Recently I wanted to play with solar power but I realized that they don't really make small versions of solar chargers for my favorite chemistry(lifepo4).
My idea was to use an LM2596 or similar low parts count switcher with a current sense amplifier and precision rectifier to make a CC-CV source like some of the cheap ebay regulators do, except use a digital potentiometer for the current setting, and control that in software to adjust to the maximum possible setting before the input voltage drops too low to keep turning it up. Probably wouldn't reach the real MPPT but it might be pretty decent.
My plan is to charge 2 cells in series, and to give each cell it's own op amp+transistor shunt regulator to balance and as an extra protection against overcharging. I'd probably add a (poly) fuse on the panel input too for good measure.
Does any of this sound reasonable? I'd imagine a DIY LiFePo4 charger should be about as safe as a DIY NiMH charger and people do those all the time, but I'd like to hear if anyone has done anything similar.
IIRC once you get LifePo4 cells balanced (ie before you use them) there's little reason to continue balancing them every time you charge. At least with large batteries such as in a solar system.
You DO need to monitor the voltage of each cell as going over for even a brief time is bad. I think simple systems just cut the charge to all cells if any of them reach max V, wait for a while then reapply charge. Certainly I know people on another forum that do this and it appears to work.