Part 1 of this project is here, and part 2 is here.
Yes, I'm still working on a good float charger for maintaining my batteries. See my other posts on this here and here for a recap of the project. I finally gave up on the LM7805 circuitry that comes with these devices. The Harbor Freight 42292 charger used to be controlled by a LM317 variable voltage regulator a long time ago. They probably changed the regulator at some point to lower cost or overcome a supply problem. However, it originally was a much better device; it even had a variable resistor in it to allow the end user to tweak it a bit to fit his purposes. I got a clue on this during my searches to get ideas on what to do. Here is the link to a schematic of the original device (link). Notice how different it is from the current circuitry? Just so I have a picture to talk about, here is the schematic from that site:
![]()
![]()
I still have a diode in series with the output to prevent problems with reversing the connection to the lead acid battery and used different values for all the resistors than the original had. The reason I changed the values was purely practical; it's what I had on hand. Notice that the parts count is a grand total of five? So, why doesn't the manufacturer do something like this? I don't have a clue. So, my schematic looks more like this:
![]()
Yes, I'm still working on a good float charger for maintaining my batteries. See my other posts on this here and here for a recap of the project. I finally gave up on the LM7805 circuitry that comes with these devices. The Harbor Freight 42292 charger used to be controlled by a LM317 variable voltage regulator a long time ago. They probably changed the regulator at some point to lower cost or overcome a supply problem. However, it originally was a much better device; it even had a variable resistor in it to allow the end user to tweak it a bit to fit his purposes. I got a clue on this during my searches to get ideas on what to do. Here is the link to a schematic of the original device (link). Notice how different it is from the current circuitry? Just so I have a picture to talk about, here is the schematic from that site:
Back in those days, the wall wart was just a transformer so the diodes to rectify the supply were in the little black box. IC1 is a LM317 and notice how they set up the resistor array to limit the range of available voltage and have a transistor to limit current? Nice design that you can find all over the web.
I just duplicated part of this on a small piece of protoboard that I cut to match the size of the existing board. It came out like this:
The input on the left is the wall wart's 19 VDC (apx) output; VR1 is a 1K, 25 turn potentiometer I picked up on eBay; R2 is 1.2K; and R3 is 180 ohms. This gives the charger a range of about 9.6 to 16 or so volts. To calibrate it, I put a 1K resistor across the output, set it for 13.4 volts and then took the resistor off. When I hooked it to the battery, it floated the battery up to 13.37 volts over an hour or so and just held it there.
I rely on the LM317 to take care of itself. It has over current and over temperature protection built in and regulates the voltage quite nicely. If I discharge the lead acid battery by pulling a lot of current such that the voltage drops down to 12.4 or so, the IC will get hot, but not dangerously so. Then over time, it charges the battery back up to the 13.4 volt level and just holds it there.
So, since I'm only using the wall wart, plastic enclosure, and heat sink of the original device, did I save any money? Yes. The wall wart alone is worth the price I paid for the original charger and the rest of it serves as a platform for the charger I ended up with.
So far, I've converted three of these devices to the new design and they are chugging away keeping batteries charged. I chose 13.4 V as the float point based on experience and a suggestion from a helpful person on the Arduino forum; this seems to be the best point to hold the battery without losing water to the charge current. I haven't had to add any water to any of the batteries yet. Over time, I'll know if I need to make adjustments to this, but at least now I can.
Thinking about this a bit, it brings up an interesting point. This wall wart can supply an amp easily and the output seems to be very good, that means I have just made a variable voltage 1 A supply that can serve various general purpose uses around the house. The next time they have these on sale with a 20% off coupon I may pick up a bunch of them and stockpile them for later use. Heck, since the LM317 can work as long as the difference from voltage in to voltage out isn't more than 40 V, I could use it to control the voltage from a 24 VAC transformer like I needed way back in the thermostat project. I needed 5 VDC from a rectified 37 VDC taken from an air conditioner power supply; this would have saved me about $20 in power supply cost.