 100% to 0%, allowing sufficient time for stable measurement of electrical and meteorological parameters. At each step, the PWM signal is applied to the electronic load, modifying the electrical load at the photovoltaic module terminals to obtain various operating conditions. The third stage, data acquisition, involves measuring instantaneous values of current, voltage, temperature and irradiance whenever the electronic load varies. These measurements are recorded in data tables, associating each voltage and current value with the corresponding duty cycle and power. In the analysis and extraction of parameters stage, once the duty cycle reaches 0%, a control signal is sent to an electromagnetic relay to disconnect the generator from the dummy load. The algorithm then analyzes the recorded data to determine the maximum power point and its coordinates, the open-circuit voltage, the short-circuit current and other parameters such as series and parallel resistance and form factor. Finally, in the display, storage and transfer of results stage, the extracted key data is displayed on an LCD screen for real-time visualization. The data is then stored on an SD card and transferred to a smartphone application, allowing remote viewing of the I-V curve and extracted data.

3.2. Approach Used to Extract Series Resistance

This is the internal resistance of the solar panel due to the materials and connections. A high series resistance reduces the panel’s efficiency by causing an internal voltage drop when a large current is generated. Series resistance mainly affects the I-V curve in the region where the voltage is high and the current is low (close to V_oc) [4,21,22]. This portion of the I-V curve can be approximated by a linear relationship between V and I. Thus, the series resistance can be estimated from the derivative of the I-V curve in this region. Figure 5 displays the influence of series resistance for irradiance values of 1000 W/m², at a temperature of 200 °C and a shunt resistance of 300 Ω.

In practice, the series resistance R_s can be approximated by the slope of the current–voltage characteristic curve at points close to V_oc according to the relationship in Equation (4), where ∆V represents the variation in voltage in a small portion close to V_oc and ∆I denotes the corresponding variation in current in the same portion of the curve.

$$R_S\approx {\displaystyle \frac{∆V}{∆I}}\approx {\displaystyle \frac{V_{i+1}-V_i}{I_{i+1}-I_i}}$$

(4)

To refine the estimation of the series resistance R_s, several points close to V_oc are taken into account. The idea is to improve accuracy by using a statistical approach rather than relying solely on a single pair of points on the current–voltage (I-V) characteristic curve. To achieve this, the proposed approach is to apply a linear regression to all points close to V_oc. This method models the relationship between two variables using a straight line. In our case, the I-V characteristic curve becomes approximately linear in the region close to V_oc, where the current is low and the voltage high. So a straight line in the form of Equation (5) is fitted to this part of the I-V characteristic curve, where V is the voltage, I is the current, and R_s represents the series resistance, corresponding to the slope of the straight line.

$$V=R_{S}I+V_{Oc}$$

(5)

In this case study, the relationship in Equation (6) is used to determine R_s (the slope) from the points (I_i, V_i) obtained by the proposed device, where n is the number of points selected, I_i and V_i are the current and voltage values, respectively, for each point and R_s represents the series resistance. Thus, instead of arbitrarily selecting a pair of points, linear regression exploits all the data available in the linear zone, thus optimizing the information extracted from the current–voltage (I-V) characteristic curve.

$$R_s={\displaystyle \frac{n\sum _{i=1}^n{I}_i{V}_i-{\sum }_{i=1}^n{I}_i\sum _{i=1}^n{V}_i}{n\sum _{i=1}^n{I}_i^2-{\left(\sum _{i=1}^n{I}_i\right)}^2}}$$

(6)

3.3. Approach Used to Extract Parallel Resistance

The shunt resistor represents the losses due to leakage currents through the solar panel. It is placed in parallel with the panel junction, and its influence is shown in Figure 6. A low shunt resistance indicates high losses, resulting in lower panel efficiency. To extract this parameter, the same method as that used to extract the series resistance was used, with a few differences. Parallel resistance mainly affects the current–voltage (I-V) characteristic curve in the region where the voltage is close to zero and the current is maximum (close to I_sc) [4,21,22]. This portion of the I-V curve can be approximated by a linear relationship between V and I, given by Equation (7). Thus, the shunt resistance can be estimated from the derivative of the I-V curve in this region.

$$I=-{\displaystyle \frac{1}{R_{sh}}}V+I_{sc}$$

(7)

By applying the method of least squares to the data points close to the short-circuit point, the relationship for estimating the parallel resistance R_sh can be accurately formulated according to Equation (8).

$$R_{sh}={\displaystyle \frac{n\sum _{i=1}^n{I}_i{V}_i-{\sum }_{i=1}^n{I}_i\sum _{i=1}^n{V}_i}{n\sum _{i=1}^n{V}_i^2-{\left(\sum _{i=1}^n{V}_i\right)}^2}}$$

