|DIY Solar Charge Controller|
Solar Charge Controller
I've owned a small 5W, 12V solar panel for a number of years. When I lived in an apartment, there was little use for it, but once I moved into a house I started thinking of practical uses for it. Illuminating an exterior storage shack seemed like a good use for the panel so I set out to build the electronics around this small "Solar System". As in several other projects in this website, I like the challenge of reusing stuff I already own, rather than seeking the "perfect" component . Though this approach doesn't often result in the optimum design, I find that one of the rewards of doing Electronics as a hobby is that you need not worry about the optimum design at the best price-point. And so it was that I ended-up also reusing a 6Ah lead-acid battery for this system; not that this was the better choice for the project, but that it was readily available. The two major components of the project thus selected (battery and solar panel) I started thinking about the battery charge controller.
Why is a charge controller needed?
The "rule" of thumb is as follows: if you keep the charging current below 3% of the batterie's capacity, then it is safe to charge the battery without a controller. So, for the example at hand, a 6Ah battery can be safely charged with a current of up to 6 x 0.03 = 180 mA. My 5W panel can deliver up to 325 mA according to the spec sheet, so clearly some sort of controller is needed. Put another way, we need to "divert " up to 145 mA (325-180) from the battery to another load during peak charging conditions. The circuit in Figure 1 accomplishes this task.
Figure 1 - Circuit Schematics
The main component is a TL431 "shunt" regulator. (Tip: if you have an old PC switching supply board laying around, you will likely find one or two TL431s inside that can be "salvaged" for this purpose). As depicted in it's internal block diagram, this IC has a fairly precise 2.5V internal reference that is compared against an external voltage. Whenever the voltage at the REF input pin exceeds 2.5 V the output NPN transistor is turned ON and the power transistor T1 also turns ON thus " diverting" some of the panel's output current towards the LEDs. R1 and R2 are designed so that this happens when the battery voltage exceeds about 14.2V. I decided against including a trim-pot here because of reliability concerns. Trim-pots can open over time and in this application the box is subject to some significant temperature variations. If an adjustment is needed to set the exact battery charge voltage, it is preferable to add a resistor in parallel with R1 or R2. This type of charge controller is known as a "shunt" regulator because the panel is "shunted" when the battery voltage exceeds a certain threshold. The "shunt" in this case is a set of 10 LEDs in parallel drawing 15 mA each for a total of 150 mA (thus meeting our initial 145 mA minimum requirement as calculated above). You can use a power resistor instead with the appropriate power rating. However, I wanted to put the excess power from the panel to some useful use and this is why I chose the LEDs. These are 20 mA high-brightness white LEDs running at 15 mA. I chose a current below their rated spec to help improve the life of the circuit. The series limiting resistor is calculated as follows: R = (14 - 2 - VCEsat) / 15e-3 = 753 -> 820 Ohm. T1 should have a small heat-sink to improve it's reliability. Though it operates in saturation most of the time, it will still heat-up slightly under normal operating conditions. D1 is a low-drop Schottky diode and prevents the battery from discharging through the panel at night. Fuse F1 protects against Mr. Murphy and his laws.
You will notice that the LEDs will conduct a little even as long as there's some sunlight. This is because the TL431 isn't a perfect comparator as the diagram would suggest and allows some current through the base of T1 even when VREF is slightly less than 2.5V. This small base current is amplified by T1 and lights-up the LEDs. This is normal operation for this circuit. Once the VREF voltage exceeds 2.5 V (VBat > 14.2V) the T1's base current increases dramatically and saturates the transistor. At this stage, the LED current reaches it's maximum of about 15 mA as calculated above.
The circuit has been in operation in my shed for a few weeks now without any issues. See photo below showing the LEDs ON while charging.
I've been monitoring the battery voltage to make sure it is not overcharging and the circuit seems to be doing the job it was designed for. I'm keeping an eye on the LEDs also as I'm slightly concerned with their long-term reliability. If some of the LEDs fail, I might replace them with a simple power resistor. As a parting note, I must emphasize that this is a design for a very low-power solar installation. Though the concept can be extended for larger panels (and larger capacity batteries) I would recommend a more "conservative" approach in that case, possibly with two transistor controlled loads in parallel for redundancy protection. Be safe, and enjoy the free solar energy!