I have a 4.2V 120mA max solar panel and 3.7V 1200mAh battery. I'm trying to make a solar power system for my arduino.

I search through battery management IC's and generally understand the charge concept but at somepoint I'm confused. Lipo battery has a 4.2V overcharge level and 2.75V discharge level. Most of the documents connect battery and system load together on same line but do not explain well, how to cut off at 2.75V low voltage disconnect level.

What is the correct approach for this subject?

  • \$\begingroup\$ Is this question anything to do with solar cells? \$\endgroup\$
    – Andy aka
    Apr 3, 2014 at 9:09
  • \$\begingroup\$ Isn't charging a part of of solar cell topic at all? \$\endgroup\$
    – cheour
    Apr 3, 2014 at 9:12
  • \$\begingroup\$ Charging is charging whether it's from a solar cell or from an AC wall-wart or from another battery or from an alternator.. \$\endgroup\$
    – Andy aka
    Apr 3, 2014 at 9:14
  • \$\begingroup\$ At that perspective you are right, but I tought it would be helpful for who suffer same problem in the future and find it easy. \$\endgroup\$
    – cheour
    Apr 3, 2014 at 9:19
  • \$\begingroup\$ And possibly exclude those who are charging batteries but not from a solar cell? \$\endgroup\$
    – Andy aka
    Apr 3, 2014 at 11:05

1 Answer 1


unless the battery management/charger IC specifically does it, your system load WILL drain your battery to death. I have that issue on my BQ24090 single cell lipo charger i put on my little combat robot. The output of that charger goes to the lipo, but that it also the system load rail!

You actually need to monitor this issue yourself in most cases: one nice way to do it is monitor the battery voltage using a micro controller (your arduino!) ADC or even an opamp (probably better, then it's not relying on software to do it) which sees that the Lipo is now at 3V (safer to stop here, dont go so low as 2.75V) and then cut off/turn off the system. At this point you assume "the battery is flat" and you deal with it - in your case you would disconnect the load (system, not the lipo) until the battery voltage has been beefed up enough by your solar charge system to turn on the system again. Maybe the "turn back on" voltage is 4V?

do some investigations into battery monitoring/protection ICs, but basically you need a more complicated one to recover after the solar cells have brought up the battery level enough to continue again. Otherwise you can assume that a human can just replace the battery, and deactivate the "load disconnect" switch, and let the system continue now on a full battery?

  • \$\begingroup\$ Good answer, maybe I can put a fet to system load path and cut it's gate if battery connected node's voltage level below 3V's. I think I can use some kind of diode mechanism to do that. \$\endgroup\$
    – cheour
    Apr 3, 2014 at 11:58
  • \$\begingroup\$ some battery charger ICs have a power good output, you can use that to drive the logic to 'reset' a latching "off" logic circuit. Or you could make an SR-flip-flop. Where the output disconnects the system from the lipo while it recharges, and the Power Good output can then reset the condition. I guess you will have to look at whatever charge IC you have, and work with it. It will be hard to make an Opamp based approached, if the system load also drops power to the opamp haha! perhaps a resistor divider (as a threshold sort of thing) to the base of a transistor combination would work. \$\endgroup\$
    – KyranF
    Apr 3, 2014 at 13:08
  • \$\begingroup\$ have a look at implementing a schmitt trigger with hysteresis, out of two transistors - set the upper and lower voltage thresholds to be your "oh no! turn off!" and "oh, its okay now, turn back on" points. These should drive on/off a Pchannel FET or other form of high-side load switch. Make sure any FET you get has as low Gate turn-on voltage threshold as possible, seeing as you wil be dealing with low (~3V) voltages during these cases. \$\endgroup\$
    – KyranF
    Apr 3, 2014 at 13:10
  • \$\begingroup\$ @cheour - your solar panel is just too small for your battery. It will take way too long to fully charge your battery with that panel. That panel's specs are under full perfect sunlight conditions, anything less and the output will drop and charging will slow a lot. Even at full sunlight conditions it will take more than 12 hours to charge. (Even though your battery is 10x the output of the panel ao logic says it should take 10 hours, as the battery fills up, the charging slows to less than the output of the panel, so will take longer than 10 hours). \$\endgroup\$
    – Filek
    Apr 3, 2014 at 20:41
  • \$\begingroup\$ Many batteries have a PCB built in that will automatically shut the battery off when the battery gets to a certain low voltage - does your battery have this? If not, you can add one. Not the best thing to rely on, but it is another source of protection so that you don't ruin your battery. \$\endgroup\$
    – Filek
    Apr 3, 2014 at 20:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.