Like many of us who are interested in solar power, I bought a Harbor Freight 45 Watt Solar Panel kit for about $150 which came with three solar panels, and a charge controller. It was exciting times when I set it up several months ago to see the front panel lit up. I thought that I was getting enough solar juice to charge my SLA batteries. However, the excitement was short lived. I was foolish enough thinking that the light was all powered by solar for the very first night, and forgot the fact that the battery was fully charged already before I connected it to the solar panel. Just like many found out, the charge controller doesn't appear to be working as we expected. For example, my SLA battery originally had 12.5 Volt. During the day, the front panel shows about 13-14 volts so I leave the charge controller on, but when I come back from work, I found that the battery wasn't charged at all. In fact, the voltage would go down about 0.1 volt to become 12.4 volt. What's going on?
As usual I searched for answers on the internet but most people simply complain that this particular charge controller is not efficient, or it can only be used as a power distribution device. To get a better charger, one would need to invest another $100. Basically, the solar panel kit is only good for demonstrating the concept (I have to say that from outside, the charge controller looks nice and solid).
Recently, I got some spare time and started making diagnosis on this charge controller and I found that while it indeed charges during the day, it discharges as the sun goes down. I suspect that when the voltage from the solar panel is below the voltage of the battery, it may be discharging the battery. I don't know this for sure since I haven't open the device to see the design. This device does discharge at night at about 30mA if you leave it on (the LED display uses power unless you turn the LED display off which is recommended).
Step 1: Add a Blocking Diode to Stop Discharging the Battery
One experiment I did was to simply add a blocking diode on the back of the charge controller. I initially placed the diode on the battery side but based on recommendations from "evilmunkey", the diode should now be put on the solar panel side. The diode I am using is 1N5822 which happens to be in my toolbox so I guess many other diodes would work as well. 1N5822 can handle 40v 3A so I use two of them in parallel to make sure it can handle the current from the solar panel. Also make sure that the cathode lead of the 1N5822 should be connected to the charge controller side while the anode should go to the solar panel side. Previously I experimented with putting the diode on the battery side but I was concerned that it might overcharge the battery so now the diode is on the solar panel side. After the modification, my SLA battery is charged better.
Step 2: Caveats With the Blocking Diode
Since I don't know the charging circuit inside the box, the effect of this added blocking diode on the solar panel and battery charging is not clear. Theoretically the charge controller should already have blocking diodes on the solar panel side, otherwise it will give all back the solar power at night. The ultimate solution would be to look at the components inside the box and see what it's doing. There are many different kinds of controllers and they all work differently. I assume that this is a PWM type but I don't really know. For example, does it have a microcontroller to control the voltage? What about current limiting? Given all the unknowns, I wouldn't risk charging any expensive batteries. In my case, I salvaged some old 12V SLA batteries by refiling with Epsom water. I also leave the battery outdoor in case of outgassing. So far it works fine and the battery is actually revived to a certain extent. This is very much an experiment so be careful not to overcharge. Finally, make sure to add an inline fuse near the battery terminal (such as 5A). It's better to be on the safe side. If anyone has any better suggestions improving this unit, please let me know. I may open the box in the future and write another instructable later. For now, I need to move on to the next project with desulfator.