(8)

3.4. Approach Used to Extract Fill Factor

The form factor is a key indicator of a solar panel’s performance, measuring its efficiency in converting the captured energy into useful power. It is defined by Equation (9) and represents the fraction of the theoretical maximum power that can actually be used. A high form factor indicates a better-quality, more efficient panel. In general, high-efficiency solar panels have a form factor of between 0.5 and 0.85 [18,19,20].

$$FF={\displaystyle \frac{V_{mpp}{I}_{mpp}}{V_{oc}{I}_{sc}}}$$

(9)

4. Implementation Aspects

To assess the effectiveness of the proposed system, an experimental test bench was designed to characterize the performance of the photovoltaic panels. The experimental campaign was carried out at the LESIA Laboratory of the University of Ngaoundéré, Cameroon, on 21 November 2024. The hardware configuration is shown in Figure 7. The solar photovoltaic Helios module to be characterized was connected to the load via a set of essential sensors, allowing real-time monitoring of key parameters such as current, voltage and panel temperature. This data was transmitted to a processing unit, which played a central role in analyzing and managing the information. Once the measurements were collected, they were displayed instantly on an LCD screen for quick reference. Thanks to an intuitive user interface, the operator can interact with the system and, with a simple command pulse, trigger the recording of data on a storage medium, such as an SD card, for later in-depth analysis. The entire device was powered by a dedicated power supply, ensuring stable, continuous operation of the various electronic components. This test bench provides a robust and flexible platform for evaluating the performance of photovoltaic panels, offering both real-time monitoring and data logging capability for optimized operation.

Figure 8 shows an internal view of our experimental prototype. It consists of an Arduino Mega, acting as the control interface, and a power interface driving the IRLZ34NPbF MOSFET (manufactured by Vishay Intertechnology in Pennsylvania in the United States of America) used as a variable load. The system also incorporates Zener diodes, a shunt resistor, various resistors, a man–machine interface button and the LM35DZ temperature sensor (manufactured by Microchip, a company based in the United States of America). The characteristics of the various components used are detailed in

Table 4. Main components used in the developed prototype I-V curve tracer.

	Description	Model
MCU		Arduino Mega
Current sensor	Shunt resistor	0.33 Ω, ±0.1%
Voltage sensor	Resistive voltage divider
Temperature sensor		LM35DZ
LCD display	20 × 4 + I2C blue
MOSFET	N canal MOSFET	IRLZ34NPbF MOSFET
Keyboard	Save Button and Test Button
Diodes	Zener diodes	2N2222A, MBR 10100
Relay	Electromechanically relay	125VDC 10A

. As shown in

Table 5. Economic evaluation of the proposed prototype compared with commercial devices.

	Price (EUR)
Proposed prototype	50.00
Proposed I-V tracer [5]	355.00
RS ISM 490A	1261.63
Seaward PV200	1720.80
Amprobe Solar-600	2048.05
DS-100C	5298.22

, the total cost of the prototype is around EUR 50, and should fall in the event of mass production. Table 5 clearly demonstrates the cost-effectiveness of the proposed card compared to that of commercial devices, which generally use bulky and expensive charging capacitors.

The sensors used in the proposed device were calibrated using measurement instruments on the TB-1200 electronic workbench, achieving an accuracy of less than 1%. This level of accuracy is comparable to, or even superior to, that of certain commercial devices used for similar applications, such as the Amprobe Solar-600 and Seaward PV200.

Several experiments were carried out to analyze the impact of sunlight conditions on the performance of the photovoltaic module. First, the photovoltaic module was fully exposed to light, as shown in Figure 7. Subsequently, some cells of the module were partially shaded in order to simulate the partial shading phenomena of 20% and 50%, shown in Figure 9a and Figure 9b, respectively.

Finally, to complete our study, the temperature at the surface of the module was measured. The experimental configuration is shown in Figure 10.

5. Experimental Results

In this section, the results obtained during the experimental campaign carried out at the LESIA Laboratory are presented. These data constitute an essential set of in-depth experiments on PV solar technology. During this campaign, the current–voltage (I-V) and the power–voltage (P-V) curves were measured with scientific thoroughness, maintaining a constant sampling frequency, thus guaranteeing the reliability and comparability of the results.

Figure 11 illustrates the evolution of solar irradiance and temperature over the experimental day. The blue curve represents irradiance (W/m²), which rose gradually in the morning, reaching a maximum around midday, and then fell gradually in the afternoon, following the solar cycle. The orange curve, representing temperature (°C), showed a more stable evolution: it rose slightly at the start of the day, remained relatively constant at around 35–40 °C despite irradiance variations, then began to fall later in the afternoon. This difference can be explained by thermal inertia, which means that temperature did not fall immediately after irradiance decreased. The figure shows the relationship between these two parameters, highlighting the time lag between peak irradiance and peak temperature.

5.1. Performance Under Normal Operation

Figure 12 show typical plots of the current–voltage (I-V) and power–voltage (P-V) output characteristic curves of the prototype device for the clean solar photovoltaic Helios module under different weather conditions. The difference between the performance and the effect of varying weather conditions combined with the panel’s manufacturing quality can be seen in the curves. This is to be expected, as the electrical characteristics of a panel are highly dependent on the characteristics of the materials, the climate and the quality of the manufacturing process.

Figure 12a,b illustrate tests carried out under clear, unclouded skies, characterized by very high irradiance. Figure 12c, on the other hand, shows a test carried out under cloudy skies but with high irradiance. Finally, Figure 12d shows a test carried out under cloudy skies with medium irradiance.

The

Table 6. Parameter characteristics extracted at normal operation.

Extracted Parameters
Curve	Pmppt (W)	Imppt (A)	Vmppt (V)	Isc (A)	Voc (V)	Rs (Ω)	Rsh (Ω)	FF
(a)	57.38	3.33	17.25	3.39	20.36	0.93	263.4	0.83
(b)	54.56	3.16	17.26	3.29	20.25	0.95	133.49	0.82
(c)	49.03	2.29	16.80	3.00	19.63	0.97	205.00	0.83
(d)	37.93	2.20	17.22	2.35	19.75	1.1	194.91	0.82

present the parameter characteristics extracted at normal operation. From Figure 12 and Table 6 it is clear that when our dummy load was fully powered, the MOSFET in it did not fully reach or break the maximum open-circuit voltage, as indicated in [11]. This was due to the contribution of the combination of the MOSFET’s drain-to-source resistor, R_DS, and the 0.33 Ω current-sensing resistor at its source, i.e., (R_DS + R_S = 35 mΩ + 0.33 Ω = 365 mΩ). One of the plots shows that V_pv = 0.7 V when I_sc = 3.37 A and V_oc = 20.36 V when I_sc = 0.03 A. To guarantee the reliability of the measurements made by our device, one of the things we did was calibrate our various sensors (voltage and current) using the instruments on the TB-1200 electronic workbench, which guarantees deviations of no more than 1% from the equipment’s instruments.

Additionally, the temperature of the module’s underside during these tests ranged from approximately 36.5 to 50 °C, representing an increase of about 68.5% to over 100% compared to the module’s temperature under STC conditions (STC = 25 °C). This may also have contributed to the slight voltage drop compared to the manufacturer’s V_oc specification of 21.5 V.

The current–voltage (I-V) and power–voltage (P-V) characteristic curves of the tested module reveal its electrical performance under various weather conditions. The key parameters extracted (maximum power, voltage and current under short-circuit and open-circuit conditions, fill factor, series and parallel resistances) presented in Table 4 allow for evaluating the module’s quality and its associated losses.

The series resistance is an indicator of ohmic losses through the contacts, internal connections and semiconductor materials of the module. The R_s values varied very slightly, ranging from 0.82 to 1.1 Ω. This narrow range indicates better electrical performance as it minimizes power losses caused by Joule heating. Additionally, we observed that as irradiance and temperature decreased, the series resistance increased, and vice versa, which aligns with the findings of [22,23].

The parallel resistance reflects current leakage through junctions or defects in the module. A high R_sh value is desirable as it indicates minimal leakage current losses. The Rsh values ranged from 194.91 to 263.40 Ω, indicating a module with minimal leakage losses and good electrical insulation.

Modules with low series resistance (R_s) and high shunt resistance (R_sh) exhibit superior overall performance, characterized by higher maximum power (P_mppt) and a better fill factor (FF) [4], resulting in P-V and I-V curves that closely resemble ideal conditions. Notably, the fill factor of these modules remains nearly constant despite variations in irradiance and temperature, as observed in [23]. In contrast, modules with high R_s and low R_sh experience significant reductions in P_mppt, short-circuit current (I_sc) and FF due to increased ohmic resistances and current leakage, leading to substantial performance losses.

Comparing the results of simulations (see Figure 3) with those of the experimental measurements (see Figure 12), the following differences can be observed:

• The simulated open-circuit voltage (V_oc) was about 9% higher than the measured value.
• The measured short-circuit current (I_sc) was 6.7% lower than that obtained by simulation.
• The measured maximum power was 13% lower than the simulated one.
• The measured voltage at the point of maximum power (V_mppt) was 5.5% below the simulated value.

