How to set a test bench for solar cell characterization, measure it correctly, and plot the power and I-V curves.
It may seem straightforward, but it took some time until I realize many things about solar cell characterization. I will try to summarize in this article all the important stuff you need to know before you start measuring a solar cell. I also wrote another small tutorial for measuring TEGs.
To obtain a nice power-curve like the following KXOB25-05X3F solar cell, we need to measure one by one all the operation points at different load and light conditions. Then plot in Matlab or excel and…
Is this information not provided by the manufacturer in the datasheet?
Mmmm…. Mainly NO! I found only basic information for 1 operation point, but rarely for different illuminations like the previous graph.
Also when you want to compare several light transducers, all must be characterized equally under the same conditions unifying the criteria.
Test Bench Set up: The circuit
Let’s start with the schematic we are going to use. In the next Figure, the energy transducers (solar cell) act as a voltage source while the load is a variable resistor, which forces a current on the circuit.
To measure output power from the source, two of the three basic electrical magnitudes (voltage, resistance or current) are required, as shown in the next Equation \ref{eq: power}.
The easiest and more straightforward method is to place a voltmeter in parallel with the solar cell and an ammeter in series with the potentiometer. You can also, measure the resistance and the voltage, but this implies disconnecting the circuit for each point.
Light Configuration
The amount of light seen from the light transducer in each experiment must be standardized to ensure repeatability and equal conditions for all the devices under test.
Case A:
The solar cell is lighted with a mix of a bulb and the natural light of the laboratory. By varying the distance between the bulb and the transducers, the light intensity is modified. You place a luxmeter nearby to measure the light and that’s it all!
Case B (better):
The solar cell is encapsulated inside a dark box to avoid external light disturbances and to ensure that the cell is illuminated exclusively by 3 white LEDs attached inside (Figure below). The leds are supplied by a controlled source, so the exactly brightness can be adjusted (and repeated).
I started using the Case A, but later I realize there were many variables while moving the lamp, if it is dark or daylight. Also it was not easy to adjust to an excatly lux quantity. So, then I built the dark box… and it worked pretty well 🙂
Correct Light Calibration
To adjust the exact amount of light in each experiment, and ensure repeatability, you need to measure it with the luxmeter, which is kind of unpractical and variable.
During my work, I take some measurements of my system with specific light parameters among others, let’s say 600 lux. If I repeat the experiment 2 weeks later or the day after, I need that all results match. moreover, if another person read my report 2 years later, he would be able to repeat the experiment and obtain the same numbers as I did.
A solution is to calibrate the voltage source of the LEDs. With this method, I had small deviations when I set the voltage to the LEDs, sometimes, the LEDs moved or they are warm or cold, or the voltage source is not well calibrated.
A much better option is to calibrate the open circuit voltage of the solar cell. For example, I know that at 800mV, my cell is under 680lux. That will be kind of fixed, although it also depends on other parameters, but not significantly.
If you measure the amount of light in lux versus the open circuit voltage, you will get a beautiful linear graph. To be more precise, I take the same data on different occasions, and I’ve got the average value of all of them. So, from now on, when I measure the open circuit voltage, I know the illumination in lux. I don’t need anymore the luxmeter!! 🙂
Note that this calibration is specific for each solar cell!!
The calibration table is the following:
Instrumentation
I used the following instruments, which I have available, but you may use other equivalent measurement devices:
- Hand multimeter ISO-Tech IDM 72
- Precision bench multimeter Keithley 2000 Multimeter
- DVM hand multimeter DVM645BI
- Fluorescent bulb light Philips PL 15W/865
- Light meter ILM 1332A
Ideally, ammeters have no voltage drop by measuring current. But real devices have a voltage drop due to the internal resistance of the instrument and they do alter the measurements, especially when the current is low. Live is tough!!
To minimize the interference to the measurements due to the internal resistance of the ammeter, I used different ammeters I had available in-house:
- Multimeter (IDM) was utilized for currents lower than 20 uA
- DVM precision device for higher currents.
Why?
The digital bench multimeter (DVM) has a resolution of only 1uA but a lower internal resistance, meanwhile the hand multimeter has a better resolution (0.1uA) but with higher internal resistance.
Lab Setup
With this setup is how I measured several solar cells, as I reported some of them here:
Data Acquisition
Once everything is settled, you only need to write down the data and organize it. I personally use a table for each lighting condition. Later, I take the matrix import in Matlab to a unique variable. To finish, I run a Matlab script to plot all the beautiful curves.
Below I post my Matlab template. Maybe it can help or inspire you to create your power and I-V curves. Feel free to copy and modify everything!
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Solar cell plot template
% Alberto Lopez
% MisCircuitos.com
% 14-nov-2022
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
close all;
f= figure(1);
%open('solarcell_sept-Ixys.mat');
data = lux500;
data2 = lux640;
data3 = lux680;
data4 = lux800;
data5 = lux900;
data6 = lux1000;
hold on
grid on
grid minor
%POWER CURVE
% plot(data6(:,1), data6(:,3)/1000,'o-');
% plot(data5(:,1), data5(:,3)/1000,'o-');
% plot(data4(:,1), data4(:,3)/1000,'o-');
% plot(data3(:,1), data3(:,3)/1000,'o-');
% plot(data2(:,1), data2(:,3)/1000,'o-');
% plot(data(:,1), data(:,3)/1000,'o-');
%I_V CURVE
plot(data6(:,1), data6(:,2),'o-');
plot(data5(:,1), data5(:,2),'o-');
plot(data4(:,1), data4(:,2),'o-');
plot(data3(:,1), data3(:,2),'o-');
plot(data2(:,1), data2(:,2),'o-');
plot(data(:,1), data(:,2),'o-');
%% LEGEND & TITLE
title({'Solar Cell',' IXYS- KXOB25-05X3F' })
xlabel('Voltage (mV)')
%ylabel('Power (\mu W)')
ylabel('Current(\mu A)')
legend('1000 lux','900 lux','800 lux','680 lux', '640 lux', '500 lux','Location','northeast')
%guardarImagen('IXYS_KXOB25-05X3F_Curve'); %Function to auto save the image in different formatsI hope this tutorial has been helpful, any questions, suggestions, or comments are more than welcome!!