The 10–13% overall difference in maximum power can be attributed to several factors, including:

• Losses due to parasitic resistances (wiring, connections), which were not always accurately taken into account in the simulation.
• Temperature variations, which influence solar panel performance and may differ between actual conditions and model assumptions.
• Uncertainties and inaccuracies in experimental measurements, which can introduce a bias into the comparison.

Furthermore, these deviations are strongly influenced by the values of the series (R_s) and parallel (R_sh) resistances used in the simulations. As shown in Section 3.2 and Section 3.3, these parameters directly affect the shape of the I-V and P-V curves, and hence the estimated performance of the solar panel.

These differences highlight the fact that, although modeling can provide a satisfactory estimate of general trends, the adjustment of model parameters to experimental data remains essential. A more rigorous approach would involve extracting R_s and R_sh from actual measurements, or explicitly incorporating losses into the model in order to improve prediction accuracy and significantly reduce these discrepancies.

5.2. Experimental Result for 20% Partial Shading

To assess the impact of partial shading, tests were carried out by progressively covering different sections of the panel. Once the characterization test was launched, 20% of the panel surface was shaded. The results of this experiment are shown in Figure 13, while

Table 7. Parameter characteristics extracted at 20% partial shade.

Extracted Parameters
Pmppt (W)	Imppt (A)	Vmppt (V)	Isc (A)	Voc (V)	Rs (Ω)	Rsh (Ω)	FF
46.09	3.87	11.91	3.98	18.69	1.27	203.11	0.62

presents the parameters extracted for this level of shading.

5.3. Experimental Result for 50% Partial Shading

The panel was then progressively shaded until 50% partial shading was achieved. The results of this experiment are shown in Figure 14, while

Table 8. Parameter characteristics extracted at 50% partial shade.

Extracted Parameters
Pmppt (W)	Imppt (A)	Vmppt (V)	Isc (A)	Voc (V)	Rs (Ω)	Rsh (Ω)	FF
35.06	3.13	11.2	3.37	18.93	1.79	185.70	0.55

shows the parameters extracted for this level of shading.

From Figure 13 and Figure 14 and Table 7 and Table 8, it can be seen that the series resistance (R_s) increases and the shunt resistance (R_sh) decreases as a function of the intensity of the partial shading applied to the cells of the panel tested. In addition, this partial shading has a significant effect on the fill factor (FF), leading to a reduction in the efficiency of the panel and, consequently, the efficiency of the system in which it is used.

Analysis of the series and parallel resistances highlights the manufacturing quality and state of health of the modules tested. Regarding the fill factor, it can be noted that it was significantly impacted by partial shading but remains within the range reported in the conclusions of [7]. This analysis allows us to conclude that this panel is of good quality if we exclude the effects of simulated shading.

During these partial shading tests, increases in panel surface temperature of 1.5 °C and 2.5 °C, respectively, were observed. This was probably due to the activation of bypass diodes when certain cells are shaded. These diodes protect the shaded cells and allow the current to bypass the shaded areas. However, this process introduces additional resistance into the overall circuit, resulting in a temporary, localized temperature rise (the “hot-spot” effect).

6. Conclusions

This article aims to propose an automated tool for characterizing and evaluating the quality of solar panels, optimizing their operation under real-world conditions while ensuring quality assurance upon acquisition. The primary issue addressed is the limitations of traditional methods for measuring photovoltaic (PV) module performance, which are often costly, complex and poorly suited to the user’s needs. To overcome these challenges, an automated solution was developed to capture current–voltage curves with enhanced precision while minimizing costs and system complexity, as well as maximizing durability and adaptability to various applications. The results demonstrate that this tool enables a fast and accurate analysis of PV module performance, highlighting its potential to improve the energy efficiency and longevity of photovoltaic systems, particularly in critical applications such as solar-powered water pumping systems in rural areas. The successful implementation of the device on a 60 Wc PV module confirms its feasibility and paves the way for broader applications. In the short term, future work will focus on integrating predictive diagnostic modules based on artificial intelligence to anticipate failures and estimate the actual lifespan of panels according to site-specific weather conditions. In the long term, efforts toward miniaturization and integration into IoT systems will enable real-time, remote monitoring, enhancing the attractiveness and reliability of photovoltaic technologies in the global energy transition. The fact that the system is very inexpensive and can be built on the site makes it easy to use for quick estimates in projects that do not require great precision. The proposed kit is also extremely useful for students’ practical work, enabling them to visualize different PV and IV profiles and assess the impact of faults or shading on PV systems